Report No. 8041 The World Bank and Senegal, 1960-87 August 31, 1989 Operations Evaluation Department FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. LIST OF ABBREVIATIONS AND ACRONYMS BCG - Boston Consulting Group BNDS - Banque Nationale du Développement Sénégalaise CVCCEP - Commission de Vérification de Contrôle des Comptes des Etablissements Publics CEAO - Communauté économique des Etats de l'Afrique de l'Ouest (Economic Community of West African States) CEM - Country Economic Memorandum CPP - Country Program Paper CSA - Commission de la Sécurité Alimentaire (Food Security Commission) DAC - Development Assistance Committee EFF - Extended Fund Facility (MF) GDI - Gross Domestic Investment ICOR - Incremental Capital-Output Ratio ICS - Industries Chimiquas du Sénégal ISRA - Institut Sénégk!Ris de Recherches Agronomiques MADIA - Managing Agricultural Development in Africa MOF - Ministry of Finance MOP - Ministry of Planning NPA - Nouvelle Politique Agricole OED - Operations Evaluation Department OMVS - Organisation pour la Mise en Valeur du Fleuve Sénégal ONCAD - Office National de Coopération et d'Assistance pour le Développement PCR - Project Completion Report PE - Public Enterprises PIP - Public Investment Program PPAR - Program/Project Performance Audit Report PREF - Plan à Moyen Terme de Redressement Economique et Financier (Medium-Term Recovery Plan) RCFS - Régie des Chemins de Fer du Sénégal RDA - Rural Development Agency SAED - Société d'Aménagement et d'Exploitation des Terres du Delta SAL - Structural Adjustment Loan SEFICS - Société d'Exploitation Ferroviaire deb Industries Chimiques du Sénégal SEIB - Société Electrique et Industrielle de Baol SENELEC - Société Nationale de l'Electricité SMIC - salaire minimum interprofessionnel garanti SODAGRI - Société de Développement Agricole et Industrielle SODEFITEX - Société de Développement des Fibres Textiles SODEVA - Société de Développement et de Vulgarisation Ag:icole SOFIDI - Société Financière pour le Développement Industriel SOFISEDIT - Société Financière Sénégalaise pour le Développement de l'Industrie et du Tourisme SOMIVAC - Société de la Mise en Valeur de la Casamance SONACOS - Société Nationale de Commercialisation des Oléagineux de Sénégal SONAR - Société Nationale d'Approvisionnement Rural USB - Union Sénégalaise de Banque FOR OFFICIAL USE ONLY THE WORLD BANK Washington. 0 C 20433 USA OffCe Co 0DrectoW-Cel~11 O)peratinns fvaluattum August 31, 1989 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: The World Bank and Senegal, 1960-87 Attached, for information, is a copy of a report entitled "The World Bank and Senegal, 1960-87" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY THE WORLD BANK AND SENEGAL, 1960-87 Table of Contents Page No. SUMMARY AND CONCLUSIONS ......................... .............. I: THE SETTING ................................................ 1 A. Lconomic Developments and Bank Strategy at the Macro-Level ......................................... 1 B. The Lending Program and the Bank's Sectoral Strategy .. 11 (1) Size and Composition of the Lending Program ...... 11 (2) Bank Sectoral Strategy ........................... 12 II: STRUCTURAL ADJUSTMENT AND CONDITIONALITY ................... 17 A. Structural Adjustment Objectives and Constraints as Perceived by the Bank and the Government ............ 17 B. Brief Description of the Three SALs ................... 20 C. Structural Adjustment and Conditionality within a Policy Framework .................................... 25 D. Analysis and Typology of SAL Conditions in Senegal .... 31 E. Basis for SAL Conditions and Impact on Performance .... 35 (1) Introduction .................................... 35 (2) Agriculture ...................................... 36 (a) Agricultural Price Policies ................. 37 (b) Agricultural Research ....................... 46 (c) Agricultural Performance and Its Determinants .............................. 48 (3) Improving Private Sector Efficiency .............. 52 (4) Public Investment and Finance .................... 59 (5) Public Sector Strengthening and Disengagement, and Private Sector Engagement .................. 64 (6) The Social Costs of Adjustment ................... 70 F. Foreign Assistance, and Implications of Foreign Debt and Debt Servicing on Investment and Growth ......... 73 G. Conclusions ........................................... 78 III: SECTORAL ISSUES AND AID COORDINATION ....................... 83 A. The Agricultural Sector ............................... 83 (1) Introduction ..................................... 83 (2) World Bank Lending to Agriculture, 1969-79 ....... 84 Senegalese Agriculture: Strategic Issues ........ 87 (a) Diversification ............................. 87 (b) Technology .................................. 91 (c) Producer Prices and Agricultural Taxation ... 92 (d) Institutions ................................ 94 (e) Conclusion .................................. 97 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Table of Contents (cont'd.) Page No. III: SECTORAL ISSUES AND AID COORDINATION (cont'd.) (3) Structural Adjustment and Agriculture, 1980-87 ... 99 (a) Producer Prices and Supply Response ......... 101 (b) Input Subsidies ............................. 108 (c) Institutional Reforms: The Role of the RDAs .................................. 109 (d) Conclusion .................................. 114 B. Industrial Sector ..................................... 115 C. The Transport Sector .................................. 121 (1) Introduction ..................................... 121 (2) Bank Group Assistance ............................ 122 (3) Roads ............................................ 123 (4) Railroads ........................................ 125 (5) Ports and Airports ............................... 127 (6) Conclusion ....................................... 129 D. Education ............................................. 131 E. Aid Coordination ...................................... 133 (1) Introduction ..................................... 133 (2) Problems and Constraints ......................... 134 (3) Bank Role ........................................ 136 Annexes 1: Policy Matrices and Conditions Attached to SALs I, II and III ............................................... 139 2: Development Objectives and Corresponding Performance Indicators ............................................... 166 3: Agricultural Output and Productivity of Major Crops, 1960/61-1986/87 .......................................... 167 4: Agricultural Production and Rainfall Data, 1960-86 ......... 168 51 Projects Approved by Board by Calendar Year, 1966-87 ....... 169 6: Net International Financial Flows to Senegal ............... 171 Map IBRD 21021R Table of Contents (cont'd.) Page No. List of Figures in the Text 1: Senegal: GDP and Primary Sector ........................... 5 2: Classification of Variables and Relationships among Constraints, Institutions and Structural Changes. Policy Measures, Performance Indicators and Development Objectires ................................... 30 List of Tables in the Text 1: Components of GDP and Total Expenditures, Selected Periods.................................................. 2 2: Five Selected Periods between 1960161 and 1985/86 .......... 3 3: Percentage Shares of Consumption and Investment in GDP, 1960161 to 1985/86 ....................................... 6 4: World Bank Comitments to Senegal, 1966-87 ................. 12 5: Suggested Classification of Variables for Policy Analysis ................................................. 27 6: Structural Adjustment Loans I-III: Typology of Conditions ............................................... 33 7: Classification of SAL Conditions Relating to Agriculture ............................................. 36 8: Official Producer Prices, 1970-86 .......................... 39 9: Agricultural Output and Productivity of Major Crops, 1960/61-1986/87 .......................................... 40 10: Relative Producer Prices (CFA/kg), 1970-86 ................. 42 11: Cumulative Annual Rinfall, 1961-87 ........................ 49 12: Selected Indicators of Agricultural Performance, 1981-87 ... 51 13: Classification of SAL Conditions Relating to Improving Private Sector Efficiency ................................ 52 14: Selected Industrial Indices of Production, 1979-86 ......... 54 15: Classification of SAL Conditions Relating to Public Investment and Finance; and to Public Sector Strengthening and Disengagement .......................... 60 16: Evolution of Parapublic Sector, 1962-82 .................... 64 17: Parapublic Sector, Aggregate Net Income and Government Subsidies, 1977178-1981/82 ............................... 65 18: / Key Indicators of Public Enterprise Sector and First Ten Enterprises to be Privatised ................... 68 19: Minimum Wage Rates in CFA/hour (SMIG) and Intersectoral Terms of Trade, 1979-86 .................................. 71 20: Balance of Payments Highlighting Net Foreign Assistance, 1981-87 ...................................... 75 21: Balance of Payments Projections Highlighting Net Foreign Assistance, 1986-92 ............................. 77 22: IBRD/IDA Project Disbursements ............................. 86 23: Consumer Nominal Protection Coefficients and Relative Nominal Protection for Cereals, 1970-85 .................. 106 24: Financial Position of Extension Agencies with Regard to the Budget and Disaggregated Figures for Particular Rural Sector Institutions. 1981-85 ............ 110 PREFACE This study is the third of a new series of country studies initiated in 1985. The first two reviewed the World Bank's relationship with Pakistan (Report No. 6048) and with Sri Lanka (Report No. 6074) and were issued to the Executive Directors and the President during 1986. A fourth study, of Tanzania, is expected to be issued shortly. The purpose of this series of reviews is to look at the inter- action between the Bank and selected borrowing countries over an extended period of time. Taken singly, each study should help the Bank and the borrowing country identify the strengths and weaknesses of their inter- action and contribute to making their relationship more fruitful in the future. Together the several studies should allow us to point to ways of making the development assistance the Bank provides more effective. The study of the interaction between the Bank and Senegal covers the period since the beginning of the relationship and is selsctive in its analysis. The focus is on those sectors and areas where the Bank has been most active, i.e. agriculture, industry, transport and education as well as structural adjustment. The study is based on a review of World Bank files, discussions with Bank staff members, and field missions. The kind cooperation and valuable assistance of Government officials, representatives of the private sector and of other donor agencies in Senegal are gratefully acknowledged. SUMMARY AND CONCLUSIONS Introduction i. In the quarter century since Independence, Senegal's GDP has grown at barely 2.Q? per annum, the lowest growth rate of any African state not affected by war or civil strife. At the same time population growth has been high and rising, currently estimated at 3Z per annum, so that real per capita income is now below the level reached in 1960. The primary sector, which still constitutes about 202 of total GDP, has shown the lowest growth rate at 1.7% per annum, but growth in the rest of the economy has also been sluggish at 3.9% per annum for the secondary sector and 2.12 per annum for the remainder. The period witnessed only a small structural shift in employment from agriculture to industry and an estimated 70% of the labor force continues to be principally employed in agriculture. ii. The agricultural sector has continued to play a predominant role in the dynamics of the development process. Between 1960 and today, wide fluctuations in agricultural output (the major component of primary sector GDP), largely a result of the vagaries of the weather, pulled overall GDP correspondingly up and down. Rainfall, while being highly variable, has also shown a sharply downward trend. Compounded by a very fragile resource base, soil degradation, swings in the international terms of trade and a poor domestic economic policy environment, agricultural output grew little in spite of all development efforts, and so did GDP. In a real sense, since 1960, some of the most basic constraints facing policymakers and donor agencies in Senegal in their attempts to develop the economy -- the fragile resource base, uncertain rainfall, high population growth -- have become even more binding over time. iii. Throughout the period 1960-87, Senegal has been the recipient of very large amounts of foreign assistance. The country is generally regarded as a model country in Africa, with political stability, continuity of administration and close relations with the donor community, including the Bank. Total net capital inflows from all sources averaged half a billion US dollars per annum during 1980-86, about three-quarters of it in the form of grants or on highly concessional terms. On a per capita basis this net flow amounted to $80 per annum, equivalent to about one-fifth of per capita GNP, and nearly twice the average of $45 per annum for the rest of the Sahelian countries. While a total of about 50 countries and organi- zations are involved in providing assistance to Senegal, the bulk of it originates from about 12 major donors, with France accounting for around 402 of the total during the past 1! years and IBRD/IDA for about 8.5Z. iv. These generous levels of external assistance combined with limited economic progress raise some fundamental questions about the effectiveness of foreign aid in Senegal and about the complexity of the development challenge. The focus of this report is on the effectiveness of the assistance -- in all its forms -- provided by the World Bank. Although the Bank's share in total aid to Senegal has been relatively small, and dwarfed by that of France, there has been some increase in recent years with the initiation of structural adjustment lending. More importantly, - ii - however, the Bank's role and indirect influence in economic policymaking in Senegal, through its economic and sector work and continuous dialogue, have always been substantial and therefore not necessarily related to the size of its financial contribution. And, since the late 1970s the Bank has clearly taken the lead role in providing policy advice, with the encourage- ment and support of other donors. In addition, on a country comparative basis, Senegal has been a preferred client of the Bank: with just 2.52 of the total population of all countries in "low-income Africa," it has received 4.7Z of total World Bank commitments to those countries during 1970-86. An Overview v. Senegal became a member of the Bank on August 31, 1962. The first Bank operation in independent Senegal was an IDA credit for railway development in 1966. Since then and through 1987 there have been a total of 61 lending operations for a total commitment of about $800 million. IDA credits account for 80% of this total; the bulk of IBRD lending went for projects in the transport, industry and tourism sectors. The last IBRD loan was extended in 1981, after which the country was judged not to be creditworthy for further IBRD lending. Structural adjustment lending has accounted for one-third of the total volume of commitments; project lending has been fairly heavily concentrated in the agricultural and transport sectors -- each accounting for 21? of the total volume of commitments -- although the Bank has been active in most sectors with a number of relatively small operations. After a slow start during 1966-71, Bank lending to Senegal grew rapidly to a yearly average of $85 million during the last four years. vi. From a macroeconomic standpoint, the Senegalese economy went through several distinct periods since 1960, with the country's economic fortunes generally fluctuating with rainfall and changes in the inter- national terms of trade. These fluctuations appear to have had little impact on the nature and sequencing of Bank projects, however, with one major exception: towards the end of the 1970s, a major turning point in Government and Bank strategy took place. Before that time the Bank was, in varying degrees, generally optimistic about the country's future, and its project lending covered most key sectors. The 1979 Economic Report, based on the findings of an economic mission in November 1976, presented a much more sober picture of the economic outlook and called into question previous Bank assessments. The country's growing financial difficulties and economic stagnation, aggravated by the 1979 oil price increase, together with often unsatisfactory past project experience led to a major shift in strategy with the recognition that the general policy framework in Senegal was at least partly responsible for this weak performance and therefore a Bank decision to support the change in strategy through introduction of structural adjustment lending in 1980 and reduced emphasis on projects in subsequent years. vii. Although the Senegalese economy has made little progress over the last quarter century (with per capita income actually declining), it would be wrong to assume that the sizeable investments, generated both through domestic savings and foreign borrowing, have yielded little or no return. A number of areas in which significant advances have been made can be identified and includes the development of cotton cultivation in Eastern Senegal; the growth in marketed agricultural output in both the Casamance and Fleuve regions; the development of maize and cowpeas in the more arid tracts of the Groundnut Basin, the development of transport infrastructure and in particular the strengthening of the road maintenance organization; and, more recently, cereals market liberalization; the reform of the agricultural cooperative system; and the whole set of structural adjustment policies aimed at improving the overall efficiency of the economy. The fact remains, however, that this progress has been both partial and insufficient and has, for the most part, been offset by other adverse developments. viii. While the Bank's activity is clearly only part of the overall economic policy and foreign assistance environment, it is, nevertheless, an important part. On the one hand, Bank support has played an important positive role in most of the success areas listed above. On the other hand, many Bank-supported projects were beset with major problems that compromised the overall effectiveness of the interventions. Moreover, while it is too early to judge the impact of the recent structural adjust- ment initiatives on economic performance, a critical ex-ante evaluati%r the possible impact of these initiatives suggests that the results are likely to fall considerably short of current expectations. ix. It is illustrated below that if any one factor can explain both the rather poor record of past Bank project assistance and the likely shortfalls in the growth objectives embedded in the current structural adjustment program, it is that the Bank has tended to overlook, downplay, or assume away major problems and issues. The Bank has also continued to provide advice and to make recommendations on the basis of very inadequate information and less than full understanding of the relationship between policy measures and policy objectives. This has not only often resulted in unfounded optimism about future outcomes, but in this process the Bank has made little progress towards the design of a long-term development strategy and perspective for the Senegalese economy. While the Bank's economic and sector work often did correctly identify problems and issues which were besetting the country's economy, in the end, the Bank, for a variety of reasons including, presumably, the desire not to interrupt the continuity of the lending program, would choose to ignore these problems and "hope for the best." This is not to say that there has been no improvement in this approach over time, the introduction of structural adjustment lending being a prime example, but rather that the learning process has been slow and that the strategy at the basis of the Bank's lending was, and remains, weak and incomplete. Bank Strategy and Impact x. During the first few years of the Bank's involvement in Senez' it took some time to establish a close relationship with the Gove rnum_. It proved to be extiamely difficult to find well-prepared project, selectioa of the first projects was more the result of their availabilit,% - iv - and suitability for Bank financing rather than of any strategic considera- tions. By about 1970 a Bank lending strategy was beginning to be formu- lated. Senegal was then judged by the Bank to have overcome the structural problems that followed Independence and the future seemed to hold promise. Bank strategy documento emphasized the need to strengthen the country's competitive position and improve production incentives, especially in the rural sector. Agricultural diversification into rice, cotton, vegetables, livestock, as well as geographical diversification became major objectives. The focus of the lending program was to be on directly productive activities (agriculture, manufacturing) and supporting infrastruct:re. This Bank strategy mirrored and supported government objectives and plans, and for the most part, continued unchanged throughout the 1970s. However, while the Bank's economic work correctly identified a large number of major problem areas -- high price and cost structure, overvalued exchange rate, large number of civil servants and emerging public finance difficulties -- little was done to remedy any of these, and there was no reason to assume them away, as they would not be resolved by themselves. xi. Furthermore, the "diversification" strategy was not all that well focussed. Internal documentation throughout the 1970s suggests that the Bank did not have a clear vision of which areas and sectors of the economy had most potential and were most in need of support. As a result, one overriding objective of the Bank's assistance was to be involved in a large number of sectors "as an experiment in determining the comparative advantage of Senegal" (internal memo). At the industrial sector level, the Bank did, in 1975, launch a major research study on industrial competitive- ness and incentives in four West African countries, including Senegal: some of the results of this research were later incorporated in the struc- tural adjustment loans. While the Bank's lending program eventually did touch most sectors, the strongest concentration and most meaningful inter- ventions were in agriculture, transport, industry and education before 1980, supplemented by structural adjustment lending during the 1980s when the overall strategy changed radically. Sectoral Experience xii. Agriculture. Policy toward the Senegalese agricultural sector has to contend not only with a poor natural resource endowment but also with highly variable and deteriorating climatic conditions. Experience over the past quarter century suggests that these constraints have not been overcome to any significant degree. Between 1960 and 1986 agricultural value add3d grew at around 1.3Z p.a. and at under 0.5? since 1970. While the policy environment has certainly played an important role in this development, agricultural sector output trends in Senegal can primarily be explained as a function of rainfall. There were practically no counter- measures taken as by the early 1980s less than 12 of the total arable area was irrigated. The bulk of agricultural producers have been faced with strongly declining rainfall trends. In addition, extreme year to year variation in climatic conditions has been the rule. xiii. At the present time, the broader outlook for Senegalese agricul- ture remains bleak. Export prospects for the traditional tradable -- groundnuts -- are poor and productivity in the sector has failed to -v - improve. Likewise, under current conditions including the small area under irrigation, the prospects for a significant increase in domestic cereals production to replace food imports appear very limited. This makes the reliance on agriculture in the development strategy look unrealistic. But given that the bulk of the population is still employed in agriculture and given that the country's options in other sectors are also limited, it is evident that high priority has to be placed on improving the efficiency of resource use in the agricultural sector. xiv. The involvement of the Bank in Senegalese agriculture has been considerable since the first project in 1969. However, both the modalities of lending and the content of that lending have changed significantly, particularly in the period after 1979. Before 1980, all Bank lending to Senegal was project lending: 35% of the total number of projects approved and 31% of the volume of commitments was directed to the agricultural sector. Between 1980 and the present, the bulk of Bank financing has been directed towards structural adjustment lending, with a major subset of policy actions that have directly impinged on agriculture, while direct project-based lending to the sector has been given less emphasis. xv. Despite this sustained presence, the available evidence suggests that, in general, the returns to Bank investment in Senegalese agriculture have been very low. OED audits of agricultural projects completed to date suggest that two-thirds of the monies spent in agriculture did not produce tangible net benefits. Of course, this can, in part, be explained by exogenous shocks, especially erratic rainfall patterns, but it has also to be related to inappropriate and inconsistent specification by the Bank of its objectives within a longer-term strategy, to the resulting neglect of the policy framework, and to project designs not necessarily corresponding to what would have been required in these circumstances. In addition, there is considerable evidence that the Bank has changed, often radically and to some extent unpredictably, the focus of its lending. xvi. The content of Bank lending between 1969-79 reflected fairly closely the principal emphasis of sector strategy at that time, i.e. increasing output and productivity in the core groundnut basin, and diver- sification, especially toward rice and including irrigated agriculture in the Fleuve region. Bank analysis tended to focus exclusively on short-run options, however. While it was commonly argued that groundnut export potential was held back by exchange rate overvaluation and demand con- straints, there is little evidence that the Bank seriously evaluated policy options under a range of scenarios, including exchange rate adjustment. No analysis of the latter in particular was ever made public by the Bank, because of the active support by France of a fixed exthange rate in the context of the West African Monetary Union. This limited the utility of the sectoral policy advice that the institution could offer. While the exchange rate issue is clearly a very delicate one and must be dealt with in the context of the Monetary Union, and whereas the effects of a depreciation are many and complex--improved competitiveness of the economy, but unclear impact in the short run--the exchange rate is nevertheless a major policy variable which warrants explicit analysis. It could be, in the very least, used as a "dummy variable" in the Bank's assessments of policy change requirements and their expected results, or as a "slack - vi - variable* which should be taken recourse to if other measures to achieve such objectives as exports growth fail. Neither did the Bank develop any consistent approach to the question of rice and irrigated agriculture. Emphasis on import-substituting rice production had emerged as a key feature of government policy by the end of the 1960s, largely in response to the growing food import bill. By 1970 there vas a shelf of a dozen rice projects. Production costs in the proposed large-scale irrigation schemes in the Fleuve region were estimated to exceed the landed cost of imported rice by some 30Z, however, and the Bank consequently favored and supported the development of rice in the Casamance region. Yet, by the mid-1970s, the Bank's approach changed. The Government was now urged to invest in rice schemes in the Fleuve, because it was now estimated that the projected barrage at Diama and dam at Manantali would make economic rice production possible: there was little indication, however, that these estimates were based on detailed studies. Furthermore, this apparent shift in Bank emphasis had limited translation into resource allocation by the Bank, largely because of continuing doubts about the economic viability of rice production. xvii. In any event, the focus on rice production, either in Casamance or in the Fleuve was not sustained and, perhaps more importantly, little attempt was made to incorporate longer-term considerations. Yet, in a context where agricultural as well as non-agricultural sources of growth are limited but balance of payments and creditworthiness constraints remain strong, the rationale for high investment costs in irrigated agriculture would be greater than that revealed through conventional cost-benefit analysis. This would assume, however, some weighting being given to both the positive externalities, especially the reduced risk and dependence on rainfall, and to intergenerational benefits: in a country with few investment choices, the very large and long-term irrigation scheme proposed for the Senegal valley (investments over 30 to 40 years would be required) would more than double agricultural production, hence strongly affecting future distribution of income. There is little or no evidence that any c=nsider'tions given in the past to estimating the opportunity cost of capital in Senegal have included the above. Clearly, purely conventional ane cursory estimates of such cost would have excluded any but the most straightforward and relatively unimaginative projects. xviii. The inability to generate a sustainable agricultural sector strategy can, in part, be attributed to grave data inadequacies and the failure to improve the basic information systems in Senegalese agriculture. Lack of sufficient micro-level information regarding production and distri- butional attributes resulted in attempts to apply technical packages which have ultimately proven to be inappropriate and inadequately differentiated. Nothing was done to resolve satisfactorily the basic agronomic questions regarding yield responses to a range of inputs. Widely differing opinions came to be held over appropriate levels of fertilizer application. But the response rate by producers to the technical packages offered by the exten- sion agencies and by Bank-supported projects remained poor throughout this period. The inability to develop adequate technological packages that could be extended to producers resulted primarily from insufficient research. The Bank did not direct resources to the Senegalese agricultural - vii - research effort, except in a very limited provision of vehicles and equip- ment, until the Agricultural Research Project of 1981. Likewise, the Bank approach to input subsidies was also inconsistent. A recurring theme in Bank documents was the need for more emphasis on ecological differences and targeting of fertilizer use to areas where returns were higher. During the early 1970s, input subsidies attracted qualified support on "infant industry" grounds, but as fertilizer subsidy costs mounted, and in the absence of a major output response, this approach altered. By the late 1970s the Bank came out against generalized input subsidies, but this was partly contingent on alternative measures for hedging risk, such as crop insurance. In the post-1980 period, the Bank has advised against input subsidies under any circumstances. Yet, throughout, in the absence of an adequate technical package, a positive output response could not have been expected. xix. The inputs subsidy issue raises a number of difficult questions. It would appear that at unsubsidized prices the consumption of yield- enhancing inputs will remain insignificant. Indeed, it is worth noting that even at the height of the "Programme Agricole" during the 1970s average fertilizer consumption at subsidized prices, including for the groundnut basin, remained at very low levels. Nevertheless, subsidies , at least for a period of time, can be an incentive to use inputs. The "Programme Agricole" did succeed in raising farmer exposure to and use of modern inputs. Moreover, it should also be noted that with heavy subsidies the "Programme Agricole" did effect a major technical change -- unique in West Africa -- the shift to animal traction, one key consequence of which was extensive growth. While no satisfactory case can be made for permanent input subsidies, the particularly binding constraints that operate in the Senegalese economy at this stage -- soil degradation and desertification, high climatic risk and an overvalued exchange rate -- suggest that selec- tive input subsidies are prot .bly a prerequisite for productivity increases. The Bank's current generalized aversion to input subsidies may need to be re-examined. xx. Surprisingly little attention was paid during the 1970s to pricing and taxation policy, in part a consequence of the limitations of the project lending format. Throughout the decade Senegalese farmers remained implicitly taxed through controlled prices and, in particular, by groundnut producer prices that averaged only 43% of the unit export price. While sectoral policy analysis did point to the desirability of producer price increases for major crops in the mid-1970s, there was little analysis of what should be the appropriate price levels for individual outputs and, in particular, what should be the price structure, i.e. the interrelation- ship among various output prices. At the same time, in the determination of the crucial groundnut price, which has major budgetary implications, Bank staff accepted that these should precede producer welfare considera- tions. A major consequence of the Bank's inability to deal with pricing issues in this period was the development of anomalies in relative price structures which regularly weakened project performance. Thus, e.g., falling producer prices coupled with constant input prices over the course of a project would restrict input use and compromise the potential for any productivity effect. While the scale of the disincentives effect on 1 - viii - production is unclear, price and marketing controls constrained officially marketed releases and, together with other factors such as the recurring droughts, contributed to out-migration from the sector. xxi. The Bank's approach towards agricultural parastatals during the 1970s was one of strong support. The grave inadequacies of the major institution in the sector -- ONCAD -- were skirted in the hope that the new rural development agencies (RDAs) would increasingly play the leading role in developing the sector. Given the continuing importance of ONCAD in the agricultural sector, this hope was largely unrealistic. Other institu- tions, such as the Ministry of Rural Development, were consistently bypassed during those years. Growing disillusion, in turn, with the RDAs and the heavy burden they also imposed on the government budget eventually led to a withdrawal of support from the Bank for them. In this respect, institution-building by the Bank remained limited and uneven. xxii. The post-1980 shift in strategy towards the agricultural sector was a major turnaround. In contrast to the past, primary emphasis was placed on price policy. In place of the RDAs, the Bank now supported the strengthening of the Ministry of Rural Development, an institution that was earlier deemed unsuitable for managing sectoral initiatives. Project level support was cut drastically. xxiii. Significant growth in agriculture has been postulated in Bank documents since 1980 on mere price effects. While the new emphasis on the role of prices is highly desirable, evidence suggests that the effective- ness of price policies in the Senegalese context is bound to be limited as long as some other factors affecting the performance of agriculture -- climatic trends, soil degradation, and a weak research and extension effort -- are not being dealt with. There is a substantial body of literature which estimates very low short- and medium-run supply elasticities in the Sub-Saharan African context. xxiv. For Senegal, the very limited econometric evidence available indicates the predominant influence of rainfall and the very limited impact of prices on both groundnuts and millet output. The Bank did undertake some research on the effect of price changes on agricultural output and incomes and on government finances but the results of that research were not used in the formulation of its assistance strategy. Moreover, as it turned out, in spite of nominal price increases, real agricultural prices -- or nominal prices corrected for inflation -- have not been increased during the structural adjustment period. Furthermore, clear yield- augmenting technological packages available for dissemination to the farmers have not been developed. In fact, some of the key problems to achieving agricultural growth -- institutional development, credit policy and the choice of appropriate technology -- have all been put aside in the post-1980 period, largely on account of public expenditure constraints. Under these conditions, the Bank's ex-ante expectations and projections with regard to the aggregate agricultural output response appear highly optimistic. xxv. In retrospect, Bank policy to foster agricultural growth in Senegal would appear to have shifted from one extreme to another, while tending to ignore or neglect some of the more fundamental constraints to the development of the sector, i.e. climatic factors, and research and technology. During the 1970s the focus was on institutional and project level support at the expense of the role of prices. Since 1980, liberali- zation of prices and markets has been pursued while de-emphasizing the need for institutional and other support. Pricing and liberalization measures alone, while showing impressive results in some areas such as cereals market liberalization, can -- on their own -- only play a limited part in achieving growth in the sector. They are necessary prerequisites but not sufficient conditions. Agricultural development in Senegal will require, in addition, a focus on providing suitable agronomic and technical packages to farmers backed by support institutions and a concerted attack on the climatic impediments to growth through the development of irrigation systems and soil regeneration and conservation programs. All of this implies an important role for the state and foreign project support. Only then will other policy measures become truly effective. xxvi. Industry. Ever since Independence the Government's expressed strategy for the industrial sector, as it was for the economy as a whole, has been "diversification." In practice, little was accomplished largely because the overall policy framework was not conducive to growth or diver- sification. Following Independence, which brought about the loss of a good part of the colonial French West African market for Senegal's industry, the Government established high tariff barriers supplemented by numerous quan- titative restrictions. In addition, many exceptions to the common tariff regime were granted on an ad hoc basis. This system led to the development of a generally uncompetitive industrial sector at a high cost to the economy with limited potential for growth. Between 1960 and today, indus- trial value added is estimated to have grown at slightly more than 4Z per annum which would have been a praiseworthy growth if applied to a higher starting basis or if resulting from a lower capital expenditure. None of these factors existed, however. xxvii. Industrial finance. While the Bank fully supported the diver- sification strategy, the focus of the lending program -- industrial credits -- was clearly not the binding constraint to the development of industry. The Bank's economic and sector work did dwell on the many obstacles to industrial growth including, from the very beginning, the crucial role of the exchange rate, but these concerns were not made operational until the introduction of structural adjustment lending, and then only to a limited degree. Instead, it was implicitly assumed that the provision of long-term finance would somehow lead to growth. At the Government's request, a new development finance institution, SOFISEDIT, was to be created to channel the resources. xxviii. It is not totally clear why the Bank decided to help set up SOFISEDIT in the first place. While there was a need for additional sources of long-term finance for industry and, with it, for improved appraisal capabilities, the exterlL of that need, given business prospects, appeared limited. The choice, therefore, was between a totally new insti- tution or some modification to the financial system in place, in view of the modest demand prospects. While the Bank did prepare a feasibility report, the record shows that Bank staff were highly skeptical about the - x - need for a new institution. The Bank, in any event, went ahead with the decision to help create and provide project support to SOFISEDIT in 1974. primarily because of Government insistence or, to put it differently, for country relationship reasons. xxix. Almost from the start, SOFISEDIT ran into financial and other difficulties, which grew worse over time. Nevertheless, the Bank continued its support with two more operations, in 1976 and as late as 1981. If there ever was a gap to be filled by SOFISEDIT in the market for long-term finance for industry, this gap became much smaller within the year following the institution's creation on account of new banking regulations that gave all banks access to long-term resources. Moreover, the market was by now proving to be far from dynamic as industrial growth slowed con- siderably after 1975. xxx. SOFISEDIT's record of performance has been very poor. Its operations have been stagnant, far below projected levels and its financial performance dismal. It would appear that the Bank throughout the whole period ignored the reality of the situation, i.e. an extremely small market for industrial finance, which could have been well served by the existing financial system, and poor industrial growth prospects given the economic policy environment. Starting with the SALs, especially SALs II and III (1986/87), and with a 1987 Industrial Sector Restructuring project, the Bank has begun to address earnestly the structural issues impeding indus- trial growth in Senegal. Also, SOFISEDIT has been included in the Government's efforts to restructure the public banking sector. It must be added that some of the issues -- the exchange rate, the supply of private entrepreneurs and capital -- are still not being squarely faced. The primary lesson from the Bank's past experience in industrial lending is that if issues are major, one cannot leave them unattended and hope for the best. xxxi. Transport. At independence, Senegal already had a well- developed transport network. Until that time, the system served export/im- port traffic between France and all of French West Africa. Independence brought some major changes. After 1960, the network served a much reduced area and with this the significance of Dakar as a regional center declined. Further, transit services to air and shipping lines provided in the past by Dakar gradually lost their earlier importance. In addition, the slow growth of the economy further limited the demand for transport services, and rapid population growth meant high pressure to create jobs and over- staffing in public institutions, including those in the transport sector. xxxii. There is no doubt that the Bank's strategy towards Senegal's transport sector has been sound and that its main features are as valid today as they were in the mid-1960s when they were first formulated: sound investment planning, priority to maintenance, emphasis on economy and efficiency. Between 1966 and 1987 the Bank Group approved nearly $170 million in 15 loans and credits for the transport sector; 6 projects in the roads sector, 4 for railroads, 3 port and 2 airport projects. In addition, the Bank mobilized considerable cofinancing and devoted much staff time to sector and supervision work. The Bank's achievements have varied widely. For highways, project performance has been generally good - xi - but quite the reverse for the railways subsector. Senegal now has an effective and efficient road maintenance organization; the port and airport have good physical facilities, sufficient for the likely demand generated in the foreseeable future; and, in spite of the poor performance of projects in railways, without Bank assistance the system probably would have broken down. At the same time, there are major problems: the railway has continued to show poor operating performance and mounting deficits; the road maintenance organization functions well but depends on donors for maintenance finance which gives rise to policy questions that have not received the scrutiny they deserve. The port and airport facilities are good but traffic is down. While the Bank's strategy was sound, often its tactics were wrong. xxxiii. A first important finding regarding the Bank's interventions in the transport sector is that projects are more likely to be successful when institutional change is approached gradually. In the area of road main- tenance, where the lending program was highly effective, it became so only after the Bank changed tactics and opted to work within the existing environment to strengthen and improve the road maintenance division, rather than through a drastic reorganization which was politically unacceptable. In contrast, the Bank's attempt to have Government create a new institution to deal with feeder roads and its insistence on accelerating the implemen- tation of the project in the face of staff shortages and lack of appropriate expertise led to major problems. xxiv. Second, difficult issues have to be addressed more directly and honestly. When there are well-substantiated doubts about the commitment or capability of the Government to implement the various elements of a project, the Bank should not hesitate to defer lending. To the extent that the Bank had no impact on government railway policy and could not generate any improvement in railway management or operations, it is unclear why it continued to lend. Also, experience with the first port and airport projects did indicate the importance of financial and operational autonomy but the Bank did not effectively resist the Government's subsequent encroachment upon that autonomy, which contributed to the management and cash flow problems of the two institutions in recent years. xxxV. Finally, it is worth emphasizing the obvious: good preparation and appraisal are key to the success of any project. The disastrous second and third railway projects were poorly appraised from the start. Both had to be redesigned during implementation to focus on other more urgent work, which could have been foreseen. At times, lessons from the past to improve future project appraisals are somewhat elusive. Experience all over Africa suggests that the subdivision of former colonial territories requires new approaches to transport planning and, in the specific case of Senegal, implies reduced transit traffic. Yet, overoptimistic traffic forecasts have been a feature of many appraisals. xxxvi. Education. While significant progress in expanding education has been made since independence, the results are not entirely satis- factory. Between 1965/66 and 1984/85 public spending for education rose from the equivalent of 1Z of GDP to 5Z. Enrollments increased relatively fast at higher levels of education, a continuation of the traditionally - xii - elitist emphasis of the system. This contrasts with the fact that at present Senegal still ranks extremely low in basic education indicators: by 1985, adult literacy was only 282 and primary enrollment about 50% for the country as a whole and much lower in rural areas. xxxvii. When the Bank began its dialogue with Senegal on education issues and policies during the late 1960s, the two major weaknesses of the system were identified as the poor adaptation of education to the needs of the modern economy and the low primary enrollment rates, especially in rural areas. To date, the Bank has financed five education projects. With the exception of the fifth project in 1986, which supports primary educa- tion, the Bank's focus has been on the first of the two weaknesses of the system, i.e. the need for more manpower training, in conformity with government priorities. On the whole, the Bank has been quite successful in helping to strengthen education infrastructure at these levels. xzxviii. On the other hand, although the importance of primary education as a fundamental basis for the long-term development of the country was long recognized, the tight budgetary outlook and competing priorities always held up the growth of this subsector, and the Bank unfortunately, for many years, did little to encourage the Government to increase or redirect resources towards it. While other donors were generally even more supportive than the Bank of the urban and upper level bias of the education system, it can certainly be argued that the Bank should have tried harder than it did to lend its support to primary education both through its economic and sector work and through its lending. It was not until 1986 that a Bank-supported primary education project was approved. It was designed to make the system more efficient, lower unit costs and expand primary education. This was to be financed without additional public resources but rather by restructuring intrasectoral budget allocations. Structural Adjustment Lending xxxix. By the end of the 1970s all macroeconomic indicators showed that the economic situation had become untenable. With large and rising current account and budgetary deficits, internal and external disequilibria had reached crisis levels. A cumulative process of worsening economic perfor- mance was underway. Poor agricultural and export performance resulted in GDP and output stagnation while both private and government consumption continued to grow at alarming rates (between 1977178 and 1980/81, govern- ment consumption grew at almost 52 per year and private consumption by 3.2Z compared to a much slower growth of GDP (0.8%) and a decline in investment (-4.3Z per year)). By 1980/81, consumption exceeded domestic production and domestic savings had become significantly negative. Given the depth of the crisis, some stabilization and adjustment was unavoidable. xl. The choice which the Government confronted was between: (i) taking a passive approach which would finally and unavoidably require forced adjustment or (ii) intervene actively with appropriate policy measures and reforms to reach a modicum of internal and external equilibrium. The Government opted for the second alternative which, although entailing high political risks, was still preferable. After extensive discussions with the Bank and the IMF, the Government issued an - xiii - "Economic and Financial Rehabilitation Program' in December 1979, followed in October 1980 by a "Declaration on Economic Policy." These documents paved the way for a stabilization and structural adjustment program with the IMF and the Bank which is still continuing eight years later. Bank support to date has consisted of three structural adjustment operations (SALs), in 1980, 1986 and 1987. xli. The overall objectives of the SALs were quite clear: to achieve stabilization and avoid external payments default in the short term, and, at the same time, to lay the basic conditions for growth in the longer term. These conditions, in summary form, consisted of an improvement in the incentives' framework for production, an increase in the efficiency of public activity in the economy, and a strengthening and increase in the role of the private sector. While SAL I in 1980 was only partially suc- cessful towards meeting its objectives, and while it is too early to judge fully the impact of SALs II and III (1986/87), it is clear that, to date, the principal short-term objectives have been achieved, even though repeated debt reschedulings have been necessary. With regard to the longer-term (growth) objectives, however, the Bank's approach can only be evaluated in an ex-ante sense. x1ii. The key question from the standpoint of policy analysis in any development plan is the nature of the interrelationships which exist between policy objectives, on the one hand, and the package of policy measures and institutional reforms designed to meet these objectives, on the other, while allowing for the impact of exogenous variables which by definition are outside the control of the policymaker. The objectives of the structural adjustment programs and their policy measures and reforms, i.e. the conditions agreed upon by the Government and the Bank, form an essential subset of such a development plan. Ultimate success of the structural adjustment process cannot be gauged by whether conditions per se are met but by whether both economic performance and the social situation improve (independently of the impact -- positive or negatire -- of exogenous factors). For this reason it is essential to have an accurate qualitative and quantitative understanding of the links between conditions and policy objectives. An ex-ante evaluation of the structural adjustment programs, in other words, must answer the question of what evidence did the Bank use when relating policy instruments and reforms to performance, in formulating its conditionality package, or on what basis were the condi- tions selected. The terms "conditions" and "conditionality" are used in this report to refer to the whole set of measures described in the letters of development policy, not solely to those measures which are specific tranche release conditions: it is the program as a whole, made up by the complete set of measures, that is approved by the Bank's management and Board of Executive Directors as a condition for success. xliii. First of all, it must be pointed out, that the Bank througho-at the process imposed a large and increasing number of conditions, from 32 in SAL I to 46 in SAL II and 77 in SAL III: this would indicate an ever tighter and more comprehensive adjustment program. On the other hand, an increasing proportion of the conditions (from 34Z in SAL I to 56Z in SAL III) were studies or "pre-conditions," i.e. they did not L astitute policy actions as such and could not alter in any way the status quo. In - xiv - fact, the need for a multitude of studies is indicative of a lack of knowledge on the part of the Bank and Government of the underlying socio- economic reality and relationships among variables. Added to this, the absence of any clear priority ordering among the other conditions and the incomplete nature of many of them (e.g., public sector disengagement without private sector engagement) suggest that the adjustment package was considerably less than tight. While a large number of conditions, without priority ordering, allows for flexibility, important conditions can easily be overlooked and the administrative cost on the part of the Government becomes extremely high. xliv. A review of the conditions contained in the three SALs grouped by four major domains -- (a) agriculture, (b) private sector efficiency, (c) public investment and finances, and (d) public sector strengthening and disengagement cum private sector engagement -- suggests that, in large measure, conditionality has been based on an incomplete analysis and grasp of the dynamics and workings of the Senegalese socioeconomic system. This has given rise to a continuing tendency to overestimate likely future performance (particularly with regard to growth) and to underestimate the negative impact of constraints on development. xlv. First, in agriculture, not only did the almost exclusive emphasis on prices have a weak empirical foundation, but insufficient attention was given to weather and the deteriorating natural environment which are increasingly binding constraints on improved performance. It would appear that it is only through modifying the natural environment that policy measures can eventually become effective. Second, the restructuring package to improve private sector efficiency constitutes a necessary but insufficient set of conditions: in the absence of a simultaneous devalua- tion private sector profitability would, unavoidably, decline further. Moreover, poor demand prospects, tight financial policies and a host of other constraints, which cannot be removed rapidly, cast doubts on the Bank's favorable forecast of performance. Third, conditionality imposed to improve project preparation and investment programming is a major step forward, but the limited technical competence in many parts of the Govern- ment is likely to limit the "bite" of these conditions. Finally, in the crucial area of privatization, public sector disengagement without stinuli to ensure complementary private sector engagement cannot by itself be expected to change the underlying situation. At the present time, there are enormous obstacles to private sector engagement in Senegal, ranging from the negative profitability of the parastatals offered for sale to unfavorable domestic and foreign demand prospects, an extreme shortage of credit and venture capital and a lack of entrepreneurial talent. xlvi. An additional issue of crucial importance to the future develop- ment of Senegal is whether the net foreign resources available are sufficient to provide the required investment funds for a take-off; i.e. a departure from historical stagnation trends. Evidence suggests that they are not, as the bulk of new foreign resources is likely to be needed to service foreign debt. This issue can no longer be left untouched, because the choice for the Bank and the donor community will inevitably become one between debt repayment and development. Some form of debt cancellation -- not rescheduling which only postpones the inevitable -- would appear to be - xv- a precondition to insure that a sufficient net inflow of foreign resources is available for productive investment. It is, of course, imperative that the funds thus freed up not be used to finance uneconomical projects; any alleviation of the debt burden would have little meaning in this case. On the other hand, a reduction in debt payments would not only permit the financing of good public sector projects but the accompanying easing of credit and foreign exchange scarcity in the economy is an important condition for private sector investment growth. xlvii. In sum, if the Bank is to help develop a genuine strategy for the long-term growth of the Senegalese economy, the starting point must be a full recognition of the immense constraints to development. 