Senegal - Agricultural Service Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.: 18001 IMPLEMENTAT'ION COMPLETION REPORT REPUBLIC OF SENEGAL AGRICULTURAL SERVICES PROJECT (Cr. 2108-SE) June 12, 1998 Agriculture Technical Cluster III Country Department 14 Africa Region 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.  CURRENCY EQUIVALENTS Currency Unit = CFA Franc (CFAF) CFAF per US$ 1990 - 272.56 1993 - 283.00 1994 - 546.72' 1997 - 582.96 WEIGHTS AND MEASURES Metric System FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS ANCAR National Agency for Agricultural Support (Agence Nationale de Conseil Agricole et Rural) AMRT Monthly Technical Review Meeting (Atelier Mensuel de Revue des Technologies) AVB Field Extension Worker (Agent de Vulgarisation de Base) DA Department of Agriculture (Direction de / 'Agriculture) DIREL Department of Livestock (Direction de l 'Elevage) ISRA Senegalese Agricultural Research Institute (institut Senegalais de Recherche Agricole) ITA Institute for Food Technologies (Institut de Technologie Alimentaire) MDR Ministry of Rural Development (Ministere du Developpement Rural) M&E Monitoring and Evaluation NPA New Agricultural Policy (Nouvelle Politique Agricole) PSAOP Agricultural Services and Producers' Organizations Support Project (Projet de Services Agricoles et dA'ppui aux Organisations des Producteurs) RDU Research Development Unit RRDC Regional Research Development Committee SAED Senegal River Development Authority (Societe Nationale d'Amenagement et d'Exploitation des Terres dii Delta du Fleuve Senegal et des Vallees du Fleuve Senegal et de la Falam6e) SODEFITEX Cotton Development Agency (Socijlepour le Developpement des Fibres Textiles) SODEVA Groundnut Basin Extension Agency (Soci&tj de DIveloppement et de la Vulgarisation Agricole) TS Subject Matter Specialist (Technicien Specialise) UPA Agricultural Policy Unit (Unite de Politique Agricole) Vice President Jean-Louis Sarbib Country Director Mahmood Ayub Sector Manager Jean-Paul Chausse Task Tearn Leader Abdoulaye Toure In January 1994, the CFAF parity with the French Franc (FF) was realigned from 50 CFAF per FF to 100 CFAF per FF.  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT SENEGAL AGRICULTURAL SERVICES PROJECT (Cr. 2108-SE) TABLE OF CONTENTS Page Preface ......................................................... i Evaluation Summary ......................................................... ii Part I: Project Implementation Assessment .......................................................... 1 Introduction ......................................................... 1 Project Objectives .......................................................... 1 Evaluation of Project Objectives .................. ....................................... 2 Achievement of Project Objectives ........................ ................................. 2 Implementation Record and Major Factors Affecting the Project ............ ............... 6 Project Sustainability ......................................................... 7 Bank Performance ......................................................... 8 Borrower Performance ......................................................... 9 Assessment of Outcome ......................................................... 9 Future Operation ......................................................... 9 Key Lessons Learned ......................................................... 10 Part II: Statistical Tables ......................................................... 11 Table 1: Summary of Assessments .......................... ............................... 1 Table 2: Related Bank Loans/Credits ......................................................... 13 Table 3: Project Timetable ......................................................... 14 Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual .............. 14 Table 5: Key Indicators for Project Implementation ............................... ............... 15 A. Extension ......................................................... 15 B. Main Extension Themes .................... ..................................... 15 Table 6: Key Indicators for Project Operation ....................................................... 19 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 7: Studies Included in Project .......................................... 20 Table 8A: Project Costs .......................................... 21 Table 8B: Project Financing .......................................... 22 Table 9: Economic Costs and Benefits .......................................... 22 Table 10. Legal Covenants .......................................... 23 Table 11: Compliance with Operational Manual Statements ................................. 28 Table 12: Bank Resources: Staff Inputs .......................................... 28 Table 13: Bank Resources: Missions .......................................... 29 Appendices ICR Mission Aide-Memoire .............................................. 32 Completion Report prepared by Borrower .............................................. 38 Maps IBRD 21784 IBRD 21785 IBRD 21786  IMPLEMENTATION COMPLETION REPORT SENEGAL AGRICULTURAI, SERVICES PROJECT (Cr. 2108-SE) PREFACE This is the Implementation Completion Report (ICR) for the Agricultural Services Project in Senegal, for which Credit 2108-SE in the amount of SDR13.0 million (US$17.1 million equivalent) was approved on March 22, 1990. The Credit was closed on December 31, 1997, 36 months after its original closing date of December 31, 1994. It is expected to be fully disbursed, and the last disbursement will take place in June 1998. The ICR was prepared by Abdoulaye Toure amd Leopold Sarr, AFTA3 of the Africa Region, and reviewed by Jean-Paul Chausse, Sector Manager, and Mahmood Ayub, Country Director. The Borrower provided comments that are reflected in the text of the ICR. The ICR is based on a mission that took place in January 1998 and material in the project file. The Borrower contributed to preparation of the ICR by providing views reflected in the mission's aide-memoire, commenting on the draft ICR, and preparing its own evaluation of the project's execution, which is attached as Appendix II to the report.  