Senegal - Second Small Rural Operations Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 20649-SE IMPLEMENTATION COMPLETION REPORT (IDA-19920; COFN-01870) ONA CREDIT N THE AMOUNT OF SDRs 12.0 MILLION (US$16.1 MILLION EQUIVALENT) TO THE REPUBLIC OF SENEGAL FOR THE SECOND SMALL RURAL OPERATIONS PROJECT JUNE 30, 2000 Rural Development m Country Department 14 Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (Exchange Rate Effective 1988) Currency Unit = Franc CFA FCFA 300 = US$ 1.00 US$ 0.003 = FCFA 1.00 The exchange rate at the time of the ICR mission was US$1.00 = 640 Franc CFA FISCAL YEAR January 1 December 31 ABBREVIATIONS AND ACRONYMS AGEP Project Implementing Agency (part of ASPRODEB - Agence de Gestion des Projets) ASPRODEB Senegalese Association for the Promotion of Small Rural Development Projects (Association Senegaliaise pour la Promotion des Projets de Base) CNCAS National Agricultural Credit Bank (Caisse Nationale de Credit Agricole Sendgalaise) CNCR National Counsel for the Rural Population (Conseil National de Concertation des Ruraux) ICR Implementation Completion Report MTR Mid-Term Review PMU Project Management Unit SAR Staff Appraisal Report Vice President: Callisto Madavo Country Manager/Director: Mahmood A. Ayub Sector Manager/Director: Jean-Paul Chausse Task Teamn Leader/Task Manager: Leopold Sarr  FOR OFFICIAL USE ONLY REPUBLIC OF SENEGAL SECOND SMALL RURAL OPERATIONS PROJECT (IDA Cr. 19920-SE; IFAD Ln. 01870-SE) CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 4 5. Major Factors Affecting Inplementation and Outcome 5 6. Sustainability 6 7. Bank and Borrower Performance 7 8. Lessons Learned 9 9. Partner Comments 9 10. Additional Information 9 Annex 1. Key Performance Indicators/Log Frame Matrix 12 Annex 2. Project Costs and Financing 15 Annex 3. Economic Costs and Benefits 16 Annex 4. Bank Inputs 17 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 20 Annex 6. Ratings of Bank and Borrower Performance 21 Annex 7. List of Supporting Documents 22 Annex 8. Summary of Borrower's Project Completion Report 23 Map: IBRD 29056 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.   Project ID: P002327 Project Name: SMALL RURAL OPS. II Team Leader: Leopold Sarr _ TL Unit: AFMSN ICR Type: Core ICR Report Date: June 26, 2000 1. Project Data Name: SMALL RURAL OPS. II L/C/TFNumber: IDA-19920; COFN-01870 Country/Department: SENEGAL Region: Africa Regional Office Sector/subsector: AP - Perennial Crops; AQ - Annual Crops; AY - Other Agriculture KEY DATES Original Revised/Actual PCD: 03/15/1987 Effective: 06/07/89 06/07/89 Appraisal: 03/31/1998 MTR: 09/28/92 12/21/92 Approval: 03/21/1989 Closing: 06/30/98 12/31/99 Borrower/Implementing Agency: GOVT OF SENEGAL/MINISTRY OF SOCIAL DEVELOPMENT Other Partners: IFAD STAFF Current At Appraisal Vice President: Madavo Callisto E. Edward V.K. Jaycox Country Manager: Mahmood A. Ayub Michael J. Gillette Sector Manager: Jean-Paul Chausse David R. Steeds Team Leader at ICR: Leopold Sarr Joseph Baah-Dwomoh ICR Primary Author: Dirk Nicolaas Prevoo; Leopold Sarr; Francois Gadelle 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: U Sustainability: UN Institutional Development Impact: M Bank Performance: U Borrower Performance: U QAG (if available) ICR Quality at Entry: U Project at Risk at Any Time: According to the ICR mission the project should have been rated at risk shortlyfollowing the lifting of suspension.  3. Assessment of Development Objective and Design, and of Quality at Entry 3,1 Original Objective: The project was the successor to the Small Rural Operations Project (Cr. 991-SE), which was approved on March 11, 1980 and closed on December 31, 1987. The project's broad objectives were to encourage local initiative and thereby broaden popular participation in decision-making on rural investments, and to undertake only those small rural projects that are manifestly in the interest of the beneficiaries. Within this framework, the project would encourage small groups of producers to constitute themselves into legally recognized units such as Economic Interest Groups and render these creditworthy in the judgment of the local banking system; create employment in rural areas to help decrease rural-urban migration; and strengthen the existing Senegalese capacity for identification and preparation of small rural projects. The specific projects to be financed would increase crop, fruit and vegetable production, enhance security of production and raise rural living standards. The project was a standard second generation project with no particular innovations. The objectives were in line with Government's policy at the time, which accorded high priority to private sector development, the development of farmer's organizations, and a retrenchment of Government's role. The Staff Appraisal Report (SAR) states that the implementation experience under the predecessor project was incorporated in the design of this project (clearer delineation of responsibilities between project and executing agencies and a higher beneficiary contribution and participation in the decision process). Implementation