Niger - Dosso Agricultural Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7333 PROJECT COMPLETION REPORT NIGER DOSSO RURAL DEVELOPMENT PROJECT (CREDIT 967-NIR) June 24, 1988 Africa Regional Office This document bas a restdcted diutributon and way be used by reipicnbt only In the perfonrmaee of their official duties. Its contents may nt othenrise be disdosed without World Bank authorizatio  NIGER DOSSO RURAL DEVELOPMENT PROJECT PROJECT COMPLETION REPORT ABBREVIATIONS BDRN - Banque de Developpement de la R6publique du Figer (Niger Rural Development Bank) CF.A - Centre de Formation des Jeunes Agriculteurs (Young Farmers' Training Center) CCCE - Caisse Centrale de Cooperation Economique (French Agency for Economic Cooperation) CNCA - Caisse Nationale de Credit Agricole (National Agricultural Credit Bank) CNSS - Caisse Nationale de Securite Sociale (Social Security) COTEAR - Comite Technique d'Arrondissement (District Technical Committee) COTEDEP - Comit6 Technique Departemental (Provincial Technical Committee) CPR - Centre de Promotion Rurale (Residential Training Center) DEP - Direction des Etudes et de la Programmation (Studies and Programs Directorate) EDF - European Development Fund FAC - Ponds d'Aide et de Cooperation (French Fund for Assistance and Cooperation) FNI - Fonds National d'Investissement (National Investment Fund) GM - Groupement Mutualiste (Mutual guarantee group) IDA - International Development Association INRAN - Institut National de la Recherche Agronomique au Niger (Niger National Agricultural Research Institute) MDR - Ministere du Developpement Rural (Ministry of Rural Development) MES - Project Monitoring and Evaluation Section OED - Operations Evaluation Department PPF - Bank Project Preparation Facility RMWA - Regional Mission in Western Africa SAD - ' :vice Agricoli Departemental (iPcovincial Agricultural Service) UNCC - Union Nigerienne de Cr6dit et de Cooperation (Niger Credit and Cooperation Union) USAID - US Aid Agency for International Development  FOR OVICIAL US OLY TWf WOtLD SANK W.stwton. O.C. 20433 U.S.A. O0Co Dir0.cowaW Opsata Iabkasto Jtue 24, 1988 MEMORANDUM TO TEE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report - Niger Dosso Rural Development Project (Credit 967-NIR) Attached, for information, is a copy of a report entitled "Project Completion Report - Niger Dosso Rural Development Project *(Credit 967-NIR)" prepared by the Africa Regional Office. No further evaluation of this project by the Operations Evaluation Department has been made. Attachment Yves Rovani by Graham Donaldson Thsdocument has a rstficted ditbutonand may be ued by recipients only In the pufoms_ oftheirofAcial dutis Its contents may nsot odwise be discloed without Wowd Bank authoiuatlonL  IL - . iiL X . _ E NIGER DOSSO RURAL DEVELOPMENT PROJECT PROJECT COMPLETION REPORT TABLE OF CONTENTS Page PREFA CEC..** ...... i ABBREVIATIONS ...*............................... ii BASIC DATA SHEET..... ... .. ..... **to.* ........ .. iii EVALUATION SUMMARY ...... **t ..... . vi I. BACKGROUND*.* ..... ...................... II. PREPARATION AND APPRAISAL ................... *. ... 2 Project descriptions...... a... #. ......3 Cost and financing ......4....... ..... .... Organization ....5.* .... III. PROJECT IMPLEMENTATION .......................... . 6 Changes after appraisal ............................6 Project yerformance .......... ... ..*.6 - Administration and management......***..* ....... . .7 - Recruitment of staff .... . ................ . . .. ..8 - Technical assistance. ..... . . .. . .. ...... ........ .8 - Monitoring and evaluation .... ........ .....8 - Applied research.*... ..to ........ ... ....*8 - Seed multiplication and distribution ............. .9 - Fertilizer distribution ..........................9 - Animal traction ....... *.. ... ...... .... - Credit .................................1 - Training and extension ..........................l2 - Bottomland development ................ ....... . 13 - Livestock.,.*,... #.... ^1 - Buildings, vehicles and equipment .....4 - Strengthening of farmers' associations ...l.....14 Project operational constraints.....*..**...... .14 IV. PROJECT ACHIEVEMENTS ...... ............. . 15. 15 V. PROJECT EXPENDITURES AND FINANCING ................ 16 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.  VI. INSTITUTIONAL DEVELOPMENT ................. . ...... . 20 VII. PERFORMANCE BY GOVERNMENT AND IDA ................. 22 VIII. LESSONS OF THE PROJECT .......... ................ 23 TABLES 1. Assets and liabilities 1981-1985 2. Source and application of funds 3. Projected expenditure 4. Actual expenditure 5. Actual expenditure as a percentage of projected expenditure 6. IDA disbursements 7. Project objectives and achievements Annex - Commente from Caisse Centrale de Cooperation Economique Maps: IBRD 14214, 14215  PROJECT COMPLETION REPORT DOSSO RURAL DEVELOPMEPT PROJECT (Credit 967-NIR) PREFACE This is the Project Completion Report (PCR) of the Dosso Rural Development Project, for which IDA Credit 967-NIR in the amount of US$20 million was approved on December 20, 1979. The Credit Agreement was amended on June 29, 1984, to modify allocations by disbursement categories and disbursement percentages. The project hat% been jointly financed with he French Caisse Centrale de Coopdration Economiaue (CCCE) and French bilateral aid (FAC), in the amount of 20 million and 10 million French francs respectively. The original closing date of June 30, 1985 has not been extended, in spite of a Government request. However, one disbursement category was kept open until December 31, 1985 to audit the project accounts at the closing date. Tne final disbursement was approved on December 31, 1985. More than US$16.5 million of the IDA credit was not disbursed and was cancelled. The project was the seventh financed by the Bank in the rural sector, the third and by that date so far the largest one specifically for agriculture. It followed pilot operations and was prepared by consultants assisted by Bank staff, on the basis of two studies financed by the PAC and Belgian bilateral aid. A small rural operations project, to be implemented in part in the same area was appraised in May, 1987. This PCR was prepared by the West Africa Region and is largely based upon a report of the FAO/Bank Cooperative Program, and on information available in a completion report prepared by Government, the staff appraisal report (no. 2493-NIR) and the attached working papers, the President's report (no. 11109a-NIR), the IDA Credit Agreement of June 13, 1980, the amendment to the Credit Agreement, dated June 29, 1984, various reports issued by the project man.igement, donors' supervision reports and other relevant documents, as well as upon interviews and discussions with donors' staff and Nigerien officers who have been associated with the project. A copy of the draft report was sent to the Borrower on February 17, 1988, for comments. Comments received from the CCCE have been incorporated and is attached as an Annex. This project has not been subjected to an audit by OED.  - ii - NIGER DOSSO RURAL DEVELOPMENT PROJECT (Credit 967-NIR) Basic data sheet Key project data Appraisal Actual or Actua. as X of estimate estimated appraisal actual estimate Total project cost (US$ millions) 39.3 12.2 31.0 Total project cost(CFAF millions) 8,639.0 4,203.0 48.7 Credit amount($US millions) 20.0 3.2 16.0 Credit amount (CFAF millions) 4,400.0 1,210.0 27.5 Other donors: - CCCE ($US millions) 4.5 1.5 33.3 - CCCE (CFAF millions) 1,000.0 552.0 55.2 - FAC ($US millions) 2.3 0.8 34.8 - PAC (CFAF millions) 500.0 303.0 60.6 Date Board approval 12.20.79 Date Credit agreement 06.13.80 Date Credit effectiveness 09.16.80 02.25.81 Closing date 06 30.85 06.30.85 Economic rate of return (%) 25 N.A. Financial rate of return (Z) N.A. N.A. Institutional performance unsatisfactory Number of beneficiaries 25,000 N.A.  - ill Cumulative disbursements FY 80 FY 81 FY 82 FY 83 FY 84 FY 85 FY 86 Appraisal Estimate (US$ millions) 0.10 1.50 4.20 9.80 15.80 119.60 20.00 Actual (US$ millions) - 0.12 0.82 1.76 2.27 3.16 3.20 Actual as Z of estimate - 8.0 19.5 18.0 14.4 16.1 16.0 Date of final withdrawal approval: December 31, 1985 Other project data Borrower : Republic of Niger Executing Agency : Ministry of Rural Development On-lending institution : N.R. Fiscal year of borrower : April 1 to March 31 Currency exchange rate Average exchange rates 1/ Used at appraisal 1978 1 US$ - CFAF 220 1979 1 US$ - CFAF 213 1980 1 US$ - CFAF 211 1981 1 US$ - CFAF 272 1982 1 US$ - CFAF 329 1983 1 US$ = CFAF 381 1984 1 US$ - CFAF 437 1985 1 US$ - CFAF 449 I/ From Interr.ational Financial Statistics published by IMF Related projects Preceding projects Follow-up project Name Agric. Credit Drought relief SROs Credit Number 207-NIR 441-NIR NA Credit amount (US$ millions) 0.6 2.0 NA Date Board approval 06.16.70 11.13.73 NA  Supervisions Date No.of Nd in Specializations Performance Trend Types of (mo. yr) persons field (1! represented (2) ratina (3) (4) problems (5) Identification (Rl4WA) 03/77 Preparation I (FAC) 04/78 - - Preparation II (Bank) 10/78 - - Appraisal - IDA 12/78 6 1:O A,L,T,E,AE,I - Supervision I 06/80 2 28 A,E 2 1 M,F SupervisJon Ir 10/80 2 14 A,F 2 1 M Supervision III 03-05/81 2 30 A,F 2 2 M,F Supervision IV 01-02/82 2 20 A,F 2 1 M,F Supervision V 06/82 1 3 A - - Sq Supervision VI 12182 1 16 A 3 2 T,F,M Supervision VII 02/83 1 8 L - T Supervision VITT 12/83 1 18 A 3 2 T,F,M Supervision IX 03/84 4 72 A,L,AE.RE 3 2 T,F,M Supervision X 09/84 1 13 A 3 3 N,T PCR 10/86 1 13 S - - - 4 (1) Time spent in Niger by several supervision missions also covered other activities and only part of total time was devoted to the Dosso project. (2) A: Agriculture; L: Livestock; T: Training; E: Extension; AE: Agricultural economics; I: Institutions; F: Financial analysis; RE: Rural engineering; S: Sociology (3) 2: Moderate problems; 3: Major problems (4) 1: Imprwving; 2: Stationary; 3: Deteriorating (5) M: Managerial; F: Financial; T: Technical STAFF INPUTS (Staff Weeks) FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 Pre-appraisal 16.9 12.4 19.3 Appraisal 78.1 Negociations 4.2 Supervision 8.1 19.6 9.5 10.9 14.9 5.5 1.6 PCR .1 10.1 1.1 Other .2 1.1 4.1 4.9 .4 .4 .1 .2 .3 1.3 Totals 17.1 13.5 101.5 17.2 20.0 9.9 11.0 15.1 5.8 3.0 10.1 1.1 TDTAL AFF INPUTS IN STAFF WEEKS - 225.4  PROJECT COMPLETION REPORT DOSSO RURAL DEVELOPMENT PROJECT (Credit 967-NIR) EVALUATION SUMMARY Introduction 1. This was the third project financed by the Bank addressing directly the development of Nigerien agriculture. The two other projects were for rural development in the Maradi area, still in progress under a follow-up project, and for irrigation. Two other earlier projects - Agriculcural credit and Drought relief - were clso related to some extent to agriculture. The Bank had in addition financed in the rural sector a forestry project and a livestock project. (para 1.4) Niger, which used to be self-sufficient in terms of basic foodstuffs, went through a severe drought from 1973 to 1975, which resulted in a large increase in cereal imports. This disquieting experience highlighted the importance of national cereal production, a subsector hitherto neglected. The concern did not wane in spite of good yields in 1976/77 and 1978/79. Endowed with generally poor soils and an unfavorable climate, Niger was furthermore facing a problem of declining fertility, due to high population growth, translating into pressure upon the land and shortening of the fallow period. (para 1.3) It appeared, therefore, that an intensification effort would be justified, which would mainta.- self-sufficiency in cereals w1th average rainfall, and provide some surpluses to be marketed in Nigeria in good years. Government informed a Bank (RMWA) identification mission of its interest in obtaining IDA assistance to expand an on-going FAC-financed pilot project, aiming, in particular, at increasing yields of major crops. Project preparation started with PAC financial assistance and was completed with PPF funding by a team of consultants, assisted by Bank staff. The project was appraised in end-1978 and the IDA credit was approved in December 0979. The project was financed jointly by the FAC and CCCE. ObJectives 2. Project objectives were in agreement with Government policies of (i) exploiting the potential in rainfed crops and livestock production, and increasing the area under controlled irrigation, (ii) in respect to rainfed agriculture, setting up in each province a productivity projezt, to improve all services available to smallholders, and (iii) developing agricultural production, to cover the domestic needs for foodstuffs, to raise farmers' incomes, and to improve the standard of living of the rural population. (para 2.5) More precisely, the prcject was intended to develop the production of cereals, cowpeas, groundnuts, livestock products and bottomland crops, and to reverse the declining trend in productivity. The project included (i) the construction of a training center for local staff, (ii) the construction and operation of 24 residential centers for farmers' training, and introduction of the "Training and Visit" extension system  - vi - (para 2.5 (a)), (iii) the improvement of farming practices and application of inputs on some 150,000 ha, and accelerated supply of donkey-traction and ox-traction farm equipment (para 2.5 (b)), (iv) the improvement of veterinary coverage and animal husbandry (para 2.5 (d)), (v) the development of stall fattening, (vi) the rehabilitation of a cattle market and a slaughterhouse, (vii) the establishment of a rural engineering provincial service, for construction works and development of 200 ha of bottomlands (para 2.5 (e)), (viii) applied research, particularly for protection of soil fertility (para 2.5 (f)), and (ix) a small project unit, working with the farmers through the regular Government technical agencies, but having the capability for evaluating project progress and for providing guidelines for future projects (para 2.5 (g)). The total cost was estimated at US$ 39.2 million, or CFAF 8.639 billion, to be financed by IDA (51Z), CCCE (11%), the FAC (6%), and the balance by local contributions from Government, CNCA and beneficiaries (para 5.4). The IDA credit of US$ 20 million was to be disbursed over somewhat less than five