ii IMPLEMENTATION COMPLETION REPORT SENEGAL AGRICULTURA:L SERVICES PROJECT (Cr. 2108-SE) EVALUATION SUMMARY Introduction 1. At the time of project preparation (1987-89), the main issues in the agricultural sector of Senegal were: (i) an unfavorable macroeconomic and sectoral policy environment; (ii) a lack of dynamism in agricultural technology generation and diffusion; (iii) a high level of dependence of the agricultural secfor on millet/sorghum for domestic production and groundnuts for export; (iv) the lack of a well-functioning financial intermediation system in rural areas; (v) the low productivity of public investments; and (vi) insufficient and weak agricultural producer organizations. In 1984, the Government had issued a New Agricultural Policy (NPA) focussing on lifting supply-side constraints and creating a policy and institutional environment conducive to private sector participation and a reduced role for the State. A central element of Government's NPA was the strengthening of the extension services of the Ministry of Rural Development (MDR). In 1987, Government launched a pilot extension program with support of IDA. The positive results of the pilot were incorporated into the design of the present project, which became effective in July 1990. In the early 1990's, during project's implementation, the Government of Senegal launched an economic reform programn, including an adjustment of the exchange rate and trade liberalization, that addressed some of the main constraints of the agricultural sector. Project Objectives 2. The project constituted the first tranche of a long-term program aimed at strengthening Senegal's core agricultural services. The program's long-term objective was to promote the sustained increase in farmers' and pastoralists' productivity and incomes through technological change. The specific objectives of the program's first phase (the project) were to harmonize the different extension activities in use in the country around a system based on a farming system approach and to improve the links between extension and research. In support of these objectives, the project included three components: (a) the restructuring of the agricultural and livestock extension services of the Ministry of Rural Development (MDR); (b) the strengthening of producer organizations; and (c) the strengthening of MDR's analytical and policy-making capacity.  iii Implementation Experience and Results 3. The project's Extension Component did not reach its objectives. It was successful in introducing a new extension approach, based on the Training and Visit (T&V) System, in the targeted areas. A better management of extension activities (programming, supervision), the extensive use of farmers' Contact Groups and farmer-managed demonstration plots, strengthened linkages with research through participatory on-farm research programs and the systematic training of field extension agents supported the successful introduction of a variety of improved technologies-improved varieties and farming practices. Detailed information on adoption rates is however not available because of the project's failure to establish and maintain a satisfactory monitoring and evaluation (M&E) system. The project was also successful in strengthening research- extension linkages through the establishment of a Research-Development Unit at central level and of Regional Research-Development Committees where farmers, extension agents and research staff were able to jointly plan and review annual adaptive research programs. However, it fell far short of reaching the projected number of farmers. It is estimated that its extension activities benefited, directly (through participation in contact groups) or indirectly (through diffusion), about 116,000 farmers (20% of the country's total number of farm households) against 360,000 (60%) estimated at appraisal. The main constraints to expanding farmer coverage were difficulties in recruiting suitable extension agents and the low number of farmers/contact groups per field-level extension agent. These shortcomings were identified early in project implementation but were not corrected by the Project Management Unit, which focused more on the administrative than on the technical aspects of project implementation. 4. The project's Literacy Program was successful. It provided training to about 30,000 farmers, against the 10,000 projected at appraisal. The success of the component was largely due to the transfer of its implementation from the extension services (as originally envisaged), which were ill-equipped to carry it out, to specialized private agencies/NGOs. The project however largely failed to achieve the modest objectives of its Farmers Organization Component (sensitizing leaders to winning their active support for the new extension approach). Like in the case of the Literacy Program, the Ministry of Agriculture was ill-equipped to carry out a sensitization program for farmer organizations leaders, nor was there much interest in designing such a program. The latter should have been identified at the design stage, leaving the project with only two options: either exclude the component from the project altogether, or design a comprehensive program-to be transferred to a specialized private agency/NGO-for the strengthening of farmers' organizations. 5. The strengthening of MDR's analytical and policy-making capacity was partially achieved. An Agricultural Policy Unit (UPA) was created within the Ministry. UPA played an important role in the preparation and monitoring of the sector policy reform program (supported by IDA under the Agricultural Sector Adjustment Credit, Cr. 2738- SE) and public investment program in the sector. Progress on strengthening UPA is,  iv however, still fragile, as the Unit relies on staff recruited on a contractual basis and its analytical capacity is therefore largely dependent on continued external funding. 6. The performance of the Bank is judged unsatisfactory. It was satisfactory at the preparation stage, where the main issues and approaches were identified and field tested through the pre-project pilot phase. It was also (marginally) satisfactory at the appraisal stage, during which however more attention should have been given to detailed implementation arrangements, a coherent approach to farmers organization strengthening, an efficient monitoring and evaluation system and a systematic analysis the project's financial and economic impact and viability. Bank performance during supervision was unsatisfactory. Although Bank staff correctly identified implementation problems and prepared action plans to correct them, the follow-up was largely ineffective. The Borrower's performance was only marginally satisfactory. Adequate counterpart funds were provided, but often late. Project coordination by the Project Management Unit was weak, and its monitoring of field level activities and project impact was deficient. The project' M&E system was never operational. Financial management was overly centralized and bureaucratic, and a constraint to the flexibility required for adequately supporting field level activities and changling priorities and work programs. 7. The project's outcome is judged unsatisfactory. Although the project was successful in establishing a new extension system and in strengthening research-extension linkages, farmer coverage fell far short of projections. The lack of reliable relevant data doesn't allow a meaningful estimation of its economic impact. However, from the information available, it is doubtful that it achieved a satisfactory economic and/or cost- benefit outcome. Project's achievements in terms of institution/capacity-building were mixed. Progress was achieved in strengthening MDR's policy-making capacity and core agricultural services. However, sustaining and deepening these results will require more comprehensive institutional reforms in public sector management (and this cannot be tackled in a sectoral context) and in the management of the agricultural sector itself, in particular with the devolution of a far greater responsibility in the management and financing of agricultural services to beneficiaries and their associations. Summary of Findings, Future Operations and Key Lessons Learned. 