experience showed, however, that major problems identified under the first operation -- poor technical designs and poor supervision by executing agencies -- were not adequately addressed. In addition, it is unclear to what extent beneficiaries participated in the selection and design of the subprojects. Indeed, several of these subprojects involved the rehabilitation of development initiatives that had failed previously, or were based on unrealistic project appraisals, as noted by subsequent supervision missions. The project's total cost was estimated at US$26.1 million, with IDA contributing US$16.1 million, IFAD US$5.1 million, the Government US$3.0 million and beneficiaries US$1.9 million. The project was to benefit about 4,500 farm families at an average cost of US$5,800 per family/US$500 per household member (including project overheads). Given the low expected rate of return of the subprojects (16%, excluding overheads) and the high sensitivity of this rate to changes in benefits or costs, it raises the question whether the project as appraised was an optimal use of scarce resources. This question is not addressed in the SAR. 3.2 Revised Objective: The project's objectives and rationale were judged to be still pertinent at the time of the project's mid-term review (MTR) in 1992, and again in 1996 when it was restructured after a five year suspension in disbursements (1992-97) due to financial management issues. 3.3 Original Components: The project, which was to be implemented over five years had the following components: * consolidation and rehabilitation of 12 of the 28 irrigated rice and vegetable perimeters developed under the first Small Rural Operations Project in the Saint Louis region; * construction, equipment and technical services for 10 rice and vegetable perimeters of 30 ha each in the zone lying between the Ngalenka and Senegal rivers in the Saint Louis region; * construction, equipment and technical services for four rice and vegetable perimeters of 20-30 ha each in the Matam area for four women's and youth groups; * construction, equipment and technical services for six perimeters (rice, maize, sorghum and -2 -  vegetables) of 30 ha each in the Bakel area of the Tambacounda region; * construction, equipment and technical services for a 15 ha vegetable perimeter and a 45 ha fruit tree operation for 100 families at Sirmang (Lower Saloum region); * construction, equipment and technical services for two perimeters of 60 ha each for rice, maize, banana and citrus fruit, at Fas near Medina-Gounass (Kolda region); * individual and collective equipment and technical services for 150 traditional beekeepers in three villages in the Kolda region; * construction, equipment and technical services for 2 banana perimeters of 15 ha each in Diende (near Sedhiou); * unidentified sub-projects/activities; and * improvement of the project management structure with the required technical and administrative staff and logistical support, including improvement of project accounting, monitoring and evaluation and links between the project directors and the sub-projects. 3.4 Revised Components: The project underwent a dramatic change as a result of its restructuring. Project management was transferred from the Government to a newly constituted non-governmental association, with close ties to producer organizations: the newly created Senegalese Association for the Promotion of Small Rural Development Projects (ASPRODEB - Association Senegalaise pour la promotion des petits projets de DOveloppement de Base), a Non-Govermmental Organization. This is a confederation of 13 producer organizations and committees that represent most areas of village-level economic activity in Senegal. The new set-up ensured a much more participatory approach than was possible under the previous structure.In addition ,this participatory process contributed in empowering the agricultural producers' leadership through Asprodeb/Agep and building trust between the Government and other project stakeholders. 3.5 Quality at Entry: Although project's objectives were sound and in line with Government's overall strategy as elaborated in 1984, its design was poor. The Board document for the restructured project states that dismal project performance was caused to a large degree by a lack of clarity in implementation arrangements. The relationship between the Project Management Unit (PMU) and the supervisory ministry, and between the PMU and the lnterministerial Steering Committee, was not clearly defined. This opened up the PMU to interference in its mandate and operating procedures by the frequently changing supervising ministry. To a certain extent the Implementation Completion Report (ICR) mission concurs with this assessment. The project basically adapted a model that had worked reasonably well for the predecessor project. The design team at the time had therefore no reason to drastically change that model given the information available to it at the time. The appraisal team omitted, however, the details for the implementation arrangements between the PMU and the executing agencies. The SAR