years. Implementation experience 3. The project was expected at appraisal to be implemented over five years, but the actual implementation period was substantially shorter. The Credit Agreement signed in June 1980 assumed that the project would become effective in December 1980 and be completed in December 1984. It took however longer than anticipated to meet the conditions of effectiveness and the project only became effective in June 1981 (paras 3.1, 3.2). In mid-1983, the Government submitted a request to reorient project activities and to extend the closing date. It was also upon the request of donors, and particuliarly the CCCE, which had envisaged a two-part financing plan (three years, then two years), that Nigerien authorities were led to prepare a reorientation of the project. The reorientation proposals, prerared by project staff, were only made available in March 1984. They were largely similar to those which were reviewed at the end of 1983 for the Second Maradi project and which were still being discussed with Government. In essence, they consisted of involving the rural population much more than before in the planning and implementation of development activities, and of making the farmers responsible for the maintenance and operation of investments. Project operating costs were to be reduced, by developing a new and less costly village-based extension system, and by testing alternatives to the expensive and largely unsuccessful training of farmers in residential centers. Donors, including the Bank with a four-staff-member mission, reviewed in depth the proposals in the field. They concluded that the new approach being contemplated entailed dramatic changes that would require imaginative and energetic management. The chances of achieving this successfully appeared to be reasonably good in Maradi, where there was at the time an effective project management team. In Dosso, however, where project management was particularly weak, it appeared most unlikely that the proposed reorientation would have a chance to succeed, unless the capability and attitude of local staff changed, and unless the whole project team and most chiefs of Government services were replaced (paras 3.4, 3.5, 6.1). Consequently, donors together decided to stop the project in May, 1984 and informed the Government that they saw no  - vii - justification for extending the project, which was cXosed as projected on June 30, 1985 (para 7.3). Hc-sever, one disbursement category was kept open, to allow for the audit of project accounts at the closing date, and the final withdrawal application was approved on December 31, 1985. Total project expenditures, expressed in local currency, were less than half of projections because major targets were not met, and disbursements of the IDA credit reached only 16% of projections in US dollars, but 27.5% of projections in CFAF, due to a higher than anticipated exchange rate of the US dollar versus the CFAF. As a consequence, some US$ 16.8 million of the IDA credit were cancelled at the closing date. In contrast, Government and other local sources contributed 51% of total costs instead of 32% as planned, equivalent to 77.5% of projected disbursements. 4. The appraisal mission was aware that a rural development project in the Dosso region would face multiple technical, institutional and sociological problems. During project implementation it became clear that the recommended technical packages, which had been designed during a period of satisfactory rainfall, were not sufficiently proven with the rainfall deficit which prevailed during much of the project life (para 3.24 (a)). They were also ill-adapted to the diversity of conditions in the project area and did not well enough take into account farmers' constraints (para 3.11). The packages were based upon a considerable increase in the application of chemical fertilizer, which was to be largely subsidized. In fact, fertilizer distribution through official channels remained negligible (para 3.14). Government was unable to continue subsidizing inputs as expected, due to the loss of revenues caused by the uranium crisis, and cheaper fertilizer from Nigeria was available anyhow on parallel markets (para 3.13 (a)). Furthermore, despite experience with the earlier pilot project, farmers' motivation had not been well perceived and understood. Total fertilizer consumption is not known precisely, due to unknown imports from Nigeria, but has without a doubt been much below appraisal pro'ections, which anyhow were too optimistic. The project also suffered from an unsatisfactory management structure. The Project Manager was never able to effectively coordinate activities of technical services, with a large turnover of local staff, which remained under the de facto authority of their respective ministries (paras 3.7, 3.24 (e). Furthermore, technical assistance staff, whose responsibilities were not clear, also lacked stability and some of them were not replaced (para 3.8). Finally, supervision by the Bank and other donors was inadequate at the most critical time (paras 3A.4, 7.3). Results 5. Accor-ding to Government statistics, agricultural production of the province increased during three consecutive years. It is unlikely that the project substantially contributed to such an increase. Evidence suggests, to the contrary, that the prcject had at best a marginal impact and that the recommended technical packages were little adopted by farmers. In the absence of reliable data, no meaningful rate of return can be computed. In view of the limited achievements of the project, it is however reasonable to assume that the actual rate of return is well below  - viii - the 25X projected at appraisal, and possibly close to zero or even negative. Applied research, monitoring and evaluation (MES) of some aspects of the project, and bottomland development are the project components with the most satisfactory results. Little has been achieved in the livestock sector, but vaccination coverage has remained reasonably good, although it gradually decreased. Stall fattening developed outside the project.