Madagascar - Agricultural Credit Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7768 PROJECT COMPLETION REPORT MADAGASCAR FIRST AGRICULTURAL CREDIT PROJECT (CREDIT 1064-MAG) MAY 22, 1989 Africa Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Malagasy Francs (FMG) US$1.00 FMG 210 (at appraisal) US$1.00 = FMG 627 (at completion) ABBREVIATIONS AND GLOSSARY BTM National Rural Development Bank (Bankin'ny Tantsaha Mpamokatra) FMR Smallholder lending program (Financement du Monde Rural) FMR-ODRI Individual smallholder lending program (Operation de Developpement Rural Individuel) Fokonolona Institutions: Fokonolona and Fokontany = village level local government Firaisam-pokontany - second tier local government (former canton) Fivondronam-pokonolona - third tier local government (at former sub-prefecture level) Faritany = fourth tier local government (at former province level) FISCAL YEAR January 1 - Decembet 31  FOR OMCL USE ONLY THE WORLD BANK Washington. D.C 20433 U.S.A Oike of Douctv.Cen,ti Opufatmnn[a Eutatim May 22, 1989 HD(ORNDK TO THE CUTIVE DECTORS AND THE PRESIDENT SUBJECTs Project Completion Report on Madagase-ar First Agricultural Credit Project (Credit 1064-MAG) Attached, for information, is a copy of the report entitled 'Project Completion Report on Madagascar First Agricultural Credit Project (Credit 1064-MAG)a prepared by the Africa Regional Office. No further evaluation of this project by the Operations Evaluation Department has been made. Attachment T 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 Wotld Dank authorization.  FOR OMCIAL USE ONLY PROJECT COMPLETION REPORT MADAGASCAR FIRST AGRICULTURAL CREDIT PROJECT (Cr. 1064-MAG) TABLE OF CONTENTS PaRe No. Prefacek... ........ ...............*...*.*.*.*.*.*.*...... . . *.*.*.*.....*.*.. i Basic L'sta .............. ...... ..............................l Evaluation Summary ........ ....... .......... ............. v PROJECT COMPLETION REPORT I . Introduction .. o ..... ............... ...........................I. II. Agriculture and Rural Finance..... ................... 2 III. The Project ............ . ................ 0-000 6 IV. Implementation ....... .......................... 0* ... 9 V. Operating Performance .. - ... .... ...... .... .. 20 'VI. Economic Evaluation....... .. .................. ... 24 VII. Summary Evaluation, Issues and Lessons ................. 25 ANNEXES: 1. Statistiques des octrois des credits agrinDles (clientele ganirale) 1980-87 2. Octrois FMR traditionnel 3. Octrois FMR ODRI 4. Octrois totaux au petit paysannat 5. Analyse du credit FMR par culture 6. Situation de tecouvrement 1983-87 7. Taux d'intre't de credit agricole 8. Utilisation du cridit 9. Analyse des sommes payees a Guy Barbier 10. Fonds de contrepartie 11. Bilans de 1979-87 12. Compte d'exploitation 13. Principaux ratios 14. Personnel de la BTM 15. Agences tests - Taux de recoulrrement 16. Agences tests - Octrois 17. Rapport valeur: Coat du riz 1981-88 Carte: IBRD 19406 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.  4- ~ ~ - PROJECT COMPLETION REPORT MADAGASCAR FIRST AGRICULTURAL CREDIT PROJECT (Cr. 1064-MAG) PREFACE This is the Project Completion Report (PCR) of the First Agri- cultural Credit Project in Madagascar, for which IDA Credit 1064-MAG for US$11.50 million was approved on September 9, 1980. The Credit was fully disbursed when it closed on December 31, 1986 some 2 112 years later than originally scheduled. The PCR was prepared by the World Bank's Resident Representative's office in Madagascar, and is based on data provided by Bankin'ny Tantsaha Mpamokatra (BTM), review of the Staff Appraisal Report (No. 2898-MAG) dated July 1980, the President's Report (No. P-2788-MAG) dated August 20, 1980, the Credit Agreement dated December 23, 1980, and relevant reports, corres- pondence and memoranda in Bank files, as well as interviews with Bank staff. The draft report has been read by the Operations Evaluation Department (OED), and copies were sent to the Borrower and its agencies for comments in February 1989. No comments were received.  - ii - PROJECT COMPLETION REPORT MADAGASCAR FIRST AGRICULTURAL CREDIT PROJECT (Cr. 1064-MAG) Appraisal Key proiect data estimate Actual Total project costs (US$ million) 14.20 13.50 Underrun (X) - 5Z Credit amount (US $ million) 11.50 11.50 Disbursed d do ) - 10.93 Not Economic rate of return 60Z calculated Financial performance Fair Institutional performance Fair Original Other proiect data Plan Revision Actual First mention in files 3110178 Government's application March 1978 Negotiations May 1980 Board approval 919180 Signature date 12/23180 Effectiveness date 8117181 DCA closing date 6/30184 12/31/85 12/31/86 Cumulative disbursements 1981 1982 1983 1984 1985 1986 1987 ($ million at June 30) Appraisal estimate (US$ million) 0.5 2.7 6.7 11.5 Actual disbursement (US$ million) 0.0 0.2 1.0 1.3 3.2 9.2 10.9 Actual as % of estimate 0.0 7.0 15.0 11.0 28.0 80.0 952 Date of final disbursement 7/14/87  - iii - Hission data No. of Staffdavs Specialisations Pertorrance Hission Oate persons in field .epresented rating Trend Problens 4._a o b t_c /_d L_e Identification March 78 1 7 a Prepration Km 79 2 31 a,b wraisul Sept. 79 5 120 a, b, c Sub-total 1S? Supervisian 1ov. 80 2 12 a 2 2 , n IIIar. 