merely mentions that certain clarifications need to be made, but fails to come up with a standardized terms of reference. As a result, the issue of poor civil works resulting from inadequate supervision also re-emerged. Project management costs were high relative to overall project costs (US$6.0 million for project management or 37 percent of estimated project costs). The new organizational set-up with ASPRODEP was sound and properly assessed. ASPRODEB with its roots in rural communities in Senegal was much better placed to put in motion the participatory process initially envisaged. Management costs remained, however, high relative to investments in part because of the set-up of the executing branch of ASPRODEB (Project Executing Agency or AGEP), which had 25 staff responsible for project implementation (16 in Dakar and 9 in 3 regional offices, most of them non-technical staff). An important issue facing the project was that project redesign had taken almost four years and the Bank was reluctant to change the closing date. The original closing date of June 30, 1998 was thus maintained in the revised legal document, even though the suspension of disbursement was not - 3 -  lifted until February 18, 1997. This led to difficulties in the identification of a pipeline of new sub-projects by the implementing agency and thus a low output relative to capacity/overhead costs. It also created problems with the implementation of already identified projects, because they all had to be completed by June 30, 1998. Two extensions were approved by the Bank. The first by one year to June 30, 1999. The second by six months until December 31, 2000. In each instance the extension was not approved until late. The issue of poor subproject supervision was partially addressed by relying more on beneficiaries and project staff. Technical quality did not improve much though, as it was increasingly carried out by smaller and less experienced construction companies. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: Overall, the project's outcome is judged unsatisfactory, in spite of having achieved its the majority of its identified investment targets and greatly exceeded the estimated number of beneficiaries (from 4,500 families to 10,400 families). 4.2 Outputs by components: In broad terms, the project achieved its objective of encouraging local initiative and effectively promoting beneficiaries' participation in the selection and design of agricultural investments. Although the early subprojects didn't fit this criteria entirely, subprojects became increasingly beneficiary-driven, with increased participation of beneficiaries in subproject selection and execution. The project was to finance about 85 pre-identified subprojects as well as a number of yet to be identified subprojects, to benefit about 4,500 farm families or 51,000 people. Overall, the project achieved the majority of its pre-identified investment targets and greatly exceeded the estimated number of beneficiaries (subprojects benefited about 10,400 farm families or 114,000 people). The outcome of the subprojects defined in the project's components are detailed in the tables in Annex 1. The project did fall short of the target for unidentified sub-projects, reaching only 41 percent of the expected value of total investment cost (no target in terms of number of subprojects was defined at the time of project appraisal). The shortfall in the investment level can, however, be explained by the change in the types of subprojects from high-cost medium-sized irrigated perimeters to smaller subprojects such as small vegetable perimeters and semi-intensive livestock and product transformation. Also, even though the suspension and redesign of the project had taken almost four years, with the suspension of disbursement lifted only in February 1997, the Bank was reluctant to change the project's original closing date of June 30, 1998. This led to difficulties in developing a pipeline of new sub-projects by the implementing agency and to a low output relative to capacity/overhead costs. It also created problems with the completion of several of the already identified projects, stopped during the suspension period, which had to be completed by June 30, 1998. Eventually, two extensions were nevertheless approved by the Bank. The first by one year to June 30, 1999. The second by six months until December 31, 2000. However, in each instance the extension was not approved until a few days before the closing date and this too prevented forward planning of sub-projects. The quality of the subprojects improved with the increased implication of beneficiary groups and the strong desire of ASPRODEB to associate with the most active producer groups. ASPRODEB also started to use local consulting firms to assess the capacities of producer groups. Costs were, however, in many cases higher than planned because of the long delays between subproject identification and implementation. In particular, the cost of irrigation works were high, almost twice as expensive as those in neighboring Mauritania under similar conditions. There is no satisfactory explanation for this large cost differential. In addition, several subprojects were dropped during project implementation because they were no longer -4 -  viable, partly because of the long delay between subproject identification and implementation (up to 10 years), but also because of the large size of some of the groups of beneficiaries which made it difficult to efficiently manage a subproject, and the improper selection of some groups. 