(para 4.2) Training of farmers in residential centers was not up to expectations, and the "Training and Visit" extension method was rapidly abandoned, in part because it was perceived to be too costly, and in part because it was not succeeding in getting the technical packages adopted. In retrospect, the packages were wrong, not the extension method (para 8.1 (e)). Institutional arrangements have not worked, but the provincial rural engineering service, set up by the project, has been useful. Sustainability 6. Rules and methods for developing bottomlands have been defined and continue to be applied (para 3.20). Farmers' requests are increasing for such investments, which involve beneficiary participation in cash and in kind. This is encouraging for the future small rural operations project. Trials by the applied research section have provided a number of practical proposals, which have not been used for the project but will be of interest later for the development of the area. In spite of limited impact upon the livestock sector, the project has contributed in the longer term to the better handling of cattle herds, through training of field staff, setting up a small stock of drugs, and at least starting systematic monitoring of animals (para 3.21). The many reports of MES have helped to better understand the overall rural environment, which will be useful for preparation and implementaticn of future projects in the region. Findings and lessons 7. Findings and lessons to be learned from the project include: (a) Projects covering large areas, with a wide variety of conditions, and many components, face considerable problems, in particular of management and coordination; simplicity must be the watchword. (In addition, it will be necessary to find appropriate modes of intervention by better incorporating the notion of "time" (i.e., results cannot be obtained with the same speed in the following areas: distribution of existing improved seed, perfection of new production systems adapted to farmer strategies, creation of farmer organizations, etc.), and by taking into account the links between different geographic/administrative levels (plot, farm, village, region, country, group of countries....)(para 8.1 (a)); (b) When there is a large range of ecological variations and rainfall is a crucial factor, findings and recommendations of a pilot project in the same area are not necessarily replicable; on the  - ix - other hand, experience gained with similar projects ought to be taken into account during project preparation (para 8.1 (b)); (c) Project preparation requires an in-depth analysis, not only of physical conditions, but also of all farmers' incentives and constraints in order to set realistic targets; this can best be done by representative pilot operations, not by extending the orthodox preparation process even further. (One may also wonder if, in the interest of efficiency, integrated rural development projects must not rely on one or two successful "products" and/or on a real regional unity.)(para 8.1 (c)); (d) Small rural operations, such as bottomland development, with a rapid impact upon revenues, are of interest to the farmers, even when participation in cash and in labor is being requested (para 8.1 (d)); (e) An institutional structure has to be sufficiently flexible to adjust to changing circumstances, but has nonetheless to define clearly from the start the responsibilities of all participants to avoid any risk of misunderstanding (pars 8.1 (f)); and (f) It is difficult to judge to what extent less erratic supervision by the Bank and other donors would have improved project implementation, but it does seem that, without interfering in daily management, more systematic supervision, particularly in the initial stage, would have helped in solving problems, in overcoming constraints, and in taking rapidly whatever measures were needed to improve the situation (paras 3.24 (g), 7.3).  NIGER .;OSSO RURAL DEVELOPMENT PROJECT (Credit 967-NIR) PROJECT COMPLETION REPORT I. BACKGROUND 1.1 At the end of the 1970s Niger was in a favorable economic position, and the time seemed appropriate for investments in the agricultural sector. As the world's fifth largest uranium producer, the country drew 80% of its revenues from uranium exports, which at that time commanded a high price on the world market (CFAF 24,500 per kg. fob for uranium metal). In 1979, external debt service represented only 3% of export earnings. 1.2 Government's strategy for the rural sector consisted of (i) developing the potential for irrigation; (ii) increasing both production and productivity of rainfed farming through crop- and region-specific projects; and (iil) rebuilding cattle herds that had been decimated during the period of drought. Since for practical reasons the area under irrigation could not be increased by more than about 1,000-2,000 hectares annually, the chief potential 'or development lay in rainfed crops and livestock production. 1.3 The Bank supported this strategy, which aimed at achieving food self-sufficiency while Laintaining exports of agricultural and livestock products, particularly meat. Nevertheless, at the time of the Dosso project appraisal in 1979, the risks of investing in rainfed farming and livestock were pointed out, the main constraints being inter-alia uncertain climatic conditions, the absence of adapted technological packages and poor soils. 1.4 The Dosso project was the seventh financed by the Bank, wlhich by that time was providing assistance to the major agricultural and livestock areas of the country. Between 1969 and 1979, six credits were approved for the rural sector, namely: (a) 1970: Credit 207-NIR for US$ 0.6 million, to provide smallholders with credit. The project ran into difficulties, mainly due to the 1971-74 drought, but also to the lack of technical advice and effective extension services, which should have accompanied credit provided to producers; (b) 1974: Credit 441-NIR for US$ 2 milion, to finance a variety of sub-projects aimed at assisting economic recovery after the drought. Execution was generally satisfactory; (c) 1976: Credit 603-NIR for US$ 10.7 million, for rural development in Naradi. This was the reasonably successful first phase of a project, whose second phase, including a sizable irrigated  -2- component, is still in progress after undergoing major changes at the end of 1983; this follow-up project (Credit 1026-NIR for US$ 16.7 million) encountered problems and required attentive monitoring; (d) May 1978: Credit 800-NIR for US$ 4.5 million, for a forestry project, including a pilot component of 400 hectares of irrigated tree plantations; the project was disappointing concerning estate plantings, but was promising with respect to protection of the environment and resources; (e) September 1978: Credit 8U1-NIR for US$ 15 million, for the Namarigoungou irrigation project, including a new 1,500 hectare perimeter, which was successfully completed in the end of 1983, but the irrigated perimeter was included in the "Irrigation Rehabilitation Project" (Credit 1618-NIR for SDR 9.6 million) because further support was needed; and (f) March 1979: Credit 885-NIR for US$ 12 million, for a livestock project, to provide support services and a variety of innovative technical packages to pastoralists in the semi-arid zones in the center and east of the country; in spite of problems caused by severe drought in the early eighties, the project proceeded well and a follow-up is being prepared. 