Mali - ODIPAC Technical Assistance Project

Document of The World Bank FOR OFmFCIAL USE ONLY Repot No.7565 PROJECT COMPLETION REPORT MALI ODIPAC TECHNICAL ASSISTANCE PROJECT (CREDIT 1174-MLI) December 23, 1988 Sahelian Department, Agriculture Division l This docmmeot bu a ried d dsibutdn and may be ued by reipiens mly 1 the prormance of their officid duaes Is ontets may not othrwse be disclosed without Wodd Bank authorhltion.  ABBREVIATIONS FAC Fonds d'Aide et de CcopAration (the grant mechanism for French bilateral aid) PAO Food and Agriculture Organization FAO/CP Cooperative Program of FAO and World Bank for project preparation IER Institut d'Economie Rurale (the national agricultural research institute) MF Malian franc, the national currency until Hali re-entered the West African Monetary Union in June 1984, and adopted the CFA franc (CFAF) OACV Op6ration arachides et cultures vivri&res (Groundnut and Cereals Agency, the executing agency for Cr. 491-MLI) ODIPAC Operation de developpement intdgrd des productions arachidiere et cer6aliere (Groundnut and Cereals Integrated Development Agency, the successor to OACV and executing agency for Cr. 1174-MLI) PPF Project Preparation Facility SEPAMA A groundnut crushing mill TA Technical assistance  FOR OMCLAL US, ONLY THE WORLD BANK Washnton, D.C. 20433 US.A. OCice of Oiettw -Gmwal December 23, 1988 MENORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report: Mali ODIPAC Technical Assistance Project (Credit 1174-MLI) Attached, for information, is a copy of a report entitled "Project Completion Report: Mali - ODIPAC Technical Assistance Project (Credit 1174- MLI)" prepared by the Bahelian Department, Africa Regional Office. Under the modified system for project performance auditing further evaluation of. this project by the Operations Evaluation Department has not been made. Attachment This document has a restrited dibution and may be used by recipients onbl in the peformance of their offlcial duties Its contents may not otherwis be disclosed without World Bank authorization.  I 0FR OMlCIAL USE ONLY MALI ODIPAC TECHNICAL ASSISTANCE PROJECT PROJECT COMPLETION REPORT TABLE OF CONTENTS Pre!ace ..................................................... i Basic Data Sheet ........ .................................... -i Evaluation Suumary ..... .................................... iv I. INTRODUCTION .............. .............................. 1 Importance of Groundnuts ...............................1 Development of Groundnuts .............................. 1 II. PROJECT ORIGIN AND BACKGROUND ........................... 2 Early Problems ......................................... 2 Early Amendments ....................................... 4 III. PROJECT IMPLEMENTATION ................................. 6 Findings of Early Supervision Missions ................. 6 Credit ................................................. 7 Livestock .............................................. 7 Postponement of Closing Date ........................... 8 Special Efforts to Prenare a Follow-up Project ......... 8 Abandonment of a Follow-up Project ..................... 10 IV. INSTITUTIONAL PERFORMANCE .............................. 10 Overview ............................................... 10 Economic Context ....................................... 10 Constraints ............................................ 11 Borrower and ODIPAC .................................... 12 V. CONCLUSIONS ............................................ 13 ANNEXES: I. Estimation of Project Cost II. Actual and Estimated Disbursements by category III.Yields, Outputs and Input Use IV. Project Area Statistics V. Project Accomplishments as describec by the Borrower This document has a retricted distribution and may be used by rcipients only in the petformance of their offcial duties. Its contents mnay not otherwise be disclosed without World Bankc authoritidon.  - i - MALI ODIPAC TECHNICAL ASSISTANCE PROJECT - CR. 1174-MLI PROJECT COMPLETION REPORT PREFACE This is the Project Completion Report (PCR) of the ODIPAC Technic.al Assistance Project in Mali, for which Credit 1174 in the amount of SDR 5.5 million (US$ 6.5 million) was granted in April 1982 to the Republic of Mali. The Credit was closed on September 30, 1986, one year after the original closing date, and was fully disbursed. The PCR was prepared by the African Regional Office on the basis of the President's Report No. P-2993-MLI dated July 1981, the Credit Agreement with the Republic of Mali and, in particular, the extensive project files, including correspondence with the Borrower. The project officers associated with the project, including agronomists, a financial analyst and economists, were interviewed. Some statistical material was drawn from the Borrower's project reports. Further evaluation by the Opera- tions Evaluation Department has not been made. The draft PCR was sent to the Borrower on October 26, 1988 for comments. No comments were received.  - ii1 - MALI ODIPAC TECHNICAL ASSISTANCE PROJECT (Cr. 1174-MLI) PROJECT COMPLETION REPORT BASIC DATA SHEET Key Proiect Data Actual or Original Current Plan Estimate Project Cost (US$M) 11.4 11.4 (est.) Credit A mount (US$M) 6.5 6.5 Disbursed 6.5 6.5 Cancelled _ _ Repaid Outstanding 6.5 6.5 Date for Completion of Physical Components 3/31/85 2/87 Proportion Completed by Target Date (Z) 100 70 Economic Rate of Return (Z) n.a. n.a. Institutional Performance poor Cumulative Estimated and Actual Disbursements FY82 FY83 FY84 FY85 FY86 FY87 Appraisal Estimate 1.0 2.5 4.5 6.5 - Actual 1.4 2.0 3.1 4.0 5.4 6.5 Actual as I of Est. 104 1/ 80 69 62 83 100 11 Includes repayment of PPF STAFF INPUTS 1988 1981 1982 198a 1984 1986 1986 1987 1988 LEW 4.6 LEA Co 17.1 LOP .1 6.8 .3 LENN 8.2 .8 PAD .6 .8 .2 .7 .2 SPN 18.4 14.5 18.9 13.1 14.1 9.3 .8 PCR 10.8 TOTAL 10.7 27.1 15.1 14.8 18.9 18.3 14.8 9.6 10.8 TOTAL STAFF WEEKS: 185.0  - iii - Other Project Data Original Item Plan Revisions Actttal First Mention in Bank Files 5/76 - 5176 Negotiations 5/81 - 5/81 Board Approval 9/81 - 4/82 Effectiveness - - 7182 Closing Date 9/30/85 - 9/30/86 Borrower Republic of Mali Executing Agency Office de Developpement Integre des Productions Arachidiares et C4realibres Fiscal Year of Borrower October 1 - September 30 Follow-up Project none Mission Data Month/ No. of No. of Man Date of Item Year Weeks Persons Weeks Discipline Report Preparation 1/ Appraisal 2/ Supervision 1 2/82 1 3 4 ECON(2), AGRO 03/18/82 Supervision 2 9/82 1 3 4 ECON, FA, AGRO 10/28/82 Supervision 3 10/83 1.5 2 3 ECON, AG110 12/30/83 Supervision 4 9/84 2 3 3 ECON, AGRO, FA 12/07/84 Supervision 5 11185 1 2 2 ECON, AGRO 02/11/86 Supervision 6 2/86 .5 1 .5 AGRO 03/25186 1/ Preparation was spread over several years, reflecting changing project objectives. 