4.3 Net Present Value/lEconomic rate of return: The SAR foresaw the creation of a monitoring and evaluation function at the level of the PMU. This was, however, not implemented and received little or no emphasis by Bank supervision missions. Reliable data is therefore not available to repeat the analysis presented in the SAR. Based on the information gathered during field visits, it is unlikely that the project attained its objective of a 16 percent economic rate of return. In several cases costs were higher than estimated and benefits much lower or delayed. 4.4 Financial rate of return: No financial rate of return was presented in the SAR. As the project did not track the production impact of the subprojects and beneficiaries had not been trained to keep written records to track their own performance, the impact of the subprojects on production and household incomes is not known. It is therefore impossible to assess whether the project has been able to achieve the production and income increases as detailed in the SAR. Given the basic lack of reliable data, the ICR mission could only make a rough estimate of the subprojects' impact on household incomes. The appraisal estimate called for an increase in household income by on average US$760. The ICR mission estimates the actual increase ranges from US$180-450, depending on the type of project. Actual project cost per beneficiary is about US$190. This information is presented for illustration purposes only as the data was obtained mostly from beneficiaries' recollections and not from written records. 4.5 Institutional development impact: The institutional impact of the project is moderate. ASPRODEB and its operational arm AGEP, is still a young, relatively inexperienced organization. ASPRODEB was established in February 1995 to take over responsibility for PPPR II. Its origins precede the project, however, as its core is the National Counsel for the Rural Population (CNCR, Conseil National de Concertation des Ruraux), a national lobby for rural people which had already demonstrated an active commitment to grassroots development. The CNCR continues to exert a powerful influence on ASPRODEB, as its president is also the president of ASPRODEB, and CNCR's member organizations account for 9 of the 13 member organizations of ASPRODEB. The project established and financed the AGEP within the ASPRODEB framework. Although staffed with well-trained professionals with experience in project management and implementation, AGEP's real performance has been relatively poor and its costs high, and further efforts are clearly needed to improve efficiency if it is to generate a real demand for its services and become a viable and sustainable institution. At the beneficiary level, capacity building was given insufficient priority until late in the project cycle. Initially, it mostly focused on technical support for subproject implementation, but often did not tackle financial, organizational and marketing issues. With the relaunch of the project in 1997, this deficiency was addressed and about four members of each producer group have since systematically benefited from such training. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: The main factor was the devaluation of the franc CFA in January 1994. This initially greatly reduced the cost of subprojects in foreign currency equivalent, as they mostly required local expenditures. Inflation in subsequent years erased most of these cost savings. - 5 -  5.2 Factors generally subject to government control: Project implementation was greatly hampered by the frequent change in supervisory ministry (4 different ministries in 4 years), which each had its seperate sets of demands on the project in terms of staffing, priorities and equipment needs. These changes caused the PMU to be in a perpetual stated of flux and unable to fulfill its mandate. Technical supervision of project imnplementation suffered, cooperation with implementing agencies was far from ideal, with no priorities accorded to the project, and subproject identification and implementation was not always decided on the merits of the subproject, but also influenced by political factors. Finally, financial management was weak and the PMU was unable to account for all the funds used by the project. It was for these reasons that the IDA credit and IFAD loan were suspended in 1992. It is unclear why the institutional set-up that had worked well under the first project, failed so miserably under this project. Government was unable to fulfill all conditions related to the lifting of the suspension of disbursements until February 1997. This was mostly due to the difficult negotiations between IDA and the Government on the transfer of project implementation and thus control, from the Government to an independent association. It is not clear why Government moved so slow and IDA was so patient. After the transfer of project implementation responsibility to ASPRODEB/AGEP, the Government's role in the project was substantially reduced and limited to its role in the project steering committee. As a result implementation performance of the project was improved. Beneficiary participation in selection and execution of projects increased and there was a marked change in the type of subprojects supported: away from capital intensive investments in irrigation to smaller and more directly productive investments such as cattle fattening. 