1.5 The present report deals with a project that benefitted from the largest credit (US$ 20 million) to the rural sector made available to Niger by the International Development Association (IDA). The main sources of information are the documents on the project prepared by Bank supervision missions, the CCCE (the French Agency for Economic Cooperation) and the FAC (French Fund for Assistance and Cooperation), together with some of those produced by the project itself. Government, with the assistance of provincial authorities, prepared a project completion report which served as the basis for the present mission's discussions with various technical officers. During the mission's visit to Niger, from October 5 to 18, 1986, it was not possible to meet former project officers. Since precise quantitative data are not available, this report is concerned essentially with the quality of performance. II. PREPARATION AND APPRAISAL 2.! During the years 1974/75 to 1979/80, the FAC (French Fund for Assistance and Cooperation) financed a pilot "productivity project" in the Dosso Pruvince ("departement"). The project, which was based essentially upon the cultivation of cowpeas, aimed at rehabilitating soils in areas under rainfed crops and at increasing yields. Cereals, planted after cowpeas, would benefit both from the residual effects of chemical fertilizers and from nitrogen-fixing by cowpeas. Technical messages relating to the project (pure stands, drill-seeding, application of chemical fertilizer) were based upon 10 ha blocks of land in which  individual farmers had one ha plots. This approach in itself was not too popular. Nevertheless, a considerable number of farmers voluntarily grew more cowpeas and carried over to part of their traditional fields what they had been taught in collective fields. Under the circumstances, the pilot project was quite reasonably regarded as successful, even though the high yields that were observed may well have been due in part to the favccable rainfall during the period. 2.2 Several other projects were also being undertaken at the time in the Dosso Province, including a USAID-financed seed production project, which encountered serious distribution problems. In the livestock sector, the European Development Fund (EDF) and the FAC were providing financing to improve and extend the fattening of beef cattle and to encourage the use of animal traction. At the time of appraisal of the Dosso project, early results were considered encouraging, despite a somewhat unsystematic approach. In principle, the Dosso project was expected to extend activities of the productivity project and to create favorable conditions for integrating the development of agriculture and livestock. 2.3 After the visit by a Bank project identification mission in March-April 1977, terms of reference were prepared for a feasibility study which was financed by the FAC. The team responsible for the study visited the country in September and November 1977, and suhmiitted a feasibility study in March 1978. A parallel visit by a team from Belgian bilateral aid took place in Jul,-September 1977, to examine the justification in the area of a small-scale rural development project. The Belgian team's report was submitted to the Government in early 1978. 2.4 Government decided that an additional study was needed to reconcile the approaches to development proposed in the reports financed by the FAC and by Belgian bilateral aid. A Bank mission visited Niger in May 1978 and reviewed the two reports, and in agreement with Government prepared terms of reference for the proposed supplementary study. At the request of Government, the Bank made available US$ 30,000 under the Project Preparation Facility (PPF) to finance that study. The supplementary study, prepared by consultants with the assistance in the field of Bank staff, was completed in October 1978. The project was appraised in November 1978. The appraisal report includes as annexes five working papers covering social aspects, training and extension, livestock development, project costs, and economic analysis. The authors of the first three papers raised a substantial number of technical, institutional and sociological issues, and they insisted that these issues be taken into consideration and that appropriate answers be found. The appraisal mission was consequently well aware of the difficulties of setting up a rural development project in the Dosso Province. (Furthermore, an additional mission was necessary after the appraisal mission ex-ante to work out disagreements between the donors and the Nigerien authorities on the CPRs and the extension system. Project description 2.5 According to the appraisal report, the overall objective of the  - 4 - proposed project was, in agreement with Government's policy of increasing agricultural production and farm incomes, to improve the welfare and standard of living of the rural populaLion. The project's specific target was to increase the production of cereals, cowpeas, groundnuts, livestock, and bottomland crops, and to reverse the declining trend in productivity. The project was expected to benefit some 25,000 farmers during the five-year implementation period. The proposed components consisted of: (a) constructing a training center with facilities for 75 boarders, providing courses for village extension workers and staff, introducing the "Training and Visit" system for agricultural extension, and constructing and operating 24 r-rmers' residential training centers (CPRs); (b) improving farming practices, increasing Input use on about 150,000 ha of rainfed cereals, cowpeas and groundnuts, and supplying about 6,400 donkey-traction units and about 1,800 ox-traction units; (c) strengthening the farm input and implement supply system, and constructing 30 new and improving 12 existing rural cooperative centers; (d) improving the vaccination coverage and the supply of drugs and feed supplements, and upgrading animal husbandry practices for beef cattle and work oxen; (e) establishing -a rural engineering unit to plan and supervise the construction of dispersed rural works, and to develop 200 ha of small-scale irrigation in bottomlands; (f) conducting trials to test the introduction of new techniques and farming practices, particularly with regard to soil fertility; and (g) setting up a management structure with the required technical and administrative staff and logistical support, capable of evaluating project progress and providing guidelines for future projects. Cost and financing 2.6 The total project cost was estimated at US$ 39.2 million. Base costs amounted to US$ 31.4 million. The various components listed in paragraph 2.5 represented the following percentages of base costs:  Component Percent (a) 19 (b) + (c) 61 (d) 7 (e) + (f) 2 (g) 7 Financing for the project was provided by IDA, CCCE and FAC with a substantial contribution from Government (see Chapter V - Project expenditures and financing). The economic rates of return were estimated at 25% for the project as a whole; 232 for rainfed farming (89% of project costs); 40% for livestock development (9% of project costs); and 41% for bottomland development (2% of project costs). The IDA credit was granted on standard terms, providing for repayment of the principal over the period July 1, 1990 to July 1, 2029. Organization 2.7 Unlike the Maradi project, which was provided with a conventional type of project structure with its own technical services, according to the appraisal report the Dosso project was designed "to avoid duplication of effort". The project was to be executed through the regular government technical agencies operatiag in the Dosso Province. All proposed activities were, however, to come under a single Project Manager, who would report to the Minister of Rural Development and the Provincial Governor ("prefet" of the "d4partement"). He would be responsible for planning, supervising and managing all project activities, including the activities of all project-financed staff in the provincial agencies. To support him in this task, two sections were attached to his office: (i) the Commercial, Administrative and Financial Section, and (ii) the Technical Section, responsible for monitoring and evaluation. For all matters related to project execution, the provincial officers for agriculture, livestock, input supply (UNCC), credit (Caisse Nationale de Credit Agricole - CNCA), rural engineering, and applied research (Institut National de la Recherche Agronomigue - INRAN) were to report directly to the Project Manager, and each of these agencies would receive a share of financing, in proportion to the activities it was expected to undertake. Provincial agencies were, however, to remain under the authority of their head offices in Niamey. This ambiguous situation was never clarified. Coupled with the inability of project management to exercise sufficient authority, the institutional arrangements have badly affected project implementation. 