2/ Original full scale project March 1978, subsequently converted to TA project for which there is a President's Report only, July 1981 Country Exchange Rates Name of Currency Malian Franc, then CPA franc 3/ Appraisal Year Average (1981) US$1 = MP.00 (or CFAF 250) Intervening Years Average (82-86) US$1 = CFAF 340 Completion Year Average (1986) US$1 = CFAF 345 3/ The CFAF replaced the Malian franc as the currency of the country in June, 1984 when Mali re-entered the West African Monetary Union.  - iv - MALI ODIPAC TECHNICAL ASSISTANCE PROJECT (Cr. 1174-MLI) EVALUATION SUMMARY The ODIPAC Technical Assistance Project was partially successful in that it achieved some of its infrastructural and educational targets, but not the main objectives. The latter comprised the identification through agricultural trials and research of modified crops and new varieties capable of large scale replication and cultivation, and based on this, the preparation of a follow-up project to boost agricultural development and production substantially in the project area. The plan was that this should, to a large extent, compensate for the loss experienced by area farmers when the very important cash crop, groundnuts, could no longer be profitably exported and production fell to a level to supply the domestic market. The infrastructural and educational components of the project included assistance to build feeder roads, which was achieved to some extent; provide seeds, which was partially accomplished; improve extension services, which were reformed but were only partially effective; implement a functional litetacy program, which was quite successful as was a livestock health component; and introduce an adequate accounting system in ODIPAC, the agriculture agency established to execute the project and develop the project area, which was a success. ODIPAC itself experienced many problems and functioned with limited efficiency; primarily dependent on donor-financed activity, ODIPAC is unlikely to survive the cessation of donor support in 1987. The project started in 1976 as a full-scale project, which focused on groununut production. The deterioration of groundnut production and its financial and economic viability led to the emergence of a technical assistance project in 1982, whose main objective was to identify and prepare a follow-up project to be based on farming activities other than groundnuts. Neither the Government nor the Bank was able to ensure that project staff were concentrating their efforts on their principal objective - the identification of viable farming activities other than groundnuts - instead of directing their energies to keeping the institutional ship afloat. As a result, applied research was inadequate and there was no basis for a follow-up project. When the credit was closed, the Bank provided a PPF to finance further work on the preparation of a follow-up project, but soon concluded that the costs of the continued effo:t were too high. The Government was advised of this in November 1987, when the Bank proposed that it would instead provide financial assistance for pilot operations in improved extension services concentrating on low-cost and low-risk technical improvements nationwide. the following points, highlighted in the PCR, may be of interest. the decline of groundnut production (para 1.06); groundnut debt (para 2.02); the instability of the groundnut market (para 2.06); farmer credit (para 2.09); agricultural research (para 3.04); extension services (para 3.05); livestock health (para 3.09); and environmental degradation (para 3.15).  - v Lessons might still be learned from the following issues: (a) Pitfalls of converting a "full-scale" to a TA project. This was a gradual process over a long period with the Bank's role being essentially one of accommodating changing circumstances. As a result, the TA project was assembled rather than formulated (para 4.06); (b) Lack of a sector context: generally, crop development and related research endeavors to identify modified crops and new varieties should be initially considered within the framework of a country or a sector-wide agricultural context. This was not the case here (para 4.05); (c) Early termination of a project: on a number of occasiots the early termination of the project would have been wise. This did not happen, however. for reasons which were not entirely clear (para 4.09); and (d) Risks h. establishing new agencies: because of the original agency's Lzbt, a new agency, ODIPAC, was established to take over the forv-',; groundnut activities, among others, in the project area, leaving the old agency with the debt. In fact the new agency's performuace was probably not much better than that of the old agency, and in the end its continued existence appeared unlikely, lacking donor funding. The Government is left with the political problem of what to do with the considerable number of ODIPAC personnel (para 4.01).  -1 MALI ODIPAC TECHNICAL ASSISTANCE PROJECT CR. 1174-MLI PROJECT COMPLETION REPORT I. INTRODUCTION 1.01 The project was implemented under the eight.a IDA credit for the development of the agricultural sector in Mali and the second in the western and central area. The project area, in southwest Mali, is almost 700 km from west to east with an annual rainfall between 700 mm and 1000 mm, and is very diverse, ethnically and physically. Its agriculture is considered, by Sahelian standards, to be quite good, that is, within the limits of sparse and erratic rainfall di^tating risk-adverse cultivation practices. Agriculture is the most important activity in Mali and some 85% of the population is engaged in it. Importance of Groundnuts 1.02 The project area has a population exceeding half a million and was not nearly as well developed in terms of infrastructure and social services as other parts of the country. In particular, its lack of roads has proved to be an overriding restriction. Groundnuts were both the main crop and the only cash crop. It was the main source of income for farmers, the largest source of industrial employment in the area, and the main export commodity and largest source of foreign exchange next to cotton. Sporadic efforts had been made to better the lot of farmers, but the area received limited priority compared with other areas more suited to development. The project, derived from an earlier groundnut project sponsored by the Bank, intended to assist in redressing the situation. Development of the Groundnut 1.03 The history of the project is unusual. It spans many years, going back to the successful initiation in the sixties of groundnuts as an important crop in the Western and Central provinces, under a FAC-financed project. This led to the Bank Group's Integrated Rural Development project, Cr. 491-MLI, $ 8.0 million, of June 1974, to expand activity to the eastern part of the area and improve cereal farming. It included the improvement of extension services and the provision of inputs and marketing services in the old project areas, and the extended project areas, under the direction of OACV (Operations Arachidieres et Cultures Vivrieres), a branch of the Ministry of Rural Development. 