5.3 Factors generally subject to implementing agency control: The project was in essence implemented by two agencies. Until 1992 by the PMU and after 1997 by ASPRODEB/AGEP. The PMU was weak, overstaffed and unable to carry out its mandate effectively because of continued political interference. As mentioned above, financial management and technical supervision of the suprojects was weak leading to problems related to the quality of the subprojects and its inability to justify all funds used for the project. It is difficult to seperate the impact of the changes in supervisory ministry from poor management at the level of the PMU. With the relaunch of the project in 1997. ASPRODEB/AGEP implementation performance -- planning, technical supervision and financial management -- was satisfactory. It also undertook to implicate beneficiaries in subproject design and execution, which had a clearly positive impact on subproject implementation. AGEP's operating costs were, however, high. Its set-up was too costly for the level of activities planned (25 staff). 5.4 Costs andfinancing: Planned and actual costs and financing are detailed in Annex 2. Project costs at appraisal were estimated at US$26.1 million. Actual project costs are estimated at US$21.45 million. For most subprojects, actual costs exceeded appraisal estimates by a significant margin, except for Matain and Fass, and the unidentified subprojects. Subproject costs were higher than estimated because in quite a few cases there was a substantial delay between identification and implementation and design issues had to be addressed. It is for this reason that several identified subprojects were dropped after the suspension of disbursements were lifted in 1997. Project management costs in absolute terms were slightly less than appraisal estimates (but not as a percentage of total project costs). The IDA credit was originally equivalent to US$16.10 million. In the end, US$14.68 million were utilized, in part because of the shortfall in unidentified subprojects, despite increased expenditures on subprojects that were identified before 1988. The contribution by Government was close to appraisal - 6 -  estimates according to the Government, this is, however, difficult to verify for the pre-1992 period. Likewise the contribution by beneficiaries is difficult to estimate. In principle, they have contributed to the subprojects, but this is difficult to verify ex-post. Supervision missions have underscored this problem on several occasions. The original IFAD loan was in the amount of US$5.00 million, its actual contribution is US$3.70 million. 6. Sustainability 6.1 Rationale for sustainability rating: The sustainablity of subproject is also not evident an therefore judged unlikely. In strict financial terms, few of the operating subprojects appear to be sustainable. Their capacity to generate a significant return after loan repayment and provision for capital replacement is limited, their ability to withstand adverse physical and economic conditions is weak. Half of all subprojects have been unable to meet their loan repayments, and none have managed to maintain their capital replacement accounts as envisaged in the original project design. Inadequate project training in product marketing and financial management has further reduced the short and medium-term prospects for financial sustainability. This issue was not satisfactorily addressed until after the 1997 project restructuring. However, this assessment ignores the capacity of participants to draw on other sources of income to finance loan repayment and capital replacement when required. Of the sub-sample of longer running subprojects visited, where the producer groups had benefited from longer project support, many had been able to replace their capital equipment when necessary, through recourse to cash contributions by members and use of the available funds in the capital replacement account. It is therefore imperative that beneficiaries continue to benefit from support and training for their access to agricultural production, marketing and management activities under the newly launched Agricultural Services and Producer Organizations Support Program (PSAOP). 