'While the outlook is not encouraging, little will be gained by ignoring these facts. Overly optimistic projections are misleading in many ways but especially in providing estimates of the future effort required by both the Government and the donor community. While the Bank can rightly take credit for having contributed to the creation of the pre-conditions for growth, the agenda is far from finished. In other terms, simple economic models do not work in practice. Especially so, when some principal variables, often the deter- minant ones, are missing from them. A number of very fundamental issues and constraints, both at the macro and at the sector level, cannot be ignored. They must be given priority ranking. They should then be incor- porated into a framework which allows the policymakers to estimate, as explicitly as possible, the likely impact of their resolution or allevia- tion on performance. Only then will the conditionality -- and structural adjustment -- process be made fully meaningful. Aid Coordination xlvii. The large size of external aid flows in the Senegalese economy makes aid coordination an important issue. While donors often showed little interest in coordinating their activities with one another in the past, aid coordination became a focus of attention with the beginning of the structural adjustment process in 1980, the consequent discussions and reviews of development programs, and the convening of the first consul- tative group meeting in 1984: the Bank played a major catalytic role in this process. xlix. Numerous mechanisms have sprung up to ensure improved aid coor- dination. Within the framework of the structural adjustment programs, the Government has entablished an interministerial council and several coordinating committees. The consultative group meetings -- a second meeting was held in 1987 -- have allowed the Bank to disseminate its economic analyses and policy recommendations to others, making its advice more effective, and to begin to provide systematic guidance simultaneously to the Government and the donor community on the types of assistance required and the phasing of projects. The Bank has also helped organize sector meetings to discuss programs and policies, and the Bank's resident mission convenes informal meetings at regular intervals to inform the donor community on progress in implementation of the structural adjustment program. -xvi - 1. The Bank's efforts to strengthen coordination among donors have paid off and, while important problems remain, the obvious conclusion is to try to do more of the same. A more active role for the Bank would mean a more intensive program, which clearly has limits in availability of staff resources: some division of labor among donors would seem appropriate. At the macroeconomic policy level there is considerable room for a more inten- sive exchange of views among all parties. In the case of structural adjustment policies in particular, several donors resent the fact that they are presented with a final product, agreed between the Government and the Bank, in which they have had no input. The most effective way to ensure closer cooperation in this .espect is to have the donors, at least the major ones, closely involved in the development and the design of these policies from the very beginning. Several of Senegal's donors believe that current disagreements with the Bank on important policy issues could have been avoided if the preparatory studies had been financed and executed jointly. li. At the level of projects, the Bank has made substantial efforts to help thf Government rationalize its investment programming by developing sectoral strategies and prioritizing projects, and to have donors conform to it. This is proving to be a difficult task, partly because of the extreme scarcity of good project information, and partly because aid programs carry a mixture of objectives which do not necessarily correspond to those of the government investment program. Nevertheless, excellent progress has been made. In order to get the donors to focus even more on country priorities, a strong case exists here also for the review of the investment program to be done by the Bank jointly with major donors. The practice already adopted of having different donors take a lead role in the review of these sectors where they are heavily involved should be encouraged. 1ii. In the day-to-day operation and the actual implementation of development programs, includirg in the area of technical assistance, there remain a number of problems stemming from inconsistencies in the approach of different donors to such things as remuneration of nationals, expendi- tures in local currency or disbursement procedures. The local donor repre- sentatives are in the best position to know these problems and come up with solutions which would make for a more unified approach. They also appear to be very eager to cooperate more closely with one another. Progress in this area would require devolution of some decision-making power from head- quarters to the local representation and, in addition, in the case of several of the donors, considerable strengthening of the local mission. liii. In the end, however, it must be acknowledged that there always will be limits to what the Bank can accomplish in the area of aid coordina- tion, given that the objectives of different donors are never likely to coincide fully. A genuine solution to these problems can only be arrived at over time, as Senegal's in-house capacity for economic policymaking and investment programming strengthens to the point that the Government is in a position itself to accept or reject aid proposals in accordance with its needs, its absorptive capacity and its own priorities. Chapter I: THE SETTING 1. Senegal is generally regarded as a model country in Africa. Its political stability, with continuity of administration, has been exemplary. Warm relations with the donor community, including the Bank, have meant that the country has received very high levels of uninterrupted external assistance and Bank support over many years. At the same time, per capita income has declined almost steadily over the past quarter century and is now below the level reached at the time of independence. These few intro- ductory observations clearly raise some fundamental questions about the complexity of the development challenge in Senegal and about the effective- ness of external, including the Bank's, assistance. This first chapter reviews economic developments in Senegal since 1960 and the Bank's assistance strategy in very general terms. It highlights some of the major issues and suggests some broad conclusions about the Bank's role. This is then followed by a detailed review of the Bank's approach and effectiveness in, first, the structural adjustment process since 1980 (Chapter II) and, second, the major sectors of Bank involvement since the beginning of the relationship (Chapter III). The discussion of agricultural and industrial sector issues is broken down into two parts. Chapter II covers the structural adjustment period since 1980 and, among others, reviews agricultural and industrial conditionality and performance in the light of the constraints on performance. Chapter III, apart from covering all of the Bank's project lending experience to those sectors starting with the first project in 1969, examines these constraints on performance, primarily in the agricultural sector, in some detail as well as the degree to which the Bank has helped alleviate them over time. A. Economic Developments and Bank Strategy at the Macro-Level 2. From a macroeconomic standpoint, it is possible to identify five reasonably distinct periods that the Senegalese economy went through since 1960. Table 1 provides the values of the major macroeconomic variables for the six reference years (beginning and end of each period) spanning the five periods. To reduce somewhat the large annual fluctuations, two-year averages were used for these reference years. In turn, Table 2 gives the corresponding growth rates of these same variables in each of the following five periods: (i) 1960/61 to 1966/67; (ii) 1966/67 to 1973/74; (iii) 1973/74 to 1977/78; (iv) 1977/78 to 1980/81; and (v) 1980/81 to 1985/86. - 2 - Table 1: COMPONENTS OF GDP AND TOTAL EXPENDITURES, SELECTED PERIODS (two-year averages in billion CFAF at 1979 constant prices) 196061 1968167 197/74 19777 I!mal 196518i La Is Primary Sector 86 107 117 127 111 129 lb Secoidary Sector 67 87 111 188 141 174 le Other 218 250 259 288 809 367 1 GDP 871 444 487 548 561 670 2& Exports 171 159 161 201 184 197 2b Imports 191 210 198 289 288 258 2 Resource Cap 20 6. 87 8 99 61 8 Total Expenditures Lb 891 495 524 616 660 781 4a Government Consumption s 85 78 99 114 128 4b Private Consumption 259 841 346 419 460 518 5 Total Investment s 68 101 99 85 65 6 Domestic Savings 46 18 64 80 -14 24 La Each column represents the beginning and/or end of a distinct period, e.g., 1960/61 to 1966/87 Is the first period and so on. All the figures are two-year averages, e.g., 1960/61 means the average of 1960 and 1961. l Total Expenditures (row 3) equals the sum of GDP (row 1) and the Resource Cap (row 2). It Is also equal to the sum of Government Consumption (row 4a), Private Consumption (row 4b), and Total Investment (row 5). Source: Calculated on the basis of Table SA 8 in the World Bank, Senegal: An Economy Under Adjustment, Report No. 6464-SE, February 18, 1987. 3. Even though the choice of periods always entails an element of arbitrariness, an examination of these tables confirms that Senegal under- went significant and discrete changes (and in some cases turning points) in GDP and sectoral output growth, and in the export-import and investment- savings performance which are well captured by these periods. -3- Table 2: FIVE SELECTED PERIODS BETWEEN 1960/61 AND 1986/U8 (annual growth rates) Periods: I II III IV V I-V 1960/61 1968/67 1978/74 1977/78 1980/81 1960/61 to to to to to to 1966/67 1978i74 197707. IMLSI 196/e 196/U (Number of Years) (6) (7) (4) (8) (6) (25) la Primary Sector 8.72 1.28 1.97 -4.00 3.15 1.65 lb Secondary Sector 4.88 8.68 4.71 1.64 4.80 8.88 1c Other 2.32 0.52 2.74 2.82 8.47 2.11 1 CDP 3.04 1.84 8.02 0.75 8.62 2.40 2a Exports -1.12 0.18 5.71 -2.03 0.90 0.68 2b Imports 1.53 -0.77 7.95 1.71 -1.71 1.20 2 Resource Cap 3 Total Expenditures 4.01 0.21 4.90 2.38 2.06 2.58 4a Government Consumption 4.22 -1.17 6.42 4.80 2.24 2.67 4b Private Consumption 4.74 0.19 4.91 8.17 2.41 2.82 5 Total Investment 0.40 5.82 0.59 -4.81 - 1.01 6 Domestic Savings -8.85 20.17 -11.11 -85.11 80.01 -1.58 Source: Derived from Table 1. 4. Before describing briefly the main features of each period and the Bank's strategy (including its ex-ante analysis), it is essential to emphasize the predominant role which agriculture plays in the dynamics of the development progress of Senegal. Figure 1 illustrates graphically the high correlation between the growth rates of the primary sector output and of GDP.1 It shows that the wide fluctuations in agricultural output (the major component of primary sector GDP) pulled overall GDP correspondingly up and down. This correlation has been confirmed statistically in a number of regression equations run on the basis of annual observations between 1/ Since primary sector output is a component of GDP, such correlation is to be expected. However, the share of the former in GDP was quite low, i.e. 22Z in 1960/61 and 19.3% in 1985/86, making this correlation more remarkable. 1960 and 1983.2 Given the dismal growth performance of Senegal during the period under consideration (1960 to the present), it would be misleading to suggest that agriculture was the engine of growth pulling the rest of the economy behind it. Instead, the correct inference is that GDP moved in an almost completely parallel fashion with agricultural production. Given the predominant influence of weather (mainly rainfall) on the latter, it means that the course of GDP in Senegal was largely predetermined by the vagaries of the weather. As agricultural output stagnated, so did GDP. 5. The first period (1960/61 to 1966/67) is characterized by a relatively good performance of agriculture (mainly groundnut production) with primary sector output growing at 3.7Z annually. Groundnut revenues provided the stimulus for GDP growth of slightly above 3.0% p.a. However, two negative trends are noticeable: (i) disinvestment in modern sector activities; and (ii) the growth of consumption outstripping that of GDP (the share of consumption to GDP grew from 87.62 in 1960/61 to 95.92 in 1966/67 as shown in Table 3). These were the early years of the Bank's activity in Senegal, with some limited economic work, but no project involvement as yet. 2/ The most noteworthy results are that the annual rate of change of GDP is very significantly correlated with the annual rate of change of groundnut production (t ratio = 6.3; R2 = .66) and, alternatively, with the annual rate of change of groundnut yield (t ratio = 7.4; R2 = .72). Another statistically significant correlation was found between the rate of change of GDP and the rate of change in groundnut exports. For more detailed results, see ROpublique Frangaise, Ministbre des Relations Extdrieures, Coopdration et Ddveloppement, Des6quilibres Structurels et Programmes d'Ajustement au S4ndgal, Mission d'Evaluation (Mars 1985), Volume II, Annexe au Chapitre 1.2. -5- Figure 1 SENEGAL GOP AND PRIMARY SECTOR (Annual Growth Rates at Constant Prices) 30 20 - -20 II I II I , I Ii If Ii I O - OP - - -1Primary1f 60 62 64 66 68 70 72 74 76 78 80 82 84 IMA446M5 - 6 - Table 8: PERCENTAGE SHARES OF CONSUMPTION AND INVESTMENT IN CDP, 1960/61 TO 1965/8 1960L*1 1966/67 197./74 1977/7 1980/1 1985/6 Consumption/GDP Share at Constant 1979 prices 67.6 95.9 87.1 94.5 102.8 96.4 Share at Current Prices 86.1 91.7 67.4 96.8 102.6 95.0 Investment/COP Share at Constant 1979 prices 17.8 15.8 20.7 16.1 15.2 12.8 Share at Current Prices 14.6 10.6 20.7 18.0 18.0 13.8 Sources: Derived from Table 1, and R6publique Frangaise, Minist&re des Relations Ext6rioures, Coop6ration et D6veloppement, D6sequillbres Structurels *t Programmes d'Ajustement au SWnfgal, Mission d'Evaluation (Mars 1985). 6. The second period (1966/67-1973/74) is characterized by a marked slowdown in groundnut production and, correspondingly, GDP growth -- with both primary value added and GDP growing at about 1.32 annually. The stag- nation of agricultural output was largely influenced by the loss of the French preferential price program for groundnuts. Groundnut exports suddenly had to compete in the world market and remained almost totally stagnant. At the same time, both domestic savings and investment increased substantially -- with the share of investment in GDP reaching its highest level of 20.7Z. This high investment ratio is misleading in that much of it went to purchase foreign enterprises (i.e. the Senegalization process). Likewise, private consumption did not increase (0.22 growth) which combined with population growth meant that per capita consumption was falling at about 2.0% per year. 7. The Bank's analysis of the economic situation in its 1970 Country Program Paper (CPP) is perceptive and accurate. It criticizes the Government for (i) having done little "to facilitate adjustment of the artificial pre-independence economy to a level more in line with the actual economic potential and human capabilities"; (ii) favoring import- substitution industrialization which turns the internal terms of trade against the domestic farmer; and (iii) failing to set up a properly functioning groundnut marketing system to replace the barred private traders.3 A projected annual growth rate of GDP of 5.62 between 1968 and 1976 was set by the Bank largely because of a perceived improvement in project preparation and implementation and the potential for agricultural 3/ World Bank, CPP, June, 1970. - 7 - diversification in the south and southeastern region.4 In the 1971 Country Program Note, the Bank recommends higher prices and other incentives for groundnut production, but, in contrast with its preceding year's evaluation, is very critical of the state of project preparation which it sees as a major roadblock. Furthermore, the Bank warns against the low share of investment financed through domestic savings and the consequent implications for future debt servicing. It also revises its GDP growth projection slightly downward to 4.5% p.a. to 1977. 8. By 1972 the Bank had become quite upbeat about Senegal's economic progress and prospects, especially in manufacturing. As stated in that year's CPP, "In the private sector, Government was remarkably successful in its policy of stimulating import substitution industries...0; and "the substantial expansion of manufactured exports over the last few years, reflects Senegal's improving external competitive position....I However, it continues to express concern about low savings, Senegalization and the scarcity of local entrepreneurs, i.e. "lack of formal training and practical experience amongst local businessmen are major constraints." The 1972 CPP concludes on an optimistic tone as reflected by the following quote, "over the past 5-6 years, its (the Government) economic policies became more vigorous and successful so that by now a satisfactory basis exists for faster economic growth in agriculture, fisheries, tourism and manufacturing industries.' The Bank also feels that its influence and leverage are rising, i.e. the Bank Group' is considered by the Government "not only as a major lender, but also as an important outside counsel on development matters, who can stimulate investments by other aid donors.' 9. The third period (1973/74 to 1977/78) is one of export-led growth. Prices and world demand for phosphate and groundnuts rose very substan- tially and the international terms of trade for Senegal doubled between 1973 and 1976. The volume of exports grew at 5.7Z p.a. during the third period (and, of course, by much more in terms of purchasing power) and GDP by 3.0% p.a. The export boom was used largely to finance consumption (government consumption rising by 6.4? annually). Investment grew marginally (0.6% p.a.) and domestic savings fell at a rate of 11.1% p.a. (see Table 2). The investment/GDP ratio dropped from 20.7% to 18.12 during this period (see Table 3) -- with much of the investment going towards unproductive uses -- particularly the public purchase of private enter- prises described in detail in Chapter II, section E.5 The 1974 CPP continues to be relatively optimistic regarding future prospects. It identifies four crucial factors upon which the latter will depend: weather, terms of trade, Government ability to respond and availability of additional financial help. It foresees the combined effect of these factors to develop in a satisfactory way to resume growth immediately while laying the basis for accelerated development in the 1980s. The Bank 4/ Ibid. 5/ Surprisingly, the 1976 CPP supports the establishment of a number of parastatals, such as SOFISEDIT, oriented towards the industrial sector and tourism developments. - 8 - projected a 4.4Z GDP growth rate to 1980. In particular, the 1974 CPP estimates that because of irrigation and agricultural development in the South, weather should be a less decisive factor than today. In fact, as described in Chapter II, section E, weather (in terms of continuously falling rainfall and soil degradation) became an increasingly determining unfavorable force. 10. The 1976 CPP supports the Government's development strategy which continues to give priority to rural development and agricultural diversification. More specifically, it strongly endorses developing the irrigation potential through the Organisation pour la Mise en Valeur du Fleuve Sdn4gal (OMVS). Some concern is expressed about the possibility of gross public savings falling below debt service obligations -- forcing the Government to roll over and refinance its foreign debt; the urgent need for program aid is recognized. However, the Bank was not yet prepared to comply and move up front since program assistance was available from other sources (presumably France). Surprisingly, the CPP concludes that "despite the current liquidity crisis, Senegal's long term creditworthiness is not in question...." 11. The fourth period (1977/78 to 1980/81) was one of crisis. Two drought years combined with a drastic fall in groundnut world prices result in primary sector output declining by 4.0% p.a. and a marginal GDP growth rate of 0.75Z p.a. (Table 2). For the first time, total consumption exceeds GDP -- this ratio rising from 94.52 to above 102% (Table 3). Conversely, the investment share of GDP falls further from 18.02 to 15.22. In its 1978 CPP, the Bank revised somewhat its "fairly positive assessment of Senegal's long-term growth prospects" and lowered its GDP growth projec- tions to a rate of about 4.02 p.a. at best, during 1977-81. (In retrospect, this should be contrasted with the actual 0.75 rate.) Among others, it takes a very permissive, if not supportive, view of Senegalization which is "implemented with pragmatism and at a modest pace."6 Once more it points towards devaluation as the preferred and appropriate instrument which, however, is impossible to implement. 12. The 1978 CPP raises, perhaps for the first time, the trade-off between debt servicing, on the one hand, and growth and income distribu- tion, on the other, as the following quote indicates: "The real question is therefore not one of creditworthiness but one of limits beyond which additional borrowing would create the necessity of raising local funds for debt service payments in a way which is detrimental to overall economic growth."7 6/ Considering the extremely fast process of public encroachment which took place during the 1970s and until 1977 (see Table 16), thus before the drafting of the 1978 CPP, this evaluation appears difficult to rationalize. 7/ Through its membership in the Monetary Union, the Senegalese currency is fully convertible. - 9 - 13. The fifth and current period is that characterized by stabili- zation and structural adjustment. This period is extensively analyzed in Chapter II on structural adjustment. One might simply note here that GDP growth accelerated to 3.6% a year between 1980/81 and 1985/86, fueled by increased agricultural output influenced by relatively favorable weather in 1985/86.8 A disturbing macroeconomic development -- directly related to the stabilization process -- was the further drop in the investment ratio in 1985/86 to only 12.3% of GDP (at constant 1979 prices) and 13.8Z at current prices. 14. The 1984 Country Economic Memorandum (CEM) revised downward -- once more -- the Bank's macroeconomic projections for Senegal. Three alternative scenarios are developed between 1986 and 1995 as follows: Base Case Low Growth Hlah Growth GDP Growth Rat* 8.1 1.9 3.8 GDI/GDY 18.6 12.1 16.6 ICOR 6.0 6.4 4.2 The base case, an extrapolation of present trends, is shown to be infeas- ible because it implies financing gaps which continue to widen indefinite- ly, which leaves the low growth (forced adjustment) and high growth (managed adjustment) as the only two viable options.9 The CEM proceeds to make a case for the feasibility of the managed adjustment scenario (and a 4.0% GDP growth rate). However, by early 1987 in a thorough analysis of the impact of the adjustment process, the Bank lowers its sight still further and projects a 3.52 growth rate between 1986 and 1990 and between 1990 and 1995.10 15. At the most aggregative level the Bank's case for achieving the above managed adjustment growth path (i.e. a 3.8Z growth) hinges on two key factors, an investment/GDP ratio of 15.62 and an incremental capital-output ratio (ICOR) of 4.2 (i.e. an average gross return on investment of 23.82) -- as indicated in the preceding paragraph figures. Historically between 1960 and 1983, this ratio was around 6 (implying a return on investment of 8/ The fact that the base years 1980/81 corresponded with relatively low rainfall years and the two end years 1985/86 corresponded with two relatively favorable crop years means that the primary GDP and total GDP growth rates measured during this period may convey an exaggerated impression of annual growth. 9/ The 1984 CEM appears to be slightly inconsistent. On page 95 the above three scenarios are given as listed, while on page 102 where detailed indicators are presented, the corresponding growth rates of GDP for the low growth option is given as 2.22 and 4.02 for the high growth option. 10/ World Bank, Senegal: An Economy Under Adjustment, Report No. 6454-SE, February 13, 1987, p. 36. - 10 - 16.7Z). Through better project preparation, selection and implementation and the larger share of directly productive activities in the Public Investment Program (PIP), the Bank expects the above discrote jump to occur in the productivity of investment. Given the present state of project analysis and the institutional weakness of the Ministry of Planning in Senegal which is analyzed in some detail in Chapter III, one can be somewhat skeptical of the likelihood of reaching this level of productivity. However, even if one adopts this hypothesis, the prospects of attaining a 15.62 investment/GDP ratio appear totally unrealistic. As we have seen, this last ratio amounted to a dismal 12.3% in 1985/86 (the last year for which national income account figures are available) measured at constant 1979 prices. Since the Bank's policy scenarios above are also projected at constant 1979 prices, this is the correct indicator to focus on -- rather than the 13.82 ratio obtaining at current prices (see Table 3). Should the investment share remain at its present level (12.32) and using the high ICOR of 4.2 yields a future GDP growth rate of 2.92. Even if the investment share could rise from its 1985/86 level to around 14.02 and stabilize at the level (the maximum Bank's expectation as discussed in para. 138), the corresponding GDP growth rate would only be 3.33Z (using the high ICOR ratio of 4.2). The above aggregative analysis suggests that there is little basis for expecting, realistically, that economic growth in Senegal over the next five to ten years should depart more than marginally from its historical trend (i.e. a 2.42 rate between 1960/61 and 1985/86 at constant 1979 prices, see Table 2). At best, one could perhaps expect a rate of around 3.02. The analysis of the debt rescheduling implications on the limited availability of foreign resources for investment undertaken in Chapter II, section F strongly reinforces the view that at the macroeconomic level a rupture with past growth trends is most unlikely. It will be argued that debt servicing requirements are projected to absorb a large share of the anticipated future flow of foreign assistance -- leaving very little over for investment so that short of debt cancellation (not rescheduling) the Senegalese economy is most unlikely to propel itself out of its historical stagnation path. 16. The primary conclusion suggested by the review of Bank/Senegal relations period by period at the macroeconomic level is as follows. We have documented the continuing overly optimistic expectations regarding the prospects for socio-economic growth and development in Senegal. In its first CPP (1970), as we saw, the Bank projected a GDP growth rate of 5.6% annually between 1968 and 1976. Since then the Bank proceeded to adjust this estimate downward step by step as follows: (i) 4.52 (1970-77 in the 1971 CPN); (ii) 4.42 (1974-80) in the 1974 CPP; (iii) 4.02 (1977-81) in the 1978 CPP; (iv) 3.8-4.02 (1985-95) in the 1984 CEM; and, finally, (v) 3.52 (1986-95) in the Bank's 1987 CEM. Notwithstanding this continuing downward re-evaluation, each target proved, in retrospect, to have been greatly overestimated, when confronted with the actual observed performance given in Table 2. 17. The main reason for this phenomenon would seem to be, at least before 1980, a lacking perception of how binding internal and external constraints were to the development of the economy. Poor performance was typically blamed on presumably temporary exogenous factors such as the - 11 - weather (low rainfall) and cyclical changes in the international terms of trade and world demand for Senegal's exports rather than deep-rooted structural and institutional weaknesses. Of course, as we have illustrated in the historical review above, the Bank was cognizant of many of these weaknesses yet failed to realize that such trends as the continuing (i) public encroachment and Senegalization process and consequent weakening of the already low competitiveness of the private sector and of the fostering of entrepreneurship; (ii) the secular (not cyclical) worsening of the agro-climatic conditions (i.e. downward rainfall trend and soil degradation); (iii) the dismal domestic savings and investment performance -- both in terms of low levels and poor quality, and (iv) the cumulative and pervasive burden on the balance of payments and the net supply of foreign assistance available for productive investment of the rising external debt, predetermined economic stagnation. Take-off under these conditions was impossible. Furthermore, as will be seen later on, the continued over-optimism of the Bank easily contributed to erroneous strategic decisions. 18. Even in the post-1980 period, after the Bank recognized and accepted that the causes of stagnation were structural and not temporary (cyclical), it continued to underestimate significantly the tremendous difficulty of altering the structure and institutions so as to allow a new development regime. Given the implications of the debt burden -- discussed in Chapter II, section F -- it can fairly be asked if take-off is possible, in the first place, without a fundamental change in the international rules of the game such as partial debt cancellation and future foreign grants. B. The Lending Program and the Bank's Sectoral Strategy (1) Size and Composition of the Lending Program 19. During the past two decades Senegal has been a major recipient of international donor assistance, with the World Bank accounting, on average, for less than 10% of the total. Even so, on a country comparative basis, Senegal has been a preferred client of the Bank also: with 2.5? of the total population of all countries in "low-income Africa" it has received 4.7% of total World Bank commitments to those countries during 1970-86. 20. The first Bank operation in independent Senegal was an IDA credit for railway development in 1966. Since then and through 1987 there have been a total of 61 lending operations for a total commitment of about $800 million. IDA credits account for 80Z of this total; the bulk of IBRD lending went for projects in the transport, industry and tourism sectors. The last IBRD loan was extended in 1981, after which the country was judged not to be creditworthy for further IBRD lending. 21. Table 4 shows World Bank lending by major sectors and for sub- periods. Structural adjustment lending accounted for one-third of the total; project lending was fairly heavily concentrated in the agricultural and transport sectors, although the Bank has been active in a great variety - 12 - of sectors with a number of relatively small operations. After a slow start during 1966-71, Bank lending to Senegal grew rapidly to a yearly average of $85 million during the last four years. Project lending, however, has tapered off: it declined in absolute terms in the traditional sectors of Bank involvement, i.e. agriculture and transport. Table 4: WORLD DANK COMMITMENTS TO SENEGAL, 1966-87 (in millions of USS) 5 of Sector a 1966-71 1972-75 1976-79 1980-88 1984-87 Total Total Agricul6ure (18) 11 88 80 56 89 189 21 Industry A Tourism (6) - 8 18 17 - 88 4 Transport (15) 15 21 46 67 29 18 21 Energy & Telecom. (5) - 6 - 18 42 61 7 Urban A Water (4) - 8 8 - 80 41 5 Education A Health (6) 2 15 22 15 17 71 9 Subtotal 28 88 119 158 157 548 67 SALs & Other (7) 6 76 182 264 88 TOTAL (61) 28 86 125 284 889 812 100 Memo Item: Disbursements 18 89 80 157 298 585 /a Total number of projects is in parentheses. Source: World Bank (figures have been rounded). (2) Bank Sectoral Strategy 22. A detailed analysis of changes over time in the Bank's strategy for Senegalese development at the individual sector level is presented in Chapter III, at least for the major sectors of Bank involvement -- agricul- ture, industry, transport and education. Project development in the Bank is often a long time process and tends to generate a momentum of its own. The distinct macroeconomic subperiods discussed earlier appear to have had little impact on the nature or sequencing of projects, with one major exception, i.e. the turning point in strategy, which occurred towards the end of the 1970s. Before that time the Bank was in varying degrees basically upbeat about the country's future, and its project lending covered a great variety of sectors. The 1979 Economic Report, based on the findings of an economic mission in November 1976, presented a much more sober picture of the economic outlook and called into question previous Bank assessments. The country's growing financial difficulties, aggravated by the 1979 oil price increase, together with often unsatisfactory past - 13 - project experience led to a complete shift in strategy with the introduc- tion of structural adjustment lending in 1980 and reduced emphasis on projects in subsequent years. 23. During the first few years of the Bank's involvement in Senegal, it took a fairly long while to establish a close relationship with the country. It proved to be extremely difficult to find well-prepared projects and the selection of the first couple of projects was more the result of their availability and suitability for Bank financing rather than of any strategic considerations. After several years of economic analysis and reporting on the country, the Bank did begin to develop a strategy by about 1970. Senegal was now judged to have overcome the structural problems that followed independence -- the loss of markets for its manufac- tures, the departure of the French army, the cessation of preferential groundnut export prices -- and the future seemed to hold promise. Bank strategy documents laid stress on the need to strengthen the country's competitive position and improve production incentives, especially in the rural sector. Diversification became the main objective and the focus of the lending program was to be on directly productive activities (agricul- ture, manufacturing) and supporting infrastructure. This Bank strategy mirrored and supported government objectives and plans and, for the most part, continued unchanged throughout the 1970s. 24. There were, however, two major problems with the Bank approach. First, the judgment that, by 1970, Senegal had overcome its major post- independence structural problems was wrong. While the 1970 CPP correctly identified a large number of problem areas, including major ones such as the high price and cost structure, concern with the exchange rate, the high number of civil servants and the emerging public finance difficulties, nothing had been done to remedy any of these and there was no reason to assume them away. The exchange rate is an unusually interesting case of neglect by design of a major policy variable. From the very beginning of Bank economic work it was identified as crucial: an internal document (March, 1963) states "Consequently the fundamental problem of adjustment appears to be closely related to the valuation of the CFA franc." Subsequently, the history is one of Bank management variously instructing the staff to "study the exchange rate problem' and then ordering exchange rate discussions removed from economic reporting. 25. Second, the Bank's assistance strategy for Senegal was in actual practice less well focussed than that described above. CPPs and other internal documentation throughout the 1970s suggest that the overriding objective was to be involved in a large number of sectors. On the one hand, the Bank did not have a clear vision of which areas of the economy had most potential and were most in need of support, hence the spread of the lending program over all sectors. On the other hand, the Bank wanted to maximize the flow of resources to the country both from itself and from other donors. Thus, the 1971 CPP states: "In order to achieve even moderate economic growth, Senegal, given its lack of major natural resources, has to rely on a number of relatively small projects in a variety of productive sectors. This requirement is reflected in national plans and is a determining factor for the character of the Bank Group - 14 - oprrations program which is closely related to the country's planning process....Our operations program comprises a large number of projects spread over a variety of sectors. This approach is justified in view of Senegal's limited but diversified potential., And, in 1978 the CPP says: "The strategy we have followed and propose to intensify is one of increasInS Senegal's absorptive capacity so that it can benefit as much as possible from available foreign aid. We therefore want to maintain our broad involvement in all key sectors of the economy... .For larger projects we will continue to solicit cofinancing, thus preserving the possibility to maintain our involvement in many sectors at the same time and offering to less experienced donors ready opportunity to invest money." 26. In other words, the lack of a long-term perspective for the development of Senegal's economy, coupled with the desire to increase the flow of resources and meet lending targets, led the Bank to look for oppor- tunities and eventually be involved in almost any sector. The review of project experience in Chapter III suggests that the single most important reason for the overall disappointing project performance is that, in the process of trying to reach its lending targets, the Bank has regularly tended to ignore some major issues. Although these issues were often well known to the Bank through its economic and sector work, for one reason or another they could not be dealt with. The lending program went ahead anyway with the Bank hoping for the best, mostly with poor results. 27. The Bank's lending in both the agricultural and industrial sectors suffered during the 1970s primarily because the overall policy and incen- tive framework was simply not conducive to growth, and little was being done to improve upon it. In the transport sector Bank projects succeeded when issues were faced squarely, but failed otherwise. In the education sector projects succeeded in a narrow sense but the Bank neglected to address the major bottleneck to long-term growth, i.e. the deficiencies in primary education. The Bank's involvement in these four sectors is reviewed in Chapter III. There is little to comment on the Bank's role in the power, energy, telecommunications, urban and water sectors. Bank projects have been sporadic and some are very recent: most of these have been implemented successfully. The 1972 Sites and Services project, however, is a good example of the Bank's eagerness to lend to Senegal even when major issues have not been settled. This was the Bank's very first urban project meant to be a model for replication elsewhere. Yet, the objectives and policies of the Bank and Government were simply different, from the outset: the Bank was launching its new urban program while the Government only wanted to build some additional housing units. The project encountered numerous difficulties and delays. 28. The Bank's turnaround in strategy towards Senegal by the end of the 1970s was dramatic. A comparison of the lenoing program proposed for FY79-83 in the July 1978 CPP with actual lending during those years illus- trates the point. After years of hesitation and misgivings the Benk had become, during the second half of the decade, very enthusiastic about irrigation in Senegal, and was planning to make it a major element of its assistance program. As stated in the 1978 CPP "Large-scale irrigation should be seen as the most certain and least cost way to permit this, - 15 - almost stagnant, economy to grow at a modest rate of 4 percent a year.* Two large irrigation projects were included in the lending program, one each for FY81 and FY82, making up fully one-fifth of the total lending proposed for the five-year period. In the event, neither project materialized as structural adjustment lending became the main focus of the Bank's assistance strategy.11 29. The shift in strategy was appropriate for at least two reasons. First, Senegal's internal and external financial disequilibria had become almost insurmountable and the country was in desperate need of emergency financial assistance. Second, major reforms in the policy environment were called for to permit the re-establishment of financial equilibrium and lay the basis for a resumption of growth. As project lending is not a suitable instrument to induce such changes, program assistance was the obvious choice. The strengths and weaknesses of this new strategy are discussed in various parts of this report. The primary conclusion is that, while struc- tural adjustment lending to date has succeeded in removing major distor- tions in the economic policy environment and thus helped lay some of the necessary conditions for economic growth, these conditions are far from sufficient, a factor to which the Bank has paid insufficient attention during the 1980s. Yet it is unlikely that factor productivity in Senegal will increase in response to a "better" policy environment only. Simul- taneous direct institutional and project-level support will be essential. 11/ Subsequent to an Irrigation Technical Assistance project in 1985, an irrigation project eventually materialized, approved by the Board in December 1987. - 17 - Chapter II: STRUCTURAL ADJUSTMENT AND CONDITIONALITY A. Structural Adjustment Objectives and Constraints as Perceived by the Bank and the Government 30. The chapter on the macroeconomic developments in Senegal since 1960 reveals clearly that the economic situation during 1977/78-1980/81 had become untenable. With large and rising current account and budgetary deficits, internal and external disequilibria had reached the crisis stage. A cumulative process of worsening economic performance was underway. Poor agricultural and export performance resulted in GDP and output stagnation while both private and government consumption continued to grow at frightening rates (between 1977/78 and 1980/81, government consumption grew at almost 5% per year and private consumption by 3.21). By 1980/81, consumption exceeded domestic production and domestic savinge had become significantly negative. Given the depth of the crisis, some stabilization and adjustment process was unavoidable. 31. The choice which the Government confronted was between: (i) a passive posture of allowing the cumulative retrogressive trends to lead to some breaking point which would trigger a process of forced adjustment or (ii) intervene actively with appropriate policy measures and reforms to reach a modicum of internal and external equilibrium. The Government opted for the second alternative which, although entailing high political risks, was still preferable to the uncharted course which the first alternative would have led to. After extensive discussions with the Bank and IMF, the Government formulated and issued in December 1979 an "Economic and Financial Rehabilitation Program," followed in October 1980 by a "Declaration on Economic Policy." These documents reflected the Government's commitment to major economic and institutional reforms and paved the way for a stabilization and structural adjustment program with the IMF and the Bank which is still ongoing eight years later. 32. In order to evaluate critically the structural adjustment process and the role of conditionality, it is essential at the outset to spell out as explicitly as possible the objectives of both parties (the Government and the Bank) and the constraints they faced and continue to face. The nature and evolution of the relationship bi.tween the Bank and the Government throughout the adjustment period can best be explained in terms of these respective objectives and constraints. Subsequently, the analysis which follows can form the basis for examining these interacting relation- ships within a simple game theoretic framework. 33. In general, both parties shared fairly similar objectives but, in many instances, for somewhat different reasons. In contrast, they con- fronted significantly different constraints. In the immediate and short run, both the Government and the IMF/World Bank desired the restoration of some degree of internal and external equilibrium through appropriate stabilization policies. From the Government's standpoint, an additional - 18 - flow of foreign aid was needed to avoid defaulting on its foreign debt obligation as well as to avoid too drastic a cutback on government expendi- tures and imports which were required, respectively, to reduce the budgetary dcficit and the balance of payments deficit. The capacity to continue to service the rising foreign debt and to import strategic imports (particularly food) was at the heart of the political and economic viability oi the regime and, ultimately, its survival. In addition to insisting on these stabilization objectives, the IMF and World Bank had another reason than Senegal for going as far as they could to help the latter in servicing and rescheduling its external debt. A default by Senegal, a key Sahel country, would have sent shock waves throughout the donor community and the third world and could have led to a band wagon effect which would have further endangered the already fragile interna- tional monetary system. Thus, default was to be avoided at almost any cost. 34. In the medium and long run, and conditional upon the restoration of some degree of external and internal equilibrium, the World Bank resources under structural adjustment lending were to be used to encourage economic growth and distributional objectives favoring agriculture and the rural households which throughout the period between 1960 and 1980 had been discriminated against. Whereas here again there existed a strong agreement between the two parties regarding the ultimate objectives, they faced significantly different constraints. The Government by adopting a wholR package of structural adjustment measures -- including higher arricultural prices; publ3c sector disengagement; keeping a lid on the number of civil servants and their salaries; import liberalization and the elimination of most subsidies -- was bound to hurt certain socioeconomic groups in terms of their relative, if not absolute, income position and political power over, at least, a transitional period. Civil servants would be hurt by a combination of wage restraints and higher food prices which would also affect informal sector workers' households in the urban areas, and indus- trialists and workers producing import-substitutes would lose the benefit of the contrived rent they had enjoyed in the past. Even some of the rural groups such as the large producers in the Groundnut Basin (mourides) were negatively affected by the new seed policy and the reorganization of cooperatives. For a Government which traditionally had supported these groups at the expense of the bulk of the rural population, the potential political costs of removing the artificial rents and subsidies upon which this support was based was very high. This was particularly true since from a political economy standpoint, these same urban groups by being smaller, more homogeneous, better organized and less dispersed geographically, could exert political pressure through collective action much more efficiently than the farmers. 35. The Government was confronted with the very difficult problem of making a case to these groups, if not convincing them, that their short- term losses would be more than compensated for by long-run benefits. In addition, some transitional compensation mechanisms for these vulnerable groups had to be designed to make the pill somewhat more palatable. Preventing too strong a reaction and opposition to structural adjustment - 19 - measures by these groups was a major constraint which had to be considered in the design of the new economic policies, structural changes and institu- tional reforms. Thus, from the Government's standpoint, structural adjustment funds were required not just for continuing to service the foreign debt and thereby maintaining some degree of credibility with its creditors and pave the way for major structural and sectoral changes which would hopefully generate some economic growth but, also, to provide some partial compensation to some of the more powerful short-run losers. 36. Since the Bank itself did not have to implement the structural adjustment measures which it recommended (and sometimes forced through), it was much less sensitive to negative effects of adjustment measures on the incomes and relative power on these same groups (e.g., civil servants, industrialists and workers producing import substitutes). At the limit, it could be suggested that obtaining a minimum degree of acquiescence for the structural adjustment package by these same groups was a binding constraint for the Government. Incidentally, the distinction between policy objec- tives and constraints is not entirely clear cut. When a constraint becomes binding, it is tantamount to an objective which has to be satisfied, as some subsequent examples illustrate. 37. It would be wrong to suggest that the Bank was completely insensi- tive to the reactions of the vulnerable groups. The Bank did, e.g., under SAL III, help establish a special reinsertion fund for redundant workers (para. 160). But, on the whole, the costs of adjustment were portrayed as transitional costs pending a substantial supply response from the economy. The Bank tended to be much less sensitive to the concerns of different socioeconomic groups than the Government which had to implement the package and carry the burden of its consequences. 38. On the other hand, the Bank was confronted with a delicate yet powerful constraint on its own lending behavior. Starting in the late 1970s, there was a growing perception among the donor community and the public at large in the Western World that the transfer of resources to Africa and the Sahel countries, in particular, had to increase signifi- cantly. The rationale for this view depended on a combination of factors such as (i) humanitarian concern for one of the poorest regions in the world, (ii) the provision of food either directly or indirectly to relieve drought; (iii) for many western donors (excluding France) as compensation for a biased past regional allocation of foreign assistance, benefitting Asia and Lalin America, at the expense of Africa; and (iv) a whole host of geo-political reasons including as a possible counteraction to the relatively weak posture of the West vis-a-vis apartheid in South Africa. 39. Under these circumstances, the World Bank was placed under strong pressures to lend to Senegal. It can be hypothesized that even if an explicit lending quota had not been predetermined by management, at least an implicit quota is likely to have existed. Under the weight of world opinion -- which tends to equate foreign assistance with relief and as a precondition to improved performance, but almost never visualizes aid as a - 20 - crutch impeding development -- placing a filter on the amount of resources provided to a prototype country such as Senegal would have hurt the Bank's image. In some respects, increasing the flow of lending to Senegal was a constraint which the Bank faced -- particularly in the light of the French retrenchment (para. 165). 40. In short, the large foreign assistance requirements of Senegal throughout the 1980s caused by its critical debt and resource gap situation combined with the pressures on the World Bank to transfer resources to Africa and Senegal, in particular, are crucial to an understanding of the structural adjustment process and conditionality game played by these two parties. Next, we turn to a brief description of the structural adjustment loans (SALs) analyzing, more specifically, the types of conditions which they incorporated in order to understand better the policy dialogue and conditionality game played by the two players. B. Brief Description of the Three SALs 41. Three SALs have been extended to Senegal so far. SAL I was approved in December 1980, SAL II in early 1986 and SAL III in mid-1987. The rationale and strategies underlying these loans are described in great detail in the three President's Reports which should be consulted for more detail.12 Only the essence of the three SAL programs is described here before scrutinizing, in much more detail, the recommended measures included in the strategy (i.e. conditions) and developing a typology of these conditions. The terms "conditions" and "conditionality" are used in this report to refer to the whole set of measures described in the letters of development policy, not solely to those measures which are specific tranche release conditions: it is the program as a whole, made up by the complete set of measures, that is approved by the Bank's management and Board of Executive Directors as a condition for success. 42. Annex 1 contains the complete matrix of measures regarding structural adjustment in Senegal for the three SALs. For SAL I this matrix consists of an identification of the main economic problems, their symp- toms, the Government measures taken, the Government measures planned (i.e. the conditions), the Bank monitoring of the conditions, and crucial dates. The format of this policy ratrix for SAL II is slightly different consisting of structural issues, actions taken by the Government, measures to be taken, and monitorable actions and timing. Finally, the matrix for SAL III contains three columns, respectively, for objectives, recent developments and progress under SAL II, and conditions encouraged under SAL III. Annex 1 should be consulted for specific information on these programs. 12/ World Bank, President Report Nos. P-2869a-SE, November 26, 1980 (SAL I); P-4213-SE, January 10, 1986 (SAL II); and P-4498-SE, May 4, 1987 (SAL III). - 21 - 43. The SAL I program specified a number of actions aimed at increasing public savings (as was seen in Chapter I, domestic savings had become negative in 1980/81 to the tune of 2.52 of GDP), raising the produc- tivity of investments, improving the efficiency of the parastatal sector, providing stronger incentives to producers and exporters, containing urban consumption and initiating a whole set of institutional and policy reforms in the agricultural sector. The broad macroeconomic objectives of the program over the medium term, i.e. 1981 through 1985, were to: (i) reduce the current account deficit from 15.6? of GDP to 6-72; (ii) generate net public savings from 15? of public investment in 1981 to 25? in 1985; (iii) raise the overall rate of investment from 16? in 1981 to 18? in 1985; and (iv) achieve an annual growth rate of GDP of 4?. The Program Performance Audit Report (PPAR) concluded that during 1981/82: "The Government succeeded in substantially improving its investment program, reforming a number of important parastatals, providing export incentives to domestic producers, and reducing distortions of some key commodity prices. However, the Government failed to meet the macroeconomic objectives laid down in the SAL and EFF agreements, and made virtually no progress on agricultural reform. Consequently, the EFF was discontinued (January 1981) and replaced by one-year stand-by agreements; disbursement of the second SAL tranche was postponed and eventually, the outstanding balance of the loan was cancelled (June, 1983)."13 44. There were three main contributing factors to the failure of SAL I. First, two successive bad weather years in 1979/80 and 1980/81 led to a disastrously low groundnut crop, export receipts and domestic food crops. Secondly, the structural adjustment exercise was based on incomplete and inaccurate information. In particular, the public fireign debt was much higher than had been thought by the multilateral agencies, and the Government's deficit and arrears were, likewise, significantly higher than the initial estimates. Thirdly, excessively optimistic forecasts of exogenous variables and, in particular, world gv:oundnut prices encouraged the Government to raise producers' prices by a whopping 43? in 1981. As international groundnut prices fell thruugh the floor in 1982, the margin between the high producer price and the low export price, which had to be covered by the Government, led to one of the worst hemorrhages in the Government's budget. In a thorough evaluation of SAL I, John P. Lewis, the past chairman of the Development Assistance Committee (DAC), finds that in retrospect the most worrisome aspect of SAL I was that of modalities. No fewer than 32 conditions were attached to the loan with 13 actions having to be completed within the first three months of the life of the loan. To quote him, "The procedural tone or style of the undertakings and then, on the other hand, their particulars -- their operational fine print -- were ambivalent.'14 He lays the blame on both the Government and the 13/ Operations Evaluation Department, PPA.: Senegal Structural Adjustment Loan and Credit, Report No. 5637, .ay 9, 1985, p. iii. 14/ John P. Lewis, "Aid, Structural Adjustment, and Senegalese Agriculture," August 1984, p. 43. - 22 - Bank for the failure of SAL I. Specifically, he feels that the Government waL wrong not to have consulted with the Bank about the cancellation of farmers' debts, about the level at which groundnut prices for producers should be pegged and about the winding up of the Office National de Cooperation et d'Assistance pour le Ddveloppement (ONCAD). On the other hand, the Bank was wrong to have let itself get jockeyed into putting all of its SAL agricultural reform eggs into a fertilizer and seed distribution basket. It was especially wrong to let the procedural imperatives press It into plumping for particular fine-grained policy models, such as a specific format for local cooperatives, or the rather odd seed policy solution it anointed, before its knowledge was full and reliable enough to warrant imposing such policy specifics on the client. (This is a troubling aspect of the record, because it bespeaks a kind of Washington-centered analytical arrogance taat over the years often has tainted this mainly admirable institution.)115 45. On the more positive side, Lewis gives credit to the Bank for experimentation with various organizational alternatives in agriculture. The Bank knew that decentralization was essential but did not have a sufficiently complete grasp of the cultural, tribal, statu, migratory and other sociological dimensions of Senegal's rural fabric to identify a specific institutional and policy model for the replacement of ONCAD by village level cooperatives or other institutions. 46. The Operations Evaluation Department (OED), in its 1985 PPAR (Report No. 5637), concluded that among the lessons emerging from SAL I were: (i) the need for firm political support on the part of the recipient of the structural reforms -- the issue being that while the Bank was aware of the political opposition that could be expected from urban population groups and the bureaucrats, it underestimated the opposition of the rural elite and the lack of cooperation on the part of the Minister of Rural Development; (ii) the impact that alternative sources of external finance can have on the political resolve to adhere to the conditionality of a SAL -- alternative sources of finance were available to the Government with no or little conditionality attached; (iii) the need for adequate preparation of a SAL program, especially when conditions are derived from questionable baseline data -- i.e. setting precise quantitative criteria on key indicators when a reasonable degree of confidence in baseline figures is lacking, might make such criteria obsolete from the start. We shall return to many of the issues which have been touched upon above with specific emphasis on the different types of conditions imposed and their link with objectives. 47. Two main themes underly the Government's strategy during SAL II as well as SAL III, i.e.: (i) progressive withdrawal of the state from direct involvement in production activities and the promotion of private sector initiatives; and (ii) achievement of greater efficiency of public resource management through improvements in the quality and efficiency of public 15/ Ibid., pp. 53-54. - 23 - investments and reforming the public sector. SAL II consisted of a large number of conditions (46) ranging over an extremely wide range of sectors and issues. The main features of the SAL II package can be broken down in the two following main headings and appropriate subheadings: (1) Policies for accelerated growth in production and employment (a) Incentive reforms in agriculture to rAise cereals output through: adjustments in the price of imported rice; the liberalization of cereals marketing; floor price support mechanisms for cereals; the preparation of a program of supporting meesures to improve the production potential for rainfed and irrigated cereals; and the restructuring of rural development agencies. (b) Steps to decontrol input distribution and marketing arrange- ments in the groundnut sector, increase the producer price of groundnuts to improve farmers' incomes; and reduce the cost of government support. (c) A comprehensive package of industrial policy reforms aimed, among others, to reduce the levels of effective protection, rationalize the export subsidy system and progressively eliminate price controls. (2) Public sector management (a) Implementation of a two-year public investment program which channels resources to productive sectors; reform of the public investment programming and budgeting process to develop a three-year "rolling" investment program and budgeting system; and preparation of the first three-year investment program and consolidated budget; (b) Acceleration of parapublic sector reforms, including elabora- tion of a coherent strategy to guide the restructuring and/or privatization of the bulk of state-owned enterprises; (c) Reducing financial disequilibria through measures to control the growth of expenditures, improve the revenue base and reduce payment arrears. 