1.04 Groundnut production reached its peak during Cr. 491's implemen- tation (1973-78), when, in the mid-seventies, it attained 150,000 tons, 100,000 tons as shelled groundnuts and the remainder as groundnut oil, pro- cessed at the two large privately-owned mills located in Bamako and Kica.  -2- Yields of as much as 1,000-1,200 kg/ha were obtained in the latter area for several years through the use of improved seeds, sound cultivation practices and the application of limited quantities of fertilizer, despite the unreliable rainfall pattern. However, groundnut production had re- lapsed to the pre-project level by the last project year. Although cereal production appeared to have increased, tnere was no pre-project data base against which to compare. In any case, there had been little maize production before. 1.05 The decline of groundnut production was primarily due, at least initially, to inadequate producer prices. Set by the Government, they were often half or less than the market price, leading to the diversion of groundnut production to unofficial channels, but with h'gh transaction costs; crop output declined. OACV's weak performance in the sale of inputs and their distribution, quality of seed, and incompetence of extension staff, exacerbated the problem. 1.06 Between 1975 and 1980 there was a considerable decline in the world groundnut supply due to competition from other oils. Whereas Malian groundnuts had been, within limits, economically profitable despite their high inland transportation costs (1,000 km), by the early 1980s this was no longer so. The export-parity price did not provide an adequate price incentive to farmers. In addition, OACV had to bear high and exaggerated administrative costs while world groundnut prices became increasingly low. Indeed, even in Senegal, with much larger annual production and smaller internal transportation costs, groundnuts had become only marginally economical. II. PROJECT ORIGIN AND BACKGROUND 2.01 In light of the apparent groundnut success early in the history of Cr. 491, a follow-up OACV project was initiated. By 1978 it had been appraised but was already in the process of being reformulated, with an objective of 193,000 tons of groundnut production by year 5 (year 0, 130,000 tons) and, for millet/sorghum 300,000 tons by year 5 as against 240,000 tons (year 0). In fact, the project became the ODIPAC Technical Assistance Project, Cr. 1174-MLI, and was signed in mid-1982, some years later. Project activities during this four-year period mirror the problems of the sub-sector and foreshadow the termination of the entire initiative in 1987, almost 12 years after its commencement. Early Problems 2.02 Even at the pre-appraisal stage in 1977, the appraisal team had been reluctant to proceed in the face of seemingly intractable problems: the decline in the groundnut crop and the untenable financial position of OACV. The Government provided a "bareme" to OACV each year, its main income. Government set the producer price for groundnuts while OACV contracted to buy and market the farmers' groundnut production. OACV had already accumulated a very sizeable debt from groundnut activities, much of it the result of the Government's burdening it with various levies and  taxes unrelated to OACV's groundnut operations, as OACV was one of the few sources of revenue in the country and for the Government. 2.03 It was clear all along that if the project were to be pursued, OACV would have to be relieved of its large debt by the Government. No funds under the proposed project were available to assist. Several plans to bail out OACV were considered, but eventually, the Bank had to accept the fact that Mali simply did not have the resources and no other donor would assist. The Bank nonetheless decided to go ahead on the grounds that the project was (now) needed to restore groundnut production to earlier levels, i.e., those reached in the mid-70s under Cr. 491. It proposed that the financial problem be solved by isolating the OACV debt through the creation of a new agency to assume OACV activities, but not its debt, which was left to future government resolution. 2.04 However, further deterioration of OACV's position preempted the Credit's smooth pronessing. When Bank management visited Mali in December 1978, it saw that some other course of action was necessary. This led to a proposal ir, April 1979 to mount a technical assistance project which would, in principle, promote the production of all feasible crops in the project area (not just groundnuts); reorganize extension services on the basis of the agricultural characteristics of each zone, study long-term groundnut prospects; and identify and prepare a follow-up project. A new Government institution, ODIPAC (Office de Developpement Integre des Productions Arachidieres et Cerealieres) would be responsible for the purchasing/marketing functions, including groundnuts, and the execution of the project. By June 1980, however, the Bank had to acknowledge that groundnuts were no longer an economic crop in Mali and that its future prospects were poor. 2.05 Finally, Credit No. 1174-MLI, SDR 5.5 million (US$ 6.5 million), granted under the ODIPAC Technical Assistance Project, was negotiated in May 1981, with the somewhat modified objectives of testing alternatives to groundnuts while improving the efficiency of groundnut production, and focussing on agricultural diversification and the exploration of domestic potential for groundnut products. In the process it would help ODIPAC make the transition from an export-oriented groundnut organization to an agri- cultural development agency emphasizing food crops. The testing of alter- native crops was to lead to the preparation of a follow-up project. ODIPAC would rely on an annual barAme, but the credit would provide for the bulk of ODIPAC's incremental operating costs. 