8. Project implementation highlighted the severe constraints faced by farmers in Senegal, in particular in rainfed areas, if the necessary productivity increase are to be sustained, and therefore the urgent need to revitalize the country's research and extension system. It highlighted the institutional weaknesses of public service delivery systems, as well as the potential role that farmers' associations could and should play, if they were strengthened via the development of responsive, efficient and sustainable agricultural services. In doing so, project experience was very instrumental in Government's decision to launch the in-depth institutional restructuring of Senegal's national extension and research system to be carried out under the proposed National Agricultural Services and Farmers Organization Program (PSAOP), to be supported by all interested donors  v including IDA (appraisal scheduled for June 1998). This comprehensive, long-term program will address: (i) the restructuring of MDR around its core policy-making and oversight functions; (ii) the rationalization and strengthening of the country's agricultural extension system, through the establishment of a national agricultural extension agency where producers will have a major management and funding responsibility; (iii) the continued strengthening of the national agricultural research system; and (iv) the strengthening of farmers organizations. 9. The PSAOP will incorporate the main lessons of the Agricultural Services Project, among which are: (a) Ensuring sustained responsiveness and efficiency of extension services through a top-down centralized system is hardly, if at all, possible. It requires an open system relying on multiple sources of advice and services as well as the commitment and effective participation, at all levels, of beneficiaries in decision-making and control. (b) The identification and prioritization of farmers' needs and constraints is a prerequisite for the success of an extension program. This needs to be done with the help of a multi-disciplinary team that can guide the farrners in identifying what their key needs and constraints are. An AVB may not be the most appropriate person to carry out this task as the types of problems that require attention move from relatively simple to more complex. (c) Contact groups are a very effective instruments for two-way communications in a technology diffusion system. However, the CG concept needs to be reviewed. Experience under this project has shown that criteria for the establishment of GCs need to reflect existing social structures and/or commonality of economic purpose.  IMPLEMENTATION COMPLETION REPORT SENEGAL AGRICULTURAL SERVICES PROJECT (Cr. 2108-SE) PART I: PROJECT IMIPLEMENTATION ASSESSMENT Introduction 1. At the time of project preparation (1987-89), the main constraints facing the agricultural sector of Senegal were the: (i) an unfavorable macroeconomic and sector policy environment, including overvalued exchange rate and restrictive trade and pricing policies; (ii) a lack of dynamism in agricultural technology generation and diffusion; (iii) a high level of dependence of the agricultural sector on millet/sorghum for domestic production and groundnuts for export, and limited efforts in crop diversification; (iv) the lack of a well functioning financial intennediation system in rural areas and the resulting highly restricted access to credit by farmers; (v) a low productivity of public investments in the sector; and (vi) weak agricultural producer organizations. In 1984, the Government had issued a New Agricultural Policy (NPA) focusing on lifting supply-side constraints and creating a policy and institutional environment conducive to private sector participation and a reduced role for the State. A central element of the NPA was the strengthening of the extension services of the Ministry of Rural Development (MDR). In 1987, Government launched a pilot extension program with the support of IDA. The positive results of the pilot were incorporated into the design of the present project. Also, the Government launched, at the begimiing of the 1990s, a renewed economic reform program, including trade liberalization measures and an adjustment of the exchange rate, that addressed some of the main policy constraints in the agricultural sector. Project Objectives 2. The project constituted the first tranche of a long-term program aimed at strengthening the country's core agricultural services. The program's objectives were to promote a sustained increase in farmers' and pastoralists' incomes through accelerated technological change. The program had a special focus on rural women. The specific objectives of the program's first tranche (the project) were to harmonize the various extension systems in use in the country around a farming system approach and a multipurpose service that comprehensively addresses farmers' and pastoralists main activities; and to improve the links between extension and research. In support of these objectives, the project included three components: (a) restructuring agricultural and livestock extension services of the Ministry of Rural Development (MDR); (b)  2 strengthening producer organizations; and (c) strengthening MDR's the analytical and policy-making. 3. The project was to be implemented over a four-year period and expected to cost US$20.2 million equivalent. Financing was to be provided by IDA (US$17.1 million or 85 percent of total project costs) and the Governnent (US$3.1 million or 15 percent of project costs). Evaluation of Project Objectives 4. The project's objectives were central to Government's NPA and the Bank's Country Assistance Strategy. They were clear and realistic. However, the SAR failed to adequately quantify the project's long-term and interrnediate impact on production, productivity and incomes. The extension component was well designed. The project's main design weakness was in regard to the component for strengthening producer organizations that included only very limited and vaguely defined activities aiming at the sensitizing of "farm leaders" to the new approach to extension. The component however also included a farmer literacy program that was better articulated. 