81 1 11 a 2 2 f. n IIIlug. 81 1 11 a 2 2 f, n 1UNar. 82 2 22 a 2 2 f. . UNw. 82 1 1H a 2 2 f,ll Ul Rpr. 83 1 7 a 2 2 on, r,tl UIIOct.83 1 7 a 2 2 0h, , n UIII lhr. 8i 1 6 a 2 2 0* f,r n IX NW. u8 1 7 a 2 2 r.0' X thr. 85 2 it a, c 2 n.a. 0. f Xi Oct. S 2 10 a 2 na. 0,1 KIIrelr. 86 2 10 a 2 n.a. 0,n1 NII Oct. 86 2 10 a 2 n.a. 0, 1 Sub-toW tll total 298 Infation on borrwuer Borrzer Republic of IMdagascar Executing agency OtM (Bain'nv Tantsaha patokatra or abtional Rural Deuelopnent Bank) fiscal year of borrwer Janwy I to Oecenbr 31 and executing agency 'Other Problem' refers to the balance of paynents crisis.  - iv - STAFF INPUTS (staff weeks) FY78 FY79 FY68 FY81 FY82 FY8S FY64 FY86 FY66 FY67 FY8S TOTAL Preappralsal 4.e 44.1 2.8 50.4 Appraisel 68.7 .2 68.9 Negotiations 6.0 1.9 7.9 Supervision 6.6 13.3 8.2 9.7 6.4 14.2 6.6 .9 70.1 Othe- .1 .6 .1 .7 Total 4.0 44.2 67.6 10.6 18.3 6.8 9.7 6.4 14.2 6.8 .9 188.0 Exchanze rates at June 30 each Year (US$ FMG) 1979 (Appraisal) 210 1980 205 1981 285 1982 380 1983 420 1984 600 1985 680 1986 (Completion) 627 1987 i370 Follow-up project * - Name Second Agricultural Credit Project Credit number 1804 HAG Amount (US$ million) 10.0 Agreement date 7131187 8a Actual number of days worked on project. b a - financial analyst, b - agriculturalist, c - computer specialist /c I - problem free or minor problems; 2 - moderate problems; 3 - major problems /d 1 - improving; 2 = stationary; 3 - deteriorating (Note: trend ratings were no longer provided, from early 1985) le F - financial; P - political; H = managerial; T - technical; D - Development Impact (new category introduced in 1985). Worst problem first.  !  - v - PROJECT COKPLETION REPORT IADAGASCAR FIRST AGRICULTURAL CREDIT PRJECT (Credit 1064-MAG) I. EVALUATION SUMMARY Introduction The First Agricultural Credit Project was appraised in September 1979. Project cost was $14.20 million, and IDA provided a Credit equi- valent to $11.50 million on September 9, 1980. The project was revised to focus on institutional strengthening and the financing of key imports for the agricultural sector. The Credit closing date was twice extended and the Credit was finally closed in December 1986, instead of June 1984 as originally foreseen, with cancellation of US$0.7 million. Oblectives The main component of its design was funding for an increase in BTM (Agricultural Development Bank) lending to small farmers, largely for export crops but alsc for industrial and food crops. Other components provided consultancy, training and equipment to build up BTM as an insti- tution. IMplementation Experience The Credit became effective in August 1981. Project implemen- tation coincided with a catastrophic period for the Malagasy economy which had repercussions in every sector. Agriculture suffered from a battery of controls at the same time as investment, input supply and producer services collapsed. As a result, supply and demand for credit--which formed the basis of the project--dwindled sharply. The credit was amended three times to allow the use of Credit funds for import of vital inputs, in order to unblock the input supply constraint, but this had little impact on lending levels. The banking sector also suffered a severe financial crisis of low profitability and illiquidity resulting from unwise lending, largely to the public sector, but also reflecting the poor performance of the economy as a whole. The institution building components of the project were expanded to help BTM through this difficult period. Results The project did not achieve its objective of financing extra lending to small farmers (only 2% of IDA funds for the purpose were dis- bursed). The real value of smallholder lending fell during the project period and the number of beneficiaries did not grow. However, the project  -vi - did sustain the smallholder program through a period in which it might otherwise have collapsed and helped BTM to test alternative approaches. The institution building components contributed to BTM's financial and institutional rehabilitation and pointed the way to financial sector reform, as well as leading lito the Second Agricultural Credit Project, with a credit of US$10.0 million signed on July 31, 1987. The import financing (which absorbed most of the original credit funds), paved the way for the Agricultural Sector Adjustment Credit in 1986. Sustainability The sustainability of the institutional developments will clearly depend on future government support for credit operations, together with the future course of the country and economy after a period of severe decline. The bulk of project expenditures, however, went eventually for critical imports to meet immediate needs, where the sustainability concept is of little application. Fndingst At the project level, the main lesson is that the formula for providing financial services to small farmers has not yet been found in Madagascar. The project pointed up some conditions that will be required for success: a conducive policy context, availability of other services, and the need to base interventions on a complete financial circuit, includ- ing savings, not just lending.  