48. In terms of macroeconomic objectives, two projection scenarios were prepared: one, portraying what might take place without the conscious implementation of structural adjustment policies, i.e. "a forced adjustment" scenario, and a second set of projections illustrating a growth path considered feasible and consistent with the estimated financing resources and the application of adjustment policies. Under the second "managed adjustment" scenario, the following macroeconomic targets over the period 1986-90 were specified: (i) a GDP growth rate of 3.4Z per year - 24 - (compared to a rate of 1.42 under the forced adjustment alternative); (ii) a growth rate of exports of goods and services of 5.12; (iii) the achievement of a ratio of domestic savings to GDP of 10.7Z (compared to 12 in 1984) and an investment ratio of 17.32 (compared to 132 in 1984); (iv) the achievement of a debt service to export ratio of 202 and a ratio of debt outstanding to GDP of 66.12 percent by 1990. In retrospect, these targets appear quite optimistic and consistent with a tendency to look at the future with rose colored glasses -- an issue already brought up in Chapter I. 49. SAL II was considered a success by the Bank. The loan was fully disbursed and many of the conditions were implemented by the Government. Among the most important ones were: (i) the adoption of a cereals plan, with a view to promoting food security and reducing reliance on imported foodstuffs and the abolition of the monopoly on rice imports formerly held by the Price Equalization and Stabilization Fund; (ii) the progressive rationalization of the system of effective protection, a phased elimination of quantitative restrictions on imports, a revision of the export subsidy scheme, the preparation of a three-year "rolling" public investment program and, finally, a program of public disengagement of a number of parastatal and public enterprises. Notwithstanding the positive appraisal of SAL II by both partners, the critical evaluation of the SAL II process which follows in section E raises a number of pertinent issues. 50. The relatively good economic performance in 1985-87 during and following SAL II was partially caused by exogenous factors (mainly the favorable weather) and partially influenced by the Government's policies. This raises methodological issues regarding the effects on socioeconomic performance of exogenous factors, not under the control of the policymaker, on the one hand, and the effects of policy instruments and institutional reforms, at least partially under the control of policymakers, on the other hand. This question is discussed in section C which follows next. In any case, the favorable economic performance between 1985 and 1987 led to SAL III which is by and large an extension of the strategy and program designed under SAL 11.16 The basic objectives of the program for 1987-88 to 1989-90 are to: (a) achieve an average annual rate of growth of real GDP of about 3.82; (b) continue to curb the annual rate of inflation from 5.62 in 1986-87 to some 3.42 in 1989-90; and (c) reduce further the external current account deficit, excluding official grants, from the estimated 10.8Z of GDP in 1986-87 to around 7.3% in 1989-90. Particularly ambitious is the 4.72 growth rate projected for the primary sector. In agriculture, the expansion is expected to stem largely from major increases in cereal production, while groundnut production is projected to remain around present levels. The target growth rates for industry and services are 3.72 and 32, respectively. 16/ SAL III did enter some new ground, especially in the areas of labor laws and the financial sector. - 25 - 51. To attain these macroeconomic objectives, the Government intends to continue to implement two key strategies consistent with SAL II, i.e.: (i) the promotion of private sector initiative and a more efficient utili- zation of resources through appropriate pricing and other incentive policies and (ii) reducing the direct involvement of public entities in production activities and achieving greater efficiency in public resource management. SAL III contains 77 conditions which are shown in Annex 1. C. Structural Adjustment and Conditionality within a Policy Framework 52. In order to evaluate the structural adjustment process, it is essential to analyze it within a policy framework. At the more general level the Government formulates a development plan which specifies a strategy in terms of a set of policy objectives to be achieved through a package of policy measures and institutional reforms. The structural adjustment objectives and conditions agreed upon by the Government and the Bank form an essential subset of the plan and at the limit, as will be seen later, may even coincide with the plan. Ultimate success of the structural adjustment process cannot be gauged by whether conditions per se are met but by whether socioeconomic performance improves. Improved performance, in turn, depends on (i) an accurate qualitative and quantitative under- standing of the effects of the policy measures and reforms (conditions) on the desired policy objectives; and (ii) the avoidance of unexpected unfavorable developments in key exogenous variables affecting performance, such as the weather and the international terms of trade. The first set of variables is, at least partially, under the control of the policymaker; whereas the second set consists of uncontrollable factors. The difficulty in evaluating the structural adjustment process is that good performance can be caused by favorable exogenous factors (such as an excellent harvest brought about by an unusually high rainfall) rather than by the right set of policies and conditions or that, alternatively, poor performance can result from unfavorable external or internal shocks notwithstanding the implementation of an appropriate development strategy and conditionality. It would be as unfair to credit the structural adjustment process for the favorable outcome in the first case above as to blame it for the lack of development in the second case. 53. A conceptual framework is needed to clarify and distinguish the complex impact of interacting policies and exogenous factors on perfor- mance. Such a framework is presented briefly next before applying it more specifically to the case of Senegal.17 The approach underlying this frame- work entails three steps. The first one consists of identifying the major 17/ This framework was developed in J. Lecaillon and others, Economic Policies and Agricultural Performance of Low-Income Countries, OECD Development Centre, Paris, 1987; Annex to Chapter 1. It was applied there in a somewhat different context to evaluate the performance over a 20-year period of six poor countries. - 26 - developmental objectives which can be taken as economic growth (G); a more equal (equitable) income distribution and/or poverty alleviation, which in the case of Senegal means reducing the large disparity between urban and rural incomes (D); an improvement in the balance of payments (B); reduced dependence on the rest of the world reflected by, e.g., food security and greater food self-sufficiency (S); greater price stability (P); and other socioeconomic including non-material (possib1xideological) objectives (0). In both the short and the long run, conflicts may prevail among and between objectives. It is important to keep this in mind when evaluating the overall performance of the economy. 54. The second step, after having identified the major objectives, and perhaps their relative importance, is to select a corresponding set of performance indicators reflecting these objectives. These performance indicators act as proxies reflecting the objectives themselves. Changes in these indicators over time can be used to ascertain whether performance has improved or worsened and, in many instances, the extent of it. By way of illustration, Annex 2 gives a set of possiole agricultural performance indicators at the aggregate and disaggregated levels corresponding to the above objectives. 55. The third and final step to evaluating performance is to attempt to estimate the effects of policy measures and structural and institutional changes -- and more particularly in the present context, the SAL conditions -- on performance indicators reflecting the various development objectives. However, as was just pointed out, exogenous variables -- amounting to internal and external constraints -- affect performance, in addition to policies and institutional changes implemented by the policymaker. It is therefore important to try to distinguish and separate, at least to some extent, the effects of these exogenous factors from those of policies and institutional changes per se. 56. Thus, from the standpoint of policy analysis, it is useful to subdivide variables into (1) uncontrollable variables; (2) variables partly under the control of the policymaker; and (3) performance variables which are affected by the above two categories. Table 5 presents a suggested classification of variables for policy analysis. The uncontrollable variables consist of constraints and shocks which can be either of internal or external origin. Examples of internal constraints might be the initial socioeconomic structure as well as the soil, weather, and agro-ecological conditions. (In particular, the amount and pattern of rainfall in Senegal is by far the most important determinant of agricultural output.) Examples of external constraints are changes in exports or imports such as the OPEC oil shock and increasing protectionism abroad. - 27 - Table 5: SUGGESTED CLASSIFICATION OF VARIABLES FOR POLICY ANALYSIS Uncontrollable Variables: I. Constraints and Shocks A. Internal 8. External Variables Partially Under Control of Policymaker: II. Institutional and Structural Changes III. Policy Measures A. Macroeconomic Policies B. Agricultural Policies Performance Variables Affected by I, II and III, above: IV. Performance Indicators A. Agricultural Indicators 1. Aggregate Indicators 2. Disaggregated Indicators B. National Indicators V. Development Objectives Source: J. Lecallion and others, Economic Policies and Agricultural Performance of Low-Income Countries, OECD Development Centre, Paris, 1987. While the focus of this OECD study and of the above table is on agriculture, the table Is, of course, equally applicable to other sectors. Instead of OAgricultural,' please read 'Sectoral.0 57. Variables partly under the control of the policymaker can be subdivided into two types based on the actual degree of control. Institu- tional changes and reforms usually entail changes in legal and property rights; they tend to have discrete effects on the socioeconomic structure and their impact on performance is notoriously difficult to predict. In contrast, policy measures or instruments (such as the money supply, tax rates, and price policies) can be manipulated by the policymaker much more flexibly and gradually within a given institutional and socioeconomic structure. - 28 - 58. Under a structural adjustment program, conditions are embedded in the larger package of policies and institutional (structural) changes which constitute the government development strategy. In the context of Senegal, it will be seen that the conditions which were part and parcel of SAL II and III represented a large subset of the measures included in the recent development plan so as to almost overlap with them and be indistinguishable from them. 59. The key question from the standpoint of policy analysis is the nature of the interrelationships which exists among these various types of variables. Figure 2 illustrates graphically the direction of influence among and between the five classes of variables defined above. Arrows 1 and 2 represent the effects of policy instruments mainly at the macro- economic level but also, to some extent, at the sectoral level on national and sectoral performance indicators.18 Institutional changes and reforms and, in particular, organizational changes within a sector, on either the production or distribution sides, would be captured by arrow 3 in Figure 2. In turn, any evaluation of performance has to take into consideration the impact of internal and external constraints and shocks over which the policymaker has no or very little control (arrows 4 and 5). 60. Arrow 6 reflects the interdependence among performance indicators, i.e. the contribution which sectoral indicators can make to overall performance at the national level and how, in turn, it can be affected by the latter. Next, arrow 7 indicates that the choice of performance indicators at the national and sectoral levels is derived from the major development objectives. By the same token, performance indicators are used to judge progress in achieving development objectives. (This explains why arrow 7 is bidirectional.) Finally, arrow 8 indicates that the design of institutions and the choice of policies depend, in the first instance, on the development objectives pursued by the state which in the context of a structural adjustment program would be significantly influenced by the World Bank and the donor community. Furthermore, the policymakers will modify their strategies, namely the combination of institutions and policy measures chosen, in response to development success or failure as judged by the performance indicators and, likewise, the World Bank and donors' evaluation of performance. This relation, which is represented by arrow 8a, reflects the feedback from performance to changes in policies. 61. The conceptual framework described above (and captured graphically in Figure 2), besides providing an operationally useful classification of policy variables, indicates the direction of the major causal relationships among the different types of variables. As such, it is useful as an organizational device in scrutinizing the impact of individual as well as combinations of policy measures and institutional changes (and particularly 18/ In Figure 2, these two sets of arrows were combined together for expositional purposes but, in fact, arrow 1 should link macroeconomic policies to indicators and arrow 2 should link sectoral policies to indicators. I - 29 - structural adjustment conditions) on performance and separating, at least conceptually, their effects from those ot uncontrollable factors and con- straints. However, Figure 2 remains a black box as long as the various interrelationships among variables are not explicitly and quantitatively specified by way of some general equilibrium or even partial equilibrium models. 62. Clearly, even though it would be quite unrealistic to expect to capture and explain the combined direct and indirect effects of the set of all policy measures and institutions on performance indicators within a comprehensive general equilibrium model, it is important to know the implicit or explicit partial models and approaches which were used by the Bank in support of its recommendations. 63. More pointedly, one can ask what theoretical or empirical evidence the Bank had in support of specific recommendations. Thus, one of the arrows under 1 in Figure 2 would link agricultural price policies to crop output; how did the rank estimate supply response to changing support prices and other complementary measures and how did it estimate the impact on domestic demand? Undoubtedly the underlying evidence linking policy measures and performance indicators was more robust in some instances than in others. It is easier to estimate the marginal impact of a given policy instrument such as a change in price or tax rate, within a given structure, than it is to estimate the joint impact of a package of policy instruments -- let alone organizational changes which alter the prevailing socio- economic structure. 64. Before applying the preceding conceptual framework to evaluate selectively a number of actions and conditions underlying the three SALs, it is essential to look at these conditions concretely. - 30 - Figure 2 CLASSIFICATION OF VARIABLES AND RELATIONSHIPS AMONG CONSTRAINTS, INSTITUTIONS AND STRUCTURAL CHANGES, POLICY MEASURES, PERFORMANCE INDICATORS AND DEVELOPMENT OBJECTIVES Uncontrollable Variables I CONSTRAINTS A Internal Perfomance Variables IV PERFORMANCE 4 &5INDICATORS B External 8 ExtrnalA In Agriculture .1 Aggregate . Indicators -j V DEVELOPMENT Variables Partially OBJECTIVES Under Control of Policymaker 7 STRUTURA 3 L 2 Disag gregated II INSTITUTIONS Indicators CHANGES --0) 19B At National Level Ill POLICY 8 MEASURES A At Macro Level 8 Wh8 8 B Wihin Agriculture Source J Lecaillon and others. Economic Policies and Agricultural Performance of Low income Countries. OECD Development Centre. Parts. 1987 While the focus of this OECD study and of the above table is on agiiculture. the table is. of course, equally applicable to other sectors Instead of 'Agricultu'al. please read 'Sectoral - kgrw x44605c -31 - D. Analysis and Typology of SAL Conditions in Senegal 65. The main form of leverage which the Bank can exert in its policy dialogue with the Government is through its conditionality package (in a much more limited way the Bank can influence the Government through the quality of its socioeconomic analysis, project lending and technical assistance). A number of questions suggest themselves in this connection, such as: What form do these conditions take? How specific are they? Are they expressed in qualitative or quantitative terms? How intensive are they and how much "bite" do they carry? How many conditions were imposed conditions within any given SAL? 66. To answer at least some of these questions, we turn next to a more concrete examination of the conditions underlying the three SALs. The first and foremost observation which is suggested by this examination is the large and increasing number of conditions imposed on each SAL (see Table 6)19 and the great diversity of conditions ranging from simple studies to rather major institutional changes -- touching on a multitude of issues and sectors. 67. Because of the diversity of conditions and the fact that many of them do not constitute or entail policy actions as such -- in the sense of a change in a policy instrument or an institutional (structural) reform -- it is essential to develop an operational classification which should also be consistent with the above conceptual framework. The typology which is proposed here contains five different types of conditions, i.e.: (1) studies; (2) preconditions to actions, or pre-actions, including the formulation of action programs, government announcements, the issuance of decrees, "administrative strengthening," reviews, and plans of actions; (3) changes in policy instruments within an existing structure, e.g., changes in producer and consumer agricultural prices, changes in export subsidies, changes in import duties and quantitative restrictions, tax reforms, limits on budget expenditures by category, elimination of marketing monopolies, changes in monetary instruments such as credit control, and settlement of government's indebtedness towards the private sector; (4) intermediate targets which are conditions over which the Government only has limited control because they are affected by exogenous factors, e.g., increase in the proportion of public investment- to be financed from net public savings; and finally (5) institutional and structural reforms, such as abolishing or restructuring existing public enterprises and parastatals or establishing new ones, privatization of public-mixed enterprises, implementation of a reorganization program within 19/ These are the numbers of conditions appearing in the policy matrices in Annex 1. It should be noted that the list of conditions may differ very slighzly from one World Bank document to another as can the exact number of these conditions. - 32 - a ministry such as in Planning and in Rural Development, implementation of rehabilitation plans for major banks and testing new organizational structures in agriculture. 68. Clearly, "studies" and the whole set of measures included under "preconditions" above do not, by themselves, alter in any way the existing policy and institutional status quo. This is not to say that these two types of conditions are unimportant; in fact, they may be essential pre- conditions to the identification of appropriate policy actions. But it is only if and when they are embedded into changing policy instruments or reforms that they can affect performance. The shelves of ministries in LDCs are full of past studies, action programs and reviews which have never been implemented. 69. In contrast, the other three types of conditions have "bite"; their implementation can have an impact on performance -- although not necessarily in the predicted or desired direction.20 Their ultimate effects depend partially on exogenous factors and partially on the validity and accuracy of the underlying "models,' approaches or frameworks used by the Bank in linking policy variables with performance indicators (i.e. arrows 4-5 and 1-3, respectively in Figure 2). This issue is discussed later in section E. 70. An attempt was made in Annex 1 to classify each of the SAL condi- tions into one of these five categories. Even though some of the conditions did not fall unambiguously into one of these categories -- and other observers might have made somewhat different choices -- it can be argued that the breakdown in Annex 1 is representative of the types of conditions underlying the three SALs in Senegal. It can be seen from Table 6 that both the total number of conditions imposed rose sharply over time from one SAL to the next from 32 in SAL I to 46 in SAL II and 77 in SAL III, but also the proportion of conditions consisting of studies and preconditions increased from 34Z in SAL I and 52Z in SAL II to 56Z in SAL III. 20/ Incidentally, an action or institutional reform which is only incompletely implemented will have proportionately less impact on performance. - 33 - Table 6: STRUCTURAL ADJUSTMENT LOANS I-III: TYPOLOGY OF CONDITIONS L SAL I SAL II SAL III Studies (S) a 9 is Preconditions (P) 6 15 80 Actions (A) 15 19 28 Intermediate Targets (IT) 4 1 - Institutional Reforms (IR) 2 2 6 Total Number of Conditions 82 46 77 / The complete list of condit. ne is given in Annex 1. Source: Prepared by OED. 71. This brings us to an important concept in an evaluation of the structural adjustment bargaining process between the Bank and the Government, that is the "tightness" of the conditionality package. It can be argued that two of the dimensions of tightness are the number of condi- tions in a SAL as well as the proportion of hard conditions (actions, intermediate targets and institutional reforms) as opposed to soft condi- tions (studies and preconditions). An additional dimension of tightness is the range and breadth of the policy areas and issues covered by a SAL. Mosley in a comparative analysis of conditionality including all countries receiving SALs between 1980 and 1986 identifies 19 different policy areas under four different headings, i.e. trade policy, resource mobilization, efficient use of resources, and institutional reforms.21 He defines a tightness score exclusively in terms of the number of policy areas above included in any given SAL. Thus, SAL II in Senegal included 11 of these areas and, accordingly, was given a tightness score of 11 which is relatively high compared to other countries receiving SALs (the range went from 4 in Niger and 5 in Chile to 15 in Turkey).22 In SAL III (not included in Mosley's analysis), it appears that at least 15 of the 19 policy areas are covered which would yield the tightness index of 15. 72. There is still another dimension of tightness which is the inten- sity of a particular condition in terms of the magnitude of its potential effects on performance indicators. A major institutional reform (such as the abolition of ONCAD and replacing it by another institutional arrange- ment, or a major import liberalization program) can have a significant 21/ Paul Mosley, "Conditionality as Bargaining Process: Structural Adjustment Lending, 1980-86," Essays in International Finance, No. 168, October 1987, Department of Economics, Princeton University, see Table 2. 22/ Ibid., see Table 3. - 34- impact on efficiency and other objectives, while being much more difficult to carry out and entailing short-run costs to powerful socioeconomic groups. Intensity, in the above sense, is very difficult to measure accurately but should be kept in mind. It also bears some relationship to the government degree of difficulty in implementing the measure and the political resistance that it would engender. 73. The course of conditionality in Senegal throughout the 1980s suggests some important observations which are crucial to a better understanding of the bargaining process between the two parties. On the one hand, the significant increase in the number of conditions and in the range of policy issues covered suggests a tightening, more comprehensive adjustment program. On the other hand, the significant reduction in the proportion of "hard" conditions over time, combined with a lack of a strong priority ordering among the latter and a number of incomplete "hard" conditions (see para. 75) implies a relative weakening of the "tightness" index and a desire for greater flexibility in determining whether the Government had actually complied with the terms of the SAL. It is true that a few conditions were specified as binding constraints to the release of tranches but, outside of these, it would appear that the Bank could, to some extent, pick and choose which subset to base its disbursement on and which subset to overlook and use its own discretion in determining compliance or non-compliance in specific instance. We shall return to these key issues when scrutinizing the conditionality bargaining process more concretely. 74. One additional remark of a general nature is suggested by looking at the list of conditions appearing in SALs I-III (in Annex 1). This list does not contain any performance indicator reflecting the actual development objectives (possible examples of agricultural indicators are given in Annex 2) -- notwithstanding the fact that each SAL specifies such objectives at the macroeconomic and sectoral levels. Presumably the reason for this is that socioeconomic performance is partially affected by exogenous, uncontrollable variables (see arrows 4 and 5 in Figure 2) and, consequently, that it would be unfair to penalize governments for outcomes over which they have limited control.23 However, such indicators could still be used for monitoring the evolution of the economic and social situation and comparing it with original expectations. While this would not be considered as fulfillment or non-fulfillment of conditions, it could still be very instructive for both the country and the Bank about divergencies between expectations of outcomes of structural change and reality. 231 Somewhat surprisingly the Bank does use intermediate targets (such as an increase in the share of public savings to public investment from 15Z to 25Z) as conditions even though they are only partially controlled by the policymaker. - 35 - 75. A final general remark, before embarking on a more specific evaluation of the three SALs and their impact on performance, is that a number of the selected conditions are incomplete, in the sense that complying with them is insufficient to affect the socioeconomic structure. They need to be complemented by other measures to be effective and have a policy impact. Examples of such conditions, which will be discussed subsequently, are public disengagement (putting up parastatal and public enterprises for sale) which could only lead to a completed action through the engagement of the private sector (through the actual purchase of these enterprises); and the extreme reliance on agricultural price policies to affect output in the absence of crucial complementary actions on the marketing and distribution fronts. E. Basis for SAL Conditions and Impact on Performance (1) Introduction 76. By 1980, it was clear that a cumulative process of economic decline and retrogression was underway in Senegal. Even though this process was exacerbated by exogenous factors such as consecutive droughts and falling world prices for groundnuts, the underlying causes were structural and not temporary. To cope with this critical situation, the Bank and the Government, to their mutual credit, realized that to break this vicious spiral of worsening performance, a comprehensive macroeconomic approach was called for. Intervention was required on many fronts simul- taneously to help remove existing disequilibria and bottlenecks. The underlying logic was that a set of complementary and reinforcing measures, applied simultaneously, would, after a short stabilization phase, bring about the necessary structural and institutional changes to induce the economy to move along a sustainable growth and development path. 77. In this section we attempt to answer a number of questions relating to the basis upon which conditions were selected. More specifically, what theoretical, conceptual or empirical evidence relating policy instruments and reform to performance did the Bank use ex ante in formulating its strategy and conditionality package? Granted that a multi- sectoral comprehensive package of measures was called for, how did the Bank go about estimating the likely joint effects on different development objectives of a combination of measures and separating these effects from those of exogenous variables? In other words, how far was the Bank able to unveil the black box appearing in Figure 2? Are there apparent inconsis- tencies within the Bank's aid package (including structural adjustment, project lending and technical assistance)? 78. Still another key question relates to the process of converging on a strategy and conditionality package within the Bank. Were issues debated internally and how much controversy prevailed? Did the Bank learn from past experiences and failures (e.g., what did the lank learn from the SAL I episode in formulating SAL II?)? - 36 - 79. The analysis which follows is selective concentrating on a few crucial policy issues. One qualification which should be made at the out- set is that it is too early to judge the impact of SALs II and III on performance ex-post. Structural adjustment measures take time to work themselves out through the system and affect the economy. Therefore, it is premature to judge in any definitive way the degree of success the strategy endorsed by the Bank and the Government may have had. To the extent that perceptible trends are noticeable in the last two or three years, these will be scrutinized. However, the emphasis in the present evaluation is on the validity of the analytical and empirical basis for policy recommenda- tions ex ante and the process generating these recommendations. 80. The multitude of conditions contained in the three SALs (see Annex 1) fall into four major domains, as follows: (1) agriculture; (1) improving private sector efficiency (trade liberalization, export promotion, removal of price distortions, incentives for industrial develop- ment and employment creation, reform of the financial sector); (3) public investment and finance (improving the effectiveness of the public investment process, improving the budgetary process); and (4) public sector strengthening and disengagement, and private sector engagement. Each of these domains is evaluated, in turn, next, followed by a brief discussion of the social costs of adjustment. (2) Agriculture 81. This first domain of structural adjustment can be broken down into the broad areas as given in Table 7 with the corresponding SAL conditions. We focus the discussion in this section on the first two areas since the agricultural sector section in Chapter III contains a detailed analysis of institutional changes. - 37 - Table 7: CLASSIFICATION OF SAL CONDITIONS RELATING TO AGRICULTURE Conditions L SAL I SAL II SAL III Agricultural Prices and 11-18, 17, 2, 8, S 89, 41, 42 Credit 25, 27 44, 46 Research 22 Institutional Change, 21, 23, 24, 1, 4, 6-11 87, 88, 40 Reorganization and Planning 26 48, 46-58 L The numbering of the conditions corresponds to that given in Annex 1 which should be consulted for more detail. Source: Prepared by OED. 82. The issues which we want to highlight here fall within three main headings: (1) agricultural price policies; (2) agricultural research; and (3) agricultural performance and its determinants. (a) Agricultural Price Policies 83. Over the seven-year period 1980-87, the Government significantly raised all nominal producer prices as well as the retail price of rice. Highlights of the Bank's role in the evolution and design of price policies ranged from (a) the need to study cereal and export crop price policies in SAL I (conditions 11 and 12 in Annex 1), to (b) the introduction of a producer price support mechanism for cereals and the need to maintain a nominal protection rate of at least 25% for domestic cereals by adjusting the retail price of rice24 (conditions 1-3 in SAL II), to (c) the preparation of an agricultural prices and incentives study during 1987 (condition 41 in SAL III). 84. An observation suggested by the Bank's documentation, before analyzing the rationale for the Bank's recommendations in this domain, is the constant and continuing refrain for more studies on the effects of agricultural price policies. Even though this refrain had been whistled in the 1960s and 1970s, it became louder in the 1980s as judged by the following quote: 24/ The intent appeared to be to set the retail price of rice at a level 25Z above the retail price of millet. In fact, however, it seems that this margin of protection was applied to the price of unprocessed millet which resulted in the presumably unintended outcome that the retail price of millet was higher than that of rice. - 38 - "The prices of four commodities -- groundnuts, cotton, millet and rice -- are key parameters in the economy in that they affect output, rural incomes, the cost of living, and also (especially groundnut prices) the level of public revenue and savings. The program will seek to rationalize State intervention in these areas (partly on the basis of a study on the optimal relative prices for agricultural products commissioned by the Government and now nearing completion).... *25 Seven years later, in May of 1987, this same study was still awaited judging by the following quote: "An agricultural prices and incentives study is being conducted by foreign consultants... .The study should allow the Government to determine the relative prices of various cereals and cereal-based products and help decide whether the price of groundnuts should be the basis on which other agricultural prices are decided. The results of this study, to be available in August 1987 will lead to the preparation of an action plan before end-October 1987 (second tranche condition). The plan will be implemented for the 1987-88 campaign (third tranche condition).w26 85. Given the critical importance of this question, it is very unclear why such a study was still being expected by the end of 1987. The above quotes suggest that the Bank could not have felt very confident about the analytical basis for its recommendations in this area. It also raises the question whether its fine-tuning of agricultural prices (which became incorporated as conditions in the SALs), in the absence of a good study, did not amount to "putting the cart before the horse." We shall come back to this point shortly. 86. We turn now to the rationale underlying the agricultural price strategy. First, the objectives were clear. The Bank and the Government wanted to encourage a higher degree of food self-sufficiency by replacing rice consumption by millet-sorghum consumption and also by the substitution of domestically produced rice for previously imported rice. In addition, the high consumer price set on rice (at least double the world price) would provide an important source of revenues to the Government which would receive the margin between the high retail price and the low import price. Likewise, groundnut production was to be encouraged to increase exports and contribute to the earnings of rural households. 25/ President's Report, SAL I, p. 23. 26/ President's Report, SAL III, pp. 27-28. This study had not yet been completed by the end of 1987. - 39 - 87. To achieve these objectives, the Government -- prompted by the Bank -- raised all producer prices in nominal terms. This raises two sets of fundamental questions: (1) what kind of aggregate supply response could be expected at the level of the agricultural sector as a whole? (i.e. how responsive could total crop output be to higher nominal prices?); and (2) how could it be determined whether the set of complementary measures required to generate a higher level of output, in addition to price incen- tives, were included in the agricultural strategy? 88. In order for total agricultural output to grow, two conditions have to be satisfied: (1) real agricultural prices must go up, not just nominal prices; and (2) excess (unused) land or labor resources must be available, and/or there must exist a yield-increasing technical package and the mechanisms of distributing it and diffusing it to the farmers in the very short run. Since farmers respond to real prices rather than nominal prices, one should examine the trend of the former to determine whether the first condition above was indeed met. Table 8 provides a time series of official producer prices in constant terms (i.e. deflated by the consumer price index of a Senegalese family). This table shows clearly that no significant change in real price occurred between 1970 and 1986 for ground- nut, maize, and millet. If anything, the evidence suglests a slight fall in real prices for these three crops during the 1980s. 7 Thus, the first condition for an aggregate increase in agricultural output (an increase in their real price) does not appear to have been met. 27/ Considering the fact that fertilizer subsidies were greatly reduced and fertilizer distribution hampered, these factors would reinforce the above observation regarding real producer prices. Indeed, the ratio of nominal producer price to fertilizer price is often used as a proxy for the real producer price. Of course, when the availability of fertilizer is negatively affected by institutional changes, this is tantamount to a very significant increase in the price of fertilizer. I - 40 - Table 8: OFFICIAL PRODUCER PRICES, 1970-88 CONSTANT 1970 VALUES (CFAF/kg) Cotton Groundnut Maize Millet 1970 88 20 18 17 1971 80 28 18 17 1972 27 20 17 15 1978 25 24 20 20 1974 28 26 21 18 1975 24 22 19 16 1976 25 21 18 18 1977 28 20 18 17 1978 22 19 17 16 1979 28 19 15 17 1980 28 20 14 16 1981 24 26 16 18 1982 21 21 14 15 1988 19 19 13 16 1984 18 18 15 15 1985 28 21 16 16 1986 22 20 15 15 Source: World Bank, OSenegal: Agricultural Sector Strategy Brief,' Volume II: Statistical Annex, (draft), September, 1987. 89. This also appears to be the case regarding the second condition. The evidence available shows that both land under cultivation and yields have fluctuated widely -- largely influenced by rainfall patterns -- around stagnant, if not declining, trend lines. Table 9 and Annex 3 show that the area under cultivation was significantly lower in 1985-87 for groundnuts than it was at any time throughout the period 1961-87; while the yield in these years was about at the same level as at the beginning of the period. The cultivated area and yield for millet and sorghum appear slightly higher in the recent period. In the case of paddy, there is very little change in the area under cultivation but a significant increase in the yield while maize area and yields show growth. The bottom line is that for the two key crops, groundnuts and millet, constituting 86Z of the total area under cultivation, a completely static picture is apparent. A fundamental cause for this is the weakness of the research effort and the absence of technical packages which would have affected yields positively. This is a question we return to shortly in the next subsection. It is difficult to understand the basis upon which the Government and the Bank could expect a significant growth in aggregate agricultural output with neither of the above two conditions having been fulfilled. To compound this problem, our evaluation of the agricultural sector (see Chapter III) brings out that the - 41 - institutional weaknesses on the marketing and distribution sides act as an important brake to increased output so that even in the presence of real price incentives and the availability of viable technological packages, significant aggregate output response seemed quite unlikely. (Optimistic agricultural growth targets were set by the Bank, i.e. 3.5% annual growth for the next 10 years and a growth of cereal production between 1991 and 1995 of 4.9Z per year.)28 If total agricultural production could only grow marginally, the Government and the Bank could, at best, expect a relative improvement in rural incomes vis-a-vis urban incomes because of the changes in the internal terms of trade. Table 9: AGRICULTURAL OUTPUT AND PRODUCTIVITY OF MAJOR CROPS, 1960/61-1986/87 (yearly averages for subperiods) Yearly Groundnuts M I I*/Sorghum Averages Production Area Yields Production Area Yields (tons) (ha) (tons/ha) (tons) (ha) (tons/he) 1960/61-1964/65 942,180 1,031,000 0.914 442,960 671,600 0.150 1965/68-1969/70 918,640 1,109,240 0.828 644,620 1,062,500 0.518 1970/71-1974/75 761,287 1,052,712 0.728 474,401 959,987 0.494 1975/76-1979/80 962,187 1,224,987 0.788 569,147 969,104 0.687 1980/81-1986/87 703,032 916,559 0.767 68,236 1,077,888 0.591 Rice Maize Production Area Yields Production Area Yields (tons) (ha) (tons/ha) (tons) (ha) (tons/ha) 1960/61-1964/86 91,820 74,660 1.227 29,280 84,800 0.841 1965/66-1969/70 116,840 90,140 1.291 42,720 54,820 0.787 1970/71-1974/75 88,804 74,928 1.112 88,066 43,896 0.762 1975/76-1979/80 100,821 72,161 1.390 46,601 57,787 0.805 1980/81-1986/87 119,774 69,280 1.729 87,189 78,682 1.106 Source: Annex 8. 90. The fact that aggregate supply response is constrained by the availability of inputs, technology, the profitability of farming as reflected by the real price of output, ac well as the efficiency of the marketing and distribution system means that even under the best of circum- stances, the elasticity of total output to prices tends to be quite low. However, the same reasoning does not apply at the individual crop level. Changes in relative producer prices can lead to significant shifts in resource use and the allocation of cultivated land among different crops. 28/ See President's Report, SAL III, p. 7, and Senegal: An Economy Under Adjustment, p. 35. - 42 - Within a given resource endowment and technology, changes in relative prices can induce changes in the output mix along the production frontier. Although not explicit on this issue, one gathers the impression that the goal of food self-sufficiency became relatively more important over time than that of promoting groundnut exports in the Government/Bank agricultural strategy so that one would expect the relative producer price of the latter to the former to have fallen. The evidence is not entirely clear-cut on this issue, as discussed in detail in Chapter III. Table 10 shows that the ratio of groundnut to millet prices exhibits no clear, consistent trend after 1980. In contrast, the new price policy entailed a significant fall in the relative price of millet relative to domestically produced paddy from around .96 in 1980-82 to .82 in 1986. This suggests that, in terms of relative priority, the price instrument favored first paddy followed next by millet and groundnuts. To estimate the effects and changes of relative prices on the output mix, one has to examine the existing evidence on supply response at the individual product level. The limited econometric evidence available -- presented in some detail in the agricultural sector section in Chapter III -- points to the predominant influence of rainfall and the limited impact of prices on both groundnut and millet output. - 43 - Table 10: RELATIVE PRODUCER PRICES (CFA/KG),/a 1970-38 Groundnuts/ Oroundnuts/ Oroundnuts/ Millet/ Maize/ M*ize/ Cotton Maize Millet Paddy Cotton Millet 1970 0.84 1.00 0.90 0.95 0.64 0.90 1971 0.79 1.22 1.29 0.81 0.84 1.00 1972 0.78 1.22 1.29 0.81 0.60 1.06 1978 0.80 1.26 1.87 0.8 0.63 1.09 1974 1.08 1.40 1.40 1.19 0.74 1.00 1975 0.85 1.14 1.83 0.78 0.74 1.17 1976 0.86 1.14 1.38 0.72 0.74 1.17 1977 0.82 1.08 1.14 0.84 0.78 1.06 1978 0.82 1.08 1.14 0.84 0.76 1.06 1979 0.88 1.16 1.08 0.96 0.76 0.98 1980 0.88 1.28 1.14 0.96 0.70 0.98 1981 1.02 1.62 1.50 0.96 0.68 0.98 1982 0.89 1.28 1.20 0.97 0.69 0.94 1988 0.74 1.06 1.00 0.97 0.69 0.94 1984 0.86 1.20 1.09 0.92 0.71 0.91 1985 1.29 1.50 1.50 0.91 0.86 1.00 1988 0.90 1.29 1.29 0.82 0.70 1.00 La Prices adjusted to be net producer prices, I.e. not Including notional costs of seed, etc. Groundnuts/ Groundnuts/ Millet/ Millet Maize Paddy 1970-74 1.25 1.22 0.92 1975-79 1.20 1.12 0.82 1980-86 1.25 1.31 0.98 Sources: Institut S6nAgalais de Recherches Agronomiques; and Ministry of Rural Development, Government of Senegal. 91. We will next attempt to reconstruct the evidence available to the Bank and the (decision-making) process which ultimately converged into its recommendations and conditionality package in the area of agricultural pricing. When analyzing the impact of changing producer prices as well as consumer prices (e.g., rice), it is well known that a general equilibrium framework, allowing for a joint determination of the supply and demand effects is likely to generate significantly different and more accurate estimates than a more limited partial equilibrium approach. In fact, such a "multimarket analysis of agricultural pricing policies in Senegal" was - 44 - developed within the Bank by A. Braverman and J.S. Hamer in 1984.29 In addition to being multimarket, the model was also multiregional including four regions: (1) the Groundnut Basin; (2) the South; (3) the Fleuve region; and (4) Dakar. Two types of policy interventions were simulated and analyzed within this model. First, changes in four individual prices analyzed one at a time ceteris paribus30 and secondly, three integrated policy packages entailing multiple price changes.31 92. A brief description of this model is in order since (a) it provides a good analytical framework to follow through the complex interaction of varying multiple prices; and (b) it was considered in the design of the Bank's price strategy. The output of the price simulation scenarios yielded estimates of their joint effects on the supply of and demand for the major crops, real incomes in the four regions, agricultural export earnings, the government deficit in agriculture and rice imports. While a main purpose of the model was to be an analytical tool and to demonstrate that a range of outcomes is possible depending on elasticities, about which not enough is known, the basic results indicated a trade-off between certain benefits and costs. The benefits were that (i) millet production could be stimulated by iieans of an increase in rice prices and was sensitive to the assumption concerning cross-elasticities; and (ii) the public deficit is substantially reduced mainly due to the gain in tariff revenue from rice imports. On the other hand, the costs were that (i) the net foreign exchange earnings (very sensitive to the elasticity assumption) could be expected to fall because additional millet production comes at the expense of groundnuts; and (ii) real incomes fall substantially in all regions.32 Two factors contribute to the fall in real incomes, i.e. lower groundnut prices and output, and lower returns for millet (this is because demand is inelastic and the higher output generates lower prices). 29/ A. Braverman and J.S. Hammer, "Multimarket Analysis of Agricultural Pricing Policies in Senegal," in Singh, I., L. Squire and J. Strauss (eds.), Agricultural Household Models: Extensions, Applications and Policy, Baltimore: The Johns Hopkins University Press, 1986, pp. 233-254. 30/ I.e.: (1) reductions in the price of groundnuts received by farmers of 15% and 35%; (2) increases in the price of rice of 10% and 502; (3) increases in the price of cotton of 50% and 1182 (world prices in 1981); and (4) increases in the price of fertilizer of 100%, 200% and 3002, respectively. 31/ (1) joint reduction in the price of groundnuts and fertilizer; (2) reduction in groundnut prices while increasing the price of rice; and (3) simulating a devaluation by raising the prices of traded commodities. 32/ These effects hold for a price increase of rice, alone, and are strengthened when combined with the reduction in groundnut prices. -45 - 93. The model generated much controversy within the Bank as reflected by the following quote: "The results of your analysis which indicated the detrimental effects of increasing the imported cost of rice were based on a economic and political framework still required (sic) much discussion....Clearly any public discussion of the effects of increased rice prices on the domestic market would have unsettling political consequences if that discussion highlights that the conclusions of your analysis suggest that there would be a number of detrimental effects on both consumer and producer incomes. (Regional staff] also indicated that there was strong support for increases in domestic rice prices within both the IMF and the region. This support, he suggested, was based on the strong conviction that the budgetary benefits of such a move were of primary importance..33 In another internal document we find the fnllowing quote: "The question also remains of what we plan to do with the Braverman price policy model, whose inconvenient results we have temporarily suppressed pe*nding further examination by Projects of the validity of the model's data and specifications."3 94. The fact that the validity of the results obtained from any model would be questioned is hardly surprising. However, what is totally unclear is why none of the policy scenarios explored through the model included the one which, in fact, the Bank adopted, i.e. a significant increase in the groundnut price combined with an increase in the rice price. An increase in the groundnut price is likely to generate some supply response contributing thereby to export receipts, and to reduce somewhat the growth of millet output and consequently increase the returns to millet producers (because of the inelastic demand) compared to the alternative scenario of a lower groundnut price. It would thus appear that the negative impact on real incomes and expart earnings generated in the model following a policy scenario of a combined higher rice price/lower groundnut price would likely have been reversed with an alternative scenario based on higher rice and groundnut prices combined. The trade-off under this last scenario being the large budgetary cost of supporting higher groundnut prices. 95. In retrospect, the Bank and the Government might have hit on a reasonable (defensible) price scenario without, however, having tested it on an explicit analytical framework and, perhaps, largely under pressure 33/ Internal document (September, 1984). 34/ Internal document (September, 1984). - 46 - from the IMf and the Government.35 However, as argued above, the aggregate output effects of price policies were clearly overestimated. 96. In any case, the recent reduction by the Government of the consumer price of rice from CFA 160 to CFA 130 appears inconsistent with the goal of encouraging the substitution of rice by millet in the diet and thereby increasing the degree of self-sufficiency. 97. Before leaving the issue of price policy, an additional observation should be made. One of the objectives of the Bank and the Government was to improve the nutritional status for the rural households through increased consumption of millet -- partially through substitution away from rice consumption. A higher millet price is likely to encourage the marketed surplus much more than total production. A marginal response of the latter to a higher price combined with a high response for the marketed surplus -- phenomena characteristic of subsistence crops -- could lead to a reduction in the amount of millet available for and consumed out of own production and thereby affect nutrition negatively. (b) Agricultural Research 98. A second refrain which comes through in a retrospective evaluation of Bank documents and files is the recurrent observation about the inadequacy of the ongoing research effort in agriculture and the consequent lack of yield-enhancing technical packages ready to be distributed to and adopted by farmers. Thus, the President's Report of 1980 contains the following assessment: "Some basic deficiencies in the technical packages recommended by agricultural research programs became apparent. Ongoing research has been overly concentrated on a single zone of the Groundnut Basin, is not up-to-date with the current findings of international research, and is focused too narrowly on improving yields of specific crops, without proper consideration for the overall technical, economic and sociological constraints of family farming. There is clearly an urgent need for a program to diversify agricultural research in order to cover all the main ecological regions of Senegal, together with efforts to better identify the many constraints of the Senegalese farming systems."36 35/ Privately, Bank staff on the whole felt that Senegal increased the price of rice for budget reasons only and that the Senegalese believed that the supply of domestic foodgrains would not respond to prices. 36/ President's Report, SAL I, p. 6. - 47 - Surprisingly, only one condition in the three SALs relates directly to research (condition 22 in SAL I, see Table 7). The Bank's projects in the area of agricultural research are evaluated under the agricultural sector section in Chapter III. 99. The Bank undertook a major review of the agricultural research capacity in Senegal.37 Among the conclusions reached by this document are the following: (a) the most important lesson learned in early Bank lending to Senegalese agriculture (1969-79) is probably that inadequate and/or faulty technical recommendations, unless promptly recognized as such and rectified, can be a recipe for failure of investment effort; (b) the Bank was more concerned about insuring the availability and adoption of technical packages than to determine whether the existing technical packages were adequate. In the words of the authors, this amounted to "putting the cart before the horse.038 (An expression we used previously to characterize the design and implementation of a price strategy without an adequate conceptual and empirical understanding of its impact.) 100. Jammeh and Lele bring up two other possible inconsistencies between the ongoing agricultural strategy and, respectively, research funds allocation and scientists' judgments about the technical feasibility of that strategy. They note, first, that the current allocation of research funds appears highly biased in favor of rice and maize and against millet/sorghum and groundnuts (although this can be explained in this last case by the fact that past research efforts were predominantly directed to that export crop). Secondly, the Institut Sdndgalais de Recherches Agronomiques (ISRA) argues that the Cereals Plan is deficient from a scientific point of view for two fundamental reasons: (1) it has no precise definition of what constitutes food self-sufficiency (recall that under this Plan the country should reach 80Z of food self-sufficiency by the year 2000); and (2) the Plan shows little appreciation of the technical constraints on goal realization. 101. The final conclusion of the authors deserves to be quoted since it unveils a key bottleneck between research results and increased output: "ISRA's experience has shown that the production impact of research results transferred to farmers has generally been modest. According to the scientists, such an outcome has frequently been attributed to inadequate extension and producer responses to innovation. The underlying reason, argues ISRA, is the persistent lack of appreciation of the distinction between a research result on the one hand and technology or practical recommendation on the other." (p. 82). 37/ S.C. Jammeh and U. Lele, "Building Agricultural Research Capacity in Senegal," World Bank (MADIA), August 25, 1987, (mimeo). 38/ Ibid., p. 17. - 48 - 102. The French Ministry of Foreign Affairs in a major evaluation of structural adjustment in Senegal concluded likewise that research weaknesses have been underlined by all observers for the last 20 years. Research has not been capable to come up with new varieties adapted to the changing climatic conditions (lower and more irregular rainfall). Research remains confined to experiment stations and does not integrate socioeconomic constraints characteristic of the production systems of farmers. In particular, this evaluation observes that the results are particularly weak in the domain of cereals and vegetables. Although there would appear to be some potential for cereal yield increases through a better adapted research effort, these results cannot be expected in the short run.39 103. The main conclusions which were brought out by the discussion above are that: (1) the analytical and empirical basis upon which the Bank developed its agricultural strategy was inadequate and controversial; and (2) considering that the strategy entailed (a) no increase in real agricultural prices; (b) no clear yield-augmenting technological package available for dissemination to the farmers; and (c) in the light of strong evidence coming from a number of studies reporting the limited supply responsiveness to prices, the Bank's ex ante expectations and projections with regard to the aggregate output response appear highly optimistic. (c) Agricultural Performance and Its Determinants 104. There is overwhelming evidence that a strong correlation exists between agricultural production and the amount and distribution of rainfall in Senegal as the econometric studies reviewed in Chapter III, section A confirm. Annex 4 presents on the same graph annual production and rainfall amounts from 1960 to 1986 for groundnut, millet/sorghum and paddy production, respectively. This correlation is clearly demonstrated in the case of the first two crops but is, of course, less marked for rice, which is partially cultivated on irrigated land. It will be seen in the agricultural sector section in Chapter III that neither price nor fertilizer application seem to have more than a limited effect on output. The weather variables eclipse the impact of all other variables. 105. The interrelationship between groundnut and millet production is a complex one which needs to be analyzed further, to draw policy inferences and design an appropriate strategy. These two crops are grown on the same land by the same farmers. For agronomic (soil fertility) reasons a crop rotation process is desirable over time. Groundnut is the cash crop and millet/sorghum the subsistence crop. The availability of labor-saving equipment (such as the stock of seeders) used in groundnut production releases farm household labor time which can be applied to increase the cropped area devoted to millet/sorghum. At the same time, millet yields in 39/ Rdpublique Frangaise, Ministbre des Relations Extdrieures, Coopdration et D6veloppement, Evaluations: Ddsdquilibres Structurels et Programmes d'Ajustement au Sdndgal, Chapter 5, p. 31. - 49 - the previous year are negatively associated with current millet output. During years of good harvest, subsistence stocks can be built up reducing the pressure to allocate land to food production in the next year and vice- versa. The combined effects of the above factors lead to a pattern of production from year to year which appears somewhat erratic at first glance. 106. A very serious secular development over the last three decades is the sharply downward annual rainfall trend (see Annex 4). Table 11 highlights this trend more dramatically by presenting the annual average cumulative rainfall for five consecutive five-year periods from 1960-65 to 1980-85. It can be seen that the average has fallen monotonically in each successive period from 799 mm/year in 1961-65 to 448 mm/year in 1981-85. Furthermore, Table 11 shows that the number of years, in each five-year period, where annual rainfall dropped below 600 mm/year increased likewise, from zero in the first period, to five in the most recent period. The main implication of this downward trend is that just to maintain a constant level of output, yields have to increase in a compensatory fashion -- a most difficult task in the light of the previously mentioned weaknesses on the research front. Even though it is too early to conclude that this trend is irreversible or, at least, that it cannot be arrested, it is clear that it constitutes another major obstacle to output response in agriculture. Table 11: CUMULATIVE ANMUAL RAINFALL, 1961-87 Number of Years Where Five*-Year Average Rainfall was: In mnyear > 800 M < 600 m 1981-65 799 2 0 1968-70 685 1 1 1971-75 668 0 8 1976-80 548 0 4 1981-85 448 0 5 1988-87 La 489 0 2 /a Two-year average. Sources: OECD, Club du Sahel, D6veloppement des Cultures Pluviales au S6n6gal (Mal 1988) for 1960/61 to 1980/81; and OSenegal: Agricultural Sector Strategy Brief,8 (Volume II) for post-1980/81 years. 