5. The Staff Appraisal Report listed three risks: (i) ineffective management and training of extension personnel; (ii) low adoption rates by farmers; and (iii) failure by Government and/or other donors to sustain the required (in particular financial) long-term commitment. None of these risks materialized. Achievement of Project Objectives Overall Objectives 6. The project partially met its objectives. It was able to harmonize the different extension approaches in the country and lay the groundwork for improved collaboration between extension and research. It was, however, able to reach only about a thir d of the planned number of farmers; it failed to significantly strengthen producer organizations; and it had only limited success in strengthening the analysis and policy- making capacity of the Ministry of Rural Development. 7. No attempt was made at appraisal to carry out an Economic Rate of Return (ERR) analysis. Instead, a cost-benefit estimate showed that, with the project covering-directly and indirectly-about 60% of Senegal's total farming population, its cost could be covered by a 0.33% increase in agricultural value added in the project area (or by a 0.2% increase in the country's total agricultural value added). The SAR concluded that it was a readily achievable target (however, the SAR lacked an in-depth analysis of how this increase would be achieved). The lack of a base-line survey and of an adequate monitoring and evaluation system under the project would have prevented any meaningful ERR on cost-benefit estimation during the ICR mission. None was  3 attempted. However, it seems clear that, with the project reaching only one-third of the number of farmers that were projected at appraisal to benefit from improved agricultural services, its economic impact will fall far short of what was anticipated. Strengthening of Agricultural Extension Services 8. The component's objectives were partially achieved. Prior to the. project, agricultural extension services were largely ineffective, except for those of Senegal River Development Authority (Societe Nationale d'Amenagement et d'Exploitation des Terres du Delta du Fleuve Senegal et des Vallees du Fleuve Senegal et de la Falamee, SAED) and the Cotton Development Agency (Societe pour le Developpement des Fibres Textiles, SODEFITEX). In addition, only about 10-15 percent of farmers had access to extension services. The project was able to mar]kedly strengthen the services provided by the Department of Agriculture (Direction de l 'Agriculture, DA) and Livestock (Direction de I 'Elevage, DIREL) of the Ministry of Rural Development, as well as those of SAED and of the Groundnut Basin Extension Agency (Societe de Developpement et de Vulgarisation Agricole, SODEVA). Activities under this component included: (i) developing a common extension strategy based on the Training and Visit (T&V) approach, focussing on a farming system approach and a close interaction between Field Extension Workers (Agents de vulgarisation de base, AVB) and farmners, and increasing collaboration between the different agencies involved in providing extension services; (ii) putting at their disposal the resources necessary to execute their work programs (field allowances, motorcycles, vehicles, rehabilitation of buildings, etc.); (iii) implementing a well-defined training program for extension staff; and (iv) strengthening research- extension linkages. By 1995, the project had achieved a coverage of 116,000 farms against a total of 360,000 planned at appraisal. Expansion of extension coverage beyond this level was not possible because of difficulties in recruiting qualified candidates at the offered level of salary, and also because the number of farmers per extension staff remained too limited (para 11). 9. Programming. The project introduced an effective diagnostic tool that was instrumental in identifying needs and constraints at the farm level. The programming of extension activities on the basis of this analysis, undertaken in collaboration with farmers, was supposed to make extension messages more responsive to actual needs. Actual use of this instrument was however limited to the filling out of relatively simple questionnaires by AVBs, and provided only an incomplete knowledge of the constraints faced by farmers. The approach, which iis sound, needs to be reviewed and revised if the extension system is to efficiently tackle the increasingly complex problems that need to be addressed at the farm level. 10. Training. An extensive three-tier training program was put in place. The first tier consisted of fortnightly training sessions for AVBs (formation de quinzaine, FQ) that focused on the themes to be extended diring the following two weeks and was provided by Subject Matter Specialists (Techniciens Specialises, TSs). The second tier consisted of Monthly Technical Review Meetings (Ateliers Mensuels de Revue des Technologies,  4 AMRTs) between TSs and research staff, aimed at strengthening the training content of the FQs. The third tier consisted of a targeted training program based on specifically identified needs. The FQs were regularly held during the first four years of the project, but subsequently less frequently as financing was no longer made available in a timely fashion. The AMRTs never came off the ground for two reasons: (i) the novelty of the concept for extension and research staff, who gave it insufficient priority; and (ii) difficulties in mobilizing research staff because of a lack of financial means. The targeted training plan was executed as planned, with staff participating in workshops, study tours, etc. Although difficult to assess precisely, it is apparent that the impact of the training program was less than expected. 11. Extension Methodology. The extension approach was based on the T&V system and used two main tools: Contact Groups (CGs) and demonstration plots. Farmers' participation in CGs declined with time because their initial expectations were not met, i.e. extension messages did not fully address their main constraints (in particular, they were more interested in access to inputs such as fertilizer and credit), and also because the CGs were not based on existing natural groupings in the communities. Moreover, in many cases, the approach was not properly implemented by AVBs, who worked with individual farmers and not groups, thereby limiting the transfer of knowledge. Early on during project implementation, AVB's also failed to follow-up with farmers on the impact of the many new techniques that were extended, and to help farmers alleviate related constraints (mostly inputs and credit). After 1994, the project made an effort to limit the number of extension themes (improved seeds, improved cattle feed, soil fertility and natural resources management) and AVBs were then better able to follow-up on impact with farmers. As a result, a number of messages (15) were successfully introduced to a larger group of farmers (see Table 5.B). Finally, in 1996, a start was made to restructure CGs along natural social/economic affinities. 