PROJECT COMPLETION REPORT. MADAGASCAR FIRST AGRICULTURAL CREDIT PROJECT (Cr. 1064-HAG) I. INTRODUCTION A. The Project As Appraised 1.01 The First Agricultural Credit Project was appraised in September 1979. The main component of its design was funding for an increase in BTM (Agricultural Development Bank) lending to small farmers, largely for export crops but also for industrial and food crops. Other components provided consultancy, training and equipment to build up BTM as an institution. Project cost was $14.20 million, and IDA provided a Credit equivalent to $11.50 million. In practice the main thrust of the project turned out to be institutional strengthening and the financing of key imports for the agricultural sector. B. Implementation and Revision 1.02 The Credit beeame effective in August 1981. The project period coincided with a catastrophic period for the Malagasy economy which had repercussions in every sector of the economy. Agriculture suffered from a battery of controls at the same time as investment, input supply and producer services collapsed. As a result, supply and demand for credit-- which formed the basis of the project--dwindled. The credit was thrice amended to allow the use of Credit funds for import of inputs, in order to unblock the constraint to input supply, but this had little impact on lending levels. 1.03 The banking sector also suffered a severe financial crisis of low profitability and illiquidity resulting from unwise lending, largely to the public sector, but also reflecting the poor performance of the economy as a whole. The institution building components were greatly expanded to help BTM through this difficult period. 1.04 The Credit closing date was tw,:e extended and the Credit was finall- closed in December 1986, instead of June 1984, as originally foreseen. C. Obiectives. Results and Lessons 1.05 The project did not achieve its objective of financing extra lending to small farmers. rhe real value of smallholder lending fell during the project period and the number of beneficiaries did not grow. However, the project did sustain the smallholder program through a period in which it might otherwise have collapsed and helped BTM to test alternative approaches. The institution building components contributed to BTM's financial and institutional rehabilitation and pointed the way to financial sector reform. 1.06 At the project level, the main lesson is that the formula for providing financia'l services to small farmers has not yet been found in M A r VI TS t r n I ir n^A44 q 9th v w411 b i r n, 4r d  for success: a conducive policy context, availability of other services, and the need to base interventions on a complete financial circuit, including savings, not just lending. 1.07 A second project, effective in November 1987, is largely a DFC- type credit for agricultural SMEs. However it provides funds for continuing tests of formulae for rural financial services, and funds for a study. These modest actions may lead to a more 3uccessful d-itign for the future. In the meantime, the report concludes, BTM is providing its most effective aid to agricultural production through its extensive financing of marketing of agricultural produce. II. AGRICULTURE AND RURAL FINANCE A. Importance of Ari'culture and Subsector Description 2.01 Agriculture is Madagascar's dominant economic sector; it employs 85% of the population and accounts for more than 802 of export earnings. Production and farming systems vary widely, with many commodities produced, but production in value terms is dominated by paddy (302) and beef (162). A small group of export crops, (coffee, cloves, and vanilla) accounts for about 15% of production but some three-quaxters of merchandise exports. Industrial crops, notably sugar and cotton, contribute about 72 of agricultural production value. Approximately 802 of total production comes from smallholder activity, and small farmers (in the 1 to 2 ha. range) are the backbone of Nadagascar's agricultural economy. Over half of agricultural production by value is for subsistence and the share of subsistence production has slowly increased over the past decade. In contrast to most sub-Saharan African countries, irrigation is widely practiced in Madagascar, with water control systems in use on about 1 million hectares, one third of cultivated land. Overwhelmingly, irrigated land is used to produce rice, the staple of the Malagasy diet. The country's rich and varied agricultural potential offers excellent prospects for raising incomes, improving the supply of industrial inputs, and for increasing foreign exchange earnings, but this potential is far from being realized. B. Policy and Developments 1980-87 2.02 The history of this project perfectly reflects the difficult history of the Malagasy agricultural spctor during the period when the project was carried out. The Government's policy during the first part of this period, up to 1983, was to nationalize marketing and processing activities; to dictate internal terms of trade via monopoly marketing circuits; to centralize input supply activities; and, contrarily, to  -3- decentralize other producer services - extension, credit - to regional and even local levels through links with the "fokonolonal system. 