6.2 Transition arrangement to regular operations: No specific transition arrangements were made to ensure the continuation of support to beneficiary groups after the end of the project. Arrangements are, however, being made for support for production and marketing activities to be provided under the IDA-financed PSAOP. ASPRODEBI/AGEP's future is also not ensured. ASPRODEB/AGEP has undertaken a management review to improve its efficiency and reduce the cost of its services. It is currently implementing activities for FAO and it may be involved in the implementation of certain activities for the PSAOP and the new National Rural Infrastructure Program. 7. Bank and Borrower Performance Bank 7.1 Lending: Bank's performance during preparation and appraisal was unsatisfactory. As mentioned in the project restructuring document, project design had several weaknesses, including the continued poor definition of relations between different project agencies (inter-ministerial steering committee, PMU and executing agencies) which created problems during project implementation, and the lack of specificity in the criteria for selection of sub-projects. 7.2 Supervision: Bank supervision must also be judged unsatisfactory. Even though supervision missions were adequate in fiequency, during the first two years of project implementation, they were deficient in terms of skills mix. The first three missions were staffed by an agricultural economist and IFAD's staff member responsible for Senegal. It was not until 1991 (and only with IFAD funding) that irrigation and credit specialists participated. The Bank, although making several recommendations against doing so, was unable to stop the Government from continuously transferring the PMU from one supervisory Ministry to another (4 times -7-  between 1998 and 1992), or ministry officials from interfering in subproject selection. Bank missions also failed to identify financial mismanagement and the deficiencies in subproject design and supervision, and to take the necessary corrective actions, until the mid-term review (MTR) mission. Following this MTR in 1992, IDA and IFAD took the right decision and suspended their disbursements. Agreement was rapidly reached with the Government on the repayment of funds that could not be accounted for and on the need to drastically change implementation arrangements. However, it took four years for the Government to fulfill all conditions for lifting the suspension of disbursements, in particular for effectively putting in place ASPRODEB/AGEP as the new project implementing agency. This does raise the question why it would not have been simpler to cancel the credit and design a new project that would properly and comprehensively address the implementation issues raised by the MTR, rather than trying to fix a large number of design problems. After the suspension of disbursements was lifted on January 17, 1997, project activities resumed only slowly, in particular because of the close oversight of the AGEP by the Bank, leading to delays in the sub-projects approval process (poor quality of proposals and delays in the task manager's response). The inportance of a monitoring and evaluation system was recognized but not seriously insisted upon. The decision to transfer responsibility for project supervision to the Resident Mission in Dakar early in 1998 did much to improve communication between Bank and project staff and the review process became more efficient. Technical field supervision by the Bank was satisfactory. However, no ex-post procurement and financial audits were conducted as part of regular supervision missions (the Bank had agreed with AGEP to limit the ex-ante review of the procurement process to subprojects expected to cost more than US$100,000). The ICR mission conducted the first ex-post procurement review and uncovered a number of irregularities in procurement that underscored the need to strengthen procurement procedures at the level of AGEP. Finally, failure to take pro-active action on the extension of the closing date in 1998, even though it had already been agreed upon, caused funds to dry up for the AGEP and kept the project from developing new subproject proposals. During the extension period, the project executed only subprojects identified with the help of FAO under the auspices of its Food Security Project. 7.3 Overall Bank performance: Based on the above, the overall performance is given an unsatisfactory rating. Borrower 7.4 Preparation: Borrower's performance during preparation was satisfactory although its implication in the preparation process was limited to general oversight and guidance, the bulk of the project preparation work was done by consultants, with little direct input by the Borrower as it lacked the capacity to do so. 7.5 Government implementation performance: Govermment's performance during project implementation was unsatisfactory. The change in the PMUs supervisory ministry four times in a relatively short period of time caused interruptions in implementation and was costly as each new move required the replacement of equipment and led to poor financial management practices. In addition, undue political interference led to the selection of below par subprojects. The Government, although agreeing in principle to take the measures necessary to take the measures necessary for lifting disbursements suspension, took four years to actually meet all conditions. In the end, however, it agreed to transfer project implementation to a non-Governmental agency, which at the time was a novelty, at least in the case of Senegal. Following the lifting of the suspension of disbursements, -8 -  Government's role in project implementation was reduced to final oversight through its steering committee. 