12. Demonstration plots were used on a large scale by AVBs. About 110,000 plots were established between July 1990 and 1995, allowing farmers to test a large variety of new techniques (120 new innovations were demonstrated over that period). Unfortunately, the demonstration plots' positive results were not sufficiently disseminated because, as already noted, of the poor functioning of the CGs and of the failure of the project to address farmers' difficulties in accessing fertilizers and credit. In addition to demonstration plots, the project made use of weekly rural radio broadcasts, in which farmers participated, to highlight the anticipated results of technical messages. 13. No systematic collection of adoption rates was carried out under the project. When adoption rates were collected, they varied from a low of 30 percent for improved seeds for rice, to 40-50 percent for miller seed selection, to a high of 90 percent for new varieties of sweet potatoes and onion (see also 5.B, Part II). 14. Research-Extension Linkages. The establishment and improvement of these linkages was one of the project's main successes. A scientist was detached from the Senegalese Agricultural Research Institute (Institut Senegalais de Recherches Agricoles,  5 ISRA) to work closely with the project. A Research-Development Unit (RDU) was established which organized regular meetings between the heads of the two research organizations (ISRA and the Institute for Food Technologies - Institut de Technologie Alimentaire, ITA), extension services, and the National Consultative Committee for Farmers (Comite de Concertation et de Cooperation des Ruraux). The RDU's mission was to support the generation and diffiusion of new technologies through the newly created Regional Research-Development Committees (RRDCs), comprised of research and extension staff and farmers and responsible for carrying out joint research-extension on-farm experiments. 15. The joint research-extension initiative through the RRDCs was very successful in fostering collaboration between research and extension staff and in heightening researchers' awareness of on-farm constraints. RRDCs systematically organized meetings to discuss test results with farmers. The main constraint to the effectiveness of RRDCs was the availability of adequate financing and of qualified staff. 16. Monitoring and Evaluation. The project's central Monitoring and Evaluation (M&E) Unit was unable to design a satisfactory system for the evaluation of project activities. At first, the system relied on monthly information collected by AVBs. Delays in the collection of the data and concerns about their reliability led to the redesign of the system in 1992. The new system was, however, far too complex (64 indicators) to be successfully implemented. With the exception of a 1992 study in the commune of Ndiendieng, a beneficiary evaluation in 1994 and an evaluation of extension tools between August 1996 and January 1997, the M&E failed to assist project management in tracking project implementation; nor was there ever a full impact evaluation of the project undertaken. Strengthening of Producer Organizations 17. The objectives for this componernt were partially achieved. The training of farm leaders sub-component failed to achieve its (very limited and ill-defined) objective of sensitizing farm leaders and producer organizations to (and winning their active support in favor of) the new extension methodology. Like in the case of the Literacy Program, the Ministry of Agriculture was ill-equipped to carry out a sensitization program for farmer organizations leaders, nor was there much interest in designing such a program. 18. However, the functional literacy sub-program greatly exceeded the project's objective. From 1993 to 1997, a total of 1,115 literacy centers were created and 30,904 farmers received literacy training against an appraisal target of 10,000. Literacy training was initially carried out by extension services (DA, SAED, SODEVA, DERBAC) but the decision was taken in 1994 to transfer the implementation of the program to private organizations specialized in literacy training. The latter had the advantage not only of being better qualified to carry out the programs but also of being able to pre-finance their work. As a result, the literacy program was executed with much greater efficiency.  6 Strengthening of MDR 19. The objectives of this component were partially achieved. An Agricultural Policy Unit (Unite de Politique Agricole, UPA) was established with the MDR, which played an important role in the elaboration of the Agricultural Development Policy Letter of 1995. UPA also successfully adopted tools (analysis of costs/benefits, domestic resource costs, etc.) that enable it to play a greater role in the identification and preparation of investment projects 20. However, UPA's ability to effectively supervise projects remained weak, as did its ability to do impact and financial evaluations. This is due to a variety of factors, including: (i) its weak hierarchical position within the MDR; (ii) a failure to attract qualified economists; (iii) long delays in adjusting the training program, established at appraisal, for its staff to reflect actual needs (not done until 1994); (iv) weak internal organization (no properly defined work programs or performance indicators). Inadequate supervision by IDA also contributed to the poor implementation of this component. Project Costs 21. Actual project costs are estimated at US$20.9 million, very close to the appraisal estimate of US$20.2 million, despite a much longer than planned implementation period. The lower than expected expenditures on on-farm research trials (US$1.1 million less) and changes in exchange rates provided the resources to finance project activities during the period (January 94 to December 97). Government financed US$2.9 million (US$3.1 planned) and IDA US$18.0 million (US$17.1 planned). Implementation Record and Major Factors Affecting the Project 22. The principal factors directly affecting project implementation were under the control of Government and the implementing agencies. Factors under Government Control 23. The planned institutional arrangements foresaw the establishment of a small unit within MDR to coordinate the activities of the implementing agencies. For this purpose, each implementing agency entered into a contractual arrangement with this unit. In fact, the absence of appropriately defined obligations of each party in the contract and the Unit's lack of hierarchical authority, or real influence, over the implementing agencies led to implementation difficulties. This issue was identified early on during project implementation but never corrected. 24. A similar problem arose in the case of the UPA. In spite of its attachment to the cabinet of the Minister, UPA never acquired the authority necessary for the execution of  7 its mandate, especially where it concerned policy formulation and project supervision. These problems were not adequately addressed during project implementation. 