2.8 By the fifth year, the project was expected to have a local technical staff of 222, 144 (65%) of whom would be village extension workers. There would be six expatriate specialists. By the end of the project, it was to be operating in all five administrative districts of the Dosso Province, making gradual progress in each of the three ecological zones (north, center and east) described in the appraisal report.  -6 - III. PROJECT IMPLEMENTATION 3.1 The IDA credit was approved by the Board on December 20, 1979, the Credit Agreement was signed on June 13, 1980, and eight months later, on February 25, 1981, the credit became effective. Effectiveness was conditional on the establishment of three separate revolving funds to prefinance project activities as follows: (a) a Farm Supplies Fund with CNCA for prefinancing procurement of inputs and farm implements; (b) a Special Project Account with the Niger Rural Development Bank (BDRN) tar prefinancing all project operating costs; and (c) an account with the Treasury in the form of a line of credit, to be used for prefinancing all project investments. In the course of negotiations, Government undertook to replenish these accounts at the beginning of each fiscal year. 3.2 The start-up of the project was delayed by the time needed to complete the lengthy legal procedures for meeting the conditions of the Credit Agreement and by the difficulty of rapidly recruiting local and expatriate technical staff. The delays upset the projected timetable. The estimates of operating and infrastructure costs had to be revised upwards. The project actually started In June 1981. Changes after appraisal 3.3 In the course of implementation, changes were made both in the Credit Agreement and in certaeu proposals and objectives outlined in the appraisal report. The Credit Agreement was modified to take account of the financial problems with which Niger was faced from the second year of the project. The disbursement schedule was adjusted: a new category of operating costs was created, and donors financed a larger share of investments, and a revolving fund. Continual delays made it necessary ta review periodically the program of operations and the budget allocations. Local technicians, expatriate specialists and donors had different ideas about the training and extension system, and the proposals put forward by the appraisal mission were modified accordingly. The purpose of the changes was obviously to improve project performance. However, they also contributed to make the task of the Project Manager even more difficult, given his uneasy relations with government technical agencies. Project performance 3.4 As stated in paragraph 3.2 above, the project began operating in 1981. It was closed in June 1985. During the 1983/84 and 1984/85 rainy seasons, project activities were virtually limited to bottomland development and very little was achieved in terms of rainfed crops, which  -7- suffered from inadequate rainfall. Local technicians reappraised the project in 1983, hoping to request an extension of the project's closing date. It was also upon the request of dnors, and particuliarly the CCCE, which had envisaged a two-part financing plan (three years, then two years), that Nigerien authorities were led to prepare a reorientation of the project. Their proposals were made available to the donors in March 1984. They were based upon the recommendations of the Zinder seminar. The seminar was organized in 1982, because Government had become disappointed with large area projects, which were costly and yielded limited tangible results, and was looking for new ideas to make projects more cost-effective. The proposals were very similar to those which had been submitted to the donors at the end of 1983 for a reorientation of the Second Maradi project, and which at that time were still the subject of discussions with the Government. Basically, they intended to substantially increase the role of the rural population in decisions concerning the future of their village. Farmers would be closely associated with the concept and execution of development plans, which were to include an expansion of small rural operations requested by them, and they would be fully responsible for operating and maintaining the investments. The proposals also recommended a village-based extension system, which would have relied upon the group of educated farmers in the village, with advice :being provided by extension officers of a higher level than the usual i extension agents. Finally, various alternatives were to be tested on the method of farmers' training in residential centers (CPR), which had proven to be costly and rather inefficient. Donors reviewed the proposals carefully. The Bank in particular sent a four staff-member mission, which spent more than two weeks in the field. Following these investigations, donors agreed that the proposed measures made sense. They had supported them in Maradi, where at that time a strong management was in charge. They reached a consensus however that the measures implied a totally new approach to the problems of development, which was well beyond the capabilities of the particularly weak Dosso management team and inefficient Government services. They agreed that, in order to put them possibly into effect successfully, a radical change of attitude towards the farmers would be needed, and all project staff together with most Government staff in the project area would have to be replaced by qualified and motivated cadres, with a different mentality. This was unlikely to be accepted by Government, and, in any case, would have been virtually equivalent to initiating a new project. As a consequence, donors did not approve the reorientation proposals and rejected the Government's request to extend the project's closing date. Implementation of the various project components and activities is described briefly in the following paragraphs. 