7.6 Implementing Agency: The performance of the implementing agencies was globally marginally satisfactory. The performance of the original PMU was clearly highly unsatisfactory. However, although ASPRODEB/AGEP was an untested concept and was only given a short period of time to come to full maturity, its overall performance can be rated satisfactory. It succeeded in carrying out a relatively imnportant investment program with minor technical problems. It greatly improved selection of subprojects and prepared standardized succinct evaluations and developed a strategy for evaluating beneficiaries' capacity to implement subprojects. Adherence to procurement procedures was, however, not perfectly respected. Procurement of subprojects did not always properly follow either the projects implementation manual or the procedures outlined in the bid documents. Procurement of office supplies and other small items did not follow guidelines and there was a discrepancy between the (Bank approved) implementation manual and Bank guidelines on local shopping procedures. The computerized financial management system had limitations that made it cumbersome to use. The quality of financial management was, however, satisfactory. 7.7 Overall Borrower performance: The overall performance of the Borrower was rated unsatisfactory. 8. Lessons Learned The project's poor performance illustrate a number of well-known generic lessons, such as the need to have a design suitable for local implementation capacities; the need to have the full participation of beneficiaries in design and implementation; clear responsibilities, accountability and detailed implementation plan from the start of the project; the critical need to have adequate management, monitoring and reporting system to ensure efficient implementation and accountability; the critical need, on the Bank side, to have implementation support missions with the required skill mix. Two important specific lessons can in addition be highlighted from this operation: (i) first, that the success of a project, or sub-project, in particular when it depends on mobilizing the interest and participation of beneficiaries, critically depends on the efficiency and speed of its implementation: the long delays between identification of some of the sub-projects under this operation not increased their investment costs, and therefore reduced their financial and economic returns, but also had a demotivating effect on their beneficiaries; (ii) secondly, that it is indeed almost always better to close a project with major design flaws and implementation problems and start anew if there is serious commitment to do so from the government side, than to try and restructure the operation, if only because of the need of a different mind-set of the supervisory and implementing agencies/staff. 9. Partner Comments (a) Borrower/implementing agency: Below are the specific comments made by the Borrower on the draft report. Several of these were incorporated in the final report. Page 3 - paragraphe: 3-5 "Qualite a l'entree du portefeuille" Ecrire en premiere ligne qu'il s'agit bien de resultats insuffisants, notamment en matiere de gestion financiere. 9 -  Page 4 - paragraphe 1 La CNCAS avait bien une experience en matiere de ligne de credit de projet domicilie en son sein (conf6rer: lignes de credit du PRIMOCA, du DERBAC etc...) Page 4 - paragraphe 2 Le probleme de l'insuffisance de la qualite technique des infrastructures realis6es reside dans le fait que l'unite operationnelle de l'AGEP, ne pouvait A elle seule se demultiplier dans les actions de contr6le des prestations de service depuis la phase identification, etude technique d'execution jusqu'A la realisation effective. II s'y ajoute que le Comite de surveillance de I'ASPRODEB n'a pas pleinementjoue son role. Page 5 - paragraphe 2 * Des sessions de formation en gestion financiere (compte d'exploitation) ont bien ete dispensees au ben6fice d'agents membres de GIE. * Le non-versement des provisions d'amortissements dans les comptes bancaires ouverts a cet effet, ne concerne pas tous les sous-projets. * Du fait des retards enregistres dans la realisation effective de la plupart des sous-projets selectionnes A partir de la levee totale de la mesure de suspension des credits, la mise en exploitation moyenne de ceux achetes est de 2 A 3 ans ce qui est suffisant pour evaluer l'impact sur les revenus des ben6ficiaires. Page 5 - paragraphe 4-3 Le systeme de suivi-evaluation n'a pas ete mise en place par l'AGEP en depit de la dfinition des outils y afferents. Page 6- paragraphe 4-5 (suite) - Le C.N.C.R. n'est pas un <