2.06 There was a long delay in the signing of the Credit because of the uncertainty surrounding the availability of IDA funds. Concurrently there was another unfavorable development. The two major groundnut proces- sing mills found that they could not obtain traditional credit (in effect, working capital) from the Central Bank to enable them to purchase the groundnut crop because of the mounting liabilities of the industry. The usual scenario was for the mills to borrow on a short-term basis from the Central Bank, advance the borrowed funds to OACV (now ODIPAC) who would buy the farmers' groundnut crop and market it, and then repay the loan to the  -4 mills with the proceeds, which the latter used to settle the loan from Central Bank. In December 1981 ODIPAC, lacking credit funds from the mills and the Government (whose provision of a first capital tranche to ODIPAC was tied to Credit signing), expressed its urgent need for funds to purchase groundnut seed by January 1982. A snowball effect ensued: by mid- January ODIPAC had no funds to purchase the 7000 tons which the Government had authorized it to buy in the meantime. The Resident Mission, fully involved in the efforts to resolve the situation, emphasized that if farmers were unable to obtain a fair price for their groundnuts (or otherwise market them), they could not repay the 700 million MF debt incurred the previous year, and, presumably, would not grow groundnuts as a cash crop the following year. 2.07 During this turbulent period there was much confusion within the Government; there were conflicting actions of different bodies and thus no consensus. As a result, it was difficult to get information about what was going on. 2.08 Soon after, it was learned that the Government would not provide the current bareme due ODIPAC, but was leaning towards the conclusion that there was no future for groundnuts under the worsening circumstances, and was anxious for the project to pursue diversification. At this point the Bank, after some delay, advised the Government of the release of the funds held in the PPF (Project Preparation Fund) provided in anticipation of Credit 1174 for project start-up costs (but long held in abeyance) for immediate use. In January 1982, a Bank mission recommended the cancella- tion of the Credit if substantial amounts of groundnuts were not purchased by end-March when the crop would rot. It also recommended that if the crop were marketed, the Credit would continue, but the terms would be amended so that ODIPAC would no longer be responsible for purchasing and marketing groundnuts. It proposed that the precise form of amendment be considered during the next supervision mission. In any event, a groundnut crop of 14,000 tons was marketed in March-May. As a consequence, groundnut output in 1982/83 was down by only a relatively modest 15,000 tons. Early Amendments 2.09 Credit 1174 was finally signed on April 30, 1982, when the avail- ability of IDA funds was assured, the Government providing ODIPAC a sum representing one-third of its capital contribution of MF 300 million. The Credit became effective on July 7, 1982. During the supervision mission of October, 1982, the Bank agreed wfth the Government on amendments to the Credit Agreement which would: (a) remove the purchasing/marketing function of crops from ODIPAC; (b) modify the credit terms: farmers would have access to inputs for cash only, but village groups with 50% cash down and 50% on credit;  - 5 - (c) provide funds for the purchase of other materials, including building and educational materials, as well as for agricultural inputs and rural works, including equipment for the construction of feeder roads; (d) provide more funds for consultants, reflecting a greater focus on extension and regional improvement; and (e) transfer the responsibility for the functional literacy program in the project area from the traditional agency to ODIPAC. The Government set out to provide ODIPAC, as part of its per- manent resources, MF 100 million upon signature of the Credit Agreement, MF 100 million by June 30, 1982 or six months after the Effective Date, and MF 100 million by December 31, 1982, one year after the Effective Date. 2.10 The project continued to have the follow'.ng broad objectives: the organization and strengthening of ODIPAC; seed multiplication; the establishment of a program for applied agricultural research; the promotion of Village Groups; the study of oilseed marketing policy; and, above all, the realization of a feasibility study to prepare a follow-up agricultural project. The Government would no longer provide a bareme for ODIPAC. The research program itself was not stated in detail. The program was instead to be identified and directed by the agricultural experts provided under the project (para 3.02). 2.11 The project cost (including US$ 0.9 million for the PPF advance) was estimated at US$ 11.4 million (Annex 1). The project period was set back because of the delay in credit signing to April 1982. The Fonds d'Aide et de Cooperation (FAC) co-financed the project and the Government financed 50% of local costs, as follows: Local Foreign Total US$ million IDA 2.0 4.5 6.5 FAC 0.4 2.1 2.5 Government 2.4 - 2.4 4.8 6.6 11.4 2.12 ODIPAC's existence was virtually dependent on donor financing under the new arrangements. This was the result of the fifth major change in the project since its initiation, starting with the main emphasis on groundnut, and ending without a groundnut component at all. Essentially, the Bank's role had been one of accommodating to changing circumstances in response to the financial, economic and/or organizational problems of groundnuts, ODIPAG, Gove.nment and oth 2.13 The Credit was fully disbursed. Annex 2 shows the proposed allocation of the credit proceeds (after amendment in March 1983) and the amounts actually disbursed in each category.  -6- III. PROJECT IMPLEMENTATION Findings of Early Supervision Missions 3.01 The supervision mission of September 1982 reported that project management had improved considerably with the recruitment of a new ex- patriate Senior Technical Advisor. Project liquidity was under close control, and ODIPAC field staff was substantially reduced, reflecting the shedding of responsibility for marketing activities. The withdrawal of credit and the increased use of cash for farm input purchases had brought some order to the chaotic farm credit accounts during the 1981-82 season, but input use had declined as a result. 3.02 The October 1983 supervision mission which took place at the end of the first project year, recorded that trial research had been initiated and that the fertilizer effect appeared good, but that there was only a notional improvement in extension services. There was a delay in the functional literacy program and the groundnut marketing study. The follow-up project had not been identified and ODIPAC was concerned that its continued existence could be threatened in view of this. There had been a resurgence of the groundnut crop to nearly 22,000 tons in 1982/3 (Annex 3), purchased by private traders at up to 120 HF/kg-- a good price --resulting in some restoration of farmer confidence in this crop. However, as credit had not been forthcoming, limiting the production of crops, particularly cereals, the mission (and ODIPAC) now urged the relaxation of credit regulations. With respect to research, while trials had been initiated, there appeared to be limited action to initiate a comprehensive program addressed to developing crops suitable for large scale growth or to suggest that a coherent research program compatible with project objectives was in the making. The research/testing program had been placed in the hands of ODIPAC and the IER (Institut d'Economie Rurale), which, with the support of French institutions, was responsible for the coordination of agricultural research in Mali. To assist project implementation, including agricultural research, the project provided for six experts, three under the credit and three by FAC, the latter taking the research lead. 3.03 In late 1983 Mali was stricken by drought. Crop output was much reduced (and livestock depleted) as indicated in Arnex III. Another mission at this time found ODIPAC's management very weak. While the functional literacy program had started, training of its literacy extension agents was progressing poorly. 3.04 The fourth supervision, a composite of several missions devoted to ODIPAC and other projects which took place in April 1984 (short mis- sion), July (livestock) and August (general), was discussed in a single, full supervision report distributed in December 1984. The Bank letter making recommendations to ODIPAC was forwarded in August, 1984. Since the appointment of a new ODIPAC director in March its administration had improved. While the agricultural research trials were affected by the drought, some favorable results had been obtained. Below the 800 mm isohyet, local varieties of groundnuts out-performed those imported, but  -7- there were encouraging possibilities for higher-yielding varieties from Burkina Faso and some evidence regarding optimal dates for sowing groundnuts in the two zones was obtained. However, fertilizer trials on maize varieties and combinations of maize with millet or cowpeas suffered from the drought and/or inadequate supervision, response levels of groundnuts to rock-phosphate applications were encouraging but not necessarily economic, given the current low price of groundnuts (30 CFAF/kg), OJ and cereal response to this fertilizer trial was not favorable (Annex V). 3.05 The need to further reorganize the extension service was recognized, since the pyramidal system whereby the top man trains his three assistants, these assistants in turn train those at the next level, etc., had proved ineffective. An FAO/CP mission report in June 1986, observed that with this system the long chain of command and the low level of expertise in some extension workers led to extension messages being passed on uncritically rather than interpreted in light of farmers' needs. An effort was made to revitalize the functional literacy program. The completed groundnut market study was disappointing, but apparently con- firmed the weak market prospects for groundnuts. The mission emphasized the need for an agricultural resourcs inventory to provide a better information base tor the follow-up project. Credit 3.06 ODIPAC again requested the relaxation of the credit conditions. The mission agreed, up to a point, and this led to some easing in September 1984. Whereas previously village groups could borrow up to 50% of the cost of agricultural inputs and equipment, they could now borrow 100%, still subject to constraints relating to outstanding debt obligations prior to ODIPAC's existence. But individual faimers still had to pay cash. Livestock 3.07 A report by a livestock mission in 1984 led to the establishme2t of a livestock component in the Kita region, which covers some 35,000 km , under Cr. 1174, to provide funds for one year for vaccination campaigns, parasitic control, pharmacies, fodder reserves and livestock infrastructure at a cost of about US$ 80,000. It proved successful and was continued in Kita throughout the duration of the project. The rural roads component of the project was not proving successful at this (late) stage due to poor management and lack of equipment, much of which had not yet been procured due to the management changeover at ODIPAC. 1/ The Malian Franc was replaced by the CFAF as the Malian unit of currency in June 1984.  -8- Postponement of Closing Date 3.08 The Bank's July 1985 letter to ODIPAC noting the follow-up project's delay, proposed the extension of the credit's closing date by one year to September 30, 1986, the funds remaining in the Credit being adequate. The closing date was subsequently extended. Regarding the follow-up project, the letter proposed a content similar to that of Cr. 1174, and stressed that ODIPAC should commence the identification process and formulate the follow-up project as soon as possible. The Bank's letter further urged ODIPAC to continue to abstain from marketing operations, as per the Credit Agreement and to refrain from groundnuts on behalf of SEPAMA, the largest processing mill, even with funds provided in advance by the latter. The Bank's concern was that there might be a reversion to the financial situation of ODIPAC's predecessor, which had accumulated such large debt. Eventually, the Bank advised the Government to restrain from any such activity, as it would raise questions about the structural development loan which was under consideration for Mali. 3.09 Another project supervision mission in November 1985 reported a relative resurgence of credit activities (Anuex 3), including a good recovery rate (in fact, about 45% ot the due amount of loans made in 1985 had been repaid as of April 1987). It reported that alternative trials of groundnuts and millet were disappointing, but that movement towards maize cultivation was encouraging: there had been a steady increase in its production which was attributed to the rising price levels of maize. The February 1986 supervision mission reported that the 1985/86 crop had been very good, credit was continuing its recovery, and the livestock program was to be extended to the other livestock zones as a result of its success in Kita (Annex 5). Special Efforts to Prepare a Follow-up Project 3.10 FAC and the Bank had been collaborating on the project and in September 1985, FAC financed a preliminary mission to identify a follow-up project focussing on the potential productivity of the zone and transport as primary aspects. An FAO/CP mission in April 1986 was to have built on the FAC report when it identified a follow-up project, but did not. The project that FAO/CP formulated would benefit 19,000 families, half those in the project area, by providing animal traction (which FAO considered necessary where rainfall was low and uncertain) and other agricultural inputs; restructure the extension system (discontinue the pyramid organization); further promote village organization which contained the main target population; reorganize adaptive research; and investigate rice and vegetable production. The FAO/CP-proposed project was estimated to provide an increase in production: 26,000 tons of millet/sorghum; 4,500 tons of maize; 1,400 tons of rice; 4,700 tons of niebe and a substantial tonnage of groundnuts. The FAO/CP itself was obviously reserved about the project, given the necessity to obtain a proper crop mix with the rotational constraints, and it was for this reason that it limited the number of beneficiaries. The Bank was not in favor of the FAO/CP report  - 9 - and did not pursue it. The Bank questioned the report's failure to quantify statistically the area's climate, population and extent of the potential surfaces of cultivated and uncultivated land (in fact, the follow-up project study under Cr. 1174 had been intended to obtain this information.) The Bank was also critical of the increased use of animal traction and questioned FAO/CP's farm budget. However, the report indirectly highlighted the limitations of the development of more intensive agricultural activities in the project area. 3.11 It was apparent that a follow-up project could not be ready for appraisal before the closure of Cr. 1174 in September 1986. Consequently, a PPF for US$ 750,000, in anticipation of the proposed project, to continue and finalize project preparation during the subsequent year, was initiated in October 1986. Equally important, this facility enabled ODIPAC to continue its functions for another year, when it was hoped there would be a credit for a follow-up project. The PPF was approved on March 3, 1987. Credit 1174 was closed on September 30, 1986, but ODIPAC managed to continue while the PPF funds were available. 3.12 A project brief for the follow-up project, dated March 19, 1987, stated its objectives as being the intensification of the production systems and the integration of livestock with these systems; the prevention of erosion and the protection of the ecosystem; and the strengthening of farmers' associations. The brief was vague and did not provide a concrete basis for the project. 3.13 In February 1987, another project preparation mission was carried out for the follow-up project. This mission found a discrepancy between recent reports, which claimed there was largely unexploited potential for agricultural development in the project area, and reality: anarchic exploitation of natural resources was very advanced due to recent soil erosion induced by land reclamation activity in the marginal zones. According to the report, there had been over-cultivation of the traditionally cultivated areas with no replenishment of soil nutrients, leading to soil fatigue, and causing erosion and loss of vegetation. Habitual bushfires worsened the situation. The mission emphasized that it was essential to do an inventory of agricultural potential as soon as possible. 3.14 Above all, the mission advised that the technical message should not focus so much on production, but on the system of production and the maintenance of its equilibrium. Priority should be given to farm livestock, organic fertilizers, and water and soil conservation methods. A full examination of ODIPAC's technical department and a complete reorganization would be necessary to realize such objectives. (However, others who had worked on the project confirmed their opinion that the project area is the only one left in Mali with significant potential for agricultural development.)  - 10 - Abandonment of a Follow-up Project 3.15 There had been increasing doubt that a follow-up project would be identified; even the observations made in the February 1987 supervision report were not incompatible with this view. During 1987, however, no new project pLcoposal was developed. In a letter of November 11, 1987, the Bank advised the Government that much more field work would be necessary to move the project forward, despite all the efforts that had already been expended. The Bank stated that the remaining PPF resources would not suffice, and that it had no other resources beyond its own staff, which were severely limited. The Bank suggested that, if Governitent wished to pursue the apparently very elusive goal of project preparation, Government should seek grant funding from an agency specialized in pre-investment. In the same letter, the Bank also suggested that both parties would benefit from working together to improve nationwide agricultural services, in contrast to designing area development projects. IV. INSTITUTIONAL PERFORMANCE Overview 4.01 The project was completed, credit fully expended, covenants generally met, various components carried out in some degree, and nearly all the facets of ODIPAC, and the agricultural zones on which it had an impact, were probably improved as a result. Project accomplishments, as described by ODIPAC, are given in Annex 5. The principal project objective was not met: no follow-up project was identified and the Bank abandoned any further effort to develop an area development project in the project zones. To some extent, agricultural development was left in a state of uncertainty. Donor financing for ODIPAC was curtailed and ODIPAC discontinued its operations, at least on the project scale. Its several hundred employees will probably be absorbed by Government departments or through attrition. Economic Context 4.02 History shows that the project was unfortunate in having been designed at a time when the revival of groundnut production seemed to be a viable option. But in 1982, Government and the Bank agreed that groundnuts could no longer be an export crop, and subsequent events justified this course of action. By the late eighties, the export of groundnuts from inland African countries, including Mali and Niger, had ceased, while Senegal, on the coast, continued exporting them, but with great difficulty. Mali and Niger could not match the competition due to their transportation costs. A substantial groundnut crop is still produced in Mali (and Niger), to meet domestic needs, but remains a mere fraction of former production figures. In reaching the 1982 position, the Bank carried out a rapid study of groundnut costs which showed that by excluding unjustifiable Government-imposed levies unrelated to groundnut activities which OACV/ODIPAC had to sustain (accounting for some 40% of costs), groundnut  - 11 - production in Mali still cost about 20% more than world prices. In economic terms, shadow pricing of labor and other factors reduced this difference, but not by much. Thus the final project design sought alternatives to groundnut-for-export as the principal engine of growth in the project area, but a three-year technical assistance project was hardly an adequate means to solve the problem, even if institutional performance had been exemplary. 