107. To make matters worse, the impact of the declining rainfall trend is interrelated with and compounded by a desertification process which was accelerated by overgrazing, lack of a proper crop rotation pattern, -50 - deforestation and erosion. Some keen observers are concerned that the only way to reverse desertification is through a restoration of soil fertility on a large scale by massive application of rock phosphate. We were told repeatedly (particularly by French technical experts) that soil regeneration was a sine qua non to the future of rainfed agricultural production. The issue of restoring national soil fertility should be divorced from that of fertilizer application at the level of the farm. The privatization of fertilizer markets and distribution advocated by the Bank and the Government should not be carried over to soil conservation and fertility regeneration issues. In fact, this is an area where the State is likely to have a key role to play. 108. Improving soil fertility and conservation is analogous in its potential impact to the construction of a large-scale irrigation project. High costs have to be borne at the outset, while the benefits are enjoyed only in the future. There appear to be serious technical and administrative difficulties involved in implementing such a soil regeneration program in Senegal.40 If such a program would contribute a public good crucial to raising agricultural productivity, it is a proper domain for State involvement -- as it is for research. It is very surprising that the Bank appears to have been almost completely absent from this crucial debate. 109. This brings us to the evaluation of the agricultural performance during the structural adjustment period. Table 12 gives some selected indicators covering the period 1981 to 1987. The predominant impact of rainfall on agricultural output is clearly revealed by that table. Good weather in 1982 and 1983 resulted in good harvests. This was followed by two very bad years, 1984-85. (In particular, rainfall in 1984 reached its lowest point over the period for which we have data, i.e. beginning in 1960.) 401 Soils appear to be too thin, at the present time, and not to contain enough organic matter to absorb rock phosphate which, without it, would blow away or get washed out. As a possible remedy, mixed farming could be introduced to contribute organic matter to the soil so that rock phosphate would have something to hang on to. - 51 - Table 12: SELECTED INDICATORS OF AGRICULTURAL PERFORMANCE, 1981-87 1981 1982 1983 1984 1985 1988 1987 Agricultural CDP (billions of constant 1979 CFAF) 60.6 76.2 80.1 46.4 61.2 60.8 n.e. Primary Sector GDP (billions of constant 1979 CFAF) 107.6 134.3 140.9 114.3 123.4 135.6 n.a. CDP (billions of constant 1979 CFAF) 558.9 643.3 680.9 880.9 654.8 685.1 n.s. Rainfall (m/year) 438 564 492 818 438 605 478 Groundnut (million tons) 487 888 1,080 575 490 601 841 Millet/Sorghum (million tons) 541 950 555 852 471 950 6834 Paddy (million tons) 86 k22 120 102 138 148 148 Sources: Senegal: An Economy Under Adjustment; and OSenegal: Agricultural Sector Strategy Brief.0 110. In turn, weather conditions improved in 1986-87 and so did agricultural output. Pronounced fluctuations around two-year cycles triggered by weather are noticeable for agricultural GDP, groundnut production and millet/sorghum production. Incidentally, the previously noted contracyclical adjustment to stocks in millet production is visible in Table 12 (i.e. major downward production adjustment in 1983 -- a relatively good year weather-wise -- following the bumper crop of 1982 and the adjustment in the opposite direction in 1986, following the disastrous crops of 1984-85). 111. Since the new agricultural policy (see paras. 220 and 221, Chapter III) only started to be implemented in 1986, it is much too early to judge what impact -- if any -- the new measures (particularly the price policy) may have had on performance. However, what the data in Table 12 suggest is that the rainfall pattern continues to be the predominant determinant of output, as it has been historically. To return, for a moment, to the policy framework we developed in section C, this means that weather and the deteriorating natural environment in Senegal constitute increasingly binding constraints to improved performance. The effectiveness of policy measures (price policy, fertilizer distribution, etc.) and reforms, under these circumstances, is likely to be extremely limited and circumscribed. The prime objective should be to alleviate these constraints by every means possible (e.g., soil regeneration and conservation, development of more drought-resistant varieties, and possibly through the development of small- scale irrigation projects). It would appear that it is only through changing the natural environment that policy measures can become truly effective. I - 52 - (3) Improving Private Sector Efficiency 112. This second broader domain of structural adjustment, after agriculture, can be broken down into the four areas and sub-areas identified in Table 13. The table indicates the specific conditions which relate to each one of these Lreas and sub-areas -- corresponding to the numbering scheme used in Annex 1 which should be consulted for more detail. Table 18: CLASSIFICATION OF SAL CONDITIONS RELATING TO IMPROVING PRIVATE SECTOR EFFICIENCY Conditions /a SAL I SAL II SAL III 1. Trade Liberalization a. Reduction in Import Restrictions - 14-16 54-57 2. Reduction in Market Imperfections a. Phasing Out Controlled Industrial Prices - 20 69-61 b. Encouraging Job Creation and Mobility - 21-28 62, 68 c. Renounce Practice of Entering into Special Conventions with Enterprises - 17 6 8. Restructuring Industrial Sector a. Formulation of New Industri&. Policy - 12,18 - b. Promotion of Industrial Exports 14-16 18, 19 64-70 4. Restructuring Financial Sector 9, 10 - 71-77 La The numbering of the conditions corresponds to that given In Annex 1 which should be consulted for more detail. Source: Prepared by OED. 113. The industrial sector in Senegal has been traditionally heavily protected through a combination of tariffs, quotas and special government conventions, bestowing virtually monopoly powers to a number of enterprises. The industrial strategy which was followed since Independence was one of extreme import substitution which led to a highly inefficient sector. Following the crisis of 1978-81, it became clear to all concerned that without radical structural changes, and particularly a greater export orientation, future industrial growth in Senegal was doomed. This led to the design of the structural adjustment package outlined in the preceding paragraph. Whereas such a major restructuring of the industrial and financial sectors is a necessary condition to achieve a higher growth rate than the (marginal) historical rate of 3.9Z annually between 1960-61 and 1985-86, it is not at all clear that this package will be sufficient -- given the extreme constraints presently faced by these sectors. - 53 - 114. Some of the main characteristics of the industrial sector are described in the section on the industrial sector in Chapter III. Most industrial branches are highly concentrated (with two to three dominant firms); the biggest private firms continue to be under foreign ownership; enterprises typically use obsolete equipment and technology (over half the plants were set up prior to 1970); the labor market is highly imperfect resulting in relatively high labor costs and low labor productivity; and capacity utilization of existing equipment is low, averaging 60Z industry- wide (and only 30Z in groundnut milling). By 1985 the Bank had become convinced that "past trends were unsustainable because they generated a structural and widening gap on the balance of payments, implying an insoluble financing problem....and in short, there is no "trend scenario" for Senegal's economy because past trends, unadjusted, do not constitute a feasible scenario at all..41 The process of implementation of the New Industrial Policy (a phased comprehensive program of reforms aimed at improving the competitiveness of industrial enterprises) and related SAL conditions, which started in late 1986, is presently underway. The transition from the present system and regime to the new post-structural adjustment ones may take at least three years. The short-run transitional effects are bound to be disruptive and cause short-term losses which, it is hoped by the Government and the Bank, would be more than compensated by the medium- and long-term benefits. 115. Since an ex-post evaluation of the impact of SALs II and III on private industrial and financial performance is not possible at this time, we examine critically the ex-ante models and approaches (including any theoretical and empirical evidence) linking policy measures and reforms to industrial performance used by the Bank in the formulation of its conditionality package.42 In particular, we try to ascertain whether the expected targets under the so-called "managed adjustment" scenario of the Bank were reasonable in the light of the extreme constraints faced by the secondary sector. 116. Table 14 shows the recent industrial performance based on selected indicators. It can be seen that industrial GDP has remained practically stagnant between 1982 and 1986, while the index of industrial activity actually dropped from 103 to 96 over the same period.43 The sectors which were particularly hard hit in the immediate past -- as revealed by the table -- were "oil processing" and "textile, clothing and leather." No significant change is likely to have occurred in 1987 -- the index of industrial activity for the first half of the year having risen by only a marginal 1.5Z. 41/ President's Report, SAL II, p. 25. 42/ In other words, we attempt to answer some of the questions raised in the Introduction to this section. 43/ It is not clear why these two series move in opposite directions. Some French observers believe that this reflects a basic inconsistency in the data. - 54 - Table 14: SELECTED INDUSTRIAL INDICES OF PRODUCTION, 1979-86 (1976 = 100) Weight 1979 1980 1981 1982 19088 1984 1985 1988 Food, Beverages, A Tobacco 43.1 118.8 85.8 89.2 109.4 109.7 100.5 108.2 88.4 Oil Processing 12.2 n.e. n.e. n.e. 89.2 76.7 48.0 85.8 28.8 Other Foods, *e. 30.9 n.e. n.a. n.e. 125.2 122.7 122.9 180.0 112.7 Textile, Clothing, Leather 12.8 95.0 94.8 114.4 147.8 189.9 120.1 188.9 78.7 Wood Industries 0.5 141.5 188.8 188.4 144.2 182.8 125.4 119.0 100.8 Paper, Cardboard 1.8 118.0 118.6 183.1 181.7 187.5 189.4 120.1 176.5 Chemical Industries 11.4 105.1 91.2 104.8 79.4 78.1 72.1 77.9 75.8 Construction Materials 3.8 95.8 98.4 92.8 92.2 128.6 118.2 94.1 91.7 Machinery A Equipment 4.0 98.6 70.6 55.4 77.4 98.6 97.8 78.1 78.9 Mining 18.5 101.7 89.8 107.4 87.8 84.3 107.9 114.5 116.7 Electricity, Water 6.1 181.0 129.7 134.6 141.8 148.7 158.2 165.0 166.8 Total Industry 100.0 n.s. n.a. n.e. 108.8 107.1 105.0 107.6 96.8 Total Without Oil Processing 87.8 n.a. n.s. n.e. 108.0 111.4 112.9 117.6 106.0 Industrial GDP /a - 141.0 188.9 144.6 186.3 170.1 187.0 170.4 177.2 Manufacturing - n.s. n.s. (89.6) (98.8) (97.8) (98.8) (100.0) (100.1) Other Industry - n.e. n.a. (55.0) (67.6) (72.8) (68.2) (70.4) (77.1) /a At constant 1979 CFAF prices. Sources: World Bank, Staff Appraisal Report, Senegal: Industrial Sector Restructuring Project, Report No. 6947-SE, November 25, 1987, p. 38; and World Bank, Senegal: An Economy Under Adjustment, Report No. 6454-SE, February 18, 1987, SA Table 5. 117. We can now turn to an examination of the expected effects of the New Industrial Policy and SAL conditions in the short run -- concentrating particularly on the causal mechanisms through which the latter were thought by the Bank to affect performance. The combination of trade liberalization (elimination of quantitative restrictions and tariff reductions) and the phasing out of special conventions between the Government and enterprises is the typical neo-classical recipe of removing price distortions (getting the prices right) and reducing market imperfections. Under normal circumstances, the IMF/Bank would have recommended this last package in addition to a devaluation -- an impossibility in the present context. * 55 - Trade liberalization tends to hurt domestic enterprises in the short run by reducing the wedge between home prices and world prices and forcing them to meet foreign competition; in contrast, a devaluation increases the prices of tradables versus non-tradables and acts as an incentive to enterprises producing exports or goods potentially exportable. Thus, a devaluation alleviates the negative impact of liberalization per se on short-term output and employment. The average nominal level of duties was to be reduced from 47.3Z (the level existing prior to the SAL II reforms) down to 36.2% after the 1986 reform (these figures refer to the weighted average; the corresponding figures for the unweighted average would be 46Z and 35.72, respectively). 118. The presumed logic for maintaining a level of protection of this magnitude (36-37%) is that this corresponds approximately to the level of overvaluation of the CFA franc so that, had a devaluation been feasible, tariffs could have been eliminated altogether and a uevaluation of 36-37% undertaken to yield a similar result. 119. In fact, the net effect of trade liberalization measures was to reduce the prices of competitive imports by about 7.5Z nominally on average (i.e. from an index number of 147.3 to 136.2) but effectively by much more through the elimination of quotas and special conventions. In contrast, a devaluation would have increased the prices of imports. Thus, the enterprises producing tradables (import substitutes and exportables) were faced with the equivalent of a currency appreciation to meet foreign competition. Effective prices charged by these firms had to be reduced accordingly affecting their profitability negatively. 120. In addition, two other factors are expected to affect the industrial sector negatively in the short run: first, the very tight financial policies, limiting the growth in domestic liquidity to three percent during 1987-88 (well below the estimated 92 growth in nominal GDP), constraining the supply of credit to domestic firms; and, secondly, the deflationary trend affecting domestic demand, i.e. a domestic market constraint. 121. A study carried out by the Boston Consulting Group (BCG)44 concluded that, in the short run, a number of enterprises would be forced to close down, and the labor force in existing activities would be reduced by about 4,300 workers (out of a total formal manufacturing sector employment of 27,000). The study also estimated that the full restructuring of the Senegalese manufacturing sector would, over time, require overall investments of about CFAF 13 billion (about $45 million). The agro-industrial and textile activities would account for over 90% of financial requirements and suffer an estimated 73% of expected job losses 44/ R4publique du Sdndgal, Minist&re du Ddveloppement Industriel et de l'Artisanat, Impact de la rdforme de la politique industrielle. Study prepared by the Boston Consulting Group, January, 1987. - 56 - from the manufacturing sector. Clearly the above estimates suggest that the short-term transitional costs of adjustment on the manufacturing sector are very high. 122. The key question is whether these short-term effects are the necessary transitional and frictional costs which have to be borne in order to allow the industrial sector to become competitive and enter a new regime of efficient and export-oriented growth. To answer this question, one has to explore critically the medium- and long-term constraints impinging on this sector as well as its future prospects. At the outset, it should be noted that the Bank was reasonably optimistic -- projecting a growth rate of the secondary sector of between 3.5Z and 4.OZ between 1987 and 1990, and between 4.OZ and 4.5Z during 1990-95 -- under its managed adjustment scenario. In light of the very limiting constraints faced by industry, it is not clear how a convincing case can be made for the above expected performance -- a point we return to shortly. 123. The main constraints to industrial development in Senegal are well known. First, historically, macroeconomic performance has been highly correlated with agricultural performance, as described in Chapter I. Thus, given minimal prospects for agricultural supply response over the foreseeable future, one should not expect more than marginal growth in internal effective demand for consumer goods and other industrial products. The foreign demand for Senegalese industrial exports appears, likewise, cloudy. Except for processed fish products and phosphate, the bulk of Senegalese industrial exports are directed towards the CEA045 markets where they are highly protected. In general, exports are not competitive in other markets despite preferential tariff arrangements in the EEC. The CEAO markets are presently undergoing a severe recession. Under the circumstances, the prospects for industrial exports except perhaps for fish exports, also appear unfavorable. To make matters worse, fish exports have been stagnating recently. 124. A second important constraint to industrial development is the imperfect state of the labor market. A recently completed study of the industrial labor market and economic performance in Senegal highlighted a number of serious problems.46 Some of the relevant findings of this study in the present context are: (1) philosophically the Government is in a loose alliance with the "cooperating" trade unions and the behavior of Government officials can perhaps best be described as one of creating an environment which allows for the protection of workers' 45/ Communaut6 6conomique des Etats de l'Afrique de 1'Ouest (Economic Community of West African States). 46j See Jan Svejnar and Katherine Terrell, "Industrial Labor, Enterprise Ownership and Government Policies in Senegal," Revised Draft, December 23, 1988 (prepared for the World Bank). - 57 - interests without endangering the existence of firms. This was achieved through a highly interventionist government approach relying on various labor market regulations. (M) Government officials are still apprehensive about undertaking major reforms of the industrial relations system. They feel that these reforms might disrupt the fragile balance in the adversarial union-management relationship. (3) lack of worker effort, motivation and discipline is cited by managers as a major reason for low labor productivity in Senegalese industry -- a phenomenon which appears to be brought about at least partially by the lack of a direct link between performance and remuneration in most enterprises. (4) another reason for the low level of labor productivity appears to be the long-term limited supply of skilled workers and a secular decrease in the provision of on-the-job training of workers by firms. (5) the current regulations seem to result in underutilization of labor so that a relaxation of the regulations concerning (paid) hours of work and/or layoff, as well as those of temporary (daily) workers would be Pareto improving for all the parties. (6) managers feel particularly constrained by their inability to lay off workers without passing through an elaborate and time-consuming process of government approval. The difficulties associated with extending the contracts of temporary (daily) workers are also cited as significant constraints on enterprise behavior. Unfortunately the implementation of the proposed labor reforms under the New Industrial Policy is proceeding more slowly than anticipated and is running into some opposition from the National Assembly. The syndicalist tradition is engrained in Senegal, and it appears very difficult to remove or relax workers' privileges conflicting with efficiency. 125. A third set of obstacles to industrial restructuring and development relates to the very inelastic supply of entrepreneurs and capital within Senegal. Historically, Senegal has fostered traders rather than businessmen-entrepreneurs. French observers, who know the local scene well, have consistently maintained that this is a handicap in the present context. Likewise, the lack of any capital market combined with credit restrictions arO the weakness of the banking system (to be discussed next) and pessimistic Lature business expectations all converge to limiting greatly the availability of growth capital. - 58 - 126. The extreme fragility of the banking system constitutes a final key obstacle to industrial progress. The financial sector comprises 14 banks and 5 financial institutions with 5 banks accounting for approximately 85% of total banking assets. The conditions within the financial sector have seriously worsened in the last 15 years, or so, for a host of reasons among which an excessive issuance of largely unproductive credit; frequent State interference in banks' lending decisions to support ill-conceived agricultural schemes; poor management and lack of internal control in the banks; and low profitability caused by inadequate lending spreads. The distribution of private sector credit shows that commerce is the biggest borrower. In 1985 commerce absorbed 44% of private sector credit, industry 23%, while tourism and agriculture each took 2.52. The share of industry in private sector credit reached a peak of 352 in 1975, fell to 14% in 1979 and has since staged a partial recovery. Only 17Z of all private sector credit in 1985 was for long-term (more than 10 years) loans. In fact, banks have limited experience dealing with long-term development loans as opposed to short-term trade credit. 127. The SAL III measures relating to the financial sector which are presently being implemented include the reorganization of two major banks -- the Banque Nationale du Ddveloppement S4n6galaise (BNDS) and the Union Sdn4galaise de Banque (USB) -- strengthening supervision over banking activities and increasing the minimum capital required to establish new banks (conditions 71-77 in SAL III, see Annex 1). In addition, the Bank is financing an Industrial Sector Restructuring Project in support of the above package of actions which would provide some investment funds and technical assistance, and assist displaced workers. There appears to be much skepticism within Senegal -- particularly among knowledgeable officials of the Finance Ministry that these measures, by themselves, will be sufficient to "sanitize" the financial sector. And, if not, whether the absence of a viable financial and banking sector might not present an insurmountable bottleneck to any sustainable take-off by industry. 128. The Bank is perfectly conscious of all the above-mentioned constraints and obstacles to industrial and financial development. In fact, one finds these arguments cogently and critically articulated in the Bank's internal memoranda. Thus, to cite a few examples: (i) the Policy Framework Paper for 1987-90 (September 28, 1987) warns against the possibility of weaker than expected investment and production response to the new policy environment; (ii) in an internal document (December, 1987), deep concern is expressed about the lack of progress in the implementation of the Labor Code, i.e. "The National Assembly diluted substantially the changes concerning the temporary employment contracts. The Government has been very concerned with this setback and is attempting to facilitate labor mobility under existing laws. The high cost of labor remains a major issue - 59 - in Senegal.";47 and (iii) in an internal document (January, 1988), we find the following quote "There were indications that considerable private capital outflows had taken place in the first half of 1987-88...combined with the reluctance of local commercial banks to repatriate their export proceeds, the outflow has further aggravated the already very tight liquidity in the economy... .The banking sector is facing serious and worsening liquidity problems stemming from sizeable non-performing loans and management deficiencies which have led to a severe loss of confidence in the majority of domestic commercial banks." 129. One is at a loss to reconcile the Bank's own critical evaluation of the constraints plaguing the industrial sector with its relatively favorable forecasts of performance in the foreseeable future (i.e. to 1995). The only possible analytical and empirical basis for the latter assessment is the BCG study which after having estimated the transitional costs of industrial adjustment -- discussed above -- suggested that the two most vulnerable subsectors in the short run, agro-industry and textiles, could, if they ever became competitive, provide the bulk of the industrial export potential upon which the Bank's projections are predicated. It appears extremely unlikely that these two subsectors can go through the rationalization and restructuring process required to make them competitive. Not only are the estimates of export potential arrived at by the BCG study duly qualified (i.e. conditional upon achievement of competitiveness) but, in addition, the study explicitly indicates that it was not in its terms of reference to judge the long-tem impact of the New Industrial Policy. This brings us back to the black box -- described in section C -- linking policy measures and exogenous constraints to performance. The Bank would appear to have more confidence in the favorable impact of the policy reforms than seems warranted by the evidence available and to underestimate the extreme difficulty of relaxing some of the currently binding constraints. (4) Public Investment and Finance 130. This third domain of structural adjustment can be broken down into two areas and sub-areas as shown in Table 15. The table indicates the specific conditions which relate to each one of these areas and sub-areas -- corresponding to the numbering scheme used in Annex 1 which should be consulted for more detail. Developments in each of these are evaluated critically next. 47/ This refers to conditions 62 and 63 in SAL III (see Annex 1), which could be considered as examples of incomplete conditions, in the sense that the Government cannot vouch for what the National Assembly will or will not do. However, because the Government controls the large majority of seats in the National Assembly, the very validity of these conditions comes into question. - 60 - Table 15: CLASSIFICATION OF SAL CONDITIONS RELATING TO PBLIC INVESTMENT AND FINANCE; AND TO PUBLIC SECTOR STRENGTHENING AND DISENGAGEMENT Conditions La SAL I SAL II SAL III Public Investment and Finance 1. Improvement in the Quality of Public Investment a. Project Analysis, Identification and Management 8, 4 28 15-17 b. Investment Programing 1, 2 24-27, 18, 14, 45 29, 18-20 2. Public Finance a. Foreign Debt Management 7, 8 48 33 b. Budgetary Process 5, 6, 84-44 21-28, 28-82 30-32, 84-36 Public Sector Strengthening and Disengagement 1. Strengthen Parapublic Sector 19, 20 80-38 1, 4-12 2. Public Sector Disengagement 18 29 2, 8 Le The numbering of the conditions corresponds to that given in Annex 1 which should be consulted for more detail. Source: Prepared by OED. .31. Thus, SAL I called for "the establishment of project identification, preparation and evaluation unit and the introduction of a minimum rate of return standard" (conditions 3 and 4). SAL II required -tricter technical, economic and financial project analysis criteria (condition 28) and SAL III established as conditions the completion of a project appraisal guide and the transfer of responsibilities for project identification and preparation to the various technical ministries (conditions 15-17). If the project analysis methodology is inadequate and the selection of investment criteria is too loose, this will open the door for unproductive and marginal projects either not contributing to, or actually deterring from, economic growth. 132. The methodology presently used in Senegal is based on the French (Caisse Centrale) "m4thode des effets." This method uses market prices and attempts to capture the indirect effects of a project through an - 61 - intersectoral input-output framework. This method contrasts sharply with the more conventional approach followed by the Bank and most Governments, which relies on shadow prices, and assumes that if the latter reflect appropriately the marginal value product of the factors and the border prices of tradables, they also incorporate indirect effects. The "mothode des effets" is controversial on empirical and theoretical grounds. Empirically, given the notoriously weak database in Senegal, how much confidence can one have in the accuracy of the underlying input-output table? Key officials in the Ministry of Finance feel that the "m4thode des effets" is theoretically flawed, while French experts in Senegal defend it, claiming that it represents an altogether different yet valid approach to project analysis. 133. Given the fundamental role of investment as a determinant of the rate and quality of socioeconomic growth, it is most surprising to note how slowly the Bank and the Government have proceeded in comparing critically the merits and implications of the two project analysis methods. In January 1988 (hence eight years after the above-mentioned SAL I conditions) such an evaluation had not yet been completed as the following quote indicates: "I also organized a one-day workshop led by (the consultant], to help finalizing a draft project appraisal guide prepared by the Government of Senegal.. .The final draft of the guide is expected to be available shortly. As a follow-up of his assignment, [the consultant] will prepare a short note, expected in early February, in which he will provide a synoptic comparison of the effect methods (m6thode des effets) of project analysis and the shadow pricing approach adopted by the Bank.w48 134. Local observers (both within the Government and French experts) expressed concern at what they perceived to be a low and declining quality of many feasibility studies.49 135. Relatively more progress has been achieved in improving the process of investment programming. A rolling three-year PIP has been prepared by the Ministry of Planning (MOP) -- consisting of 384 projects, amounting to 379 billion CFAF and representing approximately 8Z of GDP. The share of directly productive activities within the PIP is said to have reached 56% -- a significant increase over the historical share. To the extent that the estimated productivity of investment depends on the project analysis methodology which is employed, an important question left unanswered is whether the composition of the PIP would be significantly different if the conventional Bank (shadow pricing) analysis were applied. 48/ Internal document (January, 1988). 49/ Thus, we were told that the economic analysis in the Bank appraisal of the Irrigation IV Project was significantly weaker than that of previous projects such as Dagana and Lampsar. - 62 - 136. The preparation of the PIP has been institutionalized within the MOP. The main task of MOP consists of gathering and computerizing information on individual projects including future recurrent charges. In principle, no project is to be chosen unless the current charges are explicitly budgeted up front. It is widely asserted that the MOP simply does not have the technical competence to do more than act as a central depository for projects and catalog and classify them. The MOP, as such, does not undertake any critical evaluation of individual projects; its main function appears to be to package and keep track of projects. 137. A serious obstacle to a meaningful dialogue between the Bank and Government on investment matters is the much to be desired relationship which exists between the Ministry of FIance (MOF) and the MOP. The general perception is that the MOP has be: given much too much authority in the preparation of the PIP -- a task which, given their very limited technical competence, it can only carry out partially; and that MOP is totally incapable of being the main interlocutor of the Bank on investment questions in the structural adjustment game. Many observers feel that the main way of improving the quality of projects and investment planning and programming, in general, is to strengthen the presently weak technical capacity of the specialized ministries such as the Ministry of Rural Development, Public Works and Industry, where most projects originate; and to give a much greater role to MOF in this general domain. As the situation presently stands, one cannot really determine the quality of the great bulk of the projects incorporated into the PIP. Hence, the condition regarding approval of a rolling three-year investment program appears to have very little bite to it. At best, the number of white elephants and unproductive (donors' babies) projects ("projets pirates") is being reduced. One often heard observation is that the Bank should move towards, or at least test, smaller projects. 138. We turn next to the area of public finance. IMF-SAL conditionality was instrumental in helping to reduce the current account deficit from 26Z in 1981 to about 1O% in 1986 and the budget deficit from about 8.2Z of GDP in 1982-83 to 2.3Z in 1985-86. However, a general observation at the outset is the continuing tendency on the part of the Bank to set and often periodically reset budgetary targets significantly in excess of actual performance. Thus, a prime example is provided by the investment (i.e. gross domestic investment) to GDP ratio. In SAL I, at the time this ratio stood around 16Z, the target for the mid-1980s was to raise it to 18Z.50 In fact, this ratio fell to about 13.8% in both 1985 and 1986. In SAL II the target was significantly scaled down to a more realistic 14% through 1992 to rise gradually thereafter.51 (The negative implications of this scaling down on growth were discussed in Chapter I.) Overly ambitious budgetary targets engender unrealistic and infeasible growth objectives. 50/ President's Report, SAL I, p. 20. 51/ President's Report, SAL II, Ainex IV, p. 4. - 63 - 139. The other main issues with conditionality relating to the budgetary process are as follows. First, there appears to be an inconsistency between the Bank's SAL conditions aimed at centralizing the management of, and control over, the foreign debt (within the MOF) and some Bank projects which grant (or attempt to grant) decentralized budgetary authority to host enterprises. Thus, funds meant for Socidt6 d'Am6nagement et d'Exploitation des Terres du Delta (SAED), S6cietd Nationale de 1'Electricit6 (SENELEC), Soci6td Nationale des TOl4communications (SONATEL) and ISRA should be channeled through the MOF rather than going directly to these enterprises -- with consequent loss of central control over disbursement. Another example of a conflict between the Bank's SAL program and projects consists of the special considerations and exceptions requested by and often granted to public enterprises receiving project help (for example, Senelec benefits from a special subsidy on the price of oil). Programs and Projects staff within the Bank sometimes work at cross- purposes. 140. Secondly, an almost universal reaction among Senegalese officials and knowledgeable expatriate experts (excluding Bank staff) is that (a) the number of SAL conditions is much too large; and (b) the content of many of them is not really meaningful and not clearly connected with development objectives. We were told, and could observe, that keeping track of the degree of compliance with the extensive number of conditions in SAL III was extremely time consuming on MOF and MOP and staff from other ministries (weekly meetings are scheduled just to ascertain progress on the implementation of these conditions). One senior MOF official expressed this frustration colorfully by saying that "keeping track of SAL conditions was like driving an automobile with fifty turn signals." Furthermore, many of the conditions are perceived as unconnected, or only marginally connected, with the achievement of development objectives and performance (e.g., publication of a decree, preparation of national income account statistics).52 141. Finally, it appears that the IMF continues to exert pressure on the Bank to release SAL tranches when they come due. The IMF Resident Representative indicated to us that a major problem for him and the Fund was that tranche release (and therefore determination of compliance with SAL conditions) was itself a condition (a performance criterion) for standby credit disbursement by the Fund scheduled on a quarterly basis. Even as short a delay as two weeks in the Bank's timing of the release created havoc with the IMF program. If the Bank is truly serious about conditions being met, it should stand up to these pressures. At the very least, it should be possible to modify somewhat the IMF calendar to provide it and the Bank with additional flexibility. 52/ A word which was repeatedly used to describe the conditionality package was "pointillisme, which connotes marginalism, superficiality, and cosmetic formalism. - 64 - (5) Public Sector Strengthening and Disengagement, and Private Sector Engagement 142. Of all the structural adjustment issues, this is the most difficult one to achieve and a sine qua non to the future growth and development of Senegal. Over the two decades preceding 1982, the public sector has increasingly encroached on private sector activities. This encroachment process can be verified from Table 16 which shows that the number of wholly or majority government-owned enterprises increased from 21 in 1962 to 86 in 1982 -- extending across all sectors. By 1982, these enterprises represented an estimated 20X of modern sector GDP and employed an estimated 35,000 workers (about 301 of modern wage employment). The performance of public enterprises, which had been relatively good during the mid-1970s -- fueled by the phosphate and groundnut boom -- deteriorated drastically after 1977 as Table 17 reveals. By 1981-82 the parastatals had become a critical drain on public finances (government operating subsidies to these entities represented 40Z of the public sector deficit in that year). To make matters worse, public enterprises continued to absorb an increasing share of national investment, from 30Z of gross fixed capital formation in 1977 to 40Z in 1981 -- with approximately three-fourths of total investment into this sector financed with external aid. The essence of the problem was that much of the increasingly scarce external aid was being squandered on trying to keep afloat bankrupt companies instead of allocating those funds to directly productive activities. The Government was burdening itself with a fast rising foreign debt largely imbedded into unproductive activities and projects incapable of generating the future net output flow required to service that debt. Table 16: EVOLUTION OF PARAPUBLIC SECTOR, 1962-82 (number of enterprises L, by sector) 1962 1972 1977 1982 Agriculture 4 8 9 10 Industry 4 9 17 17 Comerce/Services /b 10 26 43 42 Financial Institutions 2 6 6 7 Administrative 1 2 8 10 TOTAL 21 50 88 86 L Government majority-owned enterprises only. L Includes transport. Source: Rapport G4n6ral sur Is Gestion des Entreprises Publiques, (CVCCEP: August 1983). - 65 - Tsble 17: PARAPUBLIC SECTOR, AGGREGATE NET INCOME AND GOVERNMENT SUBSIDIES, 1977/78-1981/8k (in current CFAF billions) 197717 1978/79 1979ee0 19o0o81 1981/82 Surplus Entitles +16.91 +18.51 +14.00 +5.49 +11.04 Deficit Entities -10.76 -10.94 -17.05 -25.23 -28.18 Net Position of Sector /a +6.15 +7.57 -3.05 -19.74 -17.14 Government Operating Subsidies 6.99 7.65 8.48 15.89 19.71 Memo Item: Number of Entities with Operating Deficits 85 31 85 40 37 La The CPSP and CSS, both transfer agencies, are excluded from all data. Source: World Bank, Senegal: Country Economic Memorandum, Report No. 5248-SE, November 5, 1984, SA Table 8.1. 143. The key questions in the present context are "how was this situation allowed to evolve historically?" and "how did the position of the Bank vis-a-vis parastatals evolve over time?" In particular, since public sector disengagement-cum-private sector engagement is the principal pillar of the present SAL strategy, it is important to evaluate how the Bank's views and philosophy regarding private versus public sector emphasis have varied over time to reach the almost doctrinaire present position. Does the World Bank shoulder some of the blame for tolerating -- if not supporting -- many public enterprises in the past? Alternatively, did the Bank provide sound advice on this issue without having the necessary leverage to effectuate a change until recently? 144. It is clear that the Bank has not followed a consistent approach on this issue. Until the mid-1970s, it either passively went along with the dominant role of the public sector when it did not actively support it. The history of the Bank/Government relationship regarding the role of parastatal agencies in the agricultural sector is analyzed in detail in the agricultural section in Chapter III and need not be rehearsed here. Until it moved into program lending, the Bank could exert influence mainly at the sectoral, as opposed to the macroeconomic, level through projects and the quality of its economic analysis. 145. Given the predominance of agricultural projects in the Bank's portfolio prior to 1980, this is the sector in which the Bank could have attempted -- at least potentially -- to use its leverage to help slow down or even reverse the inefficient State control over the marketing, distribu- tion and processing system for agricultural products and inputs. Instead, the Bank went along with the prevailing trend until at least 1977. - 66 - 146. The same obsezvation applies as well to other sectors. In industry the Bank used the Socidtd Financibre Sdndgalaise pour le Ddveloppement de l'Industrie and du Tourisme (SOFISEDIT) as a conduit for industrial credit through several projects and the tourism projects were with State enterprises. In addition, the Bank's loans to roads, railroads, ports aad telecommunications, as expected, went to State enterprises, as well. 147. By 1977 the Bank was sufficiently concerned with the implications of the growth of the parastatal sector to undertake a thorough study of it.53 Among the findings of this study were the following: (1) between 1971 and 1975 about half of direct government equity investments in the parapublic sector went to purchase existing foreign interests rather than to promote new projects (i.e. the Senegalization process); (2) the rapid growth of the parapublic sector has led on occasion to an overextension and suboptimal use of scarce public resources and limited credit facilities; and (3) major improvements were needed in screening parapublic investments, assessing its performance and establishing greater government control to supezvise the sector. At the same time, the report tones down its critique of the public sector as the following quotes suggest: "Government participation in the ownership, management and control of enterprises is only one of a number of ways to promote development, but the para-public sector, if well-run, can be used inter alia to: control key sectors of the economy and services; encourage investment where private initiative is lacking; acquire new technology and managerial expertise; attract foreign financing....While the transfer of funds from Government to the para-public is substantial, it is not abnormal, within the limits of public finance resources, for a growing nation to invest in productive 5rojects through public enterprises and mixed companies." These remarks connote a rather permissive and tolerant approach of the Bank vis-a-vis public enterprises even as late as 1977. 148. It is only with the first SAL that the Bank started to push seriously for public sector disengagement (see conditions 18-20). With the breakdown of SAL I and the consequent loss of leverage, very little was achieved on this front in the first five years of the PREF.55 The 1984 CEM makes this point cogently as follows: 53/ World Bank, Senegal: The Para-Public Sector, Report No. 1619a-SE, June 1, 1977. 54/ Ibid., p. i. 55/ Plan & Moyen Terme de Redressement Economique et Financier (Medium-Term Recovery Plan). - 67 - "In the five years since the government's explicit commitment to parapublic sector reform almost nothing has been achieved. Reformv have been pursued unsystematically and without clear political direction and support from the highest levels of Government. First, the basic step which should be a prelude to any serious program of reform -- a systematic review of the scope of the parapublic sector and clear political decisions about the role of the state in the productive sector -- has never been taken in Senegal."56 In particular, the disillusion with ONCAD and a handful of other unviable enterprises and their ultimate elimination were considered ad hoc responses to untenable situations. The 1984 CEM contains an excellent analysis of the problems plaguing the parapublic sector (Chapter VI). It spells out six areas for action thereby setting the stage for the conditions imposed in this domain in SALs II and III. 149. Progress during SAL II was relatively slow: program-contracts were signed with six public enterprises (PEs) and rehabilitation programs were adopted by two PEs; direct subsidies to PEs in the 1986-87 and 1987-88 budgets were set 52 and 152, respectively, below their 1985-86 level; and specific proposals for the sale of shares in the ten PEs with the best sale prospects were prepared. The SAL III program was a continuation of actions undertaken under SAL II and consists mainly of: (1) strengthening the management of the reform process through the establishment of an appro- priate administrative and coordinating mechanism; (2) a divestiture program offering for sale shares in 10 mixed enterprises while identifying 10 to 17 additional PEs to be partly or wholly divested and the completion of liqui- dation procedures for 7 mixed enterprises, public agencies (6tablissements publics); (3) improving the performance of PEs through program-contracts; and (4) continuing to reduce direct subsidies to PEs and submitting a new legal framework for the parapublic sector to the National Assembly. 150. Progress is occurring on all the above fronts. In particular, the Government is proceeding with the divestiture and public sector disengage- ment process. The first group of ten enterprises is up for sale. Table 18 provides key indicators on these PEs. It can be seen that they represent only a relatively modest 8-9Z of total PE sector capital and value added, respectively. An issue of concern highlighted by the table is that these ten presumably most profitable enterprises within the parastatal sector are presently operating at a loss (net profits are negative). Under these circumstances, how attractive can these ten PEs and other even less profit- able ones to follow be to the private sector? Putting up enterprises for sale is only half the story. If there are no private takers, the latter action does not alter the status quo. To succeed, public sector disengage- ment has to be complemented by private sector engagement -- otherwise, it remains an incomplete -- although necessary -- condition. Even if the disengagement condition is fulfilled by the Government, it remains an incomplete action which by itself cannot change the underlying situation. 56/ Senegal: Country Economic Memorandum, p. 56. - 68 - Table 18: KEY INDICATORS OF PUBLIC ENTERPRISE SECTOR AND FIRST TEN ENTERPRISES TO BE PRIVATIZED Indicators PE Sector Ten Enterprises (1988 data) ---------(CFAF billions)---------- Capital, Nominal Value 86.4 7.1 State's Share of Capital 88.1 4.6 Shares to be Sold 8.6 8.2 Employ**. t (88,924) n.e. Turnover 489.4 28.4 Value Added lb 127.8 11.0 Gross Investments c 56.5 8.7 Wage Bill /d 61.8 5.0 Gross Profits -8.6 0.7 Operating Subsidies 18.5 - Net Profits -24.5 -0.8 Corporate Income Taxes 4.4 ( 1.0 Other Taxes 6.8 ( Customs Duties L 2.6 n.a. La Total employment in the modern sector 1s estimated at 210,000 of which about 70,000 are civil servants. A GDP In 1984 was CFAF 1,015 billion, of which CFAF 280 billion in the secondary sector. Ac Private sector investment estimated at CFAF 108 billion. Ld Government wage bill totalled CFAF 103 billion in 1984. L In 1984, total customs revenues were CFAF 78.4 billion, total Government revenues CFAF 189.4 billion. Source: Contr81 Financier de Ia Pr6sidence and Direction de Ia Pr6vision, Ministry of Economy and Finance. 151. There appear to exist enormous obstacles to private sector engage- ment in Senegal at the present time. In addition to the negative profit- ability of the parastatals offered for sale, as mentioned above, the following main obstacles and constraints prevail. First, the deflationary pressures resulting from the stabilization and adjustment process, in the short run, lead to a compression of domestic demand by reducing domestic consumption and absorption. In the medium run, these deflationary pressures are compounded by the unfavorable prospects for (a) agricultural and industrial supply response; and (b) for agricultural and industrial exports. Secondly, the combination of the breakdown of the banking system, the large internal government debt and the weakness of the whole financial - 69 - intermediation system results in an extreme shortage of credit and venture capital. At least one keen observer described this situation as one of internal asphyxia. Unless the banking system goes through a radical restructuring and rationalizing process and a permanent solution is found for the settlement of government arrears, it is almost impossible to visualize how the private sector takeoff could take place. It has been estimated that at least CFAF 80 billion would be required to revitalize the banking system. 152. A third crucial constraint to private sector engagement is the previously discussed relative lack of an entrepreneurial tradition in Senegal. Historically and traditionally, Senegal has been, and continues to be, a nation of traders rather than entrepreneurs. The entrepreneurial initiative in the past came from French colonialists and other foreign groups such as the Lebanese. This relative lack of interest in and ability for business ventures on the part of Senegalese individuals has been further enhanced by the increasingly dominant role of the State in economic activity since Independence. The previously described public encroachment process including the takeover of foreign firms by the State, fostered largely inefficient managers who did not have to face the competitive pressures of the market and, thereby, greatly reduced the scope for providing training and experience to potential entrepreneurs in the private sector. In addition, the small number of potential entrepreneurs currently in Senegal faces a bleak business outlook and cannot realistically have rosy expectations with regard to the future in light of the above-mentioned low profitability of PEs offered for sale, the lack of credit and the deflationary pressures currently underway. 153. The fourth obstacle to privatization consists of the possible conflicting interest of the civil servants. We were repeatedly told by senior officials (particularly in the Finance Ministry) that bureaucrats had little or no interest in privatization, that they would continue to protect their own class interests and, at the limit, asking them to actually sell PEs was tantamount to asking them to commit Ohara-kiri". One denotes a cynical attitude among a number of civil servants that a game is being played which consists of putting companies up for sale but that as long as these companies are not sold, the status of civil servants remains unaffected and yet the Bank is satisfied because a key condition, i.e. disengagement, has been fulfilled. This reluctance among civil servants to push through the selling of PEs through a variety of administrative and bureaucratic hurdles could ultimately sabotage the process. 154. The ceiling on the number of public servants and the fall in their real salaries which are components of the budget retrenchment process act as further disincentives to "working themselves out of a job." 155. While the Bank recognizes the inherent uncertainty "regarding the speed at which private entrepreneurs adjust to the new industrial environment and foreign investors respond to the new opportunities in - 70 - Senegal,"57 and lists areas in which it perceives obstacles which could jeopardize the adjustment process and future development such as the weakness of the country's institutions, weaker than expected private sector response, and socio-political, climatic and external market risks, on balance it continues to remain optimistic.58 The following quote reflects this "In the medium term, however, new employment opportunities and improved earnings are expected as manufacturing and processing industries respond to the new situation and as controls on internal trade are relaxed. This more favorable environment for private investment and entrepreneurship is likely to encourage a faster departure of government employees into private sector activities, which would in turn permit some improvement in rewards for performance in the civil service, compensating partly for its overall loss of remuneration since the adjustment process began.*5 (6) The Social Costs of Adjustment 156. Estimating the impact of adjustment measures on the welfare of Senegalese producers and consumers remains extremely difficult given the absence of basic income, employment and consumption data. Indicative of the problem is the fact that the last household survey was conducted in 1960. This lacuna was noted in the President's Report on the third SAL, "there is also insufficient basis to identify and design operational programs and projects aimed at alleviating the transitional adverse impact of structural adjustment on vulnerable groups." However, Bank officials have tended to make a number of assumptions regarding the distributional impact of policy reforms. 157. In the first place, it has been assumed that raising nominal producer prices for agricultural producers would raise both income levels and labor retention in the sector. Second, it has been assumed that the costs of adjustment are largely transitional costs and that measures designed to boost supply-side responses would have the expected positive employment and income effects. As suggested earlier, both assumptions are open to question. 158. The argument that the adjustment period has not seen welfare gains can be bolstered by reference to a number of factors. First, the predominant characteristic of the Senegalese adjustment episode has been reduction in aggregate absorption. One component of this has been the strong downward pressure on public expenditure as part of the objective of reducing the fiscal deficit. Between 1980/81 and 1985/86 total expenditure fell from nearly 32Z 3f GDP to under 22t, with a 45Z real reduction in capital outlays and a 13% real reduction in current expenditures. The share appropriated by wages and salaries fell from 12? of GDP to under 10% in the same period. 57/ President's Report, SAL III, p. 10. 58/ Ibid., pp. 48-49. 59/ President's Report, SAL II, p. 31. - 71 - 159. Available evidence regarding trends in real wages indicates a significant fall in the adjustment phase. Using the minimum wage data -- the salaire minimum interprofessionnel garanti (SMIG) -- it can be seen (Table 19) that there has been a real decline of over 20Z between 1979/80 and 1986. Moreover, wages in the informal sector tend to be significantly lower than the minimum wage and subject to greater flexibility. As the brunt of the labor market adjustment has fallen on the informal sector, this would suggest that the already large supply of unskilled labor in that subsector has been increased forcing down nominal wages and generating greater open unemployment, as well as increased underemployment. This argument cannot be validated empirically due to the absence of satisfactory employment data. But given the insignificant supply response to adjustment policy measures in all core sectors, the hypothesized employment effect associated with those policies has not materialized. Table 19: MINIMUM WAGE RATES IN CFA/HOUR (SMIC) AND INTERSECTORAL TERMS OF TRADE, 1979-88 Year SMIC SMIG Index Terms of Trade L (current) (constant) L (1979=100) 1979 107.05 107.05 100 100 1980 133.81 122.76 115 82 1981 137.16 119.28 111 93 1982 148.25 108.88 101 95 1983 169.25 112.09 106 76 1984 175.00 103.65 97 85 1985 183.75 98.20 90 102 1986 188.75 90.07 84 111 L Weighted agricultural producer prices/SMIC. L 1979 CFA constant prices. Source: Based on Government data. 160. The burden of adjustment has thus been distributed, in flex price markets -- such as .haracterize the informal sector -- in terms of both wage reductions and lower labor absorption. In the formal sector where labor legislation remains an important factor, wages have fallen as the SMIG has lagged behind the rate of inflation. But labor market rigidities imposed through legislation while limiting the scale of redundancies have also severely constrained, over a longer period, recruitment. In the Government and public sector, wages have likewise fallen in real terms but there has also been greater direct retrenchment. In the first place, closure of ONCAD and its successor body, the Socidtd Nationale d'Approvisionnement Rural (SONAR) -- the former agricultural sector - 72 - parastatals -- led to major job losses. In the case of the rural development agencies (RDAs) also there has been some retrenchment. The Socidtd de Ddveloppement et de Vulgarisation Agricole (SODEVA) staffing has been cut by 55Z between 1982 and 1987 with a total of around 1,000 redundancies. SAED staffing has declined by 9Z in the last two years and the pressure on the entire parapublic sector has been to reduce the workforce. For the civil service, aggregate employment has been held constant at around 68,000 since 1985. Howexer, between 1980 and 1985 employment in this category grew by over 3.5Z p.a. This suggests that the direct employment loss has not been great but that wage levels have declined significantly. In short, combining the civil service and parapublic sector, both have been marked by falling wages and, at best, over the period 1980-87 minor growth in total employment. This was not sufficient to act as a counter-cyclical employment strategy. Since 1985, however, there has been some reduction in employment. Under SAL III, a special reinsertion fund for redundant workers has been established but, as yet, it is too early to comment meaningfully on its coverage and impact. 