3.5 Administration and management. There have been two Project Managers, neither of whom ever had all the technical staff which had been projected to carry out the planned activities. Local staff, both technical and administrative, were all too frequently transferred to other posts, contrary to the terms of the Credit Agreement, which stipulated that they should remain at post for at least three years. The Project Managers were neither able to coordinate the Government technical services nor able to establish themselves as the authority responsible for all project  - 8 - activities. By the end of the first year, the two sections attached to the Manager's office (financial management, and monitoring and evaluiation) were fully staffed, but this did not last, since both the local technicians and the two expatriates were replaced several times during project implementation. It took the expatriate technical assistants almost 15 months to make effective contact with the provincial technical agencies they were supposed to assist, mainly because of the lack of clarification regarding the respective responsibilities of the Project Manager and the Chiefs of provincial agencies. 3.6 The project suffered financially from Niger's economic difficulties, with the Government no longer able to replenish project accounts and to pay the input subsidies. Moreover, the IDA disbursement procedures were not properly followed, resulting in delays in making funds available. Progress reports were not prepared in time (and indeed were non-existent for the third year), so that donors were unable to monitor operations adequately. 3.7 Recruitment of staff. The project had considerable difficulty in finding local technical staff with sufficient training for such posts as counterparts for the expatriate specialists, agricultural district chiefs, training officers, cooperative warehouse managers, and extension officers. Salaries and benefits for project staff were to be in line with government rates, which were insufficient to attract well-qualified personnel. A more flexible policy on pay scales and benefits would probably have helped to reduce staff turnover and to produce better service. 3.8 Technical assistance. The expatriate staff consisted of the Chief of the commercial, financial and administrative section, a training specialist, an agronomist, a rural engineer, an applied research specialist, and the head of the Monitoring and Evaluation Section (MES). These posts were filled in the first year of the project but replacements were necessary and early in project life several posts remained vacant. Combined with the rapid turnover of local staff, this prevented proper follow-up of operations. 3.9 Monitoring and evaluation. This unit performed well and the number of studies undertaken (use of inputs, feedlots, migration, etc.) is impressive. Unfortunately, bacause of the lack of coordination between the project and provincial agencies, the conclusions and suggestions of these reports had little effect on the conduct of the project. Nonetheless, this section did contribute to a better understanding of farmers' conditions and attitudes, and of farming systems in the region. 3.10 Applied research. This section was very active during the first three years of the project. From the fourth year onward, however, it was virtually paralyzed by the departure of the expatriate specialist. The section worked with INRAN and tested the project's proposals for improving farming techniques. These were presented to the farmers as a package: fertilizer recommendation per crop, frequency of weeding, pure stands and improved seed varieties, seeding methods, etc. The package also included  - 9 - the purchase of animal-drawn farm equipment. As presented, the proposals implied that the validity of the improved techniques depended on the adoption of the package as a whole. Applied researct was conducted mainly in the residential training centers (CPRs), with the farmers who were following the training courses; it was also carried out in a number of villages, sometimes with former trainees. 3.11 Based upon available reports on fertilizer technical and economic justification, and on the variety of agricultural conditions and farmers' attitudes, it appears that: (i) there was no technical package for rainfed crops well suited to the diversity of physical and human conditions in the Dosso area; (ii) the technological recommendations were too rigid, they offered no alternatives and had not been sufficiertly proven; and (iii) the farmers react to other factors (such as land tenure, labor availability, and climatic uncertainties) in deciding whether to adopt one or the other of the technical recommendations. Both in design and implementation of applied research, too much emphasis was given to response to fertilizers, and not enough to comprehension of farming systems. Research was conducted in isolation and the results could not be extended to modify the original technological package, because of both organizational problems and the limited life of the project. 3.12 Seed multiplication and distribution. The seed multiplication program was conducted both at the residential training centers (CPRs) and on the land of volunteering farmers. Tha program covered millet, maize, cowpeas and groundnuts. Production figures in the reports are not vsry precise, and are sometimes contradictory. They are in any case rather irrelevant, due to problems with seed distribution: frequent delays in making seed available to contract farmers caused by poor organization; in farmers' view, excessive cost of selected seed; financial difficulties of CNCA, which was to provide marketing funds to the project and credit to farmers; and increasing disappointment of contract farmers, who were faced with marketing problems of their seed production and received little help to that effect from government agencies. In addition, yields from the so-called "Lproved" seeds were often about the same as or even lower than those from the traditional varieties because selected varieties required better than average ecological conditions, and technical husbandry which disregarded farmers' constraints. As a consequence, it was impossible to achieve one of the project's main objectives, which consisted of providing farmers with large quantities of improved seeds in order to substantially increase average yields. 