4.03 The approach to agricultural credit reflected the Bank's perceptions that farmers should not be granted further credit until previous loans had beer. settled, and that farm credit will only succeed if it is administered by a non-political body. On this basis, and given the decline in groundnuts, the Bank considered that there would be little effective demand for credit. This view was borne out by the events but, eventually, the credit terms were relaxed. By then, other crops were developing well, especially maize, and ODIPAC'S new, competent manager was determined to improve ODIPAC in every sense including its credit administration. Constraints 4.04 The Government did not want roads built to evacuate farm produce, fearing that traffic, particularly trucks carrying groundnut, would be diverted from the railroad which ran through the area. Initially, the full scale project was to go ahead without a supporting road component, but road equipment was ultimately included in the TA project and eventually one feeder road was built. The lack of roads had always been a major obstacle for agricultural development, and the very limited provision for road-building in such a large area belied the seriousness of the project. 4.05 Another constraint for more effective implemertation might have been the lack of a broad agricultural context in which to consider project status and orientate activities. There is little or no reference to economic or sector studies in the project file, although a national program did exist. The follow-up project, agricultural trials and variety tests, and actions proposed thereto, appear never to have been discussed in a country or region-wide framework. 4.06 Past experience has shown that when a TA project is put together from an abortive full scale project, its content may be fragmentary and the objectives may not be entirely coherent. On the one hand, this TA project provided for inputs, operating costs, services and credit, which would improve crops, accounting for some 50% of project cost. Such provision was then standard and suited the original full-scale project with its specific crop objectives. However, the TA project had no specific crop objectives, and incremental crop production was not distinguishable. 4.07 On the other hand, the TA project provided for the preparation of a follow-up project, including trial and research activities, the main objective of the project. However, Bank efforts, limited in any case, were largely focused on the production aspects. Insufficient attention could  - 12 - have been given in the field to the definition, organization and execution of an agricultural research program, which should have activated at the outset of the project. Although there was an annual Bank review of the agricultural research program, it was of limited usefulness, and certainly no substitute for periodic, if not almost continual in-field direction, which the field personnel did not have the resources to provide, but which is necessary in such a program. More broadly, the framework of the project was inappropriate for a research project, and the timeframe was too short. Borrower and ODIPAC 4.08 The prolonged period of project activity together with the drastic change in project conception and composition, make it difficult to discuss institutional performance and development. While the project was moderately successful in achieving infrastructural and educational targets, it was not successful in other respects. ODIPAC performed weakly despite the efforts of management and the results were limited. The extension services were reorganized but they were never quite effective; a portion of the feeder road component was carried out; the functional literacy component was delayed but became quite effective; ODIPAC's insistence on selecting a particular consulting firm (over strong Bank objections) led to a delay of the groundnut market study, and the study itself was not well done. What is more important, however, is that no follow-up project was prepared, although this outcome reflected as much the weakness in research planning as any ODIPAC failings. 4.09 An important question is raised as to why project activity was not terminated earlier. In mid-1977 it was suggested that an interim project be implemented because of Government failure to address basic issues. In December 1977 the appraisal mission (full Project) found that the ongoing project, Credit 491-MLI, had so greatly deteriorated that appraisal was postponed. But, while OACV debt was becoming increasingly massive, the situation was not an isolated one. The public financial situation was in a chronic state of debt resulting, among other things, in the breakdown of the postal checking system, which had been starved of liquid assets for years because of the diversion of its funds by the Treasury to pay for current expenditures. Settling OACV's debt was not, under the circumstances, one of the Government's priorities. This situation persisted and in the ensuing two years, the project might have been terminated with good reason. It was at this time that ODIPAC, emerging from the project was established. In January 1982, the credit was nearly cancelled because of the groundnut marketing cr.sis. 4.10 Based on its predecessor, OACV, the institutional development of ODIPAC was at best fragmentary. More villages were exposed to functional literacy training; extension services may have improved in the end; a proper accounting system was adopted by ODIPAC, and livestock health maintenance was improved. But ODIPAC's continued existence and the sustainability of these initiatives became problematic once the project ended; it is unlikely that ODIPAC will for long survive the termination of donor funding in 1987.  - 13 - V. CONCLUSIONS 5.01 CR. 1174 MLI, ODIPAC Technical Assistance Project, was a partial success, achieving some of its physical and educational goals, but not its main objectives. New crops or varieties amenable to large scale cropping were not identified, and plans for a follow-up project had to be dropped. No further Bank agricultural initiatives were planned for the project area. The Government institution, ODIPAC, established to execute the project, is not likely to survive without donor funding. 5.02 First designed as a full scale project supporting the development of the groundnut industry, the project was beset by major problems. These were, namely the massive debt burden of the groundnut agency, the decline of the groundnut crop and, eventually, the latter's discontinuation as a prime cash crop for farmers and a major foreign exchange earner. The technical assistance project comprised of selected elements of the original project was mounted instead. 