161. From the above it can be seen that for both urban formal and informal sector labor markets the adjustment phase has been marked by declining employment and falling real wages. This outcome was explicitly recognized by the Bank60 but posed more as a transitional problem that could be corrected through greater labor absorption in agriculture and general output growth consequent upon a more efficient allocation of resources. Yet, to date, this required supply response has not been present. There appears to have been no significant area or intensity response in agriculture, largely because of climatic and technical constraints. Moreover, although the intersectoral terms of trade have moved in favor of agriculture after 1985 (see Table 19) real producer prices for both export and cereal crops have not risen. Further, because a significant proportion of farm households are net food purchasers -- not only of domestically produced coarse grains but also of imported rice -- the parallel upward adjustment in food prices, particularly for ri:e, will have dampened any possible income effect through nominal producer price increases. Since 1982 food prices have increased on average by around 12% p.a. 162. In conclusion, any assessment of the impact of the adjustment program on social welfare in Senegal has to be both tentative and provisional. Available indicators are very limited and not necessarily reliable. However, it can reasonably be argued that the overall character of adjustment to date has been deflationary. There has be%.i little if any supply response. This has- been associated with downward pressure on the real wage and direct employment losses. One manifestation of this has been growing political discontent, as manifested in the recent Presidential election. Although social sector outlays have not fallen to any major extent as a share of total government expenditure, the aggregate size of 60/ President's Report, SAL III, p. 40. - 73 - public expenditure has been much reduced. There is increasing evidence that the quality and volume of services offered by the Government in health and other social sectors has been declining.61 163. The Bank's analysis of the distributional and welfare implications of the adjustment program has been selective and, in certain respects, incorrect. Partly because the Government has been prepared to comit itself to a program of difficult reforms, the Bank has appeared loath to raise many of the complex issues that arise with regard to the impact of those reforms on poverty groups. The costs of adjustment have been portrayed as transitional costs, despite the fact that there could be no real optimism concerning the size of the supply response from the economy. Moreover, the Bank has been tardy in encouraging the collection of basic data that could throw proper light on the social impact. After eight years of adjustment it is only now that the Bank is pushing for studies on employment in the crucial informal sector and for a more comprehensive household survey. F. Foreign Assistance, and Implications of Foreign Debt and Debt Servicing on Investment and Growth 164. The pattern of foreign assistance to Senegal up to 1982 has been analyzed in some detail by John P. Lewis (ex-chairman of DAC).62 His main findings in the context of this section are the following: (1) together official transfers paid for all of the country's investment and 6Z of its consumption in 1981. This pattern of counting on aid for investment funding as well as gap closing reaches all the way back to 1960; (2) the share of France in Senegal overseas development assistance (ODA) was historically very large but falling from 68% in 1967-71 to 372 in 1976-82; conversely the share of the World Bank Group was strikingly low amounting to only 6.6% of total official net transfers (i.e. ODA and other official flows) cumulatively for the three-year period 1980-82; (3) during the 1960s and early 1970s, France vas the classical 'residual" donor. Instead of contributing to investment what French support 'really shored up -- smoothed and accelerated -- was consumption*; (4) the other side of the medal was that the Government diverted scarce foreign exchange from 'investment' to 'participation' (i.e. buying out expatriates' equity in existing facilities; (5) official transfers -- which came to account for about one-fifth of GDP in 1981 and 1982 -- may actually retard development if, instead of accelerating development, they become 'a crutch for habituating an economy to unsustainable levels of-consumption'. 61/ Nemat Shafik, 'Public Finances in Senegal," (USAID: Dakar), 1986, (mimeo). 62/ Lewis, op. cit. - 74 - 165. It is clear that, with the onset of the structural adjustment period, the Bank replaced France as the major interlocutor in the policy dialogue with the Government and its increasing share of incremental program assistance gave it more leverage than at any time previously. It appears that the French Government was only too glad to let the Bank take the front row and push hard for major structural and institutional reforms while providing some relief to the French Treasury from a continuing and increasing drain on its resources. 166. In order to ascertain the net foreign assistance flow available after servicing the external debt, Table 20 was prepared. The breakdown of the table should be explained since it is not the conventional way of presenting the balance of payments. Row 1 gives the current account balance excluding interest payments on the foreign debt, while row 2 presents the net capital inflow excluding foreign assistance (thus including private capital, short-term capital and other MLT flows); row 3 is the sum of rows 1 and 2 and can be thought to be the net state of the balance of payments exclusive of foreign assistance and debt servicing flows; rows 4a and 4b sho, respectively, official transfers and MLT loan disbursements which are summed up in row 4 to yield total foreign assistance; rows 5a and 5b indicate, respectively, the interest and principal payments on the foreign debt -- these two flows are added up in row 5 to yield total debt service payments; row 6, in turn, presents net foreign assistance, i.e. total foreign assistance minus debt servicing (row 4-row 5); and, finally, row 7 gives the overall balance (row 3-row 6), i.e. the net decrease (-) or increase (+) in reserves or what is sometimes called the gap. - 75 - Table 20: BALANCE OF PAYMENTS HIGHLIGHTING NET FOREIGN ASSISTANCE, 1981-67 (in current CFAF billion) 1981 1982 1988 1984 1988 196 1987 La 1 Current Account (excl. Interest payments on foreign debt) -162 -150 -168 -154 -18 -92 -91 2 Other MLT Inflows (net) and Capital Flows, n.e.i. 24 19 15 27 85 29 - 3 Net Balance (excl. Foreign Assistance and Debt Servicing) -138 -182 -158 -127 -148 -68 -01 4a Official Transfer* ."TI :w "w 2U! 5 if 4b ULT Loan Disbursements 60 93 122 89 97 01 79 4 Total Foreign Assistance 108 158 182 150 157 188 181 So Interest on Debt -11 -11 -16 -22 -19 -84 -85 6b Debt Amortization (Principal) -18 -8 -6 -14 -20 -39 -51 S Total Debt Service -24 -14 -22 -86 -39 -78 -86 6 Net Foreign Assistance Flow (row 4-row 6) 84 189 160 114 119 as 45 7 Decrease in Reserves (-) (row 3-row 6) or Gap -54 7 7 -18 -80 2 -36 La Estimate. Sources: Rows 1 through 4: World Bank, Senegal: An Economy Under Adjustment, Report No. 6454-SE, February 13, 1987, SA Tables 18 (for 1981-85) and 86 (for 1986-87). Rows Sa and 6b: World Bank debt tables. 167. What appears strikingly out of this presentation of the balance of payments is the sharp decline in the net flow of foreign assistance after debt servicing (row 6) in 1986 and 1987. Thus, it can be seen that the net inflow which reached 160 billion CFAF in 1983 fell drastically to about 65 billion CFAF in 1986 and 45 billion CFAF in 1987. To place the last figure in perspective, it represented only about one-fourth of gross domestic investment in 1987. In fact, the situation is even more serious in the sense that in 1987, according to these estimates, a gap (decrease in reserves) of 36 billion CFAF appears. If it is not filled by the French Treasury, it will reduce the amount of net foreign assistance available for investment to close to zero. - 76 - 168. It can reasonably be asked if the Senegalese economy has the capability uver time to finance from internal sources (i.e. domestic savings) three-fourths or more of its investment requirements. An answer to this question necessitates projecting the components of the balance of payments. Essentially, we took over the Bank's projections in its 1987 CEM with some modifications. We used debt servicing projections on the existing debt as of the end of 1987 and made some assumptions regarding export growth (5% p.a. between 1986 and 1992) and import growth (4Z p.a. growth).63 Again, the balance of payments projections are presented so as to separate and highlight net foreign assistance. The results are given in Table 21. 63/ Senegal: An Economy Under Adjustment. We assumed significantly lower growth rates of exports than contained in the Bank's projections (see SA Table 36) and slightly lower growth rates of imports. - 77 - Table 21: BALANCE OF PAYMENMS PROJFrTIONS HIGHLIGHTING NET FOREIGN ASSISTANCE, 196-92 (in US million) Me6 196? 1963 1989 1990 1901 1992 la Exports 656 691 725 762 800 840 882 lb Imports 681 896 931 969 1,007 1,048 1,069 Ic Resource Balance -208 -204 -206 -207 -207 -206 -207 Id Rest of Current Account (excl. Interest on debt) -82 -86 -68 -70 -78 -81 1 Current Account (exci. Interest payment on debt (Ic + id) -265 -269 -272 -275 -277 -285 -288 2 Other MLT Inflows and Capital n.e.i. 8 -1 -8 6 24 18 15 8 Net Bal*nce (exct. Foreign Assistance and Debt Servicing) (1 + 2) -182 -270 -275 -289 -253 -267 -278 4a Official Transfers 166 176 183 190 198 211 225 4b MLT Loan Disbursements 238 261 291 818 887 854 875 4 Total Foreign Assistance (4a + 4b) 899 487 474 506 585 565 800 So Interest on Debt -98 -116 -148 -182 -118 -102 -68 Sb Debt Asortization -112 -171 -284 -260 -249 -270 -244 5 Total Debt Service (Sa + Sb) -210 -287 -410 -892 -3867 -872 -327 6 Net Foreign Assistance 189 160 64 116 168 193 273 7 Decrease In Reserves (-) or Gap (8 - 8) 7 -120 -211 -148 -81 -74 0 8 Net Foreign Assistance after Gap Closing 196 80 -147 -17 91 119 273 Sources: Row ls: growth of exports is projected at 55 p.a. from 1986 on. Row 1b: growth of Imports is projected at 4% p.a. from 1988 on. Rows Id, 2, 4a and 4b: World Bank, Senegal: An Economy Under Adjustment, Report No. 6454-SE, February 18, 1987, SA Table 386. Rows So and Sb: World Bank debt tables. 169. Row 6 of Table 21 yields an estimate of the net foreign assistance (after debt servicing) available for investment. It should be compared to - 78 - estimates of gross domestic investment (GDI) to determine what share of the latter it could finance. In 1986, GDI amounted to 179.4 billion CFAF (or US$518 million at the rate of exchange of 346.3 CFAF - US$1); assuming GDI grew at 12? at current prices, it would equal 201 billion CFAF in 1987 (or $670 million at a rate of exchange of 300 CFAF - $1). Thus, in 1987 net foreign assistance financed one-fourth of total investment requirements; in 1988, it is projected to finance only 10Z (assuming GDI grows at 5Z); in 1989, less than a fifth; in 1990, about one-fourth; and in 1991, about 18?. According to these projections, even by 1992 net foreign assistance would only cover one-third of total investment needs which is very significantly less than the historical share. 170. In fact, the situation is even more unfavorable for two additional reasons: first, it should be noted that the projections in Table 21 leave a gap to be filled (row 7) so that net foreign assistance after gap closing appears marginal in 1987 and is large and negative in 1988 (-$147 million) and is still negative in 1989. The gap only disappears according to the projection in 1992; secondly, the projection of debt servicing excludes interest and principal payments on the new loans extended after 1987 -- if the debt servicing were to be included, it would worsen further the part of net assistance available for investment. 171. The implications of the above exercise are clear. Short of debt cancellation or a long-term moratorium, the investment funds required for any take-off in Senegal will not be forthcoming. Given the multitude of constraints which the country faces, even if a continuous flow of foreign resources for investment were available, it would be an uphill battle to overcome these constraints. Without such a flow, the best that can be hoped for is continuing stagnation -- if not decline. G. Conclusions 172. A number of important and interrelated conclusions emerge from the preceding evaluation. First, it was documented in detail that the Bank throughout the structural adjustment process imposed a large and increasing number of conditions (from 32 in SAL I to 77 in SAL III). Whereas it is clear that the adjustment process required the general and simultaneous implementation of measures affecting both the macro-environment (e.g., by moving resources into the export sector, by reducing government expenditures and private consumption and encouraging investment) and the micro-environment (through increasing the efficiency of the price system and of all resources) and extending to all sectors, it does not follow that a long laundry list of conditions without any clear priority ordering is the answer. As Stanley Fisher has argued "If the Bank is to be effective, it will have to establish prioritias among these micro policies, for by - 79 - insisting on too many of them it may get none,"64 and "the more ambitious and detailed the structural adjustment plan, the greater the strain on economic analysis and decision-making capacities in the affected country. Economic expertis. and political will are also scarce resources to be allocated efficiently in the execution of economic policy.165 173. Secondly, our own classification of these condizions into (a) studies, (b) preconditions, (c) actions, (d) intermediate targets, and (e) institutional reforms revealed that the share of conditiuns consisting of the first two types above increased from about one-third in SAL I to 56% in SAL III. 174. Studies and "preconditions" do not, by themselves, alter the prevailing situation. It is only when they are embedded into concrete actions that they can affect performance. The high proportion of the latter two types of conditions in the SAL conditionality packages -- combined with the incomplete nature of many conditions which we defined as "actions" (e.g., public sector disengagement without private sector engagement) -- implies that the Government could, in fact, comply with (fulfill) a large part of the adjustment program without any change in policy measures or reforms having occurred and, therefore, with no resulting impact on performance. In fairness, it has to be recognized that a number of "hard" conditions were implemented at the Bank's urging -- particularly those dealing with budgetary retrenchment (e.g., limits on the size of the civil service and their salaries); trade liberalization and agriculturml price policies and institutional reforms. However, because of inadequate, or incomplete understanding of the impact of many of these measures on performance, some of these "hard" conditions were not appropriate or their effects on performance marginal, as is illustrated in the following paragraphs. 175. Thirdly, we were made aware of and identified some inconsistencies between SAL conditions aimed at centralizing the management of, and control over, the foreign debt (within the MOF) and some Bank projects which grant (or attempt to obtain) decentralized budgetary authority to host enterprises. Another example of a conflict between the Bank's program conditions and project conditions consists of the special considerations and treatment (e.g., relating to subsidies and pricing) requested by and often granted to public enterprises receiving project help (see para. 139). A major effort should be undertaken within the Bank to achieve internal consistency between its program and project components. 176. Fourth, the need for a multitude of studies and empirical information in a variety of areas (e.g., agricultural price policies, project analysis methodology, operation of labor market) reflects the basic 64/ S. Fisher "Issues in Medium-Term Macroeconomic Adjustment," The World Bank Research Observer, Vol. 1, No. 2, July 1986, pp. 179-80. 65/ Ibid., p. 181. - 80 - lack of knowledge on the part of the Bank and the Government of the underlying socioeconomic reality and relationships among policy variables. The conceptual framework linking (a) exogenous variables and constraints and (b) policy instruments and reforms, jointly, to socioeconomic objectives and performance -- described in section C -- remains largely a black box in the context of Senegal. What is surprising and difficult to understand is how irresponsibly paasive the Bank continued to be throughout its policy dialogue with the Government in insisting on better data. information and studies upon which to make policy recommendations. Numerous examples have been documented in this evaluation of areas where the Bank continuously identified crucial areas of ignorance (e.g., project analysis, impact of price policies, agricultural research, socioeconomic status of vulnerable groups, operation of informal sector) calling for surveys and studies without following through and demanding prompt results.66 The main consequence of this shortcoming was that the Bank continuously had to make policy recommendations on the basis of very inadequate, and often faulty, baseline information and without a clear understanding of the causal and quantitative mechanisms relating policy variables to policy objectives. In other words, the Bank consistently was forcea "to put the cart before the horse.' Ad hoc considerations, ideology and wishful thinking often replaced sound technical economic analysis. 177. Fifth, a typical symptom and manifestation of the Bank's lack of an adequate conceptual and quantitative grasp of the dynamics of the socioeconomic system throughout the period under consideration in the context of Senegal has been the continuing and pervasive tendency to overestimate the likely future performance (particularly with regard to growth) and to underestimate the negative impact of constraints on development. In particular, failure to recognize sufficiently the structural and deep-rooted nature of many of these constraints, such as the secularly worsening rainfall pattern and soil degradation and desertification process, repeatedly led the Bank to explain away or excuse poor performance as being caused by cyclical and temporarily exogenous factors. An important conclusion which was reached at the end of the evaluation of structural adjustment in agriculture (para. 111) is that the foremost objective should be to alleviate the binding physical constraints resulting from the deteriorating natural environment by every means possible. More specifically, soil regeneration and conservation measures (such as the application of rock phosphate in a large scale) and development of more drought-resistant varieties appear almost as necessary preconditions to improved performance. 66/ For example, we have seen that in 1970 the Bank was already concerned with the weakness of the project analysis and preparation methodology and continued to express concern about this issue without any tangible results until 1988. Likewise, the Bank was continuously concerned about improving technical packages in agriculture and understanding better the impact of agricultural price policies -- yet very little progress has occurred so far. - 81 - 178. Sixth, the all-embracing nature of structural adjustment requiring concerted action on many fronts and involving the complex interaction of many policy variables simultaneously can only be very imperfectly analyzed without some intersectoral framework or model. Two recent studies of the structural adjustment process and conditionality emphasized this point, i.e., "the models are useful precisely because they force the analyst to set out the structure of the economy and to focus on the relationships that determine the outcome of policy changes,.67 and wthe alternative of forsaking models altogether and relying on general principles to guide advice giving and policy evaluation is not feasible: policy cannot be evaluated without counterfactuals, and counterfactuals require the use of either an explicit or an implicit model.w68 Likewise, on the basis of a comparative study of all structural adjustment loans extended by the Bank between 1980 and 1986, Mosley concluded that 'the World Bank's forecast of the consequences of complying with its conditions are based on faith and guesswork, not on formal reasoning" and 'My first recommendation is therefore that, in future, a donor should base its predictions of the effects of its conditions on a formal model that links the policy instruments it wants the recipient to change and the targets at which the instruments are aimed. This forecast, which should be shared with the recipient...will necessarily depend on a number of factors the recipient cannot control, such as weather, crop disease, and world demand for the country's exports. It should therefore typically take the form not of a simple point estimate...but rather of a frequency distribution of outcomes. It should consist of statements like 'If this package is implemented, there is a 50 percent chance that it will raise export volume by between 1 and 2 percent, a 20 percent chance that it will raise export volume by between 0 and 1 percent,, and so on..69 179. A final issue of absolutely crucial importance to the future development of Senegal is whether the net foreign resources available are sufficient to provide the required investment funds for a take-off; i.e. a departure from the historical stagnation trends. The question here, as the analysis of projections in section F demonstrated, is whether the bulk of the gross inflow of foreign assistance is to be used to service the foreign debt, leaving a marginal amount for investment as is currently Lhe case or 67/ Fisher, op. cit., p. 167. 68/ Ibid. p. 168. Fisher argues, furthermore, that "the modeling of macro structural adjustment programs poses no particularly difficult conceptual issues. The lack of data and the possibility that a large variety of models will be needed will make the job long and difficult in practice. Key parameters, such as the supply and demand elasticities of imports, adjustment lags in investment, and the nature of wage behavior, may be difficult to pin down with existing data. The best available may in the end not be very good. But models can be estimated and will be useful in medium-term policy evaluations." p. 178. 69/ Mosley, op. cit., p. 29. - 82 - whether some fundamental changes in the present Orules of the game* can be devised allowing for a significantly larger net inflow of foreign capital. In theory two solutions suggest themselves: (i) a large increase in the gross amount of foreign assistance (in the form of grants) destined to Senegal; and (ii) some form of debt cancellation (not rescheduling which would only postpone the inevitable). The first alternative appears most unlikely given the present mood of the donor community which leaves the second option -- however difficult -- as the more realistic one. The Bank and the donor community have to decide between debt repayment and development. Given the multitude of constraints which Senegal faces, the conflict between these two objectives appears unavoidable. A continuation of the present emphasis on debt repayment will entail stagnation if not decline. If the Bank decides that the socioeconomic development of Senegal is the objective, then it must be prepared to work with other donors to converge on a solution to the debt servicing problem. It is, of course, imperative that the funds thus freed up not be used to finance uneconomical projects; any alleviation of the debt burden would be meaningless in this case. On the other hand, a reduction in debt payments would not only permit the financing of good public sector projects but the accompanying easing of credit and foreign exchange scarcity in the economy is an important condition for private sector investment growth. - 83 - Chapter III: SECTORAL ISSUES AND AID COORDINATION A. The Agricultural Sector (1) Introduction 180. Policy toward the Senegalese agricultural sector has to contend not only with a poor natural resource endowment but also highly variable and deteriorating climatic conditions. Experience over the past quarter century suggests that these constraints have not been overcome to any significant degree. Some basic statistics provide telling evidence. Between 1960 and 1986 agriculture grew at around 1.32 p.a. and at under 0.52 since 1970. Real per capita incomes in the rural sector appear to have remained constant between 1970/71 and 1983, even despite specific policy measures designed to boost rural sector incomes. 181. While such poor performance is often attributed to the application of unsuitable and unsustainable policies,70 past agricultural sector output trends can, above all, be explained as a function of rainfall. By the early 1980s less than 1Z of the total arable area was irrigated. The bulk of agricultural producers have been faced with declining rainfall trends. In the northern Groundnut Basin, rainfall levels currently are at around half the level that was attained between 1930-60. Though less pronounced elsewhere, this decline has been a national phenomenon. In addition, producers have been faced with extreme inter-annual variance in climatic conditions. Regressing groundnut and millet -- the major crops -- yields on rainfall and time indicates that around 60Z of yield variation over the past 30 years can be attributed to rainfall. 182. Despite the fact that the share of the primary sector has declined from around 252 of GDP in the early 1960s to under 20Z in the 1980s and agriculture from 162 to below 102 over the same period, the sector remains central to the economy. An estimated 70% of the labor force continues to be principally employed in agriculture, while between 10-121 of industrial output originates from the oil mills subsector. As significantly, agricultural exports which accounted for over 852 of the total value of merchandise exports in the period 1960-69 still accounted for around 412 between 1980-86. However, the external trade profile clearly illustrates the declining share of groundnuts in total trade. In the 1960s groundnut products comprised nearly 75% of export goods earnings. This share fell to around 422 in the 1970s and to 17% between 1980-86. In addition to the importance of the sector in terms of employment, income and tradables output, a further factor has served increasingly to highlight the need for a coherent sectoral policy. That is, the growing inability of Senegalese agriculture to generate adequate food output levels to satisfy domestic demand. By the mid-1980s Senegal produced only just over half its domestic 70/ See, for example, "Senegal: Agricultural Sector Strategy Brief," 1987. - 84 - food requirements, the remainder being satisfied by food imports (40Z) and food aid (82). In the context of declining or stagnant exports, this has meant that the primary sector trade balance turned negative between 1983 and 1986. Rice and wheat imports amounted to just under 90Z of export revenues from groundnuts, while total food and beverage imports significantly exceeded earnings from fisheries and agriculture. 183. The problems of low productivity, of excessive output and price variability, of uncertain external market prospects and pressure on the balance of payments, though particularly pronounced in the period since 1979, have remained almost a constant feature of the Senegalese economy in the period under discussion. Accordingly, these issues have, in various ways, remained at the heart of both government policy interventions and donor activity in the last 25 years. However, there has been considerable variation in the emphasis placed on particular policy objectives, let alone instruments, in this period. This is clear from any examination of the Bank's lending for Senegalese agriculture from the first project in 1969 through to the present. 184. The World Bank's lending program in Senegal dates back to 1966, but it was not until 1969 that the first agricultural project loan was approved. Since that time, the Bank has financed 18 projects in agricul- ture, with 12 of those being approved prior to 1979. Bank lending to Senegal has continued to claim a relatively modest share of overall ODA, although a significant increase has occurred after 1980. For the period 1976-85, IBRD/IDA disbursements amounted to 7.32 of gross ODA. Disaggre- gating Bank Group lending to Senegal, it appears that between 1966-87 total approved lending amounted to just over $800 million, of which 212 was directly allocated to agriculture. However, this disguises the fact that non-project lending, which has become increasingly pronounced since 1980, has contained a number of agriculture-specific actions. Although six agricultural sector projects have been approved since 1980, it is important to note that adjustment lending comprised 45Z of total approved lending to the country between 1980 and 1987. Agricultural project lending accounted for 17% in the period. 185. Bank assistance to Senegal has comprised a relatively small share of total lending when compared with the major bilateral donor, France, or, equally, EEC aid. One implication is that Bank lending to agriculture has remained relatively limited in relation to overall external assistance to the sector despite the number of projects. Nevertheless -- and particularly post-1980 -- the Bank has played an important role in defining and realizing the parameters of intervention in the sector. (2) World Bank Lending to Agriculture, 1969-79 186. The content of Bank lending over this period reflects fairly closely the principal emphasis of sector strategy that existed at that time. In particular, projects were designed as a means for raising output and productivity in the core groundnuts and millet-producing regions, - 85 - primarily through credit programs and access to modern inputs, especially fertilizer, and agricultural technology. Both the Agricultural Cvedit projects and the Sine Saloum project -- all of which were concentrated on the core Groundnut Basin -- broadly conformed to this characterization. In addition, the Casamance Rice project and its sequel, the Second Sedhion project, aimed not only at diversifying production toward rice, but also at raising output levels through promotion of more intensive production tech- niques and animal traction. A second major orientation was towards irrigated agriculture in the Fleuve region (e.g., River Polders and Debi- Lampsar projects). With both Terres Neuves projects, the Bank also sought to intervene in the resettlement program where households shifted from the relatively densely populated Groundnut Basin to Eastern Senegal. Lastly, an attempt vas made with the Eastern Senegal Livestock project to improve pastoral resource use through the establishment of communal grazing grounds, as well as improved animal health care facilities. Table 22 provides a breakdown of Bank projects in agriculture for the full period, 1969-87. However, it can be seen that of the projects approved before 1980, nearly 40Z of disbursed resources were allocated to the Groundnut Basin with a further 352 directed towards irrigated agriculture In the Fleuve. The remainder was equally split between the Casamance and Eastern Senegal regions. As importantly, it can be seen that the last productive investment was approved for the Groundnut Basin in 1975, and for Fleuve in 1978, with the exception of Irrigation IV which was approved for the Fleuve region in December 1987. In the Casamance, with the exception of the Eastern Senegal project, approved in 1983, and with partial coverage of the region, the last major investment was made in 1977. - 86 - Table 22: IBRD/IDA PROJECT DISBURSEMENTS (USS million; as of December 31, 1987) Regional Allocation Estimated Projects Groundnut Eastern Non- Rat of Return (Board Approval Date) Fleuve Basin Sensagl Casamance Regional Appraisal PCR 1. Agricultural Credit - 6.93 - * - 25 n.a. (1/14/69) 2. Casamence Rice - - - 3.69 - 18.6 28 (5/11/71) 8. Terres Neuves Resettlement - - 1.86 - - 18 59 (5/11/71) 4. Senegal River Polders 4.50 - - - - 14 -6 (12/19/72) 6. Second Agricultural Credit - 8.16 - - - 22 15-20 (6/19/78) 6. Drought Relief - 1.60 1.50 - - - - (11/18/78) 7. Debi-Lampsar Engineering (4/29/75) 0.18 - - - - - - 8. Sine Saloum Agricultura Development - 7.00 - - - 25 Ia. (5/20/76) 9. Second Torres Neuves Settlement - - 2.00 - - 14 18 (7/8/75) 10. Eastern Senegal Livestock * * 4.20 - - 24 18 (8/8/76) 11. Second Sedhlou - - - 4.77 - 28 6 (6/22/76) 12. Debt-Lampsar Irrigation 19.88 - - - 10 -4 (3/7/78) 18. Small Rural Operations 8.59 8.69 - 8.69 - -- (8/11/80) 14. Forestry - - 8.48 - - - - (2/10/81) 15. Agricultural Research - - - - 18.36 - - (9/8/81) 16. Eastern Senegal Rural Development - - 8.76 - - - - (8/2/88) 17. Ireilgation Tech. Asst. 1.05 - - - - - - (10/29/86) 18. Irrigation IV 88.6 - - - - - - (12/1/87) Source: World Bank. - 87 - 187. Examination of the Project Completion Reports (PCRs) and Project Performance Audit Reports (PPARs) indicates that the majority of projects were beset with major problems that compromised the overall effectiveness of the interventions. In some cases, rates of return were significantly negative and in many cases substantially below appraisal estimates. Projected output and productivity effects tended in reality to be very weak, if at all present, while institutional constraints proved, with the exception of the Settlement projects and the Eastern Senegal Livestock project, to be debilitating. In short, the overall performance of projects in the sector mirrored the generally weak performance of Senegalese agriculture in this period. While a fair measure of this weakness can be attributed to a disastrous set of seasons, with major droughts occurr'ng six times between 1969 and 1980, it is also evident that inadequacies in project design, in sectoral policies directly impinging on projects (such as relative prices) but not tractable under the project lending format, as well as Bank and othar donor policy analysis deficiencies, have played a significant role in lessening the beneficial impact of project resources. 188. In order to bring out the underlying reasons for unsatisfactory project outcomes, it is necessary to situate those activities in relation to both Government and Bank analysis of the sector and its potential. This is particularly important given the current disenchantment with project lending as a suitable vehicle for delivering resources to Senegal. The following section attempts to situate this shift in perception. Senegalese Agriculture: Strategic Issues (a) Diversification 189. A constant feature in agricultural sector strategy has been the degree to which Senegal should continue to rely on groundnuts as the principal agriciltural tradable. Excessive dependence on one primary commodity was, in fact, one of the primary reasons for the first invest- ments in irrigated agriculture in the late 1930s.71 As then, arguments for diversification in the 1960s and 1970s hinged not simply on the problems associated with price variability but also on projected longer-term demand projections for the good. These arguments became particularly pressing following the phasing-out of relatively high guaranteed prices on the French market in the mid-1960s and, somewhat later, the growing substitution of other oils for that of groundnut in developed country markets. Despite the fact that Senegal was argued to have a definite comparative advantage in groundnuts, from the mid-1960s onwards Bank reports were urging a shift toward other crops, particularly cotton, alongside measures to enhance productivity in the groundnut sector. The latter objective was to be achieved through enhanced use of fertilizer, 711 J.R. Moris (et al.), African Irrigation Overview, Main Report, USAID (mimeo), December, 1987, p. 86. - 88 - improved seed and equipment.72 At the same time, the case for diversifi- cation was strengthened by the fact of relatively limited scope for groundnut area expansion, at least in the core Groundnut Basin. In that respect, crop diversification was generally viewed both by Government and the Bank as being in part synonymous with regional diversification, particularly with regard to agricultural expansion in Casamance, Fleuve and Eastern Senegal. This was, for example, directly reflected in the objectives of the first Terres Neuves Project which aimed to *diversify agricultural investment out of the Groundnut Basin and out of groundnutsw (PPAR, p. 1). 190. Diversification policy has had two principal components. The first has related to the promotion of tradables other than groundnuts. In particular, this occurred in the growth of cotton production in Eastern Senegal which was initiated in the 1960s. By the end of that decade around 52 of total agricultural export value was generated by cotton. While the Bank's input was very limited, projects such as the two Terres Neuves Settlement schemes, and also but to a lesser extent the Second Sedhiou project in Casamance, had a cotton component. In addition, the 1970s also saw a signilicant growth in exports from the fisheries sector. By the end of the decade fish exports had replaced phosphates as the second most important export in some years. As with cotton, the Bank played a peripheral role in encouraging, through direct project resources, the growth of this subsector. 191. The second component of the diversification strategy related to the promotion of home goods production and particularly import-substituting cereals. The rationale for this was the growing rice and wheat import bill, in part a function of the extensive consumer subsidies that existed until 1974. Augmenting domestic cereal supply was primarily dealt with in terms of boosting rice output in the Fleuve and Casamance and, to that extent, was closely linked to the question of the scale of resource allocation toward irrigated agriculture. It is notdble that neither the Bank nor the Government emphasized at this stage growth in millet output and consumption save to the extent of assuming that technical packages and extension in the Groundnut Basin would also have a positive effect on millet output and productivity.73 192. Emphasis on rice had emerged as a key feature of Government policy by the end of the 1960s, and by 1970 there was a shelf of a dozen rice projects. The Bank's attitude to this strategy was driven by the fact that with a non-negotiable exchange rate and high investment costs in irrigation and infrastructure, Senegalese rice could not be produced at economic 72/ See, for example, World Bank, The Economy of Senegal, Report No. AF- 44a, May 26, 1966. 73/ See, e.g., World Bank, The Current Economic Situation and Prospects of Senegal, Volume 1: The Main Report, Report No. AW-15a, June 8, 1970, p. 20. - 89 - costs. In 1970, it was estimated that, using shadow prices, the cost of production of Senegalese irrigated rice exceeded by around 30Z the landed cost of imported rice. By the same token, the Bank favored the development of rice and irrigated agriculture outside of the Fleuve, in the Casamance (CPP, 1970). The Casamance Rice Project was a direct consequence of this preference where the main objective was to increase the area under rainfed rice as well as, but to a much lesser extent, the improvement of existing swamp rice land. For the Bank the main attraction of Casamance was the lower level of investment cost. 193. 'While the need for higher domestic food supply levels was accepted unequivocally by the Bank, consistency over the means for achieving this was not maintained. Having viewed with some hostility the development of the Fleuve, in the early 1970s, the Bank's approach changed in the mid- 1970s. The 1975 Agricultural Sector Survey74 noted that, in part as a consequence of Bank advice, rice consumer subsidies had been eliminated, hence encouraging local production, and that import substitution was unlikely to come simply through enhanced consumption of millet. On the proviso that rice production could be justified economically, the Government was urged to invest in rice schemes in the Fleuve. At that time, it was felt that the projected barrage at Diama and dam at Manantali would allow for economic rice production. In the meantime, the development of small village perimeters was advocated. Production costs of paddy in the Fleuve were estimated at a level some 192 below the economic farm-gate price75 However, there was little indication that these results were based on detailed studies of production costs nor of the likely impact of development costs on the economic cost of additional rice output. Nor equally was there a satisfactory treatment of the investment implications of a shift to irrigated agriculture at the expense of the rainfed regions. 194. The apparent shift in emphasis by the Bank was not translated into significant support for OMVS, the major institution charged with developing irrigation infrastructure. The 1976 CPP argued that the Bank should respond positively to any request by OMVS for technical assistance and prepare for further involvement in the program. This was justified by arguing that it was no longer economical for Senegal to process groundnuts and that import substitution for high cost food was essential. The CPP also reported that the Bank had expressed willingness to consider financing the barrage at Diama -- "the economic justification for the OMVS programme should and could rest basically upon agricultural development." The 1978 CPP echoed this argument, suggesting that large-scale irrigation was "the most certain and least cost way." However, the more detailed 1979 country economic memorandum76 underlined the basic problem. Returns to irrigation 74/ World Bank, Senegal: Agricultural Sector Survey, Report No. 910-SE, November 3, 1975. 751 Ibid., Volume I: Main Report, pp. 33-34. 76/ World Bank, The Economic Trends and Prospects of Senegal, Volume II: The Agricultural Sector, Report No. 1720a-SE, December 1979. - 90 - schemes would continue, at best, to show marginal returns and an attempt at estimating domestic resource costs showed negative returns. Yet, at the same time, the report also emphasized the significant externalities from irrigation investment, not only from risk and drought reduction, but also on the rest of the economy. It was acknowledged that the inter-genera- tional benefits would not be captured in a static framework. The Senegal Valley irrigation scheme awould require investments over 30 to 40 years and more than double the agricultural production in a country with very few investment choices."77 195. This shift in favor of investment toward irrigated agriculture in rleuve had a limited translation into resource allocation by the Bank. The earlier River Polders project was followed by the Debi-Lampsar investment. Despite the latter project's generation of impressive yield levels and output growth for rice, the rate of return was negative and the combined fall in international rice prices and a failure to develop higher value crops, as well as a changing perception of the role for the parastatal agency, Socidtd d'Am4nagement et d'Exploitation des Terres du Delta (SAED), appeared to confirm the Bank's earlier reluctance to invest significantly in irrigation in the region. This judgement was based essentially on a short-run perception of the economic viability of rice production in Senegal and was not accompanied by any satisfactory examination of the longer-term potential for the subsector. 196. In other words, as was recognized in the 1979 country economic memorandum, a case for investment in large-scale irrigation works in the Fleuve appeared to warrant consideration but could not be readily ground out of conventional cost-benefit analysis. On the one hand, certain demo- graphic and economic constraints indicated a need for a different approach. For a start, nearly three-quarters of the population was in the agricul- tural sector. Labor market imperfections were profound through the economy. The agricultural sector, moreover, was marked by declining rain- fall trends across the country, increased output variance and risk and an absence of viable investment alternatives. The potential for productivity growth in the core rainfed areas was accepted to be minimal. On the other hand, however, the fact of falling border prices for the main staple to be produced under irrigated agriculture and limitations on the potential for crop diversification yielded unacceptably low rates of return. In the event, the Bank stayed out of any major involvement in the development of the Fleuve. Yet, it can be argued that the rather mechanistic application of cost-benefit techniques was shortsighted. It did not give sufficient weight to the severe degree to which output constraints are binding in other parts of the agricultural economy. In addition, no apparent atten- tion was paid to estimating the effects of different exchange rate scenarios on the long-term viability of such an investment. By staying out, the Bank also lost leverage over the type of irrigation investment and its subsequent development. 77/ World Bank, The Economic Trends and Prospects of Senegal, Volume 2: The Agricultural Sector, December 1979, p. 19. - 91 - (b) Technology 197. Inconsistency in policy advice during the 1970s was not limited to the issue of diversification and rice production, in particular. Perhaps most striking have been the set of shifts in attitude by the Bank toward the means for achieving technical advances in agriculture, the pricing of inputs and outputs and the role of parastatal agencies in the agricultural sector. At the outset of the 1970s, the Bank defined one key feature of its sectoral strategy as raising groundnut and millet productivity. This dovetailed with the Government's program for the core Groundnut Basin. One consequence, already mentioned, was the dominance, prior to the late 1970s, of the Groundnut Basin in the regional profile of Bank projects. The main mode of achieving productivity gains was through enhanced use of fertilizer, animal traction and better cultural practices. The means to this end was seen as the provision of credit, with liquidity and savings constraints the principal barrier to a technical shift. The Agricultural Sector Survey of 1975 also rightly emphasized the degree to which labor was the most binding constraint and that labor-substituting technology was critical in the Groundnut Basin. To the extent that the adoption of animal traction raised labor productivity and allowed for extensive growth, the Bank's view was that productive capacity was enhanced by the dissemination of farm equipment. At the same time, it tended to be assumed that fertilizer application could have a strong, positive effect on yields. These assumptions became increasingly open to question following the experience with a number of production-related projects. The first Agricultural Credit project, which aimed to mesh credit availability with increased adoption of machinery and fertilizer, fell considerably short of appraisal estimates. The project was closely articulated with the Govern- ment's "Programme Agricole" and was subject to many of the problems -- institutional weaknesses, inadequate technical data and packages, the effects of drought -- that were associated with the latter. In a context of exceptionally poor weather conditions, it was found that the extended farm implements had no impact on output levels. Moreover, as regards the impact of fertilizer, widely differing opinions came to be held over appropriate levels of application. There is no evidence that Bank officials were able to provide an informed judgement and the follow-on project (Cr. 404-SE) arbitrarily dropped the fertilizer component on the grounds that the BNDS could finance additional requirements. The emphasis was re-aligned and focussed on extending animal-drawn implements. Using limited field data, it was estimated that through earlier and more timely land preparation and seeding, a 1Z increase in the use of hoes could engender a 1.5-3.5 kgs/ha yield response. Nothing was done to resolve satisfactorily the basic agronomic questions regarding yield responses to a range of inputs. Yet, the Sine Saloum project essentially returned to the format of the first project, and yet again the primary objective of raising crop yields was not achieved. The PPAR noted that earlier experience meant that Bank staff already knew that with erratic rainfall, degraded soils, institutional weaknesses, inappropriate price relationships and a techno- logical package that was unacceptable to the majority of farmers, the yield and production response would be very limited, at least over the short term. - 92 - 198. The inability to develop adequate technological packages that could be extended to producers can, in part, be attributed to insufficient agricultural research and, as with irrigation, an excessive pre-occupation with short-run results. In this context, it should be noted that the Bank did not direct resources to the Senegalese agricultural research effort, except in a very limited provision of vehicles and equipment. It was not until the Agricultural Research project of 1981 that this was rectified.78 Secondly, discussions of yield responses to fertilizer tended to be based on very limited information and sample data. It is worth noting that average fertilizer applications to the major crops -- groundnuts and millet -- remained at very low levels even at the height of the 'Programme Agricole." It would therefore have been surprising if the aggregate output response had been significant. 199. Problems in the viability of the technical packages also have to be attributed to incoherence in pricing policy. This can be viewed not only in relation to relative input and output prices but also with specific regard to subsidy policy. In the latter respect, there is clear evidence that the Bank's approach to input subsidies was inconsistent. While a recurring theme in Bank documents was the need for more emphasis on ecolog- ical differences and targeting of fertilizer use to areas where returns were higher, there was a distinct shift in the 1970s over input pricing policy. Input subsidies attracted qualified support where "infant industry" arguments were used as justification.79 But as fertilizer subsidy costs mounted, with deleterious public finance implications, and in the absence of a major output response, this approach altered. The 1979 CEM (Volume II: The Agriculture Sector) came out against generalized input subsidies, but this was partly contingent on alternative measures for hedging risk, such as crop insurance. In the absence of the latter, subsidies could be viewed, within narrow and regionally targeted bounds, as a second-best solution to reducing the gap between private and social profitability. The principal spur to this shift in emphasis appears to have come from budgetary considerations. Nevertheless, it should be noted that both with regard to the means for obtaining productivity growth and the pricing rules for those means, the Bank's advice changed radically when comparing the post-1930 period with the period now under discussion. (c) Producer Prices and Agricultural Taxation 200. It is important to note that prior to 1980 the Bank paid relatively limited attention to agricultural pricing policy even if its later -- post- 1980 -- analyses emphasized the adverse, disincentive effects of controlled and discriminatory producer price levels, particularly for groundnuts. While sectoral policy analysis did point to the desirability of producer price increases for groundnuts and other major productions in the mid- 78/ See Jammeh and Lele, op. cit. 79/ The Current Economic Situation and Prospects of Senegal, June 8, 1970, Volume III: Agriculture, para. 219. - 93 - 1970s,80 there was little analysis of what should be the appropriate price level for individual outputs, let alone for relative prices. Despite the fact that it was assumed -- rather unrealistically, given other available evidence on supply elasticities in sub-Saharan Africa -- that the price elasticity of groundnut supply was high -- of the order of 0.5 -- no Bank studies were undertaken that attempted in any systematic way to measure specific crop elasticities, let alone cross-elasticities. It was assumed that producers were price responsive -- particularly following the response to the 1968/69 groundnut price decreases -- but the degree of responsiveness was left indeterminate. Likewise, the Bank failed to provide any satisfactory policy advice to Government regarding the optimal groundnut price level, given budgetary constraints and producer welfare considerations. Nevertheless, it should be noted that Bank officials accepted that budgetary considerations should precede producer welfare in the determination of the crucial groundnut price.81 201. Throughout the 1970s Senegalese farmers remained implicitly taxed through controlled prices and, in particular, by groundnut producer prices that averaged only 43Z of the unit export price. In 1970 it was estimated that the average income of groundnut growers was a mere 25% of that of the mean national per capita income and under 10 of that for the Dakar area.82 With substantial consumer subsidies -- particularly for rice, bread and sugar -- income transfers towards the urban sector existed. In part as a consequence of Bank pressures, the Senegalese Government abolished most consumer subsidies in 1974. But there is no evidence that the Bank pursued any consistent policy regarding desirable taxation levels for the sector, let alone defining the means for achieving such levels. Furthermore, this failure to deal with producer price policy issues -- in part a function of the limitations of project lending -- led to anomalies in relative price levels. 202. Bank-supported prejects were consistently compromised by price structures that were inconsistent with project objectives. For the First Agricultural Credit project (Ln. 584-SE), inputs adoption was well below target on account of a major fall in producer prices that raised the former's relative price. In the case of the Second Agricultural Credit project (Cr. 404-SE), the PPAR notes that "the project did not pay sufficient attention to the key factor: the producer price." The real groundnut price fell by 20% over the course of the project and was associated with a decline in the cultivated area and in input demand.83 203. The Casamance Rice project (Cr. 252-SE) indicates further not only the disincentive effect of a low rice price on producers in the project 80/ Senegal: Agricultural Sector Survey. 81/ E.g., CPP, May 28, 1976. 82/ The Current Economic Situation and Prospects of Senegal, Volume I, p. 19. 83/ World Bank, OED, PPAR: Senegal Second Agricultural Credit Project, Report No. 3514, June 25, 1981, p. 11. - 94 - area but also the Bank's inability to provide coherent advice. The PPAR points out that at appraisal in 1976 the Bank endorsed a farm-gate price that was even below Government fixed levels: less than a year later the Bank was strongly pressing the Government for a significant producer price increase for rice. As with the Groundnut Basin projects, inattention to relative price levels led to low input adoption rates and below-target production outcomes.84 Attempts under the two Terres Neuves projects (Crs. 254-SE and 578-SE) to introduce cropping diversification were likewise stymied by a relative price structure that yielded significantly higher returns to groundnuts than to other competing crops.85 204. A major consequence of the Bank's inability to deal with pricing issues in this period was thus to weaken project performance commonly through restricting inputs use and compromising the potential for any productivity effect. While the scale of the disincentives effect on production is unclear, price and marketing controls constrained officially marketed releases and together with other factors such as the recurring droughts, contributed to out-migration from the sector. High implicit taxation and transfers to Government had clear sectorRl and regressive income distribution implications. (d) Institutions 205. It is now generally accepted in the Bank that a key factor in limiting the effectiveness of projects was the poor performance of Senegalese institutions, particularly the parastatal agencies that were central to credit and inputs distribution, as also marketing and extension. Their chronic inefficiency and lack of adequate financial controls also had adverse public finance implications. Institutional rigidities are thus cited as one of the principal brakes on agricultural development.86 This was not a view that was generally accepted by the Bank in the 1970s. Indeed the Bank was, in some cases, instrumental in creating and supporting those very institutions that are now roundly excoriated. 206. The growth of state involvement in the provision of services and in agricultural marketing dates to the 1960s.87 The base of the system was the agricultural cooperatives. These were responsible for marketing and credit distribution at the village level. The main marketing and input 84/ Inappropriate relative price levels and returns to crops had similar effects on other donor projects. USAID's Cereals Production Projects in the Groundnut Basin achieved very limited millet output and productivity effect, largely because returns to land and labor for groundnuts were double those for millet in financial terms (see Jaeger, 1986, pp. 62 ff.). 85/ World Bank, OED, Impact Evaluation Report: Senegal Settlement Project in the Terres Neuves Region, Report No. 5170, June 29, 1984. 86/ See, for example, "Senegal: Agricultural Sector Strategy Brief." 87/ For a detailed treatment of the issue, see Abt Associates, "Senegal: Agricultural Policy Analysis," 1985, Cambridge, Massachusetts (mimeo). - 95 - distributions agency was ONCAD, created in 1966, while credit was distributed through BNDS. In addition, by the late 1960s a number of RDAs and the more project-specific Socidtds d'Intervention had come into existence. The RDAs -- such as SAED, established in 1964 or SODEVA, established in 1968 -- were originally crop-specific institutions, but this changed significantly in the course of the 1970s. The fact remaining that by 1969 the Senegalese state had established a comprehensive and intensive presence in the agricultural sector. These interventions were justified on political grounds (e.g., establishing Senegalese control over key sectors) just as much as on economic or organization grounds. Such justifications were not challenged by the Bank at the time, or indeed until the late 1970s (see para. 147 in Chapter II). 207. The Bank's approach to the parastatal sector was not, however, undifferentiated. In general, it was recognized at an early stage that ONCAD's inefficiency was a major barrier to project and sectoral performance. The First Agricultural Credit project was marked by weakiess in the institutional framework through reliance on ONCAD inputs. The failure of the latter's reorganization program jeopardized the project and shifted Bank officials toward favoring greater reliance on RDAs -- in this case, SODEVA -- and farmer cooperatives. Similar problems with ONCAD continued to affect Bank and other donor-funded projects, as with the Casamance Rice project and USAID's Casamance Regional Development project (see Jaeger, MADIA, 1986). Even so, the Agricultural Sector Survey of 1975 noted that ONCAD's performance had improved over the last five years, an opinion not shared in the CPP for 1976 where the solution to ONCAD's inefficiencies was posed in terms of transferring functions to RDAs or the private sector. The latter option was not seriously explored. The Bank's approach toward the RDAs proved variable. In Casamance, for example, the Bank strongly urged the creation of a public rice authority. "The major single handicap to the development of Casamance is the lack of an efficient, business-like development institution, similar to SODEVA.. .the Bank will contribute to the prompt establishment of a satisfactory institution" (CPP, 1974). A more general, positive perception of RDAs was held within the Bank. It was generally accepted within the Bank that traditional line Ministries, particularly the Ministry of Rural Development, were unable effectively to perform development operations and hence were to be allotted ancillary roles (e.g., the Agricultural Sector Survey, 1975). This approach in part derived from the Bank's emphasis on projects, with the limited nature of its policy dialogue and the accepted opinion of the time that integrated rural development institutions were desirable. It is clear that the RDAs were largely donor-driven institutions successful at attracting external financing but also lacking adequate financial and administrative autonomy with consequent problems of counterpart funding. 208. The sharp windfall gains from phosphate exports and high groundnut prices between 1972/73-1975/76 further accelerated the growth of the para- statal sector. By 1976/77 there were seven public enterprises and six - 96 - mixed companies operating in the agricultural sector. It was this sharp growth in the size of the sector and in the size of its subsequent claims on the exchequer that provided the main impetus for a change in the Bank's policy to the parastatals. It is notable that the 1977 para-public sector report, while commenting on the 9 billion CFAF deficit of the sector, also argued that "it was not abnormal within the limits of public finance resources for a growing nation to invest in productive projects through public enterprises and mixed companies." The report went on to conclude that the Government's strategy of establishing an RDA in each region was basically sound as such agencies "enjoy a reasonable degree of autonomy that enables them to operate outside of cumbersome financial and procure- ment procedures" (Annex 3, p. 104). This view was in part conditioned by the generally positive appraisal of SODEVA during its participation in the Second Agricultural Credit project. 