1 2.03 All this combined with severe balance of payments and fiscal crises, again largely policy-induced, to produce stagnation in the agriculture sector. Paddy production levelled off and imports of rice shot up, edible oils production slumped, coffee export quantitites fell by a third. The agriculture sector (and the economy as a whole) suffered agonizing supply shortages, collapsing marketing circuits and incentives, and a general reversiou to a lower, subsistence level of activity. 2.04 Hard times forced tough decisions and from 1983 the Government began piecemeal to dismantle the dirigiste structure and to rebuild services and incentives. The climb back is proving harder than the decline but by 1985 signs of reviving interest and activity were ev'dent. 1987 saw a bumper paddy crop, wider availability of inputs and a generally free marketing and pr4c*ing system. In general, the outlook for agriculture in 1988 is better than for ten years, and Government is investing in services, notably research and extension, to keep pace with demand. 1/ "fokonolona", traditional Malagasy concept of a local kinship community, was taken as the generic name for the local government reform that resulted from the 1972-5 revolutions. This reform replaced the colonial system of a central power hierarchy descending to village level with an ascending hierachy of institutions beginning at the the village level, each level of which elected the next level, up to provincial level. This system was intended to bring people into responsible political participation at all levels and to be the main engine of rural development. Its ability to harness people's energies for rural development has in practice been sma'll as the institutions have occupied an uneasy status. At times they lack popular support, being seen to represent narrow "political' interests. At other times they show a populist tendency to placate rather than impose. The development of the system in the late 1970s, was accompanied by a relaxation of 'colonial' fiscal and administrative rules - poll tox was abolished, health and education fees reduced, sanctions against bush fires were neglected. The result has been a hybrid structure, largely ineffectual, as it has lacked both political authority, popular support and source of funds.  C. Banking Secte'r The Banking Sector 2.05 In 1977 the Government reorganised the banking sector to create just three state-owned banks, each with a sectoral specialisation (trade, industry and agriculture). The BTM, implementing agency for the project, is the agricultural bank. The sector thus acquired two distinctive characteristics: the banks formed a small cartel that tended to collude on commercial policy and to show little competitive instinct; and the banks, as wholly-owned parastatals, were treated by Govarnment as instruments of policy. 2.06 The result was that the banks were required to participate in Government investment policy, sometimes overtly through share subscription, -more usually covertly through permanent and growing oveFdraft facilities for state enterprises. This process accelerated in the early 1980s when the banks were required to fund the unprofitable operations of the numerous ill-conceived public investments of the 1978-9 period. The result lay for sometime undetected by weak audit and poor accounting standards but emerged as a major problem by 1985. At this time, BTM had 52% of its portfolio classed as doubtful. High interest spreads, required to build up bad debt provisions, passed the cost on to the customer. 2.07 By 1985, banking sector reform had became an issue. International standard audits and provisions policies were introduced. Government and the banks set up processes to reform or liquidate dud parastatal enterprises. Separate boards for each bank were set up and interest rates freed to encourage competition. The BTM 2.08 BTM results from the fusion of two institutions and traditions. It inherited in 1977 the branch network and staff of a regular commercial bank; this was merged with a small, centralized rural development bank. This has given BTM a unique and highly advantageous heritage, at once a bank obliged to respect efficiency and profits in commercial banking and an institution with experience in, and a vocation for, term lending in rural areas. 2.09 However, BTM has never resolved the inherent conflict between these roles, nor properly exploited the advantages of each. As a commercial bank it has never generated the domestic resource base to allow it to expand lending; and in fact the profitability of term lending in rural areas has always been questionable. In practice, the obligation to tie up funds in de facto term loans to parastatals has undermined its portfolio. 