25. Overall levels of counterpart ifunding for project activities was adequate. However, counterpart funds were frequenitly made available behind schedule. Moreover, the mechanical division of counterpart finds into sixty percent for the extension related activities and forty percent for the UPA, regardless of actual work programs, appears to have strongly benefited UPA and penalized extension activities. This mechanism was put in place by Government over the Bank's objection and never modified.2 Factors under Implementing Agencies' Control 26. Three main factors under implementing agencies' control affected project implementation. The first factor was an overly centralized financial management system, which led to delays in the availability of project funds at operational level and reduced flexibility to reallocate funds to successful activities. This led to chronic underutilization of the budget and shortfalls in the implementation programmed activities. The second factor was poor technical management and supervision by project staff, which limited its supervision to administrative aspects. Technical issues and field supervision were clearly neglected, which had a negative impact in several ways: (i) the problems encountered with CG participation was not identified/addressed until late in the project's life, and only after continuous insistence by IDA supervision missions; (ii) the M&E function was marginalized because of project managers' lack of interest in its practical use for evaluating field impact and correcting operational weaknesses; and (iii) there was a conspicuous lack of vision and focus in the work of UPA, with a resulting lower than anticipated capacity to analyze and support policy-making. Project Sustainability 27. The sustainability of the agricultural extension program, given the inability of the Government to fund more than salaries and related expenditures, remains at the end of the project totally dependant on external financing and/or cost recovery from beneficiaries to the limited extent this is possible. A 'Follow-up project, the Agricultural Services and Producers' Organizations Support Program (Programme de Services Agricoles et d'Appui aux Organisations des Producteurs, PSAOP) presently under preparation, will address this critical sustainability issue by giving farmers, through their professional association, decision-making responsibilities in the inanagement of extension services and requesting them in return to finance an increasing share of extension costs. In the case of the UPA, sustainability without external financing is also unlikely, given the fact that 70 percent of 2 The Government feels that the problem causing delays in project activities was not the 40-60 split, but the highly centralized and inflexible project financial management system. The authors of the ICR agreed that this was a major issue, which is raised later in the text, but also strongly believe that the 40-60 split of counterpart funds, irrespective of needs, unnecessarily aggravated the problem.  8 the UPA's staff are contractual workers earning salaries much above those of the civil, available only in the context of externally funded projects. The long-term sustainability of UPA will to a large degree depend on the introduction of adequate career development opportunities and a performing incentives structure for civil servants (salary and non- salary benefits). Bank Performance 28. Identification. The Bank's performance at the identification stage was satisfactory. The existing constraints to efficient agricultural services provision were correctly identified, as were ways to strengthen extension services. It was also satisfactory during project preparation. In 1987, the Bank approved a project preparation facility to finance a pilot extension operation to test the new approach and design detailed implementation arrangements. In addition, it assigned a member of its staff to the Resident Mission in Dakar to assist in implementing the pilot operation. 29. Appraisal. The Bank's role at appraisal was also (marginally) satisfactory. The project was appraised in February 1990 and negotiated in April 1990. With the exception of the farmer's organization component, the project was well designed. The Staff Appraisal Report (SAR) correctly stressed the following factors: (i) the importance of clear relationships between the coordinating unit and the implementing agencies, (ii) the need to carry-out an intensive training program to help AVBs identify on-farn constraints and better understand the concepts of GCs and demonstration plots; and (iii) the need to have flexible financial/administrative management. However, as in many similar projects prepared at the time, no serious attempt was made to evaluate the project's economic and financial impact, and M&E was not accorded much importance. 30. Supervision. The supervision of the project by the Bank was unsatisfactory. Through 1994, supervision was quite intensive, with 10 formal supervision missions and for part of this period a headquarters staff stationed at the Resident Mission to provide full-time implementation support. Supervision missions were joined by staff of the MDR and project staff. The missions properly identified the problems but fell short of proposing and/or of following-up effectively on solutions, which explains the recurrence of already identified problems during subsequent supervision missions. Supervision missions were not adequately staffed. They frequently included only extension specialists and other implementation aspects (in particular management and financial aspects) were neglected. They were also frequently relatively short (about one week on average). 31. In September 1993, a mid-term review was carried out in broad collaboration with Government and project staff. An in-depth review of the project's strengths and weaknesses was done and an action plan was drawn up and agreed upon by all parties. This action plan was, however, neither properly implemented nor followed-up on by subsequent missions. Indeed, after 1994, the supervision of the project was clearly  9 neglected by the Bank as all available resources were redirected to the preparation of the follow-up operation, the Agricultural Services and Producers' Organizations Program (see para 34). As a result, many of the issues identified during the mid-term review still need to be addressed today. Borrower Performance 32. Borrower's performance during project preparation, including the pilot phase, was satisfactory. Performance during impleimentation was deficient. The Government failed to follow-up on the agreed action plans; coordination of project activities was weak; and project supervision was spotty and often limited to its administrative management. In addition, although counterpart funds were generally adequate, they often were made available with delays. The latter issue was compounded by the artificial split in counterpart funds of 60 percent for the extension services and 40 percent for the UPA, regardless of their respective work programs. Assessment of Outcome 33. The project's outcome is judged