209. The Bank's favorable perception of the performance of a number of the RDAs -- particularly SODEVA and Soci6td de DOveloppement des Fibres Textiles (SODEFITEX) -- was increasingly contradicted by direct project experience. The River Polders project (Cr. 350-SE) was marked by weak institutional performance from SAED, while the la*rr Sine Saloum project (Cr. 549 and Ln. 1113-SE) was constrained not only by ONCAD's usual inefficiencies but also by problems with SODEVA, the executing agency. Poor quality extension work, over-centralization of management, a premature proliferation in SODEVA's wider functions and its lack of an independent financial base were cited as the chief issues (PCR, pp. 42-43). By the end of the 1970s, with the exception of the Casamance Rice project, the Eastern Senegal Livestock project and the Terres Neuves initiatives, direct Bank experience with the Senegalese parastatal sector was uniformly poor. Even when agencies performed well -- as with STN for the Terres Neuves projects -- this raised the issue of their possible role following project completion.88 210. It appears that Bank policy to the parastatal sector in the 1970s was somewhat shortsighted and driven by limiting concerns, particularly the need to circumvent the traditional Government agencies in the execution of projects. The approach to the RDAs also echoed the more general Bank enchantment with integrated development projects in this period. The Bank made no serious attempt in its sectoral analysis or policy dialogues to promote greater private sector involvement nor did it seek to propose any strategy not contingent upon an expanded parastatal sector. A failure to approach these issues methodically was combined with a late and partial appreciation of the negative budgetary implications associated with the growth in the parastatal sector. Inadequate appreciation also existed regarding the degree of atrophy in the vital, constituent parts of the Government's "Programme Agricole." These were already apparent by the mid- 1970s but there is little evidence to suggest that the Bank was able to play a constructive role in its dialogue with Government for determining 88/ OED, Impact Evaluation Report, Senegal Settlement Project in the Terres Neuves Region. - 97 - the parameters of a modified system. The abolition of ONCAD in 1980 was an action taken autonomously by Governments for the most part the Bank was still proposing reform of the institution or selective divestiture of its functions. The prior effective collapse of the delivery systems for inputs and credit had likewise called forth no serious response on the part of the Bank. This meant that the modified policy framework adopted by the Bank in the 1980s was strongly pre-determined by de facto conditions, by acute financial constraints and, importantly, by a shift in ideological preferences. (e) Conclusion 211. Bank involvement with Senegalese agriculture between 1969 and 1979 appears to have been constrained by the failure to map out any longer-term strategy for the sector and hence for the directicn of Bank resources. On the key diversification issue. Bank analysis tended to focus too exclusively on short-run options. While it was commonly argued that groundnut export potential was held back by exchange rate overvaluation and demand constraints, there is little evidence that the Bank seriously evaluated policy options under a range of scenarios, including exchange rate adjustment. This limited the utility of the sectoral policy advice that the institution could offer. Furthermore, in dealing with the groundnuts market, Bank analysis paid insufficient attention to the high processing costs of the Senegalese oil mills that were subsidized by the State. Neither was attention given to the issue of market structure, particularly the dominance of France in the market for groundnut oil and, more specifically, the dominance of one group -- Lesieur -- in the purchase and marketing of the Senegalese product. On the question of irrigated agriculture, the Bank failed to develop any consistent approach or, more tangibly, to decide whether to support in a significant way the institu- tions, particularly OMVS and SAED, that were responsible for extending the irrigated area in the Fleuve. At times emphasizing irrigation potential in the Casamance, at other times more favorable to the Fleuve; the end result was incoherence in Bank advice and project lending. 212. The inability to generate a sustainable agricultural sector strategy can, in part, be attributed to grave data inadequacies and the failure to improve the basic information systems in Senegalese agriculture. Lack of sufficient micro-level information regarding production and distributional attributes led to attempts to apply inapposite and inadequately differentiated technical packages. Consequently, the response rate by producers to the technical packages offered by the extension agencies and by Bank-supported projects remained poor throughout this period. In some cases, project outcomes involved non-intentional distri- butional effects on account of informational inadequacies. In the case of the River Polders project (Cr. 350-SE) reliance on SAED determined production units -- the "groupements de producteurs" -- resulted in an inequitable direction of resources to a limited stratum of dominant groups in the Senegal Delta. The PPAR noted that these farmer organizations merely reinforced existing social structures with one result being that - 98 - merely 14Z of project participants belonged to lower income groups.89 For the Second Sedhiou project (Cr. 647-SE), despite the fact that part of the predecessor, Casamance Rice (Cr. 252-SE), project's strengths was the ability to draw in female rice growers, credit was almost exclusively directed to male heads of household due to the non-eligibility of women for cooperative membership. This was associated with lower efficiency in the use of inputs made available through the project.90 213. In marked contrast to the Bank's focus post-1980, surprisingly little attention was paid to pricing policy and the rate of agricultural taxation. This was despite the recognition that "about 30Z of the farm population needs income from other sources to avoid absolute poverty."91 Although the disincentive effect of implicit taxation through controlled farmgate prices has subsequently been overstated, the welfare costs were non-negligible. Amongst other consequences, low producer prices tended to imply heavy inputs subsidies and low general cost recovery by public agencies. While it can rightly be argued that the inability to deal with such wider policy problems jeopardized the performance of projects, it should also be noted that the Bank's interest in taxation and pricing policy only developed towards the end of the 1970s. It should also be emphasized that the Bank's leverage on the Senegalese Government through the conventional project lending format was rather limited. This resulted in an inability to raise wider pricing and taxation policy issues satisfactorily with the Government. 214. As regards institutional development, this section has demonstrated the key role the Bank played in supporting the creation and consolidation of the RDAs. Institutions -- such as the Ministry of Rural Development -- which post-1980 did attract Bank support were consistently bypassed. The grave institutional inadequacies of ONCAD were skirted in the hope that the new RDAs would suffice. Given the continuing importance of ONCAD in the agricultural sector, this hope was unrealistic. Growing disillusion, in turn, with the RDAs led to a withdrawal of support from the Bank. In this respect, institution-building by the Bank remained limited and fitful. Despite these problems, it ought to be emphasized that the 1970s was a particularly harsh decade climatically. Weather remained the prime deter- minant of sectoral performance. Moreover, there were a number of signifi- cant advances. The development of cotton in Eastern Senegal and the performance of SODEFITEX were very positive. Results from the first Livestock project illustrated the potential of the subsector, while area growth and crop diversification were fairly marked in Casamance. In the Fleuve yields of between 4.5 and 7 metric tons per hectare for rice were attained. The "Programme Agricole" did result in a major technological 89/ World Bank, OED, PPAR: Senegal River Polders Project, Report No. 2777, December 26, 1979, pp. 9-10. 90/ World Bank, OED, PCR: Senegal Second Sedhiou Project, Report No. 5122, June 11, 1984, pp. 19-20. 91/ Senegal: Agricultural Sector Survey, p. 11. - 99 - shift -- unique in West Africa -- with the widespread introduction of animal traction and new equipment. One direct consequence was extensive growth in both Eastern Senegal and the south-eastern part of the Groundnut Basin. Though not associated with a yield effect, the dissemination of agricultural material loosened the critical labor bottleneck and raised labor productivity. Lastly, the experience of this period illustrates the limitations of project lending -- the inability to correct larger *distor- tions," as also the restrictive implications for a broader coherence in sectoral lending. (3) Structural Adjustment and Agriculture, 1980-87 215. The shift away from project lending to a more concentrated use of program lending instruments cannot primarily be attributed to the intract- ability of agricultural sector issues under the former framework. While loan conditionality has increasingly been used to correct or weaken the range of constraints -- such as producer pricing and taxation policy -- that adversely affected project performance in the 1970s, the principal factor behind the shift in lending instruments was the rapid deterioration in Senegal's external accounts. 216. Program lending had, in fact, emerged as a potential need as early as 1974. The CPP of that year proposed a program loan of $20 million linked to agreement with the Government regarding food subsidy levels and recurrent expenditures. The phosphates windfall temporarily relaxed the need for balance of payments support and by 1978 it was held -- despite the collapse of the phosphates boom -- that program lending was not suitable for Senegal (CPP, 1978). This view was reversed following the Government's preparation of the PREF in 1979 and the approval of the first SAL in 1980. Since then, two further SALs have been approved. In addition, six further agricultural sector projects have been approved. These allocations have amounted to around 45Z of the total resources released under the three SALs. While the latter have been multi-sectoral, agricultural sector issues have featured prominently. Nevertheless, the "cost" of buying sectoral reforms has been high relative to the level of Bank-financed direct, productive investment in the sector. 217. The agricultural sector components of the SALs reflect, on the one hand, a continuing emphasis on the basic questions that have faced Senegalese agriculture: diversification, the role of groundnuts, 4he nature of the technical packages for raising productivity and the framework for the delivery of goods and services to the sector. Though in part continuative from the Bank approach in the 1970s, the adjustment period has been marked by a qualitative shift in emphasis and by a far greater level of enchantment with prices, markets and the private sector. Given the nature of Senegalese agriculture, this enchantment has at times failed to yield satisfactory solutions. - 100 - 218. Despite the fact that disbursements under the first SAL were terminated -- in June 1983 -- largely because of the failure to implement agreed-upon agricultural sector reforms,92 the general compliance of the Government and the scale and content of the measures taken, make Senegal an exceptional case in the context of other Sub-Saharan African economies. 219. The three SALs have been marked by continuity in their emphasis on pricing policy, particularly as a means for reducing urban/rural income disparaties through major producer price increments, as also the reduction of input subsidies, and the parallel emphasis on reforming the RDAs. In the latter context, this has involved a mixture of closures, budgetary and staffing reductions as well as a clear definition of the financial relationship between Government and the parastatals. The latter has largely been done through program-contracts, a system introduced by the Government with advice from the French. 220. Despite some basic continuities, it is clear that there have been significant shifts in Bank policy not only as regards the parastatals, but also with regard to the key diversification issue. This latter shift can be dated to the period be,.ween SAL I aad SAL II. While the former still mainly emphasized groundnuts with little attention paid explicitly to the Zood crops subsector, SALs II and III have strongly promoted a policy of import substitution for domestic cereals, partly through tariffs on rice imports, partly through relative domestic price shifts in favor of food crops. This has been further formalized in the Government's Nouvelle Politique Agricole (NPA), announced in 1984, and the Plan Crdalier that was released in 1986. 221. The NPA comprises four main aspects. First, emphasis is placed on "la responsabilisation des paysans," through a transfer of functions -- such as groundnut seed storage -- away from parastatals to producers. The beginnings of reform in the government-controlled cooperatives system can also be noted. Second, inputs pricing has been revised, with acceptance under SALs II and III of final elimination of subsidies on fertilizer by 1989/90. Liberalization of input imports has been accepted and, in theory, a transfer of input delivery functions from the parastatals to the private sector has occurred. Third, the rural parastatals have had their functions reduced and their budgetary allocations cut. Fourth, the NPA emphasized measures to attain higher levels of food self-sufficiency. In this regard, the Plan COrdalier has specified a target of 80Z self-sufficiency by 2000. Although the Bank has not wholly endorsed the policy objectives of either the NPA or Plan Cdrdalier,93 it is clear that the bulk of conditional actions specified under SALs II and III accord with the general orientation of the Government. 92/ See, OED, PPAR: Senegal Structural Adjustment Loan and Credit. 93/ See, for example the Bank-drafted Preamble to the Plan Crdalier, Ministry of Rural Development, Dakar, 1986. - 101 - 222. Any assessment of the efficacy of adjustment lending in Senegal has to contend with the fact that most of the policy measures have been very recent and the outcames consequently still ambiguous. However, some of the key measures taken under SAL I have had time to filter through. Moreover, assessment can be made of the likely effectiveness of the broader, strategic orientation of the Bank's actions in Senegalese agriculture in this period. (a) Producer Prices and Supply Response 223. In the adjustment period greater attention has been paid to pricing policy as a means for stimulating market releases and shifting income in favor of rural producers. Under the first SAL, groundnut prices were raised by 102 in nominal terms between 1979/80 and 1980/81. Prior to the second SAL (1985/86). producer prices were raised by 20-302 for millet and by 40-50% for maize and paddy. The impact of *incentive pricing' cn output levels is unclear. Fluctuations around trend have been so substantial that it is difficult to estimate satisfactorily the underlying direction of change. Annual area and output fluctuations have tended to reflect straightforward crop substitutioLs and climatic impact rather than area growth or, it would appear, productivity growth. For agriculture as a whole, this has meant that between 1980 and 1985, three years were marked by strongly negative growth rates and by negative overall growth. 224. Recent trends make growth projections under SALs II and III appear unrealistic. There, primary sector growth between 1986-90 of between 2.4-4.7Z has been projected where such growth would largely be generated by cereal output response. Aggregate cereal output would rise by one-third, with rice and wheat imports declining by at least 15Z.94 Groundnuts output would stabilize under this scenario at a level somewhat lower than chat attained between 1980/81-1982/83 when the groundnut area was roughly 102 higher than the average between 1980/81 and 1985/86. Obviously, any outcome will largely depend on climatic conditions. 225. These highly optimistic estimates appear to be at variance with the Senegalese historical experience since Independence. Value added in the primary sector grew at 32 in the 1960s, falling to around 2.3Z in the 1970s and to below 1.82 between 1979 and 1983. The latest CEM notes, moreover, that over the period 1979180-1985/86 agricultural output exhibited no trend other than that of stability in the planted area, itself implying a downward trend in the average cultivated area per head of rural population.95 Furthermore, as regards yields, those for groundnuts show a diminished mean over the past decade with higher variability. Where mean yields have increased -- as for millet, rice and maize -- this too has been accompanied by significantly higher variability around that mean. Only in the case of cotton have mean yields increased with reduced variability. 94/ President's Report, SAL II, pp. 26-27. 95/ World Bank, Senegal: An Economy Under Adjustment, Report No. 6454-SE, February 13, 1987, p. 10. - 102 - 226. Significant growth in agriculture hes been postulated in Bank documents since 1980 on mere price effects. While the new emphasis on the role of prices is highly desirable, evidence suggests that the effectiveness of price policies in the Senegalese context is bound to be limited as long as some other factors affecting the performance of agriculture -- climatic trends, soil degradation, and a weak research and extension effort -- are not being dealt with. There is a substantial body of literature which estimates very low short- and medium-run supply elasticities in the Sub-Saharan African context.96 For Senegal, the very limited econometric evidence available indicates the predominant influence of rainfall and the limited impact of prices on both groundnut and millet output. Thus, Berthdlemy97 shows that up to 1979 the variations in groundnut yields and cropped areas could be explained by climatic and technical conditions. Rainfall is represented by two variables, i.e. quantity and concentration with the latter variable having a very significant coefficient, proving the need to take account of rainfall distribution wherever the necessary data are available. Fertilizer use improves yields although the elasticity is very modest, a 10Z increase in fertilizer consumption bringing about only slightly less than a 12 increase in average yields. Finally, the use of seeders has a marked adverse effect because of more rapid soil exhaustion linked to the reduction of fallow. Against this, the stock of seeders has a beneficial effect on crop areas as this machinery enables the farmer to increase the area of land in production. The real price variable did not appear to have any significant effect on the cropped area. 227. Berth4lemy attempts to explain millet yields and cropped area through two different equations. With regard to yield, he finds that the two crucial variables are fertilizer and, more especially, rainfall. Although significant, the yield elasticity in response to fertilizer is very modest at only 0.05. Turning to the determination of the cropped areas, it appears that the previous year's yields and the stock of avail- able seeders influence it. The stock of seeders is taken as an indicator of available working time. The more seeders available for groundnut cultivation, the more time the farmer has to devote to food crops. Lagged millet yields have a negative effect on output inasmuch as a good harvest enables stock to be built up, thereby reducing output requirements the following year. Interestingly enough, all of these variables appear to be statistically significant in explaining output, in sharp contrast to the price variable. With respect to rice production, yields depended largely on rainfall and only marginally on fertilizer consumption -- a 10% variation in the level of rainfall increasing yields by 6% whereas the same change in the quantity of fertilizer per hectare had only a slight effect on yields of 12. Again, the price variable did not seem to affect output. 96/ See, for example, Marian Bond, "Agricultural Response to Prices in Sub- Saharan African Countries" IMF Staff Papers, 1983. 97/ J.C. Berthdlemy, "Construction d'un mod6le macrodconomique pour un pays d'Afrique francophone, le S6n6gal," Paris, 1979. - 103 - 228. Given Senegalese production structures and the complementarity of groundnuts and millet, in particular, in the rotational cycle, the key question concerns the likely substitution effects of relative price changes. Estimating the price cross-elasticity for groundnuts and millet supply is hampered, however, by lack of data on labor and capital inputs. Reduced estimations with prices lagged one period and with average annual rainfall for the period 1961-86 yielded ambiguous results.98 229. The results indicate that a 12 millet price increase would reduce groundnut supply by 0.82. The direct price elasticity of supply for groundnuts was estimated at 0.54 and with respect to rainfall, 1.19. The price elasticity is, in fact, high but, it must be noted, not statistically significant. At a 95Z confidence level the 0.54 estimate would be plus or minus 0.88, indicating the lack of robustness in the estimate. Only the rainfall variable was significant at a 95% level. Neither price variable was significant and was also sensitive to the time specification that was employed. The overall explanatory power of the equation was relatively low. Other estimations likewise point to the importance of rainfall in explaining area and yield. 230. Producer prices thus have only a limited role in explaining area and output levels. Moreover, in the case of millet the level of market releases has remained low, averaging no more than 15-202 of aggregate output. The bulk of production tends to be consumed on farm. Incorporating considerations of market inefficiency, risk aversion and rotational factors, it can further be shown that in an effective joint- production framework, the output of the cash crop in that joint production function in effect is defined as a residual where food stocks and on-farm consumption requirements are the determining constraints.9 As millet output accounts for roughly 752 of total cereals production, this clearly implies significant problems for any strategy based on raising its marketed output level to a major degree. 231. The limitations of price policy in stimulating output must necessarily pass doubts on a strategy that largely emphasizes prices without complementary attention to other variables, such as technology, irrigation and other factors, including land tenure.100 Moreover, it is not evident that Bank advice, which has increasingly favored the food crops 98/ Groundnut Marketed Output = 5.319 + 0.539 (Net Groundnut Price Lagged) ((1.255))) -0.812 (Net Millet Price Lagged) ((-1.64)) + 1.19 (Rainfall) ((3.21)): R2 = 0.55; F = 8.64 (T - statistic in double brackets. Equation specified in log-linear form.) 99/ See, inter alia, J. Hammer, "Subsistence First: Farm Allocation Decisions in Senegal," World Bank, CPD Discussion Paper 1984-19, June 1984, (mimeo). 100/ In the latter regard, for example, rights to land are an important issue (as recognized in the SAL III document) from both a distributional and investment perspective, particularly in the Fleuve. - 104 - subsector, has been adequately translated into a relative price structure that would favor those goals. For millet and groundnuts, real producer prices have not shown any clear, consistent trend after 1980. Moreover, since 1979180 the ratio of groundnut to iillet producer prices has increased somewhat over that which prevailed between 1975-79. However, over the longer term, despite annual fluctuations, this ratio has remained broadly constant at between 1.20-1.25 for the period 1970-86. For the groundnut/maize price relationship, the ratio of the former to the latter shifted from 1.12 between 1975-79 to 1.31 between 1980-86. 232. If one looks beyond this relative price relationship at actual net returns to the major crops, it appears that returns to groundnuta in the core producing areas have been consistently higher than for other possible substitutes. Farm budget data show that in 1986/87 net returns per man-day in three states of nature and under two current technology regimes are considerably superior for groundnuts relative to millet, but inferior relative to maize. For the critical millet/groundnut relationship, it appears that in an average year and at existing technology levels the return to the latter is between three and four times that for millet.101 233. The relative price structure that has thus faced Senegalese producers has sent signals that have not adequately reflected some of the apparent policy goals advocated by the Bank. At the same time, the key groundnut price has been fixed at an unsustainable level, given the strong downward trend in international groundnut prices. The 33Z real price increase that was introduced in 1985 coincided with a world price fall that left the price of unrefined groundnut oil in 1986 at around 45Z of its 1984 level. Consequently, for the crop year 1986/87 the deficit on groundnut operations amounted to around 1.4Z of GDP, with effective price support running to over CFAF 11 billion. This groundnut producer price level was sustained until April 1988 with both Bank and Fund encouragement, as a means for maintaining adequate incentives for producers. By 1988, however, budgetary pressures led to a reduction in groundnut producer prices from 90 CFA/kg to 70 CFA/kg. Apart from underlining the need for a concerted and systematic study of appropriate longer-term pricing rules and relative price levels (a short study has been commissioned as part of SAL III), recent events illustrate, first, the strong external constraints facing agricultural sector policymakers; second, the inability of the Bank to provide an adequate framework for determining domestic price levels and relationships. 234. The Bank's emphasis on raising groundnut producer prices closer to border price levels can be related not only to the argument of a supply response but also to the objective of reducing the implicit taxation rate. Yet, since the first SAL, it is notable that between 1980/81 and 1984/85 the producer price as a share of the unit export price remained at the same average level -- 40-45Z -- as in the 1970s. It was only in 1985/86 that this was boosted to over 80Z. Despite this increase, the groundnut area 101/ F. Martin, "Budgets de culture au SOndgal," ISRA/BAME, Dakar, 1988. - 105 - fell sharply with output over 202 below the average level over the previous five years. This can largely be explained by the abrupt changes in the groundnut seed storage system in that year. While this indicates the significance of non-price factors in area allocations, it ought to be noted that in subsequent years the groundnut area has again expanded, even if the long-run trend clearly favors growth in the millet area.102 235. The problems in any largely price-driven set of interventions occur not only through low marketed supply response, with factors such as weather and technology being important explanatory variables, but also in terms of the complex implications for the balance of payments, the Government budget and producer welfare. This comes to the fore in any assessment of the Bank's approach to cereals pricing. 236. To stem the importation of rice, SALs II and III have required the Government to maintain a minimum nominal protection coefficient of 252 in setting the domestic consumer price. The objective is to stimulate local cereals production -- principally millet. Consumption of local cereals has declined on account of low relative prices for importables. By 1982-84 rice and wheat (flour) accounted for nearly 832 of annual per capita cereals consumption in urban areas and over 30% in rural areas. Consumer nominal protection coefficients and relative nominal protection for the major cereals show quite clearly how for the period 1970-75 and again between 1979-83 rice consumers benefited from a significant subsidy. However, Table 23 also indicates that, assuming millet/sorghum to be a tradable, Senegalese domestic production generates outputs at price levels very considerably higher than the equivalent import parity cost. I U I U 102/ H. Gaye, Le Ddsengagement de 1'Etat et la Probldmatique des Intrants Agricoles au Sdndgal. ISRA, Kaolack (mimeo), May 1987. - 106 - Table 28: CONSUMER NOMINAL PROTECTION COEFFICIENTS AND RELATIVE NOMINAL PROTECTION FOR CEREALS, 1970-95 Le Wheat Rice/ Wheat/ Period Millet Rice Grain Millet Millet 1970-76 1.81 0.89 0.61 0.52 0.89 1978-78 2.01 0.97 1.01 0.50 0.52 1979-81 1.87 0.75 0.98 0.40 0.56 1982-83 1.89 0.87 0.98 0.48 0.52 1984-85 2.06 1.04 0.84 0.60 0.41 /a Consumer Nominal Protection Coefficient defined as ratio of domestic market prices to Import parity prices, here unadjusted for exchange rate overvaluation. Relative Nominal Protection is the ratio of NPCs computcd on an annual basis and then averaged. Source: C. Delgado, 'The Role of Prices in the Shift to Rice and Wheat Consumption in Francophone West Africa,* IFPRI, Washington, 1987. 237. By granting protection to domestic cereal production, consumers are effectively subsidizing producers. This is justified on spatial and income distributional grounds on the assumption that most imported food is urban consumed. This is only partly true. Official data (see Plan Cdr6alier, 1986) show that around 28% of per capita cereal consumption in rural areas is of rice. Survey data also show that the level of rice purchases in rural areas is in direct inverse association with aggregate cereals output in any given year. On average, rice purchases amount to a little under 50% of total cereals purchased and 20% of aggregate cereals availability.103 The point to be appreciated is that most farm units are food-deficitary, hence in part explaining the traditionally low levels of cereals market release. Raising the consumer price for basic cereals thus has an ambiguous outcome in terms of farm household welfare. 238. A second major point is that in spite of significant increases in the rice consumer price, equivalent processed millet products -- such as flour -- are priced at levels ranging between 95-11OZ of the retail rice price. This suggests that even with 25% protection, current processing technology cannot transform at rates that could make millet more competitive vis-a-vis rice. Third, upward adjustment of the consumer rice price in a context of falling real import prices has resulted in very significant transfers to the Exchequer. Indeed, budgetary support was the principal reason for the Government's willingness to raise the rice price. 103/ M. Benoit-Cattin, "Hypothbses sur la Consommation de riz dans le Bassin Arachidier du Sdndgal," Montpellier, mimeo, 1987. - 107 - By 1987/88, this support amounted to CFAF 16 billion. The supply response from domestic cereals output emerges as a subordinate consideration, though one emphasized by the Bank. However, rice producer price increases could be expected to have implications for export revenues insofar as area substitutions of cereals for groundnuts obviously yield a reduction in the share of exportables in total output. This result obtains because, firstly, the expansion capacity for rice production, at least in the medium term, remains very limited and, secondly, marketed surplus remains low. Thus, the direct substitution effect and hence the impact on import costs would be restricted. If other crop substitutions are accounted for, then the main consequence would be a reduction in the area devoted to the exportables, cotton and groundnuts.104 In the case of consumer rice price increases, such as have occurred in the recent period, the same multi- market analysis suggests that the effect on exportables output would not be negative but would mainly affect domestic demand and the government deficit in agriculture. As most households are rice consumers this would result in generalized real income reductions. Gaye105 has estimated a cross- elasticity of millet output to the retail rice price of 0.2. A low cross- elasticity would imply that this income effect predominates. However, contrary to the assumptions used in the Bank analysis cited above, the groundnut price increases of recent years will have tended to offset part of this income effect for groundnut producers. The strongest adverse income effects would then fall on urban rice consumers and rural households not producing tradables. For all rural households, the precise income effect would depend largely on the extent to which farm households are net purchasers of food, an area in which the Bank and Senegalese Government remain uninformed. Recent developments indicate that, if the exceptional decline in the groundnut area in 1985/86, which can be attributed to changes in seed policy, is excluded, with combined real consumer rice price and producer groundnut price increases, the government deficit has fallen but at lower than expected rates. This is because of higher than projected deficits on the groundnut sector associated with exogenous price effects. World groundnut prices have fallen resulting not only in lower foreign exchange earnings but also higher implicit subsidies to groundnut growers. Even with the continuing drift toward larger millet acreage (a function of rainfall, risk aversion and marketing constraints) the assumption of a low cross-elasticity of millet output to the consumer rice price would appear to be warranted. Consequently, the agreed producer price decreases for groundnuts for 1988/89 will tend to lower net household income. This outcome will hold particularly if, at current relative prices, domestically produced cereals remain uncompetitive with rice when taking account of processing costs. Thus, in the present conjuncture of falling export prices for the major tradable and high cost domestic production of cereals 104/ A. Braverman and J. Hammer, "Agricultural Pricing in Senegal: Their Implications for Government Budget, Foreign Exchange and Regional Income Distribution,' World Bank, Country Policy Department, September, 1983. 105/ Matar Gaye, "The Food Challenge in the Senegalese Rural Economy," Michigan State University, (mimeo), 1983. - 108 - (the consumer nominal protection coefficient for millet exceeds 2) the policy option of raising producer prices no longer meaningfully exists. Further upward price adjustment would stimulate inflation more than supply. (b) Input Subsidies 239. The shift in emphasis by the Bank toward pricing reform as a means for raising output reflects the widespread disillusionment with the earlier attempts at introducing new technical packages. Indeed, the Bank's present approach marks a complete volte-face from the policy of the early 1980s. The 1987 Agricultural Sector Strategy Brief emphasizes soil and water conservation, as well as mixed farming, rather than reliance on modern inputs for productivity growth. This can, in part, also be viewed as a de facto rationalization of the collapse in modern inputs use -- particularly fertilizer -- as a result of Bank and other donor emphasis on eliminating fertilizer subsidies. Although USAID has financed the maintenance of a decreasing subsidy on fertilizer (to be eliminated by 1939/90), apparent consumption fell from over 102,000 metric tons in 1980/81 to around 21,000 metric tons by 1983/84 and 12,000 metric tons in 1986/87. Absence of credit since the collapse of the OProgramme Agricole' in 1980 and lack of financial viability have throttled off sales of fertilizer. At 1986/87 prices, for example, ISRA-recommended fertilizer application levels would only generate for millet a benefit-cost ratio superior to 2 on the assumption of a minimum 60% yield increase. Moreover, this is with an input price carrying a 25Z USAID subsidy element. In the absence of this subsidy, yield increases would have to surpass 80% to get a comparable benefit-cost ratio. In the case of maize, the respective yield shifts would need to be slightly lower, vis. 50% and 75Z.106 Such shifts are very unlikely in the Senegalese context. This implies a continuing depressed demand for such inputs, raising in the process considerable doubts regarding the potential involvement of the private sector in fertilizer import and distribution; another feature of the Bank's overall strategy for the sector. As seriously, these facts suggest that the broad strategy documents around which the Governme- 's policies, with Bank support, have been organized -- the NPA and, more specifically, the Plan Cdrdalier -- contain unrealistic projections. Under the latter, fertilizer consumption is aimed to reach 133,000 metric tons by 1990, rising to 220,000 metric tons by 2000. With improved seed and better cultural practices, yields have been projected to increase by around 25-30% for rainfed millet and by 90% for maize. For irrigated maize, yields would double and rice yields would rise by about 30? (Plan COr4alier, 1986). Clearly, such projections are unrealistic if placed in the perspective of productivity trends over the past quarter century. But they are particularly unrealistic given the current context of very high relative input prices, a non-existent formal credit system, low savings levels and an increasingly non-functioning extension service. On these critical ingredients for any sustained initiative for raising agricultural productivity, it is fair to say that 106/ See S.J. Commander, et al., "Senegal: The Experience of Adjustment," (mimeo), London, 1987. - 109 - the Bank has remained agnostic or indifferent. Yet, past and recent experience suggests that while price measures and market liberalization, to a limited degree, can change the output mix and the volume of marketed releases, with lower overhead costs following the limitation of State interventions in markets, these measures do not provide sufficient conditions for productivity growth. 240. The inputs subsidy issue raises a number of difficult questions. It would appear that at unsubsidized prices the consumption of yield- enhancing inputs will remain insignificant. Indeed, it is worth noting that even at the height of the "Programme Agricolew during the 1970s average fertilizer consumption at subsidized prices, including for the groundnut basin, remained at very low levels. Nevertheless, subsidies , at least for a period of time, can be an incentive to use inputs. The *Programme Agricole' did succeed in raising farmer exposure to and use of modern inputs. Moreover, it should also be noted that with heavy subsidies the "Programme Agricole" did effect a major technical change -- unique in West Africa -- the shift to animal traction, one key consequence of which was extensive growth. While no satisfactory case can be made for permanent input subsidies, the particularly binding constraints that operate in the Senegalese economy at this stage -- soil degradation and desertification, high climatic risk and an overvalued exchange rate -- suggest that selec- tive input subsidies are probably a prerequisite for productivity increases. The Bank's current generalized aversion to input subsidies may need to be re-examined. (c) Institutional Reforms: The Role of the RDAs 241. Central to the Bank's perception of a necessary reform agenda for the sector has been the need to reduce and overhaul the functioning of the parastatal agencies. This component has been particularly pronounced under SALs II and III. The principal spur to reform was the growing financial burden of the RDAs. By 1979/80 the net position of the parastatal sector had turned strongly negative, and the RDAs alone accounted for over 50Z of total operating subsidies to public enterprises. By selective closure (e.g., Socidtd des Terres Neuves), divestiture or through reductions in recurrent expenditures, the Government, with strong Bank support, has aimed to arrest this hemorrhaging of resources. The degree to which this has occurred remains difficult to estimate because of lacunae in the data. Nevertheless, Table 24 presents some disaggregated figures for a number of the major agricultural sector institutions. It can be seen that the subsidy toward the extension agencies remained over 40Z higher in real terms in 1984 than in 1981 but had slightly declined when compared with the 1982 and 1983 levels. Since 1984 it appears that subsidies have been further reduced. Moreover, actual expenditures for the extension agencies in 1985/86 were less than 40% of their 1981-83 levels in real terms. - 110 - Table 24: FINANCIAL POSITION OF EXTENSION AGENCIES WITH REGARD TO THE BUDGET AND DISAGGREGATED FIGURES FOR PARTICULAR RURAL SECTOR INSTITUTIONS, 1981-85 1981 1982 1983 1984 1985 (CFA billions) Extension Agencies /a Net Subsidy -2.9 -5.4 -6.7 -6.1 n.a. Subsidy 3.1 5.6 6.0 6.5 n.o. Taxes 0.2 0.3 0.8 0.4 n.a. (CFA millions) SODEVA Subsidy 1,578.6 1,439.8 1,698.4 1,925.2 n.a. Net Financial Result -55.7 -79.6 100.3 123.0 n.m. SODAGRI Subsidy 200.8 253.4 n.a. 270.0 36.0 Net Financial Result 11.1 31.7 1.2 -0.3 9.6 SODEFITEX Subsidy 18.9 351.0 85.2 198.2 170.6 Net Financial Result 24.2 100.8 119.2 41.9 112.6 SEIB Subsidy - 4,089.8 - 2,688.4 1,888.7 Net Financial Result -828.4 -1,575.3 62.6 -1,502.6 -2,029.7 SONACOS Subsidy 4,023.9 - - 3,489.3 - Net Financial Result 26.9 1,994.1 -447.6 4,489.2 943.8 /a Extension agencies include: SAED, SOMIVAC, SODEVA, SODEFITEX, SODAGRI. Source: Financial Comptroller of the Presidency, August 1988. 242. SAL II laid out as a specific objective the withdrawal of the RDAs from direct productive activities, with their role being limited to planning of rural development operations and extension. This has been supported, in the case of SAED, by an Irrigation Technical Assistance project (Cr. 1632-SE) aimed at assisting SAED's disengagement from production-related activity, as well as implementing its program-contract. The Irrigation IV project which was approved in 1987 likewise has as one of its main objectives the disengagement by SAED from its direct involvement in production. In this framework, the main function of the agency would be the provision of better extension and training services. In the case of - 111 - the extension agency for the core Groundnut Basin -- SODEVA -- the Bank's approach has been to withdraw project support. The Sine Saloum re- appraised project was cancelled in 1982 on account of the failure to implement experiments with the extension package and the institutional framework for delivering inputs, credit and marketing. The precarious financial position of the agency, which like the bulk of other RDAs has been largely dependent on donor financing, has subsequently entailed significant budgetary cutbacks and staff reduction. Between 1969 and 1979 SODEVA's personnel had grown from 758 to 1,882, principally through the growth in middle and upper management. In that period, the share of field workers -- extension agents -- had fallen from 88Z to 622, with the wage bill rising from 282 of the budget to over 602. Since 1980, SODEVA staffing levels have been radically cut. By 1986, a 55% reduction had been achieved. However, it appears that such reduction has fallen most heavily on the extension agents, minimizing yet further the functional capabilities of the organization. This may, however, be intentional. With the suspension of the "Programme Agricole" and equipment sales on credit and, subsequently, the collapse in fertilizer distribution, the basic extension framework for SODEVA collapsed. The Bank's current emphasis away from the use of modern inputs in rainfed Senegalese agriculture and its aversion to sales of agricultural materiel on credit as well as subsidization of equipment and fertilizer, imply little, if any, role for SODEVA. Inputs distribution will be transferred to the private sector with a greater role for the "sections villageoises". Without a coheLent technical package, the options for SODEVA are necessarily restricted. 243. Yet, such an approach, though intrinsic to a reduction in the level of operating subsidy supported by the exchequer, leaves a number of critical questions unanswered. First, if reliance on modern inputs use for productivity growth is now effectively jettisoned, what scope will there be for private sector distribution and marketing of such inputs? Second, it is generally accepted that productivity advances will require improved cultural practices. This implies a continuing role for an extension agency, as presently envisaged by the Bank and Government. But it also implies some coherence as to the type of changes in cultural practices that are deemed desirable. Thus, if the fertilizer consumption estimates of the Plan COrdalier remain unrealistic and modern inputs use remains, as seems likely, at a collapsed level, a radical re-thinking of strategy for the Groundnut Basin would be in order. This is, in part, being addressed through the Bank-funded Agricultural Research project with ISRA (Cr. 1176-SE) but, as yet, there is little indication of any coherent approach, either in terms of optimal crop mix or in the systemic framework as, for instance, through emphasizing a return to a fallowing rotation or the stimulation of more mixed farming. 244. Similar issues arise when looking at irrigated agriculture. Again, the main impetus behind SAED's disengagement has been financial. This has been coupled with the perennial issue of the economic viability of rice production in the Fleuve. The present fall in the international rice price - 112 - and the appreciation of the real exchange rate107 clearly have prejudicial implications on this score. Consequently, the Bank now emphasizes the production of higher value crops (Policy Framework Paper, 1987-1990) but is mute regarding the means, let alone preconditions, for achieving these ends. Furthermore, SAED's disengagement leaves yet more muddied the question as to how to tap the irrigation potential -- possibly as much as 200,000 hectares in the Fleuve -- released by the Manantali Dam. Within Senegal the key questions concern the level of resources required to transform this potential into reality. Yet, as in the past, the Bank has been unable to provide a particularly coherent lead. On the one hand, it is argued that any strategy based on smallholders and small irrigated perimeters will have severe limitations.108 Yet, it is also widely recognized that Senegal's ability to attract foreign investment, let alone mobilize large-scale domestic investment, is limited. Further, with a bleak outlook for the core rainfed region, it is obvious that internal migration will be a necessary outcome. This would imply a continuing requirement for smallholder schemes. If this is the case, then the resource mobilization required to realize the irrigation potential of the region will, for the most part, have to be generated from donors and the Government. At the same time, it will require strong technical support and an active extension effort. The restructuring of SAED appears on the surface to be designed to achieve these ends but, under the Irrigation IV project, output growth in the Fleuve is expected to be based on improvements to existing perimeters and, perhaps most optimistically, a relatively rapid substitution by the private sector for activities previously undertaken by SAED. 245. Despite very significant weaknesses in the Bank's approach to institutional reform, a number of measures taken in the adjustment period of Bank advice clearly address the basic issues. In the groundnut subsector, for example, the collapse of ONCAD has facilitated substantive change. Responsibility for marketing of groundnuts has been shifted to the oil millers, Socidtd Nationale de Commercialisation des 016agineux de Sdn4gal (SONACOS)/Soci6td Electrique et Industrielle de Baol (SEIB), and they have also been allowed to make their own marketing arrangements. Since late 1985 private firms have been permitted to buy groundnuts directly from farmers and sell to the mills at negotiated prices. Seed storage has been transferred to producers with SONACOS retaining only a security stock. Although this led to a very substantial decline in the groundnut area in 1985/86, this shift does not appear to have reduced the grovdnut acreage subsequently. At the same time, reduction in operating suisidies to the oil mills has begun, thereby addressing one of the principal factors behind Senegal's loss of competitiveness. In the past, massive over capacity was subsidized by the State with, in addition, fixed costs averaging around 95Z of total processing costs. These features have 107/ Senegal Real Effective Exchange Rate (1980=100): 1981=89.1; 1982=91.7; 1983=92.2; 1984=94.3; 1985=103.0; 1986=111.9 (Source: IMF). 108/ "Senegal: Agricultural Sector Strategy Brief." - 113 - increasingly been addressed in adjustment lending; SAL II contained a set oF specific actions, including a 50Z reduction in subsidy on the oil mills' fixed costs. 246. Of parallel importance have been the measures to restrict the action, of parastatal agencies in cereals markets. The Food Security Commission's (CSA)109 role has been defined as a limited one: guaranteeing the floor price. Since late 1985 local cereal assembling has been completely liberalized and the official producer price replaced by a floor price with margins no longer fixed by Government. It is difficult, as yet, to assess the impact of these measures. ISRA data suggest that cereal market releases may have increased marginally as a share of total output, but it should also be noted that Government interventions in the cereal market were historically very limited.110 Grain markets were well integrated both prior to and post-liberalization, when market integration is defined as stable price spreads across markets. In the case of CSA interventions, these have remained largely ineffective. The objective of maintaining the floor price is relevant particularly in the post-harvest period, yet this is the period when the CSA has not successfully intervened. Nor equally does it appear that CSA interventions have been able to smooth consumer price fluctuations for millet. The process of State disengagement from productive activity and the reduction in levels of market intervention has yielded mixed benefits. In the latter regard, government actions and Bank support have promoted sensible and desirable re-appraisal. Some of the more extreme anomalies -- such as the provision of groundnut seed by ONCAD or SONAR -- have been removed. Greater emphasis has been placed on producers and the private sector. In the former case, actions have been undertaken to cut the financial claims of the parastatals on the budget. In 1980 the RDAs employed around 5,500 with SONAR accounting for a further 1,500 staff. At the present time, staff numbers amount to less than 601 of the 1980 level. However, in other respects, progress is less evident. As has already been argued, the lack of a coherent agricultural sector strategy on the part of the Bank has led to weaknesses in its approach to the RDAs. There has been no serious attempt to redefine functions, merely attention to reductions in present financial support levels. A general perception of RDA inefficiency -- a perception at odds with the Bank's approach at periods in the late 1970s -- has made a virtue of demolition, but has not been paralleled by any alternative strategy. Consequently, the Bank has paid inadequate attention to some of the fundamental questions that face the sector. Is there the need for a formal credit system and, if so, on what basis? What regions and crops merit the application of modern inputs and under what pricing ratios would such application be financially and economically viable? What role should the RDAs have in developing irrigated agriculture where investment costs remain high? What activities and practices should be promoted by the 109/ Commission de la Sdcuritd Alimentaire. 110/ 0. Ndoye and I. Ouedrago, "Les Politiques d'Ajustement Structurel et leur Impact sur la Performance du Syst&me Agricole au S4ndgal," ISRA, Dakar, 1987. - 114 - extension services in the rainfed areas? What is an appropriate timetable for institutional reform, given the limitations of the Senegalese private sector and village or producer cooperatives? What are the realistic possibilities for consolidating and improving an institution neglected for 20 years by the Bank and other donors, the Ministry of Rural Development? 247. To the present time, it is difficult to escape the conclusion that the Bank's approach to institutional reform has been overly guided by short-run financial questions. Such considerations provide the gravitas to an already pronounced belief in the inefficiency and inadequacy of the parastatal sector. But they do not necessarily provide alternative solutions, other than a residual and limiting predilection for the private sector. Under current conditions, activities that are in theory to be devolved on to that sector -- such as inputs distribution or irrigation perimeter development -- are unlikely to engender any notable response. (d) Conclusion 248. Since 1980 the Bank's principal objectives in Senegalese agriculture have been to raise the growth rate, to raise factor productivity and to raise rural income levels relative to their urban counterparts. To date, the degree to which these objectives have been attained remains variable. The agricultural sector growth response has been poor and continues to be largely determined by weather, rather than policy. With regard to productivity in the sector, there has, at best, been stagnation. As in the past, the most significant land productivity effects have been for rice and maize, crops that at present account for no more than 7% of the cultivated area. The collapse of the "Programme Agricole" has meant an almost complete cessation in the distribution of agricultural machinery, while at the same time the activities of the various extension wings of the RDAs have been reduced. It seems unlikely that any improvements in labor productivity have occurred over this period. As regards the third objective, it appears that rural per capita incomes were on average 5Z lower in 1982/83 than they were at the start of the 1970s111 but had improved relative to urban per capita incomes since 1979. Estimates of the trend in the intersectoral terms of trade suggest movement in favor of agriculture after 1985.112 249. This very mixed record can, in part, be attributed to inadequacies in the policy measures promoted by the Bank and other donors and implemented by the Government, but also has to be related to the deep underlying constraints -- including climatic factors -- on Senegalese agriculture, whether at a domestic or external trading level. As regards 111/ G. Durufld, et al., "D4sdquilibres Structurels et Programmes d'Ajustement au Sdn6gal," Evaluations, Minist&re des Relations Extdrieures, Coopgration et Developpement, Paris, March 1985. 112/ Commander, op. cit., p. 33. - 115 - the Bank's policy advice in the adjustment period, a number of issues stand out. In the first place, the Bank has concurred in a major shift in emphasis away from the traditional tradable -- groundnut -- to promotion of domestic cereal production. This has been despite the fact that this will entail significant economic costs. Ironically, the absence of such comparative advantage, estimated in st&tic terms, has been used by the Bank to justify an unwillingness to invest in irrigated agriculture in the Fleuve. While market prospects for groundnuts appear to be poor, Bank analysis has not satisfactorily addressed the issues of market diversification (away from France) nor the potential impact of exchange rate depreciation. At the same time, the emphasis on reducing the role of the State in the agricultural sector -- and, hence the subsidy levels borne by the exchequer -- has detracted from the basic question as to how to raise productivity rates in both the tradables and home goods subsectors. 250. In common with other Sub-Saharan African economies, the Senegalese government developed in the 1960s and 1970s a taste for direct intervention in agricultural pricing and markets. The three SALs and accompanying technical assistance projects have attempted to reverse these inclinations. In many respects they have been successful. Major subsidies have been reduced, markets liberalized and producer prices raised. Some new "distortions" have been created, particularly through protection of domestic cereals. In the case of institutions, better financial management and specification of targets has been attained through program-contracts. Budgetary transfers to the sector have been consequently reduced. Despite these achievements, as part of the more general process of financial stabilization, this chapter has argued that for agriculture, the Bank has been unable to develop a feasible, strategic approach to the sector. Although adjustment lending instruments have allowed for greater policy dialogue and a more satisfactory treatment of the wider constraints operating on the sector, it ought also to be recognized that such gains increasingly require parallel sectoral-level interventions. If measures taken under structural adjustment programs are to be fully effective, a strong argument can be made for greater sector adjustment lending with less reluctance to financing productive investments through projects. Without direct project assistance, specifically aimed at raising factor productivity, the overall response of the sector to broader policy measures will remain limited. To that extent, adjustment lending has only begun and very partially at that, to create the pre-conditions for growth. A continuing agenda still has to be set. B. Industrial Sector 251. Supported by substantial foreign and private investments, Senegal's industrial sector (mining, manufacturing and utilities) expanded rapidly during the 1950s, catering to the colonial French West African market of some 20 million people. With Independence in 1960, a good part of this market was lost, as some other former colonies developed their own - 116 - industrial base, leading to the low capacity utilization which has been characteristic of Senegal's manufacturing sector ever since. In addition, the generally high cost st ucture of Senegalese industry has severely limited its export potential, and, stagnant demand within Senegal has further curtailed industrial growth. Between 1960 and today, industrial value added is estimated to have grown at slightly more than 42 per year, much slower than in countries such as C8te d'Ivoire or Cameroon. Neverthe- less, because of the even slower growth in the rest of the economy, industry's share in GDP rose from around 122 in 1960 to close to 20Z by 1986. Most industrial growth appears to have resulted from import sub- stitution, although some new export development took place especially during the late 1960s. Growth was most rapid during 1960-75 but slowed down considerably after 1975. 252. Agro-industries (sugar manufacturing, groundnut oil, fishing and canning) dominate the manufacturing sector accounting for some 40% of industrial value added, followed by the chemical subsector (fertilizers, petroleum refinery) and the textile and leather subsector, each representing about 15Z of industrial value added. Other activities include metal and mechanical industries, construction materials and paper industries. State participation in the manufacturing sector is limited to a dozen enterprises, which together account for 302 of total value added. Most of these participations were taken by the Government during 1975-80, including majority interests in the four largest groundnut oil procensing enterprises and in the two phosphate mining companies. The manufact.ring sector remains heavily concentrated, with each subsector dominated by two or three enterprises. The largest 140 enterprises generate about 95? of production, and the largest 40, 80?. The largest private firms continue to be under foreign ownership. 253. Ever since Independence the Government's expressed strategy for the industrial sector (as well as for the economy as a whole) has been "diversification". In practice, little was accomplished largely because the overall policy framework was not conducive to growth or diversifi- cation. Following Independence, the Government established high tariff barriers supplemented by numerous quantitative restrictions. In addition, many exceptions to the common tariff regime were granted on an ad hoc basis. This system led to the development of a generally uncompetitive industrial sector at a high cost to the economy with limited potential for growth. In an effort to help industries gain export markets, the Government set up a number of export promotion schemes which, however, proved to be ineffective for a number of reasons, including excessive bureaucracy. Also, the regulatory framework (investment code, price controls, trade regulations) was highly inefficient and tended to hamper the development of the manufacturing sector. As was discussed in Chapter II, a reform process has been initiated recently, starting with SAL II in 1986, to make the policy environment more conducive to industrial growth. - 117 - 254. The Bank, since the start of its relationship with Senegal, fully supported the diversification strategy. However, the focus of the lending program -- industrial credits -- was clearly not the binding constraint to the development of industry. While the Bank's economic and sector work did dwell on the many obstacles to industrial growth including, from the very beginning, on the crucial role of the exchange rate, these concerns were not made operational. Instead, it was implicitly assumed, together with the Government, that the provision of long-term finance (and technical training supported through the Bank's education lending) would somehow lead to growth. In fact, the availability of long-term finance was not only far from being the major bottleneck to manufacturing growth, it may not have been a bottleneck at all. In retrospect, the Bank's impact on Senegal's industrial development has been minimal. 