5.03 The project was not well conceived. Among other things, insufficient time was provided for implementation. While ODIPAC was assisted by expatriate staff, the latter were too occupied to deal effectively with research activity, and the Bank itself was unable to do so, thus eliminating prospects for a follow-up project. 5.04 ODIPAC's accomplishments with respect to extension services, functional literacy, feeder roads and livestock, were considerable, al- though much delayed. The provision of inputs, equipment and credit appear to have had a favorable impact on crops in the project area, although there were no explicit production objectives. There was a notable shift in crop cultivation, the millet/sorghum and maize crops becoming much larger, mainly reflecting exogenous effects, though the promotion of maize cultivation by ODIPAC was a factor. 5.05 It might have been advantageous to terminate the project earlier. While Nalian participation was weak and the Bank input less than would have been desirable, misfortune was a very important factor, namely, the preci- pitous decline of groundnuts as an economic crop. Moreover, a TA project lacking preparation in its own right, put together from the elements of a full-scale project, had little chance of success in the particular circum- stances.  - 15 - ANNEX I ODIPAC TECHNICAL ASSISTANCE PROTECT CR 1174 MLI Estimated Project Cost / Local Foreign Total US$ Million------- Extension & Cooperative Development 0.9 0.8 1.7 Applied Research & Seed 0.7 0.5 1.2 Functional Literacy 0.4 0.3 0.7 Farm Inputs 0.9 1.3 2.2 Monitoring, Studies, Audits 0.1 0.7 0.8 Institutional Support 0.8 1.6 2.4 Contingencies 1.0 1.3 2.4 TOTAL 4.8 6.5 11.4 Actual Cost: Because of changes in project composition during implementation, detailed cost reconciliation is not appropriate. However, total cost approximates estimated cost. Includes US $0.9 Million PPF advance.  - 16 - ANEX II MALI ODIPAC TECHNICAL ASSISTANCE PROJECT (CR 1174 MLI) ACTUAL AND ESTIMATED DISBURSEMENT BY CATEGORY (Thousands) SDR (Equivalent) ALLOCATED I/ ACTUAL Agriculture Inputs, Other Materials/Supplies 1020 1406 Operating Costs, ODIPAC 850 1558 Wages, ODIPAC 890 1327 Consultants, Experts, Audits 1180 480 Initial Deposit, Special Account 260 260 Project Preparation Advance 760 469 Unallocated 540 5500 5500 1/ As amended, March 1983.  - 17 - ANNEX III ODIPAC TECHNICAL ASSISTANCE PROJECT (CR 1174 MLI) YIELDS. OUTPUTS and INPUT USE 3/ 1980283 1900/84 1884/8S 1986/890 986/87 CORN Area cultivated 1,610 11,788 14,689 16,498 19,048 Yild (ha) 1,851 766 9g0 1,476 1,793 Production (tons) 2,040 9,222 14,642 22,876 34,144 GROUNDNUTS Area cultivated 26,400 82,210 28,698 81,642 09,817 Yield (hb) 880 618 "1 994. 1,020 Production (tons) 21,912 18,608 22,126 81,462 40,103 MILLET/SORCHUM Areo cultivated 184,000 128,587 72,616 78,417 84,767 Yield (ha) 780 418 722 789 893 Production (tons) 104,520 61,041 62,429 61,871 76,e88 FARM INPUTS Fertilizer (tons) 1U4 260 810 752 578 Fungicide (groundnut.) - 14,200 16,421 25,407 42,496 Fungicide (cereals) - 16,0k0 82,489 17,942 89,246 Cultivators 25 a2 16 862 811 Ploughs - 12 14 88 82 Seedern 29 204 76 167 1,106 Ho" 101 40 87 41 18 Cart 79 62 126 223 904 Maize seed 910 1,196 980 1,082 n.a. Groundnut seed 962 1,842 484 1,084 n.a. Millet/Sorghum sd 898 971 402 691 n.a. 3J Eetimates provided by ODIPAC. Production figures are total for project area, not Incremontal.  - 18 - ANNEX IV AGRO-ECONOMIC SURVEY, 1985-86 PROJECT AREA STATISTICS 1985/1986 Season Kayes Kita Kolokani Average Population per Farm Unit 10 14 11 12 (parcel) Active in Farming 6 9 6 7 Area Cultivated (ha) 2.2 6.2 5.4 4.7 Area per (active) Person 0,36 0.69 0.86 0.65 Eguipment Number Units Equipped (%) 10 25 15 17 Number Units Semi-Equipped 31 28 32 30 Number Units Not Equipped 59 47 53 53 % Unit Cultivated under: Corn 1 4 0 2 Groundnut 2 1 3 2 Millet/Sorghum 1 3 1 2 % of Machine Seeded: Maize 14 20 10 15 Groundnut 13 20 19 18 Millet/Sorghum 19 27 10 16 % Receiving Fertilizer Maize 4 10 4 6 Groundnut 3 1 0 1 Millet/Sorghum 1 3 0 1 % Receiving Natural Fertilizer Maize 28 26 57 33 Groundnut 1 0 0 0 Millet/Sorghum 8 2 9 6  - 19 - ANNEX V Page 1 of 2 PROJECT ACCOMPLISHMENTS AS DESCRIBED BY THE BORROWER Trials, Varieties, Research ODIPAC's objectives were to establish a seed program, conduct tests of fertilizer dosage and utilization of rock-phosphates, make variety tests on cereals, groundnuts and vegetables, and carry out modification trials of soya, niebe, sesame and peas. The results demonstrated the following: - the beneficial effects of phosphates as fertilizer; - a natural fertilizer used for growing cotton together with urea seemed to give favorable results for maize, as did a combination of mineral and natural fertilizers; - sesame crops of 400-600 kg/ha could be grown with only 400 mm. to 500 mm. rainfall; and - the varieties introduced did not perform as well as those already cultivated. Seed multiplication plants (2) were not provided until 1985, at which time the production of seeds to be used for the trials was initiated. Previously seed activity was limited, diverse types of seed being obtained from vegetative materials and from the culture of soya and sesame. Agro-Economic Base From 1982 to 1985 ODIPAC conducted an annual survey to provide cumulative information on the characteristics of the project area relevant to agriculture. By the last project year the statistics had become quite comprehensive. Among other things, the 1985-86 statistics indicated the following: (a) A reduction in the hectarage cultivated per active laborer; (b) The growing preponderance of millet/sorghum; (c) The rebirth of groundnut production (but far from the level of a decade earlier); (d) An increase in yield in all crops; (e) That seeding is best done during the last half of June if rain- fall is normal; (f) That 50% of farmers weeded the main crops with manual labor.  - 20 - ANNEX V Page 2 of 2 (g) That maize yields less income for farmers than groundnut; (h) Overall, the amount of cereals ,old is small, while groundnuts are disposed of readily. Farmers find their cereals sell best in the Bamako area and they habitually purchase rice with their income. Functional Literacy In 1985/86 there were over 300 village centers functioning and about 3500 villagers had obtained some measure of literacy. Rural Roads An 80/km road was constructed, and reconnaissance of another 18 km was carried out in 1985-6, with the advice of a highway expert. Livestock The activities related to improving livestock health, increasing the number of cattle, and the recycling of natural fertilizers which began in 1984, were expanded over a larger area, and were successful.