3.13 Fertilizer distribution. This would have been the project's principal component. Objectives were very ambitious. They aimed at distributing an additional 30,000 tons of fertilizer by the fifth year of the project apart from rock phosphate, with 100% financing from the IDA credit. Fertilizer use in the year before the project was less than 1,000 tons. UNCC would continue to be responsible for distribution. Procurement of fertilizer supplies would be by competitive bidding. Subsidies would be financed by the National Investment Fund (FlI). No seasonal credit was to be provided for fertilizer. In fact, only 1,76G tons in total were  - 10 - distributed through official channels (Table 7). The fertilizer component s>uffered from the following constraints: (a) the proximit:' of Nigeria, which subsdized fArtilizer heavily, and whose currency was weak compaLed to the CFA franc. Border control was inefficient and farmers were able to purchase Nigerian fertilizer at prices 50% below official prices in Niger. This means that the volume of fertilizer actually used was certainly well in excess of that distributed through official channels, but by how much is not known (see paragraph 3.14) 1/; (b) logistical problems: trucks were ill-suited to dirt roads, warehouses were poorly managed, fertilizer was available only at canton level, at cooperative headquarters; and (c) the financial difficulties of the FNI. 3.14 Moreover, the agreements with the Bank were not rebpected insofar as competitive bidding was concerned during the first fiscal year, and there was some confusion between the respective roles of UNCC central supply office and the project. After the second year, fertilizer distribution through official channels fell to a level close to the pre-project period, and financing for this component stopped. It was observed d4ring project implementation 2/ that actual consumption of chemical fertilizer did not increase rapidly, even though it was still the most used input. This would seem to indicate, first, that the technological packages were not adequate (poor response to recommended applications) and that the extension program had little impact. and second, that a number of essential factors had not been taken into account in the project's design, such as the existing arrangements between farmers and herders for using organic manure, and market situation for cash crops, which determines the level of application and the rate ot recovery of fertilizerEi. 3.15 Animal traction. Distribution of draft farm equipment financed by the IDA credit suffered from the same logistical and also financial constraints, which limited the provision of credit and subsidies. The supply of farm implements was handled through LNCC, outside the authority 1/ In Maradi, a survey by the MES of the Maradi project has showed that imports from Nigeria on the parallel market outnumbered fertilizers purchased through official channels in the proportion of four to one. Even assuming the same occurred in Dosso, project targets were not met. 2/ Reappraisal paper, 1983. Periodical reports by the monitoring/evaluation section on farm development.  - 11 - of the project, which therefore could not adjust supply and demand. This should, for example, have been done for carts, which are essential for tran3porting organic manure from cattle yards or manure pits, and which were in much greater demand than any other kind of equipment, such as plows, base frames ("batis de base") or cultivators. Seeders were considered unsuitable by the applied research section but were not modified. Major constraints on the success of the animal traction component were: (a) farmers' obligation to purchase complete sets, which included expensive implements they did not really need, instead of letting them acquire what they wanted specifically, especially carts; (b) difficulties in obtaining supplies from Government imposed manufacturers (CDARMA), due to delayed arrival of raw materials; (c) delays in delivering equipment to the farmers; (d) restrictions on credit; and (e) the limited impact upon other farmers and farming husbandry of former trainees from CPRs, who after returning to their village with draft equipment had often problems re-integrating into the rural cominunity and were not adequately monitored. 3.16 Credit. Agricultural credit remained the responsibility of the CNCA. Credit was to be provided for draft equipment, bottomland development, stall feeding, seeds and seasonal loans. The project was to contribute financing for agricultural equipment and bottomland development. During the first project year the CNCA granted only loans for farm equipment to trainees of CPRs. In the second year, these loans were extended to some other farmers. Subsequently, loans for equipment were virtually stopped for the duration of the project: fewer than 50 loans in total were made for draft equipment. In all, about one thousand loans were made for activities related directly to the project in the whole of the Dosso Province, which has a population of about 700,000 and some 100,000 farm units. 3.17 Access to credit was very limited and strictly regulated. The project as designed was based upon the existing credit system. The CNCA delegated credit management to an intermediary, in this case UNCC. Before granting a loan, UNCC had to go throagh a selection process, involving in sequence the village mutual guarantee group ("Groupement Mutualiste" - GM), the "canton" cooperative, the District Technical Committee (COTEAR), and the Provincial Technical Committee (COTEDEP). Obtaining a loan was conditional upon belonging to a GM and to a cooperative having repaid 95% of previous loans (a joint guarantee was required), upon being sufficiently trusted by the GM, upon being on good terms with the "canton" chief -- who has administrative authority over the cooperative -- and finally, upon having reached a good level of technical expertise, which was supposed to  - 12 - have been acquired through extension, and upon having control over the land. All this was aimed at providing safeguards for the credit. Nevertheless, mainly because of Its operations with parastatal enterprises, but also because cooperatives were mainly top-down imposed fictions the CNCA is virtually bankrupt. 3.18 Training and extension. The system relied upon the UNCC extension agenta, who worked with the cooperatives, and field staff of agricultural services. Initially, extension was to be relayed by "contact farmers", the great majority of whom, it was assumed, would be former trainees from CPRs. These farmers (couples), who were chosen by the villagers, spent nine months at residential centers, where they were trained in improved farming methods, functional literacy, simple management practices, etc. Upon completion of their training, they returned to their villages, taking with them part of the crop, to which they had contributed, and an animal-traction unit bought on credit. Other contact farmers had often had some training at the Young Farmers Training Centers (CFJA). Extension staff and contact farmers together were then to be responsible for mass training of other farmers (usually all members of GMs). CPRs also served for supplementary training courses for the various categories of field officers; the building of a specific training center, considered at appraisal, no longer appeared justified and was abandoned. In summary, major training activities consisted of: (a) training of farmer groups at the CPRs, as described above. In the fourth year of the project, draft equipment could not be provided to the trainees due to CNCA's difficulties; (b) short additional course" for the trainees, on cooperative organization, community development ("animation"), health, etc.; (c) ten-day courses over three crop seasons for training officers and extension workers; (d) short courses on cooperative development at irregular intervals for UNCC officers; and (e) on-the-job training for the expatriate specialists' counterpart staff, which was more or less erratic but more consistent for the surveyors of the MES. 3.19 -Actual training and extension differed quite subs