255. Apart from two engineering projects -- a loan of $0.6 million in 1973 to study the feasibility of dry dock facilities, and a credit of $7.7 million in 1983 for phosphate industry development -- and apart from the very recent (December 1987) $33.0 million credit for an industrial sector restructuring project, all of the Bank's lending for industry in Senegal has been to a development finance institution, SOFISEDIT: a loan for $3.0 million in 1974 which helped create SOFISEDIT, a loan for $4.2 million in 1976, and a loan for $6.5 million and credit for $2.5 million in 1981. The 1974 loan was audited by OED in 1983, and SOFISEDIT was reviewed as a case study in a 1989 OED report on the sustainability of financial interme- diaries. 256. It is not totally clear why the Bank decided to help set up SOFISEDIT in the first place. On the one hand, it was recognized at that time by all parties involved in Senegal's development efforts that there was a clear need for additional sources of long-term finance for industry and, with it, for improved appraisal capabilities. On the other hand, the extent of this need for additional finance, given business prospects, and therefore the question of whecher the existing financial sector with some minor restructuring would not be able to handle that demand was not clear at all. The choice, in other words, was between a new institution or some modification to the financial system in place. The record shows that, as this issue was being discussed within the Bank, Bank staff were highly skeptical about the need for a new institution and felt that the functions of one of the existing commercial banks, the well-run Union S6n#galaise de Banque (USB), could easily be expanded to include long-term financing for industry, in view of the modest demand prospects. 257. An official Government request for Bank support towards the creation of a new development finance institution, at that time called Socidtd Financi6re pour le Ddveloppement Industriel (SOFIDI), was sent to the Bank President on September 15, 1970. Bank staff responded "We doubt that SOFIDI would have a 'market' sufficient to make it a financially viable institution: to give to USB the role assigned by the Government to - 118 - SOFIDI is hence at present a better alternative."113 The staff did suggest that, should the Government want SOFIDI anyway, they would want to see market prospect studies and would keep an open mind. While informing the Government of this position, the Bank added that it was ready to consider credit lines to USB for the purpose of providing long-term finance to industry. By mid-1971 the Government appeared to have dropped the idea of SOFIDI and in fact took action for USB to handle industrial finance. 258. On August 16, 1971, however, the Government renewed its request to the Bank President for support in setting up a new institution and asked the Bank for help in a thorough study which would allow "to justify the creation of SOFInl" (translation from French). In its reply of September 14, 1971, Bank management now suggested that "conditions change" and that it will look into the matter. As there were few indications that circumstances had changed, several Bank staff during the ensuing months remained highly skeptical about the need for a new institution, now called SOFISEDIT: "We agree, therefore, that to a large extent the 'need, for SOFISEDIT is political";114 or, "there are more efficient ways of providing long-term finance for industry in Senegal than via the proposed SOFISEDIT proposal.. .we should explicitly inform the Government.. .However, if the Government...expressed a determination to move forward...we and Senegal stand to gain more by Bank and IFC involvement than by disengagement."115 Subsequently, SOFISEDIT was established in March 1974 supported by a Bank loan and began operations in November. 259. It is hard to escape the conclusion that a project, which Bank staff in essence considered to be a bad idea, was nevertheless supported primarily because the Government insisted. The reason for Government insistence is somewhat puzzling. It was felt strongly that if USB were to be given the extra window of long-term lending for industry, USB would become too domineering, which would cause objections from the other commercial banks and therefore be detrimental to the cause of industrial financing generally. Thus, a new independent institution was needed. Yet, at the same time, the intention when creating SOFISEDIT was to give it monopoly powers (right of first refusal) in long-term lending for industry, primarily to help ensure its financial viability. This objective, however, was almost immediately defeated by the monetary reforms of 1975, to the surprise of the Bank. 260. It must be added that on the Bank side the economic work was not of much help in deciding whether or not a new development finance institution was needed. As wa3 discussed in Chapter I, economic reporting during the early 1970s was generally upbeat about the future economic prospects of Senegal, and thus about the future demand for industrial finance, although this assessment was not based on a thorough underlying analysis. During 113/ Internal document (October, 1970). 114/ Internal document (February, 1973). 115/ Internal document (March, 1973). - 119 - the 1960s and early 19709 Senegal's industrial sector grew at a reasonably rapid rate. The 1970 economic report attributed growth to import- substitution, not exports; the 1973 report says "Whereas import- substitution was a major stimulating factor during the first half of the decade, export growth has largely contributed to the expansion of industrial production in the latter part of the 1960s. New export markets were found... (para. 200), and, *the 1966 revision of the Western Africa Customs Union...was certainly a major factor explaining the export boom of the late 1960s' (para. 201). The report then projected future industrial growth at more than 6% per annum based on an "analysis of past trends and of the factors likely to affect output' (para. 204); it did not provide much analysis on likely demand constraints. Towards the end of 1976 there were indications that the Bank's optimistic assessment of Senegal's future was being questioned: "a short run problem that has been recurring over the last ten years or so is not a short run problem anymore..116 By 1978, the CPP stated bluntly, "The last Senegal CPP... in 1976.. .contained a fairly positive assessment of Senegal's long-term growth prospects...; the wild swings in Senegal's fortunes, beginning with the disastrous droughts of 1972-73 and passing through a highly favorable terms of trade shift in the middle of the decade only to return to penury again in 1976, called into question our assessment of underlying trends." Even so, the 1979 economic report again projected industrial growth at 6.8Z per year over the next several years, "if the recommended export promotion is followed." 261. One final factor which helps explain the decision to support SOFISEDIT was, of course, the Bank's explicit overall strategy during the 1960s and early 1970s to spread itself over many sectors "as an experiment in determining the comparative advantage of Senegal. Furthermore, the Bank is influential as a leading lender in some sectors and it is likely that other lenders will follow our leads. Eventually the program should be consolidated as other lenders take over some sectors.'117 262. While the prevailing optimism in the Bank about Senegal's economic future, together with the desire to be involved in most sectors, must have played some role in the decision to create SOFISEDIT, the primary factor was the insistence of the Government that the project go forward in spite of serious misgivings by many Bank staff. The conclusion, therefore, is that the decision was made primarily for "country relationship" reasons. 263. Almost from the start SOFISEDIT ran into financial and other difficulties, which grew worse over time. It is again not immediately obvious why, under the circumstances, the Bank continued its support. If there ever was a gap to be filled by SOFISEDIT in the market of long-term finance for industry, this gap became much smaller within the year following SOFISEDIT's start of operations. The new banking regulations of 1975 allowed local banks to rediscount loans of up to ten years (up from a 116/ Internal document (December, 1976). 117/ Internal document (November, 1972). - 120 - maximum of seven years in the past) thereby giving all banks access to longer-term resources. SOFISEDIT would now have to compete with well entrenched banks in a very limited market. Furthermore, this market was far from dynamic as industrial growth slowed considerably after 1975, although the tourism and fishing sectors gradually gained in importance in SOFISEDIT's portfolio. 264. It would appear that the Bank, throughout, ignored the reality of the situation, i.e., an extremely small market for industrial finance, which could have been well served by the existing financial system, and poor industrial growth prospects given the economic policy environment. By the time of the second loan, in 1976, the Bank's economic forecasti were still upbeat, but after 1978 the Bank did recognize that Senegal's future economic outlook vis indeed bleak. The 1981 project appraisal reviewed SOFISEDIT's poor past performance but concluded that new management and more active project promotion together with increased lending for tourism and fishing would turn things around. While Senegal's first SAL had been extended in 1980, it did not deal with structural issues in the industrial sector. The primary obstacles to industrial growth were, in other words, once again not being addressed. 265. The inevitable conclusion is that the Bank finds it very difficult to admit mistakes. At some point soon after SOFISEDIT's creation it must have become obvious that the institution lacked a solid basis for growth. Yet, once involved, the Bank could not extricate itself and kept on lending while ignoring the real issues, apparently hoping that somehow things would work out. Instead, as soon as SOFISEDIT's problems surfaced, the Bank should have either insisted on the structural changes necessary to generate industrial growth as a condition for further lending, or, if that proved impossible given the political/economic environment, stopped lending or at least down-scaled the operation to focus on high quality projects. 266. As it turned out, SOFISEDIT's performance record has been very poor. Its operations have been essentially stagnant over time and far below projected levels: they have supported primarily medium and relatively large enterprises although one of the objectives was always the promotion of small-scale industry. The proportion of loans in arrears has risen sharply over time and as of January 1987, principal and interest in arrears were around 40Z of loans outstanding. Also, after provision for bad debts, SOFISEDIT has incurred net losses in all years of operation. By 1981, accumulated losses had wiped out SOFISEDIT's equity capital. It was rescued by the Government only to return to a financial crisis situation more recently. While it is not possible to estimate quantitatively the benefits to the Senegalese economy from SOFISEDIT's activities, the above would suggest that they have been marginal, if any. On the other hand, the country's indebtedness has increased. 267. Starting with SALs II and III (1986/87) the Bank has begun to focus its lending on the resolution of many of the basic structural problems which impede growth in Senegal, including manufacturing growth. An - 121 - Industrial Sector Restructuring project to facilitate the adjustment of the industrial sector to the new policy environment was approved in December 1987. These are major steps forward in the Bank's efforts to help reac- tivate industrial growth in Senegal. As discussed in Chapter II, the emphasis is on rationalization of the incentives' system and on privatization. It is unfortunate that some of the major issues in this regard, i.e., the exchange rate and the supply of private entrepreneurs and capital have not been dealt with. The primary lesson resulting from the Bank's past experience in industrial lending is that if issues are major one cannot leave them unattended and hope for the best. C. The Transport Sector (1) Introduction 268. When Senegal became independent in 1960, it already had a well- developed transport network. Until that time the system served exportlimport traffic between France and all of French West Africa. A number of developments following Independence brought some major changes and go a long way towards explaining the performance of the sector since, as well as its current status. First, after 1960, the network served a much reduced area and with this the significance of Dakar as a regional center also declined; second, transit services to air and shipping lines provided in the past by Dakar lost their earlier importance because of the growth of other cities in the region and because of changes in airway and maritime routes requiring fe-4er transit stops; third, the performance of the Senegalese economy since 1960 has been poor, thus further limiting the demand for transport services; and, finally, rapid population growth meant high pressure to create jobs and overstaffing in public institutions including those serving the transport sector. 269. In 1984 (the last year for which reliable statistics are available), the road network totaled some 14,000 km, of which some 3,500 km were paved. The vehicle fleet numbered some 65,000 units, with about 801 based in Dakar where almost 75% of all traffic originated or terminated. Some 95Z of all domestic transport (excluding phosphates) was done by road, which underscores the importance of preserving the highway network and keeping vehicle operating costs to a minimum. In railways, the Rdgie des Chemins de Fer du SdnOgal (RCFS) was developed in pre-independence days as a feeder to the port of Dakar and had two main lines: one towards Mali and another to St. Louis in the north. Over the past 20 years, the railway has been losing traffic to road transport: in 1984 it handled mostly imports for Mali and short-haul phosphate exports, each accounting for some 140 million ton-km annually; general cargo and groundnuts amounted to only 10 million ton-km. Passenger traffic declined by 70% over the 1974-84 period. Dakar has a major international deep-water port frequented by many conference lines. Shifts from general cargo and bulk traffic to containers - 122 - dictated modifications in the port facilities and the necessary improve- ments have been financed by the Bank Group. Coastal shipping, serving the secondary ports of Kaolack, Ziguinchor and St. Louis, is minimal. River navigation is possible during part of the year only and, like coastal shipping, is also declining as better road connections are developed. Air passenger traffic growth at Dakar was impressive during the 1970s and Bank assistance upgraded airport facilities. The national airline, Air Senegal, provides scheduled flights to 13 domestic airports, the busiest routes being Dakar-Cap Skirring, a tourist resort, and Dakar-Ziguinchor, each with about 10,000 passengers yearly. Only two other routes have more than 1,000 passengers per year. 270. The performance of the transport sector during the 1960-87 period leaves much to be desired. In most transport modes, agencies are weak; financing of operational and maintenance costs is erratic; and the quality of transport service, offered both domestically and to neighboring countries, appears to steadily deteriorate. The railway, e.g., has major operational problems and a large staff whose salaries cannot be covered by revenues. The Government is unable to fund road maintenance activities and IDA and other donors provide assistance. The airline runs a deficit from low traffic routes imposed by the Government and from the loss of passengers to road transport. (2) Bank Group Assistance 271. Bank involvement in West African transport pre-dates the independence of most countries in the region. The first Bank operation in what is now Senegalese territory goes back to 1954 when it financed part of a modernization program for the railroad facilities of French West Africa. In independent Senegal, the Bank Group followed up with 15 more operations, constituting about one-fifth of all Bank Group lending to the country: 6 projects in the roads sector, 4 for railroads, 3 port and 2 airport projects. Only two of these are still ongoing in early 1988: the Fifth Highway Project and the Dakar Container Port Project. Throughout, Bank staff studied macro-economic developments and monitored changes in the structure and performance of the transport sector with a view to make Bank assistance more conducive to economy and efficiency. 272. From the beginning the Bank's strategy towards Senegal's transport sector has been sound. In view of the relatively well developed state of the available transport facilities in the early 1960s, the Bank strategy initially formulated in the mid-1960s focussed on three elements: emphasize maintenance; improve efficiency; and avoid overinvestment and premature projects by strengthening the Government's ability to plan and evaluate alternatives. The relevance and validity of these three basic points remain unchallenged to this day. Much effort has gone into attempts to increase railway efficiency, introduce measures to reduce railway deficits and improve transit services offered to Mali. Efforts have also been made to assist the development of thedomestic construction industry and to strengthen government procedures such as in the procurement and - 123 - payment method areas. The Bank's views concerning transport improvements were regularly discussed with the Government, which expressed its agreement and endorsement, and were translated into the lending operations discussed below. In general terms, the Bank's achievements have varied widely: for highways, project performance has been generally good but quite the reverse for the railways subsector. (3) Roads 273. On the whole, the Bank's role in the highways sector, with its strong focus on maintenance, has been very constructive. The objective of the First Highway Project, a credit of $2.3 million in 1970, was to be a pilot operation to support the expansion of the feeder road network and to study and determine the maintenance and improvement needs of roads. While the estimated economic rate of return on the feeder road construction component of the project was only 4Z at completion, much below the appraisal estimate, primarily because of lower than anticipated traffic, the project as a whole was a success because it fulfilled its purpose as a pilot project. The road maintenance and improvement study provided a good basis for future highway projects and highlighted the need to shift priorities in the highway sector from feeder roads to improvement of primary roads. It enhanced the Government's concern with and interest in the great losses to the country's economy caused by poor road maintenance. 274. The Second and Third Highway Projects, respectively a credit of $8.0 million in 1973 and a loan of $15.0 million in 1976, were implemented in close succession, were for all practical purposes regarded as one common fund, and had the same objectives: to assist in implementing a paved roads strengthening program and a highway maintenance program, both identified under the first project. Consultants investigating maintenance operations as part of the first project had concluded that the efficiency of the main- tenance organization was hampered by the regional rather than central control of operations and by unclear responsibility for the equipment. Accordingly, the Bank endorsed the recommendation that a central maintenance division be created to coordinate and control all country-wide maintenance operations and that a central equipment division be estab- lished. The second project was structured around these recommendations. The proposed reorganization encountered strong resistance, however, within the Ministry and from the regional engineers whose authority would be taken away. Little progress was made under the program between 1973 and 1975. With the third project the Bank reversed its approach. Whereas the second sought to replace an existing and firmly established system which displayed many shortcomings, the third attempted to remove the shortcomings without disturbing the established organization. In other words, the Bank assisted the Ministry in returning to the decentralized system they had before, and in improving it. Subsequently, not only did the network under maintenance increase significantly, but the rate of completion of annual programs was much improved and road maintenance unit costs were kept about constant despite high inflation. - 124 - 275. The Bank, in this instance, went through a quick and effective learning process. The approach followed in the third project was significantly more successful than that of the second mainly because, and even though profound procedural changes were introduced, the existing environment was little perturbed, and because the approach was gradual rather than all-out, all-at-once as under the earlier project. The net result was that at the end of the third project, maintenance was carried out successfully by the Senegalese with only minor assistance from outside consultants. It would seem that the success of road maintenance improve- ments introduced under the third project rested on the fact that the system attacked an area usually neglected in maintenance projects: the improve- ment of administrative procedures and the introduction of efficiency- improving management techniques. This contrasts with the usual approach followed in many other projects where the emphasis is on the formulation of detailed work programs (which are seldom implemented), increasing the number of units of equipment (which are often not required), and pressuring the Government for increased funding (which is usually hard to get). The Bank has continued to support road maintenance and rehabilitation through a Fourth and Fifth Highway Project, respectively a loan/credit for $38.0 million in 1980 and a credit for $21.5 million in 1984. 276. In contrast to the success achieved in the highway maintenance projects, the Bank's single Feeder Roads Project, a loan of $6.6 million in 1976, fell considerably short of its origintl objectives. The purpose of this project was to increase the length of the rural road network and to improve the efficiency of rural transport as a means of increasing agricultural production and the standard of living of the rural population. The project addressed both the planning process of selecting and deciding on an annual feeder roads construction program, and the process of physically carrying out what was decided. Since there was no institutional mechanism to deal with feeder roads the creation of one was included in the project, consisting of an Interministerial Committee, a technical secretariat to the committee and a new feeder roads subdivision in the Ministry's road maintenance division. It was hoped that the creation of these institutions would ensure that the building and maintaining of feeder roads would become a separately funded and permanent, ongoing feature of government activity. 277. The appraisal, however, was devoted primarily to the physical aspects and contained no assessment of whether the Government could absorb the proposed institutional planning structure or had the staff and exper- tise to do so. As it turned out, shortages of staff and expertise were major bottlenecks, especially in view of the low priority assigned to feeder roads generally. Moreover, the Bank's insistence that the project be implemented in three years rather than five as suggested by the consul- tants strained the Government's implementation capacity excessively. In the drive to achieve high targets, quality suffered. In 1979, poor perfor- mance eventually caused the Bank to threaten to cancel the loan. In response, the Government made the feeder roads subdivision in the Ministry administratively and financially autonomous. This helped speed up project implementation, but the preferential treatment given to the unit provoked - 125 - resentment within the Ministry and the arrangement could be sustained only as long as foreign funds were available to finance it. When they ran out, activities ceased and the feeder roads subdivision found itself without an immediate role: it was later on reinstated under the road maintenance division, to execute both feeder road construction and periodic maintenance of main roads depending on the priorities and funds. At project completion, the economic rate of return was estimated to be negative because of low traffic, lack of maintenance and high cost, poor quality construction. 278. Low traffic, itself a result of bad weather and poor agricultural performance generally, was clearly an important reason for the problems encountered by this project. Even if this factor had not taken its toll, however, the Bank's underestimation of the difficulty in creating a new institution in an environment of serious staff and skill shortages and its insistence, under the circumstances, on speeding up project execution and on giving the feeder roads unit special treatment would have kept this project from havirg a significant impact. (4) Railroads 279. The story of the evolution of the Senegal Railways during the past quarter century and of Bank assistance to it is primarily one of problems and failure. This is not to say that the Bank's objectives were wrong, but rather that it could not prevent, despite many efforts, the chronic deterioration of this subsector in virtually every respect. The best that can probably be said is that Bank assistance in the end helped prevent a total collapse of the system. In all, four Bank operations took place to assist railway development, the first three of which were in support of the state-owned railway, RCFS, while the fourth helped create a private railway company, Socidtd d'Exploitation Ferroviaire des Industries Chimiques du Sdndgal (SEFICS) to meet the transport requirements of Industries Chimiques du S6n4gal (ICS). 280. The Bank's First Railway Project, a credit for $10.6 million in 1966 was also the Bank's very first project in Senegal. This was followed in 1972 by a second operation in support of RCFS, a loan/credit for $9.6 million, and then in 1978 by a third, a loan for $11.0 million. While the detailed objectives of these three projects diffezed a bit from one to the next, they were similar in that all three were focussed on safety, increased efficiency and some improvements in carrying capacity, the latter especially in order to be able to offer competitive service for Mali international traffic. All three projects suffered serious delays in implementation ranging from 3.5 years for the second to 6 years for the first; all three failed to achieve project objectives and in the process the railway system as a whole deteriorated seriously. Traffic has been decreasing steadily, partly because of the railway's chronic inability to handle available demand. Transit services for Mali have declined from handling most of Mali's foreign traffic to less than half. The railway continues to show low operational performance, low productivity and large - 126 - deficits, stemming from management problems, a strong union, lack of staff discipline, continued operation of uneconomic lines, lack of autonomy and poor infrastructure. The railway serves the Government indirectly as an employment agency having many more staff than it needs with no authority to dismiss personnel. Some 652 of railway operating costs are spent on salaries which are eventually covered by the government budget. 281. Two things stand out in a review of the Bank's role and impact in the sector: poor project preparation and an ineffective dialogue. Both the second and the third projects had to be redesigned during implemen- tation. In the case of the second project, two years after signature, some 80Z of the track renewal work was eliminated, even though most of the materials had been procured, in order to include in the project other more urgent items not previously covered: as there was no indication of a sudden change in the railway's development, a good appraisal should have made this unnecessary. In the case of the third project, track renewal on the international line was replaced by rehabilitation of work badly done under the first and second projects; this should again have been avoided by either adequate supervision of the second project or good preparation of the third. Bank appraisals have also tended to be overoptimistic on traffic forecasts. A vicious circle ensued with insufficient traffic, aggravated by poor service and therefore additional traffic diversion, leading to financial problems and, in the end, to the railroad not being able to handle all available traffic. With respect to the increasing diversion of Mali traffic to roads through C8te d'Ivoire at the expense of the Senegalese railways, the Bank appears to have found it difficult to accept this fact, and has continued to assert that improvement of the Senegalese network would be partly justified on the grounds of transit services to landlocked neighbors. The 1984 appraisal report for the Dakar Container Port Project, e.g., states in paragraph 2.12: "Currently under- way are two highway and two railway projects. The railways, aviation and port projects were conceived, in addition to helping meet Senegal's domestic transport needs, to support Dakar's role as an international transit point." The Bank made in fact strong efforts to salvage some of the Mali traffic for Senegal without much success. "Great pressure was put on the Government of Mali by the Bank to restrict the number of licenses granted to truckers for the C6te d'Ivoire route. However, despite assurances that the number of licenses would be restricted, the Government of Mali has still been issuing licenses for the 'renewal of worn-out trucks,' replacing eight-ton trucks with ones of higher load capacity. Given the restricted capacity of the railway up to 1982, it is probably just as well that the Government of Mali continued to award truck licenses."118 282. Throughout its association with RCFS the Bank had remarkably little impact on the institution or its management. The political strength of RCFS and the intransigence of the union were clearly important factors to 1181 World Bank, OED, PCR: Senegal Third Railway Project, Report No. 6822, June 4, 1987, para. 8.07. - 127 - contend with. However, if the environment did prove totally unreceptive, it is not clear why the Bank did not withdraw its support. The persistence of the Bank to maintain a presence in the Senegal railway subsector becomes hard to understand in view of statements suci as the followingt "The situation was further aggravated by the loss of locomotives in serious collisions which take place on RCFS with an amazing regularity. The basic cause is lack of discipline, some drunkenness and negligence.0119 283. The Bank's fourth project, the SEFICS Rail Transport Project, a loan of $19.3 million in 1981, was designed to meet the transport require- ments of ICS by creating a private railway company, and to help streamline RCFS into a more efficient organization. The principal objective of this project was met in that transport services were made available to ICS on time at much lower costs than those of RCFS. ICS performance has been so poor, however, that SEFICS carried less traffic than its design capacity. In order to spread the high fixed costs of SEFICS over larger traffic volumes, the Bank tried to persuade the Government to throw open the national phosphate market to competitive bidding by SEFICS and RCFS. This would most likely have led SEFICS to win the additional traffic, or RCFS to improve management and operations substantially, or both. Severe labor union pressure and political deference to railway labor kept this proposal from being implemented. In hindsight, it would have been prudent to have some built-in safeguards against the contingency of the failure of the parent industrial venture. While the project, thus, in a narrow sense failed in its second objective, i.e. to turn RCFS into a more efficient organization, it did probably contribute indirectly to that objective by strengthening the Bank's dialogue on railway policy which under the SALs has led to substantial improvements in investment planning. In fact, through a very intensive exchange of views since about 1984 with the many donors active in Senegal's railroad sector, the Bank has successfully contributed to the avoidance of much undesirable investment (para. 306). (5) Ports and Airports 284. Physical facilities in both of these subsectors have been improved and expanded successfully with Bank Group assistance. The major problems of both stem largely from a 'ack of autonomy. The Port Authority does not have financial autonomy and, as a result, has serious cash flow problems and even insufficient operating funds, because its revenues are received, and mostly spent, by the General Treasury. Likewise, lack of autonomy generates management and cash flow problems at the airport: competitive road services have made some domestic air services uneconomic, which should therefore be discontinued to reduce financial losses which Senegal can ill afford. The planning of new capital expenditures in these subsectors, under the guidance of the Bank, has also not been fully satisfactory, leaving questions about the necessity of some of the investments. 119/ Ibid., para. 4.10. - 128 - 285. The 1967 Dakar Port Project, supported by a loan for $4.0 million, was largely successful. Port infrastructure was improved and the financial condition of the enterprise was considerably strengthened. Although the performance of the Port Iathority was satisfactory, with the major problems encountered being created by other agencies (overdue payments), government controls were tightened because of poor performance by other public corporations. This was contrary to what had been agreed during negotia- tions and would remain a major bottleneck to operational and financial improvements which the Bank did not manage to overcome. The 1977 Dakar Fishing Port Project, supported by a loan for $6.0 million, focussed on the development of fishing port facilities and related management services. Again the physical construction component was executed in a very satisfac- tory manner but it is doubtful whether the forecasted economic benefits were achieved. The PCR concluded that 'general cargo growth, and thereby future port congestion, was exaggerated by choosing an unusually low base year for the appraisal. Little attention was paid to increasing port efficiency before seeking to justify a new investment. Furthermore, the growth potential of the fishing industry was overestimated, it seems amongst others, by considering the fishing port as a unique venture, rather than as one of many fishing ports competing for the same catch." Also, port management has remained weak because of an acute shortage of skilled and experienced top cadres. Financial management has been particularly weak and, in spite of eight financial supervisions, only limited progress was made on proper implementation of financial covenants: the ongoing 1985 Dakar Container Port Project, supported by a credit for $7.5 million, whose primary purpose is to expand port facilities to handle the increasing container traffic, therefore also includes substantial technical assistance in the fields of general and financial management. 286. Two Bank projects have helped support Senegal's airport development: a 1972 Airport Project, supported by a loan for $3.0 million, and a 1979 Second Aviation Project, supported by a loan for $7.0 million. The Bank's involvement in airports resembles closely that in ports. Both projects were successful in improving physical facilities especially in order to ensure the continuation of Dakar's important role in transit traffic. On the other hand, after the second project financial management remains unsatisfactory and financial performance has in fact deteriorated. Also, in hindsight, the Bank misjudged the development of transatlantic air traffic by not considering the risk of a much larger proportion of flights bypassing Dakar, and this has resulted in a much lower rate of economic return than originally forecast. More importantly, however, at the time of appraisal, the technical and organizational management of the airport were judged satisfactory and the project did not include assistance for strengthening the organization. Yet, management has weakened: maintenance has not always been adequate and little effort has been made to find ways to better handle present traffic through adjustment of schedules or minor modifications of facilities, thereby decreasing the need for additional investments. - 129 - (6) Conclusion 287. There is no doubt that the Bank's strategy towards Senegal's transport sector has been sound and that its main features are as valid now as they were in the mid-1960s when they were first formulated: sound investment planning, priority to maintenance, emphasis on economy and efficiency. Between 1966 and 1987 the Bank Group approved nearly $170.0 million in fifteen loans and credits for the transport sector, 21% of all lending to Senegal. In addition, the Bank mobilized considerable cofinancing and devoted many staff-years to transport sector and supervision work. Despite all this, results are not all satisfactory: the railway shows poor operating performance and mounting deficits; the road maintenance organization functions well but is not adequately funded; the Dakar port and airport have good facilities but traffic is down. If any one factor could explain the unsatisfactory impact of the Bank on Senegal's transport sector, it would be that the Bank's tactics were wrong. All too often, and presumably in order not to disrupt the lending program, the Bank has tended to ignore or downplay the importance of major policy issues or of political or economic realities, somehow hoping that they would go away. The outcome has invariably been unsatisfactory. 288. A first important finding is that projects are more likely to be successful when institutional change is approached gradually. In the area of road maintenance, the lending program became effective only after the Bank changed tactics and opted to work within the existing environment to strengthen and improve the road maintenance division, rather than through a drastic reorganization which was politically unacceptable. In contrast, the Bank's attempt to create a new institution to deal with feeder roads and its insistence to accelerate the implementation of the project in the face of shortages of staff and expertise led to major problems. Also, privatization such as in the case of the SEFICS Railway Project is advocated as a panacea but is not likely to achieve its objectives until the political constraints militating against decisive change have been resolved. 289. Second, difficult issues have to be addressed more directly and honestly. When there are strong and well-substantiated doubts about the commitment or capability of the Government to implement competently the various elements of a project, the Bank should not hesitate to defer lending until it becomes clear that sustainable progress has been made in implementing agreed actions. To the extent that the Bank had no impact on government railway policy and could not generate any improvement in railway management or operations, it is unclear why it continued to lend. Also, experience with the first port and airport projects did indicate the importance of financial and operational autonomy but the Bank did not effectively resist the Government's subsequent encroachment upon whatever elements of autonomy remained, which contributed to the management and cash flow problems of the two institutions in recent years. - 130 - 290. Finally, it is worth emphasizing the obvious: good preparation and appraisal are key to the success of any project. As explained, the disastrous second and third railway projects were poorly appraised from the start. Apparently, lessons from the past to improve future project appraisals are somewhat elusive. Thus, e.g., experience over many years in both East and West Africa has shown that the subdivision of former colonial territories into new nation states requires a new approach to transport planning. Among others, the young landlocked countries will wish to multiply their access corridors to the sea so that their communication would not be interrupted by events over which they have no control. In the case of Senegal this translates into reduced transit traffic. Yet, overoptimistic traffic forecasts on the grounds of transit services to landlocked neighbors have been a common feature of many appraisals. Another example of how the Bank at times fails to learn from the past concerns roads. The OED audit on the feeder roads project, which concluded that the project failed to generate any lasting institutional impact, was published on December 7, 1983. It states among others "Yet, as related in the PCR and observed by the audit, building and maintaining feeder roads has not become a permanent feature of government activity. On the contrary, as soon as project funds ran out it ceased altogether.-120 At completion, the project was estimated to have a negative rate of return, mostly on account of the short life of the roads caused by their relatively poor quality and their lack of subsequent maintenance. The President's Report on the Fifth Highway Project is dated February 21, 1984, and reads, in part, as follows: "the Feeder Roads Project supported improvement and maintenance of over 1,000 km of rural roads, and established an institutional framework for continuous development of the feeder road network. The Project Completion Report for the Feeder Roads Project confirmed the economic justification of the project, provided, however, that adequate maintenance operations were carried out. The Government is intent on continuing rural road works through the Bureau des Pistes de Production created under the project, and has been actively eliciting interest from major donors for assistance...."121 291. On the positive side, it is worth repeating that the Bank's assistance to Senegal's transport sector has made some important contributions. Without it, the system and the railways especially would have broken down. The port and airport do have good physical facilities, sufficient for the likely demand generated in the foreseeable future. And, the road maintenance organization is effective and efficient. With regard to maintenance, however, the high dependence on donors for maintenance finance gives rise to policy questions that have not received the scrutiny they deserve. 120/ World Bank, OED, PPAR: Senegal Feeder Roads Project, Report No. 4810, December 7, 1983, para. 7. 121/ World Bank, President's Report, Senegal Fifth Highway Project, Report No. P-3636-SE, February 21, 1984, para. 41. - 131 - D. Education 292. While significant progress in expanding education has been made since independence, the results are not entirely satisfactory. Between 1965/66 and 1984/85 public spending for education rose from the equivalent of 1% of GDP to 5Z. During the same period enrollments increased fairly rapidly at 5.12 per year at the primary level, 7.6Z per year at the lower secondary level and 9.42 per year at both the upper secondary and university levels. The relatively faster rate of increase in enrollments at higher levels of education was a continuation of the traditional emphasis placed on secondary and higher education, as opposed to primary education, which is typical of many other sub-Saharan countries. By 1985, adult literacy was only 3Z and primary enrollment about 50Z for the country as a whole, but much lower in rural areas: while more than 60Z of primary school-age children live in rural areas, they make up only 372 of enrollments. In addition, on a country-comparative basis, per student costs are high in Senegal and given the country's overall tight budgetary situation it will be difficult to increase spending for education as a share of GDP any further: future progress will have to come from more efficient resource use. 293. To date the Bank Group has financed five education projects. The first project (1971) provided $2 million to improve the quality of technical and agricultural education. The second project (1975) for $15 million was designed to help Senegal in meeting modern sector manpower requirements through assistance at the secondary level and vocational training; it also began the establishment of rural youth training centers and helped strengthen educational planning services. The third project (1979) provided $26.5 million to strengthen technical, vocational and management training programs and to begin to lay the basis for the expansion of primary education in rural areas through planning and research. The fourth project (1985) for $5.5 million is designed to transform the management school financed under the third project into a Regional Management School for the West African Economic Community. The fifth project (1986) provides $12 million for the development of primary education. The latter two projects are still in the stage of implementation. 294. When the Bank began its dialogue with Senegal on education issues and policies during the second half of the sixties, the two major weaknesses of the education system were identified by both the Government and the Bank as the poor adaptation of the system to the needs of the economy and the low primary enrollment rates especially in rural areas. Subsequent, most attention was given to the former of these weaknesses because the system had not been able to produce enough adequately trained Senegalese to replace expatriates in the modern sector of the economy, ar this was an important government objective. - 132 - 295. Starting in 1970, the Government has considered a number of reform proposals to restructure the education system, but changes introduced have been extremely gradual and primarily at the secondary level. Government policy intentions were never very clear and this uncertainty led the Bank to, e.g., drop the secondary education component from its first project and contributed to the failure of the Village Education Centers component (post-primary non-formal rural education) of the second project. Moreover, the heavy financial burden imposed by the education sector always led to a somewhat ambivalent attitude towards primary education, an attitude apparently shared by the Bank, at least for a while. The Government's policy during the early seventies of limiting primary school enrollment growth to 3Z per annum was dictated primarily by the budgetary outlook. The Bank believed that the policy should be feasible "in view of ... the success the government has already had in controlling the rate of increase of primary school enrollments.0122 296. The Bank's three education projects during the 1970s focused primarily on technical, agricultural, vocational and secondary training and were quite successful in strengthening the country's infrastructure in these areas. Lower than expected student/teacher ratios and a high proportion of expatriates on the staff have often called into question the sustainability of several of the project components, however, because of the high costs per student. On the other hand, the issue of the role and importance of primary education was dealt with very little in either the Bank's economic and sector work or its project lending, except for the third project which contained components to study ways to reduce the costs of education, especially at the primary level. 297. Cost issues, both the high unit costs and the size of the total education bill which has been especially detrimental for the development of primary education, were and have become increasingly of prime importance. Reform was clearly needed and frequently under consideration but little happened. Nor did the Bank deal with these issues forcefully in its economic and sector work. Rather the Bank seemed to want to believe that somehow basic reform was taking place. Over time, the Bank did become increasingly skeptical, but it also continued to be ambivalent in its economic reporting, thereby obscuring the real issues. The 1976 CPP states, that "the Government has accepted our philosophy and has embarked on some major restructuring of its hitherto elitist education system". The 1978 CPP, on the other hand, comments that "we are struggling with the educational system's continuing emphasis on traditional European academics and a de facto low priority for education of the rural population"; and by August 1983, after elaborating on the virtues of primary education, the CPP continues: "Government strategy is not yet oriented in these directions. Its past policies have resulted in gross neglect of primary education". 122/ World Bank, Technical and Agricultural Education Project in Senegal, Report No. PE-30a, May 26, 1971, para 1.12. - 133 - Yet, four months earlier, a memorandum on primary education also stated: "The Government's position of placing a high priority on the development of primary education is sound.0123 298. The advent of Senegal's financial crisis and the accompanying structural adjustment programs appear to have finally brought to the fore the basic issues of the size and cost of the education sector and its excessive orientation to higher education with enough force so that action is being taken. A sectoral adjustment program designed to make the system more efficient and specifically to lower unit costs and to expand primary education, without additional public resources but rather by restructuring intrasectoral budget allocations, is now being supported by the Bank's fifth education project. 299. In retrospect, having identified the major weaknesses of Senegal's education system, i.e. the need for manpower training and the low levels of primary education, the Bank did respond to the first of these early on but only very recently to the second. At present, not only in an absolute sense but also on a country comparative basis, Senegal ranks extremely low in basic education indicators. Although the importance of primary education as a fundamental basis for the long-term development of the country was long recognized, the tight budgetary outlook and competing priorities held up the growth of this subsector, and the Bank unfortunately, for many years, did little to encourage the Government to increase or redirect resources towards it. While other donors were generally even more supportive of the elitist bend of the system, it can certainly be argued that the Bank should have tried harder than it did to lend its support to primary education both in its economic and sector work and through its lending. E. Aid Coordination (1) Introduction 300. Senegal is a very privileged recipient of foreign aid in West Africa. Total net capital inflows from all sources averaged half a billion US dollars per annum during 1980-86, about three-quarters of it in the form of grants or on highly concessional terms. On a per capita basis this net flow amounted to $80 per annum, nearly twice the average of $45 per annum for the rest of the Sahelian countries. While a total of about fifty countries and organizations are involved in providing assistance to Senegal, the bulk of it originates from about 12 major donors, with France accounting for around 40.0% of the total during the past 15 years and IBRD/IDA for about 8.5Z. Not only has this large inflow of capital had disappointing results in terms of economic growth, it has also created administrative and financial problems for the Government for a number of reasons including at times a lack of coordination among different donors. 123/ World Bank, Senegal: Sub-Sector Memorandum, Primary Education, Report No. 4226-SE, April 12, 1983, para. 4.05. - 134 - 301. While the Bank's share in total aid to Senegal has traditionally been relatively small, there has been some increase in recent years with the initiation of structural adjustment lending. More importantly, how- ever, the Bank's role and influence in economic policymaking in Senegal far outweigh the size of its financial contribution: since the late 1970s the Bank has taken a lead role among donors in many ways, including in the area of aid coordination. Aid coordination has become a focus of attention with the beginning of the structural adjustment process in 1980, the consequent discussions and reviews of development programs, and the convening of the first consultative group meeting in 1984. There has developed in Senegal, both on the part of the Government and on the part of the donor community a genuine desire to coordinate foreign assistance programs, with the Bank having played a major catalytic role in this process. 302. Numerous mechanisms have sprung up to ensure improved aid coor- dination. Within the framework of the structural adjustment programs the Government has established an interministerial council and several coordi- nating committees. At the request of the Government the first meeting of the consultative group was organized by the Bank in D.cember 1984. This has allowed the Bank to disseminate its economic analysis and policy recom- mendations to others, making its advice more effective, and to begin to provide systematic guidance simultaneously to the Government and the donor community on the types of assistance required and the phasing of projects. Following this meeting the Bank began to help organize sector meetings to discuss programs and policies, at first in the telecommunications, energy and agricultural sectors, and subsequently in others. A second meeting of the consultative group was held in the spring of 1987. The Bank's resident mission also convenes informal meetings at regular intervals to inform the donor community on progress in the structural adjustment program. A prime example of the now sometimes excellent coordination among donors is the Common Food Aid Counterpart Fund. Created in 1985, this Fund, in which the Bank has not been involved to date except as an observer, brings the Government together with about 10 donors in matters of agricultural policy: joint decisions are teken on the management and most appropriate use of counterpart funds of food aid, and the scheme is praised by Government and donors alike. (2) Problems and Constraints 303. The substantial progress towards improved aid coordination in recent years, and especially in the machinery to deal with it, does not mean that there are no important problems. First, at the macro-economic level there is clearly a great need for a much more intensive exchange of views among all parties concerned than is currently the case. For all of its merits, the consultative group meeting is too short (1 or 2 days) for a thorough discussion of major issues, let alone for a reconciliation of differences. In the case of structural adjustment policies in particular, several donors resent the fact that they are presented with a final product agreed between the Government and the Bank in which they have had no input. As a rule other donors are not associated with or even kept informed by the Bank at the stage of conception of the policy package, unless they happen - 135 - to be deeply involved in a particular sector at the same time. A good example was the set of industrial policy actions included under SAL III where some other donors strongly felt that the rate at which the Bank pushed for change was unrealistically fast and that they should have been involved in the design of the program. At times, insufficient commu- nication among donors can result in serious misunderstandings. In one particular instance both the Bank and one other major donor proceeded to propose policy changes and impose conditions on the Government which led the Government to sign two apparently contradictory agreements. While the matter was eventually resolved, it serves to illustrate not only the need for more inter-donor discussion but also the desirability of keeping conditions clear and simple and avoiding multiplicity of conditions if they are to be taken seriously. Probably the most intense dialogue and closest coordination among donors takes place between the Bank and the IMF. As argued elsewhere in this report, however, the close cooperation between the two institutions to a large extent results because both attach prime impor- tance to the restoration of financial equilibrium, at times to the detriment of growth considerations. In a number of instances, such as pricing issues, Bank staff have opted for suboptimal solutions for growth because of budgetary matters. 304. Budgetary processes in donor countries which set aside a certain amount of aid for a given country with little upward or downward flexibil- ity can also complicate the task of aid coordination. Donors will want to go ahead with their programs regardless of what is warranted by country circumstances at any given time. In the particular case of Senegal, there is the added problem of the privileged relationship between the country and many donors, and therefore the generous supply of aid. This follows especially (but not only) from the unique relationship with France. As a result it may be difficult for the Bank to persuade the country (and other donors) of the value and importance of pursuing certain economic policies if alternative sources of finance are readily available. To some extent, this appears to have been the case under SAL I, when the country "was able * to receive additional assistance from bilateral donors without conditions * on performance, at the time the Bank postponed the release of the second tranche."124 305. Second, at the sectoral level, coordination among donors differs markedly from one sector to the next. The earlier referred to Common Food Aid Counterpart Fund is indicative of the close cooperation that is taking place in the agricultural sector where a fairly large number of donors are actively involved. In some other sectors, however, coordination has been extremely difficult, sometimes in spite of very intensive efforts on the part of the Bank. In certain cases, some donors, especially smaller ones would rather be left alone and have little interest in coordinating their activities with others. There are instances of donors not even being aware of important pieces of sectoral analyses produced by the World Bank with major implications for the very sectors in which they are involved. More 124/ OED, PPAR: Senegal Structural Adjustment Loan and Credit, para. 63. - 136 - generally, commercial interests in some donor countries are very strong and at times prevail over what the local donor representation or the Government believes is in the best interest of the country. It is especially difficult to counter these tendencies in the case of grant funds. 306. In recent years, and especially with the emphasis on improving investment programming and the decision at the 1984 consultative group meeting to have the Bank play the lead role in this endeavor, the Bank has made laudable efforts to bring the donors together at the sectoral level and, though problems remain, has been quite successful. A case in point is railway investments where industries in several donor countries lobby extremely hard to sell their equipment and machinery, thereby helping inflate Senegal's investment program in railroads far beyond reasonable requirements. The Bank's concern with excessive railway investments has been continuous: it was made a major issue during the implementation of SAL II and preparation of SAL III, was the subject of important sector analyses and was discussed at two consecutive meetings with concerned donors convened during late 1986, one in Washington and one in Dakar. In the end, though some projects did slip through, much undesirable investment was avoided and funds were diverted to better uses. 307. Third, there remains one area where coordination is almost totally absent, i.e. in the area of technical assistance. In this case there are no mechanisms through which overall needs can be reviewed in a systematic manner and through which assistance and approaches subsequently can be harmonized. Donors, in fact, compete with one another in many ways: remuneration offered for the same job by different donors is seldom the same, hiring practices differ, and there never appear to be shortages of funds: when one donor for good reason decides not to fund a certain activity, another donor is usually ready to take his place. (3) Bank Role 308. As was mentioned before, in recent years the Bank's efforts to overcome the obstacles to better aid coordination have been intense: a considerable body of economic and sector work, the convening of consul- tative group meetings, ad hoc donor meetings on individual sector programs and policies. Good progress has been made and one of the obvious conclusions therefore is to try to do more of the same. A more active role for the Bank unavoidably means a more manpower-intensive program, which clearly has limits. Some sort of division of labor amongst donors would seem to be in order. 309. At the macro-economic level, better cooperation among donors will require for the Bank to have its views, economic analyses and policy recom- mendations not only known by other donors but generally accepted. The most effective way to achieve this is to have the donors,at least major ones, intimately involved in the development and design of these policies from the very beginning. A good vehicle to foster such close cooperation in the case of structural adjustment programs, e.g., is for the Bank to ensure major donor participation directly in the preparation of the program - 137 - itself. Several of Senegal's donors believe that current disagreements with the Bank on important policy issues could have been avoided if the preparatory studies had been financed and executed jointly. 310. At the level of projects, the Bank has made substantial efforts to help the Government rationalize its investment programming by developing sectoral strategies and prioritizing projects, and to have donors conform to it. This is proving to be a difficult task, however, partly because of the extreme scarcity of good project information to begin with, and partly because aid programs carry a mixture of objectives which do not necessarily correspond to those of the government investment program. Nevertheless, with Bank assistance, slow but steady progress is being made in improving project preparation and further strengthening investment programming; and, in order to get the donors to focus more on country priorities, a strong case exists here also for the review of the investment program to be done by the Bank jointly with major donors. The practice already adopted of having different donors take a lead role in the review of those sectors where they are heavily involved (the World Bank in transport, France in education, etc.) should be encouraged; still, the Bank has an added role to play in this arrangement, i.e. that of ensuring that all donors fulfill their lead role adequately. On the other hand, in some areas the Bank has clearly lagged, such as in the Common Food Aid Counterpart Fund where the Bank should be playing a more active role. 311. Finally, in the day-to-day operation and the actual implementation of development programs, and this would include technical assistance, many problems arise, especially for the Government, stemming from inconsistencies in the approach of different donors to such things as remuneration of nationals, expenditures in local currency, disbursement procedures, and the like. The local donor representatives are in the best position to know these problems and come up with solutions which would make for a more unified approach. They appear, for the most part, also to be very eager to cooperate more closely with one another, but they do not have the power to do so because "decisions of importance" are made in the capitals. Progress towards harmonization of procedures and better coordination at this level can only come through more decentralization of the decision-making process. This will require not only devolution of some power from the capital to the local representation but also, in the case of many donors, considerable strengthening of the local mission. 312. In the end, however, it must be acknowledged that there always will be limits to what the Bank can accomplish in the area of aid coordina- tion, given that the objectives of different donors are never likely to coincide fully. A genuine solution to these problems can only be arrived at over time, as Senegal's in-house capacity for economic policymaking and investment programming strengthens to the point that the Government is in a position itself to accept or reject aid proposals in accordance with its needs, its absorptive capacity and its own priorities. -- --- д t�,d'!' - АЕЬ ch д!е цL I 1I впд tI1 $�L-T Ооьвгпиn! СоvвгnвмЕ СвЧ- 8впk tltiп Есаповiе РглЫвм $утр� Мввwги Теkм Мвввигвв Рlвnnвд ypl� /1 lЬniWrine Сви�iвl Dslи 1. 5lиррiвА всоповiе - 9в11iпр ги1 iпсои Ior - мnаипсвипЕ tolbliмsl 1. Iпсиsи in пhвгв о1 дiгкеlу Р - iпвиlипt tвilinpв - lhrch 19В1 ' рвпвгsl гвviи рговЧ апд 1w игЧin populвtion АввввЫу ef 5-уввг вслnмiс ргадисlirв iтиlвмЧ lгев �9R ' гввliаЬiап of даи Ео 1ог иеаnд tгвпсhв гвlвви pгoduetivilу of ргоицв рпрги conЧininp в Ео еlои Ео 851 е1 ЕоЕв1 ц! о1 iпvввfмпЕ in ihв fо1lеввд ►у вnnивl смwl- iпьиЧмЧ ' iпсговид дрвпдвпев en гввliвlic, вогв productiaп- 2. Giliпp м мвгвll вnд puklic Р дiгкЕlу pгoductivв ЧЕiопв in 1вЧг Уввг• forв�pn иvinpв егiмЧд invиWnt рlап iпьиtвмЧ ter 8iаЧ Р1и исWгв - Msy 19�3 - firвE иьiи (Dвс. 1079) 9. ЕвWliвЬипЕ о1 ргоjкЕ IR ' еовЕ Ьмв1iЕ мвlувiв for сов8iпд riEб ввгiодiс iдмlilieвEian, рграгвЬiм впд в11 ргоjкЧ в►оьв wрвг•iвion 7А ргрjвсЕ вrаlивlion мi! t500,000 вп1 iрlвип- 1. InEroducEiм oI вiniв�w гоЧ о1 Р ЧЕiеп о1 иlвеtiап гвЬигп вЧnдвгд criЧri• II. Тоо 1ов доивlio Fu61ic,Asslg�: - liмпсiпр о1 еоипЧгрвгб 8. 1515 ef риЫiс iпвввЧмЧ W Ьв П ' гвviвв о9 дврги о1 - Msy 1903 - кгвмiпр гиli- ивiпрь соврвгвд - иvвгв liвuilitY рго6' lиnдв !о lprвipn ргоjкЧ fiпмсвд lгав пвЕ µЫ ia ивinрв веЬiивемЕ о1 pu►1ic aEienв of 19l0/Е1 ЬидрвЕ Ео iмевtипЕ 1ви ot thв Тrввwгу lгав овп иriпрв ог лЕАвг in 1980/81 ввviпрв rвliов ипдвr яввеиlian впд !hв рrороввд 1 1вvв1 spillinp оvвг оп Ее tЬв исигв воигсвв in 1980/81 6. ?tif о/ P�1ic invuWnE W Ьв 1Т 6 апд 6 1981/У1 Ьидрвt оп ргоргви � гввt of Чв всмоq ►идрвL liмпсвд lгав пвЕ puL11e иriпрв •i!А public иriпрв рвгlог- W �О - lвсk о1 соипЧгрвгL in 198dJ&S anc• fоllо.вд ►у впмлl ' funds for forвipn' еапwllвtioпs in 1вЧг уввга liппсвд ргоjвсЧ - Aikв in lогвiрп двЬЕ - lоввг соrвгсiвl Ьеповiпр 7. Свiliпрв м cowrciвl Ьоггавiпр П - forвipn дйt игviсв ' Dвс. 1983 - cмwltsiion ' игviев гвlio in 1979 !. 1b еаввге{в1 бorrwinp 6о А гвЕiо Ео М гдкд W - вliвinв6ion о1 dвficiб in ргоjкЧ ЕЬвt свппоЕ liпвnсв Чв 18i о1 вврогt игniпрв СлvвгпвмЕ'в вinkinp lund двбЕ игviев lгав ЕИвiг ом ивА апд oinЧinвd Ьв1ов EhвL in 1980/81 Ьидрвt ♦1ов 1вvв1 - гюгрвпiяtiап of вinkinp �iv■Ч исЕог: - сгиlion о1 Aouвinp Ьвпk 0. Iиимсв о1 hipfi iпЧгиЕ и•inps А lипд - сгвдiЕ вЧгvвtiоп in рвгtlу Ьвид м иrinpo евгfi4iиЧв tor Ьоив1М ввсЕог вои рвгЧ ot ЕАв вlfогЕ о1 свпдiдвЧв 1ог 10. А11 вdditianвl лгдfЕ вiпев А ргivвЧ исЕог lипв (1979) Dве. 81. 1979, W ро ввеlивi•в1У - лииivв ии of егвдiЕ - iпсгвви oI iпЧгввЕ гвЧ Ео всаnмiе prioriEy ввеlогв in оЧвг рвгЧ о1 ЕЬв вirucWr• Ьу 4.8 рвгемЧрв �, а privвN иеtог poinЧ � � s л к г.. V Cmditionв вгв еlвиiliед inW 1i•в дi0lвгмЕ евЧрегiвв, i.в.: (1) 8- 5tл+дiвв; (4) Р' РгвсаnдlЕlоnв; (s) А' AeEiмr, (�) П- IпЧгидiвЧ ТвгрвЕв; мд (8) 1Я ' InвEiWtionвl р~ гfi Rвfоги. Рог е■рlвмЕiм undвrlyinp Ni• с1ви11iсвЕiм, и• Ч■Е (СЬврЧг 1I. 5вс!{м• D мд Е). и v А �j Ccans�d.) Соvмпип0 Ооvипипе Gи- �пЧ Main Есоповiе РгеАtввr $вlдlФ�р Мвввигвв Тв4м Мввиги гlмnвд eerr /1 Моп3ler3ne Сгис�вl Dвглк III. Ргiсв дiвеогtiмв ' вргiсиlWгвl ' tев proundnue ргодисеiм - iпсгвви /вгигв' ргiевв вв f1. 8Чду м fги rrkвei�p вМ 8 - гви1Ч о1 вWду - Jим ]9!3 - дiкииiоп of ргодисвв ргiсвв о1 Мву 1У80 ргlеи fог довиеiа lоед егqв аfведу гии1Ч - вllвiпsеiм о1 иЬвiдiи ап 12. SWdy м opeirl pгicinp pilty 3 - гвw1Ч о1 sWdy ' 1lвreh 1W1 - вппивl ргiсв вргiсиlЧгвl iмиЧ. невре Ior ввреr6 егорв oonwlЧtionв Iвreilisвr 13. SWdy on ehмpinp /вгеillквг 5 - гввиlЧ ef вWду - 1lвгеh 10р1 - conwltstionв еаумiеiм впд рiаи - оvвгrкlивд ' 9вв0 р-овiпр iврог4 ' вliвinвCion о1 ввроге гвви 11. IпerodueЫon оу 10i вврогб А ' iрlвипЧеiт ' WrcA 3081 ' гвlии исопд леАвпрв гвЧ м1 вlиррiвА ирогЧ исре еhавв м рвоипдnи4 wАвiдiвв Ъгвп�в впд рАоуАаЧв in 1У7У 15. Iмввви Ьввiе iвоогб диliи А - iврlвипЧеiм - WreA 19р1 - гвlвви исапд -/iквl гдогв in 1WC lгав 10i ео 1� trвnahв - iпегии ►ввiе dutiи fгов 18. 5W1iвв м finвl воддiеiвв 5 - гввиlL о1 вЧду - Мву 19р1 - гвrlи 6! Ч 10t вв о1 1breA 19�0 iпсмlivи вувЧв iврlимиtiоn of lirac ивwгвв of b►ic iyore диеу/вqог! wbidy вувЧв - �Амв 196� - inero/исеiм о1 finвl вувЧа 1 - емwиг ргiси ' Чо hipA игЬвп - вtiвinaEiм о1 ви►вiдiквд 17. Esвinp oI dowEia рг{ев �nпЧо1• Р - дкги - Овсвiвг 19!(1 F'' � емвивреiаn иг►вn 1on1 ргlеи (1УТЬ'!0) О 1. n. Мiwмрвипе in ' исвмirв ЬвпУ егвдiе - впд оГ вЧЧ-рвгtiсiрвtiм 1�. Nвwm о1 вои р•гвµ►1iа А pnµЫie исtог вnд тгвкwry вдьмевв pliey in 10Т1 ceбlvieiи fo Чо ргiьвЧ исЧг - Ьвпkгирееiвв ot рвгв' 19. Ргоргвв of wd1G вnд tгв ninp о1 А ' видiб гврогW ' Sирвг�iид eArauph TI projкt ри61�е мЧrргiввв ввпврвгв м1 кеамЧпЧ ' роог 1iв1д рвгlегаnи oI оЧвгв - Тгввигу вдrвпеи мд вЧЧ Z0. Ргорги сапtгвсЧ ►вЧип А - пивАвг мА емЧnе о1 - Junв �0, 1Уl1 - WЧ1 ot 10 ривгвпЧи оп ►мk ar.die Оеьвгnипе вnд рвгврибliа •горгвв емtгвеЬв ргорnв са+tгвеЧ W fв drsвeically гвдисвд in 107р мЧгргiви еапеlиlд ог in ргркгвtiм (wpвrriвiм ТА ргоjвее) У. Уиkпввив in - оvвгвtвНвд Ьигввиr - сгввеiап о1 51>1iM 1ог Т1. ТиЧ вiЧ в сеорвnЫvв IR - ЧвЧ {п 8iм 5аtоив вгвв ' Ивгсh 19р1 - liгве 1меiЧЬiмв1 гвеiи .iN 1ов виррlу ef вргlеиlЧгвl вerweWrв )вsвд an v111врв ргоивв •iвеииiмв м вгоргви ие'ир in вргiаиl- рвгfогипсв мд AipA inµb (егор rrA•e1np, сгвдll, (npre `� Чгв1 исtег еиЧ wppliи вn1 вид вЬоеkв) �� К Ч r О М и w -141- Апаах 1 Pags 3 of 27 i � � '� � � 5 Ъ � 8 ._. г<< : з3 : е ;�3�s � = 3 : ; 3 � � g � : ' ♦ 3 � цУ 3 .. fi 8 � � : ! � � г � � g � _ � я = ,г � ��•-•`а � .. � : • ¢ з ���� а ! � � : � fi в ~ а о '> � � Z � • 3i 3 g о�� s w У' ^ � 3 а� � = s �sa��s`":~'s� �ъ � s : � `rsg�� ���~'ъ � g g й � � � У• 3 S У °.. � : � • • . � i : .• �� � � � : � 3 У • �У � М � ♦ ♦ • ♦ У }_= s 2 3 S� � У ` У� и � g i : М t ь л • , 1 1 1 's � ` < < а < а S _ _ • . � S � � у V � t Р • �! � � ° т � � • � р • � Z ` • С О ' ъ � � в ь" г s ^ � О г � � • � С � � �_ � °C � р а g 8 о $e g r гS� Q ii С У Л • О S �C � " • v� � � � ,У� S : : ` . � к GGY i- а 3 а� �а � �, ZЗ й с� к �' о $ S ; � ь � ё • с � -:: �ъРа � а; �.. � $�� ���' �s � �� � п = � а � �ьг в��ъ ё: г e , • � ~ л � � р � Уc У1 � � � О д Z У � н . • • У О в ♦ � � 1 ° � • ё � • � [ � �У г � ёУ У ° : � '� ♦ с • � • • ' � ; 3 ! Е � �� � � � � и � � г • о • в л • � �142- Аnпеа 1 Page 4 of 27 :' с � ' s � К'i < < t < �i е � . о а ; 3 : • '> : `о : а о=`г=- �ъ :� s у : трΡ. о а: ; � �C � ` -° i ь ° � в °л � iв ■ г � � S � � s У � : С • 5 �сс У� � L � � •о � = � � : аь с а� 3� s` оf У�= ё� о � У 'в � уΡS � : 1 i Ё � йш i д$ i 8 л" t^'• •• к к я �� g ё � i о � � � 4 � � S � i � в � 3 � 5 - . � й �. х Y f � • - с - i У � � � = ё � • � н з ± ._ в в . л � g g v °v л 0 � � _ • � в • � S S У л' � - лв 6 О° �. � М G Q�( • �q рΡ ё ii ' S i § о е е � ii с • уΡ л К Q z v вУ в � t �в О � v � � £ .а! д е в4 0 • • � � С _ 3 � = В �1 � � й � У а •= s j � MINNIMM MEMO SAL 71 Action Taken Cate- Monitorable Actions Structural Issues by the Government Klasures to h2 Taken Gory and Tinino 1. Agricultural Sector Incentives c2reals Policy Excessive dependence an food Initiated preparation of cereals 2. Identification of a coherent P Preparation of corep;s develop- Imports; subsidization of development action program with cereals development action ment action program satisfactory Imported rice; Imperfect sub- FAD assistance. program focussing on produc- to the Association (condition of stitutability between local tion factors, processing, and 2nd tranche raises cereals and Imported cereals. Increase in producer prices In price pollcies. 1996. CSA's operation is to be limited Rigid producer price structure 2. Introduction of producer A In size (about 26,M tons In and control* on marketing of Doubling the consumer price of prIco-support mechanism for 1988/87) and applied to sup- cereals. rice since IM; eliminating cereals managed by CSA. porting producer prices. consumer subsidies and providing Excessive Government inter- adequate nominal protection. 3. Maintenance of at least a 26% A Consultation with the Bank every vontion In food Imports and nominal Protection for domes- six months from date of credit marketing. Adoption of objective to tic cereals by adjusting, as effectiveness to determine need stabilize rice and wheat flour necessary, the retail price of for adjustment in vice price, Import volume at 1984 level for rico, following consultations the modalities, level and timing the 1986-89 period. with the Bank. of the adjustment. Lifting of all barriers to the 4. Progressive *tot* disengage- A Effective separation of th-s rice free marketing of cereals ment from rice Import opera- operation from the CPSP and (except paddy) - Mons and privatization of participation of private isr- rice distribution following porters/traders in Import and completion of ongoing CPS distribution of rice after Way contract (May 1988). 1988 condition of 2nd trAncho 00 X 0 ph IQ SAL It (cont'd.) Action Taken Cate- Monitorable Actions Structural Igsues by the Government Measures to be Taken Gor. and Timina I. Agricultural Sector Incentives (cont'd.) Fert1lizer Policy Elimination of fertilizer levies 5. Progresslve elimination of A Fertilizer subsidies applied in Excessive Government inter- on producer price of groundauts; fertilizer subsidies, import accordance with guidelines out- vention in distribution, sale of fertilizer on cash basis liberalization, and privatiza- lined In LDP; subsidy limited to including heavy subsidies. only; decision not to finance tion of fertilizer distribu- CFAF 24/kg during the 1988/87 subsidies through Treasury tion by 1989/96 season. season. resources and relying exclu- sively on subsidy scheme financed by external grants in 1986/88 growing season. I- Groundnut Sector Liquidation of SONAR, the pars- 6. Progressive reduction of A Central security stock limited Excessive state intervention statal managing Input distribu- central groundnut seed to 66,60W t during 1986/87 in seed distribution and tion; reduction in central security stock. season; marketing of groundnuts groundnut seed stock from leading to heavy burden on 120,086 t to 166,6 t during 7. Complete liberalization of A Discontinue free distribution of public finances; low producer the 1985/88 season; elimination sed distribution as of seed; farmers to store own seed prices; Inefficient structure of seed and fertilizer levies on 1986/97 season. or purchase from oil mills at and management of oil mills, producer prices. cost. Including heavy state subsi- 8. Elimination of all guaranteed A dies to cover direct costs. Reduction in half of the coverage of oil mills' fixed Preparation and Implementation guaranteed coverage of oi I costs. of a seeds action program mills' fixed costs. Including measures to help 9. Restructuring oil milling IR farmers store their seeds under Increase in the number of operations by merging SOMACOS proper conditions, by June 1988. private traders lIcensed to and SEIB. purchase groundnute. Finalization of contract-plan with the new company emerglag from the fusion of SOKACOS and SEIB by March 1987. 0 rf SA 1 (cont'd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken naty and TIMno 1. Aaricultural Sector Incentives (cont'd.) Institutional Reforms Liquidation of STH. 10. Continued restructuring and P Finalization of 'lettres do Institutional weakness in disengagement of RDAs. mission' for SODAGRI, SODEVA, pIe--IRg, investment program- Restructuring of SODEVA, SODEFITEX, SOMIVAC by Mar. 1988. ming, budgeting, statistics. Including 665 cut in staff. 11. Strengthening Rural Develop- P sent Ministry's abilIty to Excessive involvement of rural Restructuring of SAED, and plan, program investments, and development agencies (RDAs) In signature of a flttre de execute budgets. production activities; over- mission.' staffing; poorly defined objectives, leading to heavy burden on public finances. II. Reform of Industrial Incentives General Measures Elaboration of a comprehensive 12. Public declaration by Govern- P Announcement of the now indus- Strategy focused on import Industrial incentive reform sent of the main elements of trial policy by early Feb. 1988. substitution and high protec- program. the new Industrial policy. tion levels, discouraging exports, encouraging fraud- 13. Preparation of detailed Action P Completion of Action Program by ulent Imports, and increased Program for the period July end-June 1988. costs to the economy. 1988-Dec. 1986. Protection Proliferation of special ad- Freezing list of products 14. Reduction in absolute and A Preparation of aw draft tariff ventages (exon*rations, exsep- subject to quotas or prior relative levels of Import code by July 196, and adoption tions) distorting incentive authorization and increase in restrictions; gradual *ite- by Governmeat of the draft system, encouraging sisalloca- quotas for 1986. Instion of prior import tariff code satisfactory to the IF tion of investeents, reducing authorizations; gradual Association by Oct. 1986 (r - Government revenue, and being phasing*out of quantitative tMon of 2nd tranche release). o costly to administer. Import restrictions. SAL.11 (contd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken g2U. and Timino II. Reform of Industrial Incentives (cont'd.) Protection (cont'd.) Elimination of prior authori- 16. General reduction and harmoni- A Elimination of prior authori- zation for import of certain zatlon of tariffs. zations for categories of goods not produced in Senegal products in at least one sector starting Jan. 1986. 18. Generalized use of Common Law A (to be determined in consulta- system. tion with the Association), by Adoption by Government of July 1986. amendment to the tariff code 17. Review of Investment Code. S reducing the tariff rates on 15 Review of existing *conventions* categories of products. 18. Replacement of existing draw- A with private enterprises by end- back system by transferable 1986 with a view to their pos- Renunciation of special 'Conven- drawbacks. sible renegotiation within legal tions' system. constraints. 19. Basing export subsidies on A Adoption of principle to revise value added at border prices. Adoption of new drawback system the drawback system. by July 1988. Adoption of principle to modify Adoption of new export subsidy the export subsidy scheme so as system based on value added at to base It on industrial value border prices by July 1986 added in international prices. (condition of 2nd tranche Governmnt Control Excessive controls with regard 20. Phasing out controlled indus- P Review of the administrative to: price levels and struc- trial prices in tandem with procedures in the framework of ture; investment administra- measures to reduce protection the Labor Law regulating hiring, Mon; and labor wage legisla- In the respective sectors. firing and temporary work con- tion. tracts, by end 1986. ** 0O 0 eh ba 1 (cont'd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken ggy and Timina II. Reform of Industrial Incentives (cont'd.) Government Control (cont'd.) 21. Simpler Investment Code S Preparation of measures to administration and ocourage job-creation by labor export/import licensing Intensive Industries through procedures. reduction in social charges, by end 1988. 22. Fostering labor mobility P (direct recruitment by Further actions to be identified employers, faster action by in the Action Program. labor offices, more flexible interpretation of collective firing rules and renewal of term contracts). 28. Incentives for higher produc- P tivity (linking wage increases to productivity gains, lower social charges for labor- Intensive industries). III. Imroved Quality of Public Investment Institutional Reform Excessive emphasis on external Completion of comprehensive 24. Improved financial supervision A Extension of debt monitoring resource mobilization at the study of investment programing of investment program, and system to project level by June expense of resource alloca- and budgeting system and computerized link to debt 1988. tion. adoption of recomindations on monitoring system. Institutional reform. (I to -0 SAL1 (cont'd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken 2tZ.. and Timino III. Imoroved Quality of Public Investment (cont'd.) Institutional Reform (contd.) Inadequate policy and planning Reorganization of Ministry of 25. Introduction of now system for P environment. Planning and Cooperation, as of estimating recurrent costs. Jan. 19688. Inadequate interministerlal coordination. Investment Proaramina Overambitious Investment Reduction of public investment 26. Introduction of a three-year P Preparation of draft three-year targets and large number of targets for 1985-89 to levels rolling Investment program and program 1967/88-1989/90 by June uneconomic projects. consistent with the macro consolidated budgeting system 1988, and its satisfactory framework. as of 1987/88. review by the Association (sad- Deficient project selection Ition for 2nd tranche release. and evaluation. Use of Priority Action Programs 27. Emphasize rehabilitation and P in the preparation of the maintenance operations. Infrequent reviews leading to 1985-89 Plan to select priority build-up of distortions. project-. 28. Stricter technical, economic S and financial project analysis Unclear project ranking Preparation of 1985-87 two-year criteria. criteria. Investment program and its review by the Bank. O SL11 (cont'd.) Action Taken Cate- Monitorable Actions Struetural lueas" by the Government Measures to be Taken SAU. and T1mine IV. Accelerated Reform of The Parmublie Sector DivestIture Excessive Government partici- Adoption of a new Sector Policy 29. Oivestiture or liquidation of IR Adoption by Covernment of list pation in the economy whose Statement and Action Program. enterprises. of enterprises to be ,rivatized costs are exacerbated by or liquidated in 1986, by March special treatment awarded to Liquidation of SONAR and STH and 1988. parapublic sector (subsidies, restructuring of other par*- exemptions, debt guarantees, statals. etc.). Reduction in Indirect subsidies (e.g., fuel subsidy to SENELEC) and increase in tariffs. Rehabiithation Misallocation of investment, Introduction of 'contrat-plans 89. Improved sector data S Computerization of sector data low efficiency, poor savings system. management and reporting. base by end-196. and budgetary performance. Signature of contrat-plans with 81. Clarifying the transactions P Final reconciliation of cross- Excessive operating costs and SONEES and SOTRAC and of 'lettre between the Government and the arrears and settlement progrem losses (assumed by the do mission' with SAED. parapublic sector. by July 1988. Treasury). Restructuring of Post and 82. Improved management of P Finalization of 'lettres de High level of payment arrears Telecoemunication Sector (Oct. parastataloo contractual mission' or 'contrate-plans' for toward enterprises and the 1985). relations with the Government. SODAGRIj SODEVA; SOFEDITEX, banking sector. SOMIVAC, SENELEC; SONADIS (March Provisional reconciliation of 88. Improved control of S 198), OPCE. SONATEL, SICAP, Excessive or inadequate cross-arrears and proposals for enterprises. OHLM (end-198), SONACOS/SEIB control and supervision of the settiement. (March 198?). sector and of Its Impact on the economy. Study of modalities for restructuring public financial Poor management of enter- Institutions by end-19f8. prises. 0 ML11 (cont'd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken Gary and Timina V. Public Finance Recovery Monitored under IMF stand-by Revenues arranloement. Declining tax ratio despite Tax and price increases under 34. Simplification of the Tax Code S Submission of Tax Committee high nominal tax rates. the stabilization programs with with a view to expanding the report by July 1988 and the IMF (1988/81-198/86). tax base. subsequent implementation of Excessive legal exemptions and recommendations. exonerations. Administrative and accounting 86. Taxation of real estate in S reforms in the public sector. urban areas. Completion of feasibility study Tax evasion (notably cus6oes for fiscal cadaster of Dakar fraud). Strengthening of CPSP financial U. Reduction and harmonization of P region, by end-Sept. 1988. management to ensure timely custom tariffs to reduce fraud Weaknesses in tax adminis- payment of custom duties on and expand the tax base. Completion of program for tration (excessive number of imported rice and recovery of recovery of arrears by CPSP and low-yield taxes, inadequate arrears. 37. Computerization of custom P measures to Improve CPSP - collaboration among various statements to control fraud. technical and financial C departments, training and Discontinuing past practice of management by end-June 1988. management shortcomings, non- subsidiary agreements 8. Complete repayment of arrears A assessment of some taxes, transferring to enterprises the of customs duties by CPSP. e.g., real estate in urban benefits of debt rescheduling. areas). Current Exoenditures High and fast-growing level of Freeze of civil service staff in 89. Sectoral analysis of current S Completion of study of current personnel expenditures. 19865/8. expenditures to improve expenditures by July 1986. sectoral and functional High level of subsidies and Limiting wage bill increase to allocation. Planning for reduction of transfers to the parapublic seniority increases (4-61) in budgetary subsidies and sector. 1986/86. 40. Reduction by 6O of budgetary A allocation by beneficiery to be subsidies by 1989/98, compared submitted by June 1988. Insufficient allocation of Community participation (user to 1985/88. budgetary resources to main- fee) for primary health care, tain investments (recurrent introduced in 1985. costs). 0 M4 SAL.1t (cont'd.) Action Taken Cate- Monitorable Actions Structural Issues by the Government Measures to be Taken dor. and TlDa_ V. Public Finance Recovery (cont'd.) Current Exoenditures (cont'd.) Tighter financial control and 41. Gradual elimination of A Proposals for functional decision to reduce subsidies and Indirect subsidies. classification of direct transfers to parestatals. subsidles by end-June 1980; 42. Continuing reduction in wage A laplementation in the budget for bill and stabilizeton of fiscal 1987/88. civil rvice numbers; improving control of civil Number of civil servants not to service staffing. exceed their level of July iS, 1985 (76,114). Identification and review of Indirect subeidies by end-June 198; Otableau de bord" by end- Oct. 1986. Ratio of Wages and salaries to fiscal receipts (other than grants) not to exceed 6S8 in 1985/96, 51X in 1988/87, and 48 In 1987/88. ** 0 SL 1 (cont'd.) Action Token Cate- Monitorable Actions Str&uturl leaes by the Government Meaces. to be Taken Rory and Timina F Public Finance Recover (cot'd.) Investment Financino Inventory of Government payment 43. Settlement of Government's not A Payment of CFAF 4 billion to arreare to private sactors. indebtedness towards private private enterpries and 12 Excessive dependence on anterpris and the banking billion to the basking sector in external financing of Reduction of domestic payment sector. 5198/86 and of 10 billion and 12 Investment. arrears to enterprises and banks billion respectively in 196/87 (under Stand-by programs) and 44. Increasing budgetary resources IT In accordance with criteria Inability to servIle debt. planning for their progressive for debt servicing to improve Ilsted in LDP. Depriving the economy of liquidation. creditworthiness. financing through accumulation Increase by 191 p.a. of of arrears to enterprises and 45. Increasing role of domestic S budgetary allocation to the CAA banking sector. banks in financing investment. between 198/87-1989/90. 48. Improving debt service A Study on the ability of the 4 managIet and performance. banking sector to recycle the resources injected by Government through repayment of arrears by July 1986. Settlement of all arrears on external debt by July 19M6. Souts World Bank, President's Report (P-4213-SE) for SAL II dated January 10, 1988. *0 0 et SALJZ1 Recent Developments and Oblactives PyArCM under SAL .I SALI1I calmrv I. Public Resource Manaaement I.a. Paranubi Sntor Reduce the size of the parapublic Adoption by Government of list of 1. Strengthening the management of, and P sector and improve efficiency, enterprises to be privatized or Institutional setting (OEP and CCP) Investment and savings performance liquidated and elaboration of a for the parapublic sector reform of the public enterprises, through: methodology for privatizatlon. (June 1987). - disengagement of the State, Restructuring and disengagement of 2. Privatization of a first group of 10 IR RDAs. public/mixed enterprises (Sept. 198?) - restructuring of specific and identification of another group of enterprises, Liquidation of 8 SEMs end a EPs. 10-17 PEs to be privatized (Dec. 1987) - Improving P.E. management and Merging of SOFIDAK-CSCE for export 5. Liquidation of 9 structurally IR accountability, corporate promotion. unprofitable SEMs and EPs (Dec. 1987) planning and Internal control, and progressive disengagement of 6 Study of public enterprise tariff. PEs. - streamlining relations between government and enterprises and 4. Implementation of a program of audits A Improving transparency of and restructuring plans for several financial relati6nships, and enterprises. - reforming the government 5. Redefinition of the role and P Institutional setting for responeiblilties of public enterprise overseeing public enterprises Boards of Directors (Sept. 1987). 6. Implementation of action program to A strengthen enterprise internal corporate planning, managemnt control systems and financial accounting (from July 1987). 00 as O, SAL.III (cont'd.) Recent Developments and ablactive PrAreeas under SAL It SAL.III Catar Le. Paraubile Sector (cont'd.) 6X reduction in direct subsidies to 7. Subsidies to PEs to be reduced ta A PES In 1986/87 compared to 1986/86. FY87/8 and FY90/39 to a5 and 75 of 1985/81 levels. Transformation of Indirect subsidies into direct subsidies, whenever possible. Elaboration of contract-programs or 8. Strengthening the process of preparing P Lettre de mission' for a number of and monitoring execution of contract- enterprises. Seminar on experience programst agreed schedule of payment with contract-programs. of Government financial obligations (July 1987); compliance with financial obligations during first 6 months FY87/88 (Nov. 1988). Improvement of government sector data 9. Information system to project and P management and reporting, including monitor financial flows between computerized MIS with historical data Government and PEs (Dec. 1917). through Dec. 1986. Reconciliation of cross arrears 20. Reconciliation of cross arrears P between government and public between Government and the enterprises through June 83, 1986. manufacturing and coemercial public enterprises through J i 8, 198, and agreement to a settlement plan (Oct. 1937) Study to improve system of government 11 Acceptable plan to remove a.riori P supervision over public enterprises. controle over Etablish. Publics. by 1916 (July 1987). 12. Complementary study to reform system S of government supervision over PEs (Sept. 1987). 0 SAL Z (cont'd.) Recent Developments and Dblectives Proareas under SAL II SAL.III eatmery Lb. Public Investment Prearamiale To optimize allocation of external Preparation of a three-year public 18. Preparation and review by the P and Internal resources within a investment program (1987/68-1989/96). Association of a three-year PIP for macroeconomic and sectoral Review of the program by the 1988/89-1990/91 (Dec. 1987). framework; increase absorptive Association. capacity of investment; Improve 14. Submission to the Higher Planning P evaluation, selection and budgeting Preparation of a medium-term action Council of the proposal to reform the of projects and evaluation of plan to carry out the Institutional planning system (June 1987). recurrent expenditure; and reforms of the planning, programing strengthen supervision of project and budgeting of public investment. 16. Completion of project appraisal guide S execution. (June 1987). Improved financial supervision of Investment program and extension of 18. Transfer of responsibilities for A debt monitoring system to project project identification and preparation level. to technical ainistries in Rural Development, Water Supply, Equipment, Education and Health (March 1988). 17. Document defining project selection P criteria and methods and operating procedures of Project Selection Comittee (June 1987). 18. Implementation of the computerized A monitoring system for physical and financial execution of public Investments (June 1987). 19. Reform of taxation of externally A financed projects (part of tax reform). 20. Study of Senegal long-ter development S prospects (TOR by June 1967, completion in June 1988). SA I (cont'd.) Recent Developments and blactives Prares under SAL II SAL.1I1 CateaS Lc. Public Financ,4 Improve economic incentives. Preparation and submission to the 21. Implementation of the first phase of A Restore public savings through government of tax reform (first phase) the tax reform: application of the reducing tax evasion, increasing aiming at improving incentives to new General Tax Code (April 1987), fiscal revenue and restructuring production and investment, expanding elimination of exemption on taxes and curre:,- expenditure; develop multi- the tax base, reducing tax evasion and duties on foreign-financed projects year financial forecesting. Improve Income distribution. (July 1987). Pilot study of fiscal cadaster in 22. Preparation and lmplementatlc of the A Dakar region. second phase of the tax reform: study of general application of VAT In the Preparation of reorganization of trade section (beginning of 1988); custom administration and computeri- possible implementation of compre- zation of custom statements. hensive income tax with elimination of the scheduled tax system (end of Implementation of tariff reform and 1989). revision of export subsidy system. 23. Submission of now Customs Code to IR Inventory of cross-arrears between the National Assembly and reorganization Government and enterprises up to June of customs service (June 1987). 1988. 24. Introduction of fiscal cadester for P Daker (beginning of 1989). 25. Computerization of customs clearance P procedure (completion by Dec. 1966). 28. Annual targets for the settlement of A state arrears (June 1987, according to IMF standby arrangement). 00 OD SA I (contad.) Recent Developments and Ciectives Prearess under SAL II SALIII Catepory I.c. Pubile Finance (cont'd.) 27. Holding size of civil service at A June 30, 1987 level (68,W16) for 10987-88. Reduce wage bill as a share of current expenditures (targets for 1987/88, 1988/89 and 1989/M). Intro- duce personnel information system and improve personnel management (June 1988). Reduction of direct budget subsidies. 28. Proposals to transform indirect A subsidies into direct budget Launching of audits of indirect subsidies, or to eliminate them subsidies and of special agreements (Sept. 1987). Continue to reduce (conventions spclales) between the direct subsidies according to agreed Government and enterprises. targets. L** Actions under the IMF Stand-by 29. Undertake evaluation and programming S arrangement to reduce domestic of recurrent expenditure on new arrears, eliminate external arrears, investments launched after July 1, reduce budgetary subsidies and limit 1988. the number of civil servants and the growth of the wage bill. 89. Streamline procurement procedures P (Dec. 1987). 81. Complete review of current expenditure S to Improve sectoral and functional allocation, and efficiency (March 1988). 82. Develop multi-annual financial P forecasting (Dec. 1987 and June 18). 'a - 0 SAL.III (cont'd.) Recent Developments and Obloctives Proaress under SAL 17 SAL..1I1 CatdMi I.c. Pubicl Finance (cont'd.) Measures to improve external debt 83. Implementation of computerization of A monitoring and management. system for monitoring medium and long- term guaranteed and on-lent external debt (Nov. 1987); computerize adminls- tratlon of external grants and subsidies and of domestic debt (June 1988) . 84. Preparation of provisional national P accounts for 1982-88 in current and constant prices (April 1988). 85. Implementation of a pereanent system S of administrative statistics and conduct of two light surveys in the Informal sector (Dec. 1988). 00 88. Implementation of program to improve P foreign trade statistics (Dec. 1988). II. Aariculture 1. Pricing And Incentives Policies Cereals Reduce dependence on food imports; Preparation of cereals development 87. Implementation of the cereals A increase and diversify agricultural action program focusing on production development action program, including production and exports; increase factors, processing and prices. in particular: productivity; achieve greater - audit of CSA (April 1987); efficiency of public resource use Reduction of CSA operations and - review of performance of coarse through fostering private sector Introduction of a floor-price grains processing units before the Initiative and withdrawal of the mechanism for cereals. end of 1987; state from direct involvement in - maintain rice protection at a level production activities; and reduce Steps taken towards disengagement of sufficient to maintain Incentive to Income d1spar!tle between urban the State from rice import operations; production of local cereals. and rural areas. privatization of internal marketing of cereal except for paddy) . 0 A III (cont'd.) Recent Developments and DblactiveA Proareas under SAL 11 S I Catsoa II. Agriculture (cont*d.) CrN(cont'd.) Establishment of sufficient nominal 88. Establishment of a price information A protection for domestic cereals and system for producers and traders (July periodic adjustment of retail price of 1987). rice In consultation with the Bank. Reduction of fertilizer subsidies in 89. Fertilizer subsidy limited to CFAF A accordance with agreed timetable. 16/kg during the 1987/89 season. Steps taken towards the liberalization Agreement to complete liberalization of fertilizer import In accordance of fertilizer Imports by 1989-90. with agreed target date and the privatization of fertilizer distribution. Groundnuts Reduction of central groundnut seed 40. Groundnut Guarantee Fund to be managed A security stock. without recourse to the Budget. Liberalization of sed distribution 41. Undertake study of agricultural S and Implementation of a seed actIon incentives and price support program. determination (April 1987). Review results of the study with the Association and agree on action program (Oct. 1987). 42. Study to reduce groundnut marketing S costs (Dec. 1987). 48. Audit and plan of actions to P restructure industrial facilities of groundnut processing Industry (ier. * N 0 Ph~ iAL.II (cont'd.) Recent Developments and ablatives Progras under SASALI A I Catoar II. AlUlculturc (cont'd.) 2. Agricultural Credit 44. Covernment to define the main P principles of its agricultural credit policy (April 1987). 46. The Government will carry out a S comparative study of ongoing agricultural credit programs in the first half of 1987, to review level of agricultural interest rates, measures to reduce non-repayments and administrative costs, and review need to establish risk guarantee mechanisms. S. Diversification and Fprt Promotion 48. Agree on TOR for investors meeting P (May 1987) to stimulate domestic and foreign Investment in diversification export crops in the Senegal river valley. 47. Follow-up (June 1987). P 4. Institutional Reforms Preparation of reorganization of the 48. Implementation of the MR IR Ministry of Rural Development (DR). reorganization and strengthening of the MR capacity for investment Restructuring and/or disengagement of programing and budgeting (sa Lb. a number of rural development above) (to be done under the MDR agencies; revision and signing of Strengthening and Organization Lettres do mission' for SODAGRI, Project). SODEVA, SAED, SODEFITEK and SOMIVAC. o Mt, SA I (contod.) Recent Developments and ablactyis Proares under SAL II I.I C&tAAAr1 . griculture (cont'd.) 5. Management of Natural Resources 49. Adjustment of existing regulations on P land allocation In the new irrigated areas with a view to facilitate domestic and foreign investment and diversification crops; to foster Intensive land use; to recover part of Investment costs; and, to facilitate the reinsertion of civil servants in the agricultural sector. St. Preparation of a draft decree P codifying respective roles of rural communities and Government in allocating Irrigated land (June '97), Including regulations con4er * ** 4 access to land by outsiders of t-o community and agro-businesses. 6. Livestock Policy St. Government to define its livestock P development policy and strategy, including decontrol of meat prices, by June 1987, and Implement action program for livestock development (Dec. 1987). 52. Elimination of SERAS' marketing A monopoly of hides and skins (Dec. 1987). 58. Restructuring of SADESP (Oct. 1967). A * 0 SAL III (cont'd.) Recent Developments and Objectives Proaress under SAL II SAL III III. Industiy, Employment and Trade 1. Protection Enhance compotitiveness of in- Publication of a Revised tariff law 64. Study towards suppression of the S dustrial enterprises and their (August 1986); law calls for further Otarife de pr6clonO by 191/89 ability to reach export markets reforms in 1988. (concluelons to be discuesed with IDA through (1) rationalization of In December 1967). protection systems; (11) simplifl- cation of the regulatory framework; Lifting of quantitative restrictions 5. Lifting of remaining qusatitative A (111) Improvements in labor laws; for all products not produced in restrictions (according to agreed and (Iv) promotion of Industrial Senegal, and for metal industries calendar). exports. (July 196), packaging (Oct. 196), construction materials (April 1967) 6. Review of the system of reference S and agro-industries (April 1987). prices with a vies to eliminating them as protective instruments. Study to Agreement on a schedule for removal of be completed in May 1987 and action remaining quantitative restrictions plan to be agreed in Sept. 1987. (July 1966). 57. Reform of the investment code to: P Renouncing the practice of entering (1) eliminate current provisions which into special conventions with induce capital Intensity; (11) Intro- enterprises. duce a system of declining advantages over eligibility period; and (111) in- Undertaking of a study on existing clude clear and automatic eligibility special conventions for selected criteria. (Now Code to be published enterprises (Dec. 1986). by July 1987.) 568. Follow up on recommendations from P studies on conventions (action plan to be agreed with IDA by Sept. 197), and actual renegotiation of convention with CSS as agreed at negotiations. On 0 SAL.II (cont'd.) Recent Developments and Olectives Proras under SAL II SL.II1 Cateaor III. Industry. Employment and Ita (cont'd.) 2. Simplification of Regulatory Framework Commitment to phase out price controls 69. Continuation of the price ilberali- A In step with removal of quantitative zetion reform. restrictions. Se. Liberalization of distribution systems IR (April 187) by (1) eliminating statu- tory distinction between wholesalers and retailers; (il) allowing certain industries to directly sell to retailers; and (iii) allowing indus- tries to set up their own marketing network. 81. Streamlining of procedures to grant A Investment Code benefits and setting up of a cell within the Ministry of Finance to process benefit requests (Aug. 1987). 8. Improve Labor Laws Studies on (1) linkages between 82. Revision of the labor code to enable P remuneratiom and productivityl enterprise to (1) hire without state (1l) participation of workers in Interference and (il) renew temporary equity and profit sharing; and contract (June 1987). (111) Impact of employer's share of social charges on job creation. 88. Program of labor reforms to be P designed on the basis of studies underway. (Recomendatione to be adopted after discussion with IDA before April 1988.) I. 0 SA..M (cont'd.) Recent Developments and Oblecties Proreas under SAL II SAL.III Ca&Wmrx III. Industry. Emoloyment and Ica& (cont'd.) 4. Promote Industrial Exports Streamlining of procedures for 84. Automaticity of access to P granting importer/exporter cards. lmporter/exporter cards (May 1987). New law and decree, to rev!.. the 86. Review of experience with now system P export subsidy scheme an base It on (Sept. 1987). Industrial Value Added rather than FOS value (August 1986). New decree, to streamline the duty 66. Review of experience with new system P drawback system (April 106). (Sept. 1987). Study towards simplification of export 67. Implementation of recoemendations from A forma (SENRO committees). SENPRO commlttees (July 1987). 8. Streamlining of export procedures P (action plan by Nov. 1987). 89. Imploeentation of a program of A rationalization of the various Institutions involved in quality control (Oct. 1987). 76. Diagnostic study of ASACE (export S Insurance company) and of CICES (Senegalese Center for External Trade) (Feb. 1967), and Implementation of restructuring measures recommended in these diagnostic studies (program to be agreed upon with IDA in Sept. 1987). 80 ok N * 0 • -а .� � -�- -�. �_ _ _ se�u сиаl�а.) Rкм! Овvвlорввпlо •nd Q�jвetivи в�гив rwдвг 3AL I2 �jjj SlиярLl Ir, Ejмneial Sвеtлг RвЬвЫ 11lвls о аивЬвг ot Ргрвгаliм of rвha6ililв0iм рlвм 71. Iвploмntвtion ot гвbаЬilllвliм рlвпs IR явwгавгмL-ммд bвsks 1s егiвlв� fог l1n3� 1156 впд �G4 мд lwnehlnp of �ог B10S вдд OCS. (F1мlisвlleo of вlгивllм Lho Ьвnklпо ввеtогг гоЬвЫ 11lвt{оп/rplructuriaя tог 1йВ рlвпв for BtOS апд 8С5 Ьу Jим 1р7; двwlер гвsоигсв воЫ lisвllоа� •1!Ь в pгlvвts рвгlnвг. вхкиLlоа� 19�7•р.) гввlогв tisвae{а1 lаtвгввдlвtiм вв0 {врrоvв eontrol оvвг егвд{t Iлwntoгy о! яоvвгмм! 1iвЬi1{!1и 6о М. Аягимп! впд iрlммLвliм of А вхрsавiм впд Ьмkiпу ргвеllем. !1и Ьмklпя вкlег мд roducllos е1 ргоягsв 1ог tM иttlовмt ot RАввв 11вЧ1iliи (а1СА0 двИ!). уоwгмм! м! вггмм to мlвгргisм (owr tivв lгиг). 73. Strмplhмlпр of t1и еоnlгеl ot А рг{игу 6.aks Ьу th. Смlгвl lank (8СЕА0) вад !Ав Y1n1slry of Finsan� 1 вад of !М гоlо of t1и Caввission 1ог Воп4iпр Соnlгоl (Jим 1р7). � А �' 74. Rввctivвliм ot tбo Bвnkiny � Амк{,!{м (АРв) (.и.. 1я�7) вла гкорпlllм Ьу t�As Gмrогевва! вt АР8 вs off{e1в1 ialsгlкutor (Jим 19р). � 7�. Rвдiкliм ot tвх (TPS) м orвdit А lвtвги! (Аа. 1р7). 7А. Inerиsв 1а t�Ав в1а1w up1b1 А гвqо 1 говм! ts .вl.Ы 1sh мд врвnls в Ьм1с (Ове. 1lt7). 77. Fiмliz.Llм о* 6мklna .lud�r вiввд в! S двwlерlр довввtiе гввоьгев � вe6i11sasiм sad Г{емеiаl {пвlгиrмtв л (Арг11 19�7). Rвv1w е4 !М и г.во...ве.:1м. et !м slrep .аа " ргвввм! м в �то0nв tor lМ1г � м lврtовмlвllм (Арг11-Ове. 1рТ). Н v ��к Мbг1А бмR� Ггиbдм!'в Rврвг! (P-14DY-5� 1ог EAL III двlв4 Уву в� 1iYl. - 166 - Amex 2 DEVELOPMENT OBJECTIVES AND CORRESPONDING PERFORMANCE INDICATORS Performance Indicators In Agriculture Development ObJectivos Acaroasto Ditaggrogstod Economic Growth (0) Rate of growth of agricul- Rates of growth of production tural production per worker per worker or per capita In export/cash crops as opposed to domeati c foodcrope More equal Income distribu- tion In: (a) Relative sonso(DR) Agricultural production or Rates of growth of production Income per worker (or per worker or per capita in copits) compared to corro- traditional vs. commercial sponding variable in non- modern agriculture agriculture (b) Absolute sense, i.e. Share of population below Regional Incidence of sai- poverty alleviation food poverty line (with nutrition (DA) calorle, consumption below iscommanded daily allowance) Improved balance of payments Not contribution of situation (8) agriculture to foreign exchange earnings Reduced dependency on rest of Degree of self-sufficiency Not export performancAo by world, I.e. food security and In food consumption commodity group and not greater food self-sufficiency Import-substitution by (F) commodity group Greater price stability Composite real agricultural Price trends by commodity price index groups Other socioeconomic goals Including non-material (o.9 Ideological) objectives (0) Note: Different Investigators or portion my assign different relative weights and priorities to the various development objectives in ascertaining their contribution to overall performance. Thus, the latter can be considered as consisting of some weighted set of these objectives. The determination of what constitutes; overall performance Is essentially a normative question. Of course, a table similar to this one for agriculture would be required for the other sectors as well. Source: J. Locallion and others, Economic Policies and Agricultural Performance of Low- Income Countri", OECD Development Centro, Paris, 1987. AGRICULTMAL WTPU AND PRDUCTIVITY OF MAJOR CROPS, 196/61-1986/87 Groundnut MilletjSrahum RIcea Maira 1~at Pc"dettan åffl Y a ljd2 Prodnetlofn AtM veljde Production A Ylejdt Productlon Am8 Viajde (tons) (ha) (tons/ha) (tone) (ha) (tons/ha) (tone) (ha) (tons/ha) (tone) (ha) (tons/ha) 19/61 922,6M0 977,8ØM 8.944 892,8U 688,288 8.674 71,200 67,800 1.851 27,4N8 88,6m, 0.896 19861/62 922,8fl 1,825,6118 #.99 486,888 sa9,8me 8.484 88,$88 78,4N8 1.188 28,488 82,90N &.888 1962/68 898,908 1,813,888 8.882 428,788 804,788 8.498 89,988 71,688 1.267 26,88 81,68 8.861 1968/64 962,4N8 1,884,288 8.878 478,4M0 959,581 8.499 186,288 74,8m 1.418 28,7M8 32,788 8.817 1964/65 1,819,800 1,855,888 8.966 618,88N 1,e1,8ø 8.58 188,800 88,888 1.265 87,18 47,288 0.788 196/68 1,122,10ø 1,112,108 1.888 6,10 1,69,40@ 8.624 121,9ø 82,200 1.485 48,888 54,280 8.768 1966/87 867,188 1,114,198 8.709 428,508 996,786 8.426 124,788 87,288 1.488 41,888 68,988 8.776 19867/68 1,886,28 1,168,88 8.888 866,888 1,165,588 8.667 188,988 100,888 1.828 66,888 71,786 8.792 1968/ 9 819,68 1,198,18m 8.687 449,8G 1,068,688 8.426 81,18m 76,988 8.795 25,8Ø 80,488 8.696 199/7 788,888 968,188 8.819 684,788 1,187,888 8.612 148,100 108,788 1.861 48,988 56,401 8.888 1978/71 682,888 1,849,788 8.644 882,888 96,588 8.896 98,188 89,80N 1.848 88,1M8 5,888 8.852 1971/72 986,488 1,000,488 8.929 681,288 978,488 8.681 186,788 84,888 1.268 87,68 48,888 8.778 1972/78 670,00~ 1,871,488 8.582 822,888 986,188 8.844 87,988 63,888 8.711 28,28M 32,4M 8.628 1978/74 657,888 1,826,10Ø 8.041 618,288 1,182,988 8.468 86,288 64,888 1.886 88,8M8 89,281 8.882 1974/75 1,811,786 1,66,969 8.957 674,M84 824,884 8.697 114,619 88,242 1.877 40,629 46,778 f.888 1975/76 1,286,818 1,289,429 0.998 614,418 981,887 8.689 128,886 87,858 1.884 44,987 48,658 8.924 197/77 1,228,812 1,849,161 8.99 582,717 947,266 8.681 1ø,568 78,688 1.867 41,868 48,647 8.981 1977/78 617,196 1,182,546 8.487 417,428 916,284 8.468 88,748 28,718 1.296 46,871 76,5SU 8.881 1978/79 1,961,98 1,177,622 8.981 798,729 1,68,962 8.758 146,848 92,487 1.588 s4,948 68,587 1.810 1979/88 719,64 1,126,826 8.689 62D,443 967,178 8.688 96,567 78,718 1.227 48,296 67,688 8.684 1988/81 487,267 877,892 8.682 641,898 1,198,168 8.489 64,8668 67,174 8.988 66,777 78,969 8.719 1981/82 8M8,488 1,889,288 8.868 968,424 1,198,469 8.794 122,486 74,277 1.648 87,358 72,689 1.282 1982/88 1,8^8,868 1,182,764 8.988 654,988 1,182,627 8.687 128,865 76,687 1.692 61,388 6,85 1.818 1988/64 575,888 1,109,788 8.518 851,88 872,888 8.426 181,68 62,888 1.952 61,ffl 78,600 8.872 1984/85 498,88 874,81M 8.561 471,48 1,882,988 8.478 186,888 68,100 2.864 9,488 82,788 1.190 1988/88 081,288 684,8O0 8.994 968,888 1,386,788 8.711 147,68 78,288 1.888 147,888 181,488 1.4541 1986/87 841,288 867,768 1.841 888,788 998,15s 8.688 148,200 71,6M 2.642 167,9M8 94,748 1.189 Sousc: World Bank, Snegal: Agricultural Sector StratOgy grWf,ø Volume Il: Statistical Annøx, Septmber 1987, draft. - 168 -Annx 4 AGRICULTURAL PRODUCTION AND RAINFALL DATA, 1960-86 Groundnut Production, 1960-6 1.3 1.2 *1.00 1.a * I1,100 1 * 1.000 0.9 900 'R 0.6 Soo 0.7 700 0.6 00 0.4 *400 0.3 . . .. . . . . . .. . . . . 60 62 64 66 68 70 72 74 76 76 80 82 84 60 MiIIet/Sorghurn Production, 1980-66 09 . /1III I 0.7 *- 700 oA - I0c 0 400 60 62 64 66 68 70 72 74 78 78 60 82 84 86 Rice Paddy Production, 1960-86 goo I % I V I 1' /V , F - - 100 40 4 62 64 66 66 70 72 74 76 79 80 92 94 86 -30-0l.a 200sad r eg srf иго�«и �уРмУ./ лу вв.ге ►у с.iма.г r..г, lvas-e1 1ь� иs1м� �/сг. w lvee 1vвТ lose lоао 1оТо 1оТ1 lort 1els 1vт/ 1v7в 1аТе 1ar► юае 1vTV lveo 1а01 1+юг 1юs 1аы lоеь lvss 1гаТ тоиi --------- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ----- Аапкаии� �1е� 1уг{е. GdN Сг. 1/0 е.00 в-� Gвввикв Rieв Сг. lBiR >1.70 �.70 Твгги Иwци Сг. 48� 1.1•6 1.15 Rirвr РоtЧгв <г. 060 в.60 1.b0 Apria. G. iI Сг. /М !.'10 6.10 Dгw�A1 Rв1id Сг. 1Ч i.00 i.00 iiм 8в1оw Ln. 111J/ 7.00 7•00 Сг. Е/У 7.00 7.00 Тиги Nwvи II Сг. 670 Q.00 2•00 ОАi-Lввrввг Еу. Cr. f-O1• 1.OD 1.00 �iвиtад сг. i87 1.40 , /.40 5в/А1м Сг. НТ 6.�D t.011 Dв►i-Lвqиr Irriy. Сг. 77G 20.00 20.00 8вв11 fWrв1 О�вг. l'г. 9У1 11.00 11.00 fогввсуr Сг. 1103 У.ЭО 9.а0 Ayrie. RввввгоА Сг. t17t 1i.60 1У.60 ЕsвЧrn 1Wro1 Dвв. Сг. 1А0д 16.10 1t.f0 IгripOiм Т.А. Сг. 1бS2 1.i0 4.У0 IrгipЧm IY G�. 1066 �1.60 S1.i0 � 8йbЧ1 0.00 0.00 О.ОО {.00 0.00 6.06 ♦.b0 11.20 0.00 17.00 10.60 0.00 20.00 0.00 11.00 76.В0 0.00 1l.10 0.00 /.90 0.00 i6.t0 1N.ib Н О� IN11J;i1К'f А10 ТOtRI511 (6) �О BAiY Мувуг "'-�' Lm. 8-000 O.QO O.W {� ВОFцФП 4о. YN s.00 �.00 60F28®IT 17 4л. 1а.72 /.20 l.10 Теrгlм Ln. ц1?/ 6.00 �.00 и. 1ви f.в0 б. i0 7nvи0. Ргивовiм Ln. 1V7D/ •.80 �•60 Сг. 1 W6 4.b0 '1. 60 МкуhвЧ Iг�. Сг. 1800 Т.70 7.Т0 8r►btвl О.ОО 0.00 0.00 0.00 O.QO О.ОО 0.00 O.iO 1!.00 0.00 /.10 1s.t0 0.00 0.00 0.00 9.00 0.00 7.70 0.00 О.ОО 0.00 0 00 :q.10 ROApS, RAItR0A08. /ФRГS (1В) bi 1иу Сг. 0% а.00 9.00 ►ог0 е1 Овlгвг 1л. 4р 4.00 1.00 иt�..у сг.lав �.1о s.10 Rsilny n tл. ыа/ в.м i./o b Сг. s1t i.IIO 6.�0 � � AiгiOrs м. еа s.oo s.oo А� Иi�вву II Сг. >fдi •.00 6.00 r Рвдвг RsвЧ ln. 1711 •.W ��Ч И Иi�Авву I]I Ln. 17n ц.00 /6.00 � ОW�вг FiвA. Ро►0 Ln. 1/06 6.00 •.00 м Аilвву III L.п. 161в 11.00 11.00 N Artвtioa 2I М. 1ftб 7.00 7.00 МiуИвву 1У Iл. 1�10/ 10.00 10.00 <r. NO • 26.00 26.00 8в1iа 1bilмy 1л. 10R6 1i.30 1�.i0 ОвУм Nsrs Сг. i/M/ О.Ч =.Ч Сг. R011 /.�5 1.�6 Иiдиву У Сг. !Ч!/ 10.76 I0.76 iг. Е-016 10.7f 10.76 в�►ти1 а.оо •.оо о.оо о.ао п.1о о.�1е 1п.�о •.Оо о.0о о.ао п1.sо в.ао 11.оо Т.оо ю,ао ».so о.оо о.оо �в.ао о.оо о.оо •.оо 1а.ю и/сг. а. 1иа 1вn 1аа lвав 1т 1n1 1вп 1т 1ав tвn Im 1м 1в» 1ив 1ио 1и1 1в/о !ва 1ви 1вr � 1ва 1вn тw� гоllвг. eB1ar Aw теsссвs (в) � Т•1вs+rв � и. М/ /.�а {.та /iwвв Ewp. А Т.А. Сг. t�2{ а.а0 а.� 1Чго1. Еа�1м. Сг. lага •.а0 •.i0 {•••о Сг.1по ю.о0 ю.оо Тв1•е•вв II tr. 171/ �.ю ю.� ar►wM1 О.ао О.ао О.ао О.ао О.оо О.оо �i.яв О.ао О.ю О.оо О.ао О.ю О.ао О.ао а.ю О.оо О.оо а.ю О.оо О.оо п.оо О.ао Ч.Оf 1/rM/ А10 УАТ91 (в) sts» а в.г�. сг. т/ •.оо в.т У. fw/1/ Ел/. i ТА Сг. 1{�3 Q.ю !.ю �1гМв Мм•/. СР. 1в6�/ а.ав а.а с�.. f-ои 1г.а п.м wиг s�i, гк с.. lавв пв.оо гв.ао �►•�г�•� о.оо о.ао о.ао о.ао о.оо о.ао в.оо о.оо о.ао о.ао о.оо о.ао о.ю п.ао о.ао о.оо о.оо о.ао /.оо sв.ао о.оо о.ао м.w Ф1R'JITION q) Т•е#. ! А�. 6�re. Сг. ?ba '2.00 �.оо ( бlrсвliм II G. at0 1а.ю 1а.о0 �"' В1rсв02м I12 Cr. в0� v 11rn1 Na1W Сг. 1а10 ю.ю 1а.00 ю.ю � ц.оD ' С£.9А0 Сг. 1а!{ а.а0 а.а0 hiвry Е/�к. Сг. 17аа 1�.оо 14.о0 аr►ма� о.ао о.оо о.ао о.оо � о.оо п.оо о.ао о.ю о.оо ц.ао о.ао о.ао о.ао n.oo о.ао о.оо 1а.оо о.ю о.оо з.ю 1а.оо о.ао п.ю вАи ао оnев рр An М. акА. ТА Gr. 7М а ю { ю ТвеА. АиiвЕ•м• Сг. 10а] а.ю i.ю aAl И. 1р1/ а0.00 ао.о0 Сг. 106t ao.oU ао.00 hn 1�. 5кt. ТА II Сг. 1а0/ ц.о0 11.о0 aAL II Сг. 1/6// ю.ю ю.ю Сг. А-010 ю.ю Ч.ю ВдL 1Ii Сг. 1a0Q/ ц.ю ц.ю Сг. А-ог17 р.ю р.ю Iw�. 8•еlог Rввtг•el. Сг. 1аФа i�.00 ia.o0 ��1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 О.ю О.оо 0.о0 О.ОО а.а0 0.00 аб.а0 0.00 0.00 11.00 0.00 О.ОО Ч.ор ll{.о0 2И./о �� тотАL (ы) в.оо в.ао о.оо /.оо s.1o Т.ав а1.м ю.ео а.ао ат.ао s/.ю 1v.ю sT.so и.ю иТ.ю ат.1о 1s.ao и.sо {е.оо sв.во ив.оо 1a1.so а1п.м �� оФ0 и и о м и - � - - - - -с� � _ r1 収 →0 の1 ______ lr」一一一一 騒1. ・．.．下’柳0つ9ロ下ムて●A○ロ01→て』て’下→下』子ouoTI覇qT」4．下。つ下がり28090.&0つao &9つュnos ロ1以 ロ1崎 引山－→ー－------－ー→－-－→→－ 0.59．にIE．ヤ10-01 ヤ．029.Its.0 ヤ．0 にーヤ9.112.50.00.00.00.0 ・‘.00-00.0 羽ao L.9t6.LL．ヤL.58.E6.Ett.90.LL．ヤ9.t［ヤ．60.E[9.SL.CE.L 乞－E6.0t6.L 和11血91 日．56.Zt 乙．ケt ヤ．0t9．らE.LL.LC.90.gC.50．ヤ5.59．ヤ0.LZ．らヤ．9 ヤ．21.2[-5.a 6．乙2・ケ6”ヤ2・ヤ2．てt.22.1 ヤ．EZ・0 ヤ．tE.C6.05.Zto.E ら．ヤtg.06.t ヤ．C ん欄口り 6.EEZ．ヤ乙9.0E9.tケ0.099.tて0.955.0EI.2ヤ9.9E9．ケケ8．ヤケ8.Lてに．1ケ1．ヤり2.920．ヤケ9.5ぢ○つu&2』 talo 0.00[0.001さ．00t 0.001 0.00t 0.00て0.001 0.00t 0.00!0.00t 0.001 0.00t 0.00[0.00t 0.00t 0.00T 0.00て0.OOt 1危0エ 月自月日－－欄国田園咽h-－田りq園－日D日陣自園価りd国田日明D咽D 価巨咽日日国－ ---－田日月国月国月園田国田D国日日自qD咽Dd国日D 月自国D日園倒D り日－国り－国国咽D咽国田国柵園－図田昭曲園日咽D田国q国日国－国日－国日昭D昭P咽国嘱■日国q曲日口 996てら961 り961 E弱t 296!196t 086[6乙6[gL6t LL6t gL6t らL61 りL61 にL6T ZL6t IL61 0L61 696[ ニT’加亀’0加．つ’.d叫 ト H I t．ヤE ヤ．9りC.0L0.9ヤ6.61tL.95 [．てヤ．T6-2[0.22 て．60.00.00.00.00.00.00.0 っ記0 9．乙tT6.926.ZZE.929.2乙9.996.LZS．ヤ26.Cto.2て9.9t9.91C.11L.E 2．ヤ6.1 5．今6.C yQll朋恒1 0.6E0．りり0.0L0.9ヤ0．にro.5C0．ケEO.620.5t0.010.LO.90.90.LO.CO．ヤ0.to.9 .5．ロ 9.6！に．りt に．ヤ2 ヤー6t0.6t0.otZ．ぢ8.tIL-0 【．99.SC.IL.910．ぢに．9 ヤ．08.OL.1 式u→叩 ケ．622ヤ．乙8 2.tgて9.[6tヤ．ヤ625.1911.9ケZL.901ツ．921ヤ．69 ヤ．gL 9．ヤ9 ら．0ら1-1ヤC.52 5.L!2.ST 9.L乙．かロ’ユ』 B品10 0.9L9ヤ．0ヤC 9.16ケ2.09ケ9.999 0.6Lケ6.2ケケ2.05C 9.L6Z 9.691 L.gLt C．ヤヤ1 9．にtl 9.66 ヤ．LS !.29 ヤ．1ヤら申6りTり刀エ 9661 ぢ96t ヤ96t に96t Z96T t96t 0961 6L6t gL6t LL61 gL6t SL6t ケL6I CL6t ZL6て［L6t 0L6て696て （財Sa) 98・6961 ．田4心的明1○開T’ユq-lmヨ期5 , 国■園■． hd h目d レし一－一 5 1~ - Total mot Rocelptm. 1989-86 (UND) 19159 1970 1971 1972 im 1974 im 1978 1977 1978 1979 1990 1981 190 19M 1m4 19115 ino ---- ---- ---- --- ---- ---- ---- ---- ---- ---- Total Reccipt^ not 49.5 41.4 62.1 57.4 99.8 133.5 144.3 176.7 189.8 297.8 -.9 479.0 61101.8 480.2 491.8 340.4 676.0 ------------- ------ MA lonna, not 15.1 5.9 2.7 -21.9 4.5 22.6 20.2 32.1 34.9 34.9 63.2 201.0 111.3 114,7 123.5 105.7 283.9 o/*: Franco (5.2) (1.9) (-1.5) (-8.8) (1.8) (5.1) (8.0) (10.11) (13.2) (11.4) (94.8) (27.8) (19.% (14.7) (14.4) (MA) Oraftt4 40.3 38.8 50.4 70.3 73.8 81.8 103.4 94.7 88.0 171.0 ý99.9 199.4 173.9 207.6 241.4 188.4 283.5 ojoi Franco (25. e) (141.2) (14.3) (52.0) (27.9) (43.0) (58.2) (52.3) (ý0.8) (63. % (98.3) (99.9) (Ggj) (68_3) (67.11) (65.7) (63.1) 1 OOF. not 17.1 31.7 32.8 18.3 107.9 21.0 IN.3 115.4 79.2 37.5 57.5 Privata ~tor, not< el -5.9 -1.3 9.0 9.0 21.3 29.1 12.7 51.8 85.2 C9.1 24.5 77.2 61.8 125.8 52.5 MI.S 8.6 51.1 AD percent of Tota$ Roce;Pta 1989 1970 1971 19n 1973 1974 1975 1476 1977 1.978 1979 19% 1981 Iga 190 1%4 1985 1985 ODA logna, not 30.6 14.3 4.3 -38.2 4.6 16.4 19.6 18.3 18.4 18.4 42.5 14.3 42.0 18.9 24.9 25.7 31.1 42.0 Granta 81.4 498.9 81.2 122.8 74.1 81.8 71.7 85.9 48.4 67.5 *1.9 45.1 41.6 29.5 45.1 49.1 55.3 41.9 ------------------- 00 el For 1989-75, OOF and privat@ «ctor (not) bronkdown not, avaHmblo-. ~dual fr= total rocelpta. Sourco. 1~, 90~raphic Diatribution of Financial Floum ta, Dovalop;ng ~wi***. S8�Bt3AL ' Ивь Rвeвipb lгов Fnnd, 10вV'ав �) ivвe ivтo ivп iorг iv�a iv�a iv�s ivтe ivn ivтs ia�v ьоео ivei ivas ivвs iом iоев ivee Тоьв1 Rквiрьв, п.ь г�.ь ia.s v.s гs.а вi.i ю.ь ы.в п.♦ av.s iгв.4 iов.� пае.i isi.a sva.s ioi.e iы.г вz.в тго.в аа► с,.п., п.s в.а i.9 -i.a -�.i -s.e i.s a.i в.о io.v iа.п о.� и.в ы.в �.а iv.o iч.� и.в �a.i ог.па аа.в ie.z ц.а ь7.о 2�.v ва.о sв.г st.a ва.е ва.а п.в ов.а ev.o ве.а ве.а а.в ва.� ва.i OOF. п.ь ( 1.4 О.в 20.в 4.0 ав.а 2.1 ы.0 ва.4 ав.а 21.а вt.b Ргlrвfл исйг, not( а/ -1.9 -1 А 4.7 О.в 17.0 8.7 а.а 18J 1a.V Z7.V ?3.4 84.1 2в.1 19SA в0.V a.g -19.0 aS.7 , { h-� V W "--'°'---"�--.- { . s/ For 1овУ-7'а, OOF впд privв4s исtог (пвь) bгиkдом nоь вvвilвЫв= гиiд:св) lгав ьоьв) гвев{pb. ' For 197в ап, privnЧ исьог is • гиiдивl. Sоигсв: ОЕ�9, 'WoprnpAie Oishribu6ion о1 flг.вiкiвl F1овв ьо Овvвlорiпв Соипьгlи�. 01 �1 00 i� (D � W О� о м w 目日国■口■■■■■■■■■国■■□■■国国■■，ろーP -－一―一 図 刀 ロ り 咽国J O M 園d 刀
Groupe de la Banque mondiale · IEG Evaluation
The World Bank and Senegal : 1960-87
Voir le document original
Le texte intégral est hébergé par l’organisation qui le publie. lawenc.com indexe les métadonnées et renvoie vers la source officielle.
Texte intégral
Informations clés
Organisation
Groupe de la Banque mondiale
Type de document
IEG Evaluation
Pays
Sénégal
Source
Banque mondiale