2.10 Nonetheless, the BTM has retained a professionalism and a spirit that reflect its heritage and still offer the possibility of exploiting it. Top management is largely from the development bank stream and retains a strong interest in developing this side of the business. The commercial  - 5 - branch network has begun to participate more fully in rural lending, particularly to smallholders. At headquarters a new department was established in early 1988 in an attempt to develop term lending business. D. Rural Finance Rural Cre'it--the Formal System 2.11 BTM's unique qualification for rural finance has been outlined above. It has an extensive branch network in all the major agricultural areas of the country. It has in 1988 a total of 44 branches and 22 field offices in total (41 branches and 26 field offices at appraisal in 1979). It is a generalized bank as well as a specialised agricultural bank, but nonetheless retains a strong resource base and lending program in rural areas. The 1987 accounts show: FMG billion % of total lending Lending to agricultural production 9.8 6 Lending to agricultural marketing 45.3 29 Total lending 159.0 100 (Source: Annex 1) 2.12 Since the early 1970s, BTM (and its predecessor) has operated a smallholder lending program. With the advent of the fokonolona system (see note 1 to paragraph 2.02) this program has been called "Financement du Monde Rural' or FMR. Initially FMR operated only as group credit. This group credit uses the ufokontany', the lowest level in the fokonolona hierarchy, to absorb some of the administration costs of lending to small farmers, and to provide the moral support of the community in making the lending decision and in tecovering the loans. The basic rules are as follows:- o the fokontany authorities and the local extension agent draw up a list of applicants for loans; the BTM officer, the fokontany authorities and the extension agent go through the list and agree on a position on each applicant; BTM then finances the loans, the fokontany authorities help with disbursement and collection; in the event of the repayment rate being below 952, the fokontany is "closed", and is not eligible for finance the next year.  -6- 2.13 These loans are for farmers tiith under 5 ha (about 99% of Malagasy farmers). The interest rate is the best going rate. This has gone from 10% in 1979 to 16% in 1988 (Annex 7). The fokontany receives a small commission for its wo:. 2.14 In 1980, BTM also 'ntroduced an individual smallholder lending program, called FMR-ODRI, designed to cater for medium term lending for equipment, for land development etc. These loans are approved directly by BTM and contracted with individuals. 2.15 Performance of the FMR programs has been mixed and is described in more detail below (para 4.03 to 4.23). It has never touched more than a small fraction of the farm population, the largest program in the last years being that of 1986 when 25,000 farmers (2% of total farm families) received loans. FMR lending has never exceeded 2% of BTH's total lending or 0.5% of total bank credit in Madagascar. The FMR program has never been linked to savings or other rural financial services. 2.16 Total credit to the agricultural sector has ranged between 32% and 39% of BTM's lending since 1980. Rural credit - The Informal System 2.17 No data are available on the prevalence of informal credit but various forms of rural usury are wides)read. Most common are sales of standing crops (an illegal practice) and credit from local stores to be reimbursed by crop deliveries at harvest. No mutual savings associations exist in rural areas, although an attempt is being made by one NGO to develop this. 'II. THE PROJECT A. Identification. Preparation, Appraisal 3.01 The post-1975 revolution Government had a vision of decentralized rural development, mobilising commitment and resources at the field level through the fokonolona structure. Certain hitherto centralized services were to be linked in to this system; rural credit was one, but attempts were also made to decentralize extension, input supply and veterinary services and to make them responsible to the fokonolona strucuture. With this vision Government requested the Bank during the 1977 Annual Meetings to help identify a rural credit project. 3.02 A first identification mission in March 1978 was impressed by the BTM. The mission put its finger on several very important problems that were just germinating--notably some poor quality lending done at Government's behest and with Government's "guara.ntee", and the absence of audit to any acceptable standard. But the smallholder lending program was growing and its performance was good--113,000 loans 1971-6 with an arrears rate of 4%; and BTH looked sound.  3.03 Subsequent prepatration missions developed a project that would channel IDA funds through the BTM to a variety of investments including lending to smallholders for fooderops in selected areas where extension was operating, and for industrial and export crops; lending to agro-Lndustry; and lending for "collective" investments. 