unsatisfactory. Strengthening the coverage and efficiency of extension services and improving the links between research, extension and producers were the main foci of the project. The project did manage to achieve a increase in the number of producers covered and create a more dynamic flow of information between research, extension and producers. However, no progress was achieved to improve the long-term sustainability of extension services. The second component (strengthening producers' organizations), designed in support of the extension component, failed to sensitize farm leaders to the goals of the project, or to strengthen these organizations in any way. The failure of the project to substantially strengthen MDR's ability to effectively prepare and monitor projects did have a negative impact on the project, as it severely limited its effectiveness in supervision. Future Operation 34. The project was conceived as the first tranche of a long-term program. In 1995, the Government of Senegal requested the Bank to finance a follow-up operation, the Agricultural Services and Producers' Organizations Program (Programme de Services Agricoles et d'Appui aux Organizations des Producteurs, PSAOP). PSAOP, that will be appraised in June 1998, will build on the experience and lessons from the present project. It will take a comprehensive approach to the strengthening of essential agricultural services in Senegal. It will support: the total restructuring of MDR, refocusing the ministry on its core policy-making and regulator/supervisory functions; the comprehensive strengthening of producers associations, through a full-pledged program addressing all major economic, management and human resource issues; the restructuring of the national extension system through the creation of a National Extension Agency (Agence Nationale de Conseil Agricole et Rural, ANCAR) where producers will have a major role in decision-making and in funding extension activities; and the continuous strengthening of the country's national agricultural research system and of its linkages with both producer and extensive services.  10 Key Lessons Learned 35. The main lessons to be learned from the implementation of this project are: (a) Ensuring sustained responsiveness and efficiency of extension services through a top-down centralized system is hardly, if at all, possible. It requires the commitment and effective participation, at all levels, of beneficiaries in decision-making and control. (b) The identification and prioritization of farmers' needs and constraints is a prerequisite for the success of an extension program. This needs to be done with the help of a multi-disciplinary team that can guide the farmers in identifying what their key needs and constraints are. An AVB may not be the most appropriate person to carry out this task as the types of problems that require attention move from relatively simple to more complex. (c) Contact groups are a very effective instruments for two-way communications in a technology diffusion system. However, the CG concept needs to be reviewed. Experience under this project has shown that criteria for the establishment of GCs need to reflect existing social structures and/or commonality of economic purpose.  11 IMPLEMENTATION COMPLETION REPORT SENEGAL AGRICULTURAL SERVICES PROJECT (C:r. 2108-SE) PART II: STATISTICAL TABLES TABLE 1: SUMMARY OF ASSESSMENTS Substantial Partial Negligible Not applicable A. Achievement of objectives (X) (X) (X) (X) Macroeconomic policies X Sector policies X Financial objectives X Institutional development X Physical objectives X Poverty reduction X X Gender issues X Other social objectives X l Environmental objectives X Public sector management X Private sector development X Other (specify) Likely 7 Unlikely Uncertin (X) (X) (X) B. Project sustainability X Highly Satisfactory Satisfactory Deficient C. Bank performance (X) (X) (X) Identification A X Preparation assistance X Appraisal X Supervision X  12 Highly Satisfactory Satisfactory Deficient D. Borrower performance (X) (X) (X) Preparation X Implementation X Covenant compliance X Operation (if applicable) Highly Highly Satisfactory Satisfactory Unsatisfactory unsatisfactory (X) (X) (X) (X) E. Assessment of X outcome  13 TABLE 2: RELATED IBANK LOANS/CREDITS Loan/credit title Purpose Year of approval Status Preceding operations I Agricultural Research Strengthen agricultural September 8, 1981 Closed, December (Cr. 1176-SE) research and relevance 31, 1989 Irrigation Technical Assistance Financial and institutional December 1, 1987 Closed, December Project (Cr. 1632-SE) strengthening of SAED and 31, 1988 rehabilitation of irrigation perimeters Fourth Irrigation Project Institutional Strengthening December 1, 1987 December 31, 1995 (Cr. 1855-SE) of SAED, rehabilitation of perimeters and extension Second Small Rural Operations Support to small rural March 21, 1989 Ongoing Project (Cr. 1992-SE) productive investments to increase rural incomes Second Agricultural Research Strengthen agricultural March 22, 1990 Closed, December Project (Cr. 2107-SE) research and relevance 31, 1996 Following operations Agricultural Sector Adjustment Support to liberalization of June 6, 1995 Closed, June 30, Crecdit (Cr. 2738-SE) cereals, sugar, and 1997 groundnuts markets and I ___________________________ processing Agricultural Export Crops Support to deve'lopment of December 9, 1997 Not yet effective lProrotion Project (Cr. 3017-SE) export markets of (expected May 19, agricultural produce, and 1998) diversification cf production  14 TABLE 3: PROJECT TIMETABLE Date actual/ Steps in project cycle Date planned latest estimate Identification (Executive Project Summary) May 3, 1989 Preparation 1987-1989 Appraisal May 1989 Negotiations January 1990 Letter of development policy (if applicable) N.A. Board presentation March 22, 1990 Signing April 24, 1990 Effectiveness July 24, 1990 July 6, 1990 First tranche release (if applicable) N.A. Midterm review (if applicable) June 30, 1992 December 1992 and August 1993 Second (and third) tranche release (if N.A. applicable) Project completion June 30, 1994 December 31, 1997 Credit closing December 31, December 31, 1997 1994 TABLE 4: CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL (US$ million) FY91 FY92 FY93 FY94 FY95 FY96 FY97 FY98 Appraisal 6.20 10.40 13.80 17.10 17.10 17.10 17.10 17.10 estimate Revised 3.06 5.53 8.28 10.45 12.47 14.50 17.10 17.10 Actual 3.06 5.53 8.28 10.45 12.47 14.58 16.21 18.00 Actual as % of 49 53 60 61 77 85 95 105 Appraisal estimate Actual as % of 100 10 100 100 100 101 95 105 formally revised l___ __ _ _ __ _ _ __ __ __ _ _ Date of final disbursement3 3 The final disbursement has not yet taken place (expected in June 1998).  15 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION A. Extension Indicator Planned Actual Staffing: Extension Workers (AVB) 537 398 Supervisors 54 57 Technical Specialists (SMS) 25 42 Regional Managers 16 22 Geographic Coverage: Number of villages 13,254 Number of villages served 6,627 3,158 Rural Population Reached 362,000 116,000 Coverage (%) 60% 32% Farmers' Participation not defined 62,687 Number of Contact Groups (CGs) not defined 5,224 Number of Farms per CG not defined 12 Number of GCs per AVB not defined 13 Number of Farms per AVB not defined 156 Number of Demonstration Plots not defined 110,000 plots (120 technical themes) Adoption Rate not defined N.A. B. Main Extension Themes Crop Problem Theme Year Impact Area Groundnuts Degeneration of Introduction 1994 Increase of yields Kaoleck, Thies, seeds reducing of new from 200 to 400 Diourbal and yields variety (Fleur kg/ha in rain fed Louga (rainfed), 11) systems and 3.5 to 4 and St. Louis tons/ha in irrigated (irrigated) systems. Adopted on 2.5 ha in 1994, 12 ha in 1995 and 400 ha in 1996 Cowpeas Parasites Phytosanitary 1991 200 kg/ha increase Louga, treatments in yields. Adopted Diourbel, Fatick by 80 percent of and Thies farmers  16 Crop Problem Theme Year Impact Area Cowpeas Planting density Use of 8 1992 200 kg/ha increase. Diourbel, Fatick holed disks Adopted by 70 and Thies percent of farmers Mil Soil Fertility Organic 1991 500 kg/ha increase. Louga, Fatick, fertilizer Adopted by 80 Tamba and percent of farmers Kolda Seed quality Selection 1991 290 kg/ha increase. Kaolack, Fatick, Adopted by 40-50 Diourbel and percent of farmers Thies Planting density Use of 4 1992 350 kg/ha increase; Kaolack, Fatick, holed disks reduction in seeds Diourbel, Louga needed by I kg/ha; and Thies and one-third less l__________ time for planting Sorghum Long growth Introduction 1992 90-day growth cycle Tamba, Kolda, cycle and low of new versus 120-140 days Fatick, Kaolack yields varieties for local varieties. and St. Louis. (CE145 and Increase in yields of CE151) 450 kg/ha in rainfed areas and 1.5 t/ha in ____________ _______________ irrigated areas Rice Insufficient Introduction 1993 Shortening of St. Louis varieties with of Sahelian growth cycle for short to medium varieties Sahel 108 to 90-100 growth cycles (Sahel 108 days and to 110-120 and 202) days for Sahel 202. Increase in yields of 500-600kg/ha. Adopted on 30-40 percent of land, where it is grown mixed in with traditional varieties Sub-optimal use Application 1994 750 kg/ha increase St. Louis of fertilizers of fertilizer in in yields. Adoption two lower rate of 90-95 treatments, as percent. opposed to one high Inadequate crop Use of 1995 Adopted on 600 ha St. Louis rotation agricultural in 1995 and 7,000 calendar ha in 1997. Yield increase of up to It/ha Onion No production Introduction 1994 Production possible Niayes during off- of new between June and season variety November for 90  17 Crop Problem Theme Year Impact Area (Rouge percent of vegetable l__________ d'amposta) growing farmers Sweet Quality Introduction 1994 Increased St-Louis Potatoes of new production and varieties easier commercialization of crop. Adopted by 90 percent of potato growing farmers Jaxatu Insects Introduction 1993 Increase of lOt/ha. Niayes of insect Adopted by 88 resistant percent of farmers varieties Tomatoes Density Planting 1994 Increase from 20 to St. Louis density 50 t/ha during rainy season and 10-25 percent during off- season. Adopted by 70 percent of farmners Soil Erosion Soil fertility Ant-erosion 1990 Introduced in 30 Thies, Tamba measures villages with about and Kaoloa 500 farmers. Establishment of 400 stone barriers (40 km length) and planting of 60,000 l_______ soil fixing plants  18 C. Training Type Planned Actual/Latest Estimate In-Service Familiarization sessions for 3 days for each agent done all extension agents Retraining of extension extension agent 6 days/year; done first four years, less agents and supervisors supervisor 21 days/year during period of credit extension training for extension agents bi-monthly became twice-monthly; and supervisors in done first four years, less agricultural zones during period of credit extension training for extension agents Monthly became twice-monthly; and supervisors in sylvo- done first four years, less pastoral zones during period of credit extension Technical Specialists 12 days/year done first four years, less during period of credit extension Monthly Technical Review Monthly with assistance of not regularly held Meetings research staff Other extension training: not identified Visits by farmer number not determined at Visits in Senegal only representatives (Senegal appraisal and abroad) Senior staff visits to like number not determined at Visits to Burkina Faso, C6te projects in Africa appraisal d'Ivoire, Mali and Niger, Benin, Ghana Training of farm leaders: Introduction to extension 2,700 person-days N.A. project (participatory) Annual work programming 2,500 person-days N.A. and technical messages Participation in extension 500 person-days N.A. training sessions Basic group action skills 400 farmer groups N.A. Functional Literacy Training: Training of trainers all field extension agents done by private institutions Newly literate people 10,000 30,904 beneficiaries  19 D. Consulting Services Subject Area Planned Actual Design and establishment of 6 person months Done monitoring system within DA and DE and their Provincial Inspections Analysis of responsiveness 12 person months Done of extension service to specific needs of women farmers and traditional herders Communication techniques 6 person-months Done Design of mass-media 4 person-months Done strategy Review of design of field 8 person-months Done surveys that precede the incorporation of new areas into the project Support from external Short-termn consultancies on Done university/institution to demand MDR l TABLE 6: KEY INDICATORS FOR PROJECT OPERATION Indicator Planned Actual/Latest Estimate No indicators were established  20 TABLE 7: STUDIES INCLUDED IN PROJECT Purpose as defined Study at appraisal/redefined Status Impact of study 1. Mid-term Review of project Completed Conclusions Impact implementation over the first two (December partially adopted Evaluation years (in Ndiendieng only) 1992) 2. Beneficiary Assessment of project impact on Completed Conclusions Assessment beneficiaries after 4 years of (November partially adopted project implementation 1994)  21 TABLE 8A: PROJECT COSTS4 Actual/ Latest Appraisal estimate Estimate Items (US$ '000)5 (US$ '000) Civil Works - Extension Service 365 399 - MRD 30 26 Equipment - All but MRD 785 567 - MRD 70 73 Vehicles - Extension Service 1,500 1,537 - MRD 75 79 Training - Functional Literacy 2,800 3,098 - MRD 375 335 On-farm Research Trials 1,900 797 Salaries - Extension Service 700 1,395 - Project Unit 150 410 Operating Costs - Extension Service 7,175 8,167 - Project Unit 150 171 Consultant Services - Extension Service 500 833 - MRD 500 692 PPF 1,710 1,271 Base Costs 18,785 19,850 Contingencies/Unallocated 1,415 1,050 Total Project Costs 20,200 20,900 4 Numbers may not add up due to rounding. 5 Estimated on the basis of the Development: Credit Agreement and the Staff Appraisal Report  22 TABLE 8B: PROJECT FINANCING (in millions of US$)