3.04 From the beginning, the credit component program was recognized as high risk. During preparation it was noted that the elimination of private trade was threatening marketing services and that lack of incentives was leading to slack credit demand. The final project brief noted the inherent difficulty of supervising group credit and remarked on the "politicisation' of the process via the fokontany. 3.05 By appraisal, which took place in September 1979, the dwindling supply of inputs, the weakness of the extersion service, the diminution of incentives through price controls, together with the inevitable dependence of the smallholder repayment rate on variables like the weather, were all issues. However, the Government and BTM were felt to be essentially pragmatic. The conlusion of the appraisal mission was that the risks weru very real for fooderops and livestock, but that export and industrial crops were largely protected by their own pricing and services regimes. The mission recommended a project that concentrated on lending to smallholders for export and industrial crops and on strategic lending for foodcrops. The idea of lending to agroindustry and collective investments was dropped except for a small experimental credit component. 3.06 The decision meeting endorsed the appraisal mission's judgement and agreed on some safeguards. A general dialogue on agricultural pricing policy was required (a covenant was to provide for this). The problem of risk arising from weather was to be handled by the establishment of a crop insurance fund. However, insufficient consideration was given to the difficulties of implementing this concept in practice.2 3.07 Negotiations focussed on a traditional Bank issue, the interest rate, which at 7X - 102 was about zero or marginally negative in real terms. It was agreed to raise the schedule, with a top rate of 122. 2/ In May 1980 the Minister of Finance prompted the Government to set up a crop insurance fund using the export crop stabilization funds as financial backing. This decision has never been implemented, although the insurance companies introduced a funded scheme in 1987 on a small scale. In view of the nature of farming in Madagascar-- small, largely subsistence agriculture--the costs and risks of this kind of operation are enormous. The economic climate prevailing during project implementation would have made these risks almost impossible to evaluate.  B. Proiect Description and Costs 3.08 The project as presented to the Board was intended to be implemented over three years and comprised: (a) Seasonal credit, for the maintenance of existing plantations of coffee, cloves, and vanilla, along the northeastern and western coasts; for the maintenance of sugar cane plantations, mainly in Mahanjaga faritany; and for the cultivation of foodcrops (paddy, maize, and groundnuts) mainly on the central plateau; (b) Medium-term credit for the regeveration of old coffee plantations, the establishment of new sugar plantations, and the purchase of draft animals; (c) Long-term credit for the establishment of new plantations of coffee, vanilla, pepper and cocoa; (d) Experimental credit (short, medium and long-term) to smallholders or village-level cooperatives for various agricultural activities, e.g., vineyards establishment, steer stall-fattening, and small-scale pig fattening; (e) Staff training for BTM in financial, managerial, and computer disciplines; (f) Logistical support to BTM, notably vehicles for a pilot program of mobile banks; and (g) Consultant services to prepare a feasibility study to introduce a Data Base Management system in BTM. 3.09 Projects costs were estimated at $14.20 million, of which $12.90 million (91%) for loans to smallholders, $630,000 (4%) for the experimental credit to smallholders, and the balance of $670,000 (5%) for consultants, training and the mobile banks. Financing was to be provided by IDA ($11.50 million, 81%) and BTM ($2.70 million, 19%). These resourees were to be provided to BTM as equity and reflows from smallholder repayments would be earmarked for relending to smallholders.  -9- v. Benefits and Risks 3.10 The main beneficiaries of the project were to be 25,000 families producing export crops whose incomes would more than double. 1200 families producing sugar cane and 6000 families producing fooderops would also benefit. The rate of return from these investments was calculated at 60%. The main risks discussed in the appraisal report were, first, the provision of extension and input supply services. This risk was considered beyond project specific resolution but extension for export and industrial crops was adequate, and the BTM's lending to foodcrops was conditional on the adequacy of extension. Second, the marketing and pricing structure was a constraint; as already mentioned, this was to be the subject of a covenanted dialogue. Finally, BTM's performance and its relations with the fokonolona would require monitoring; the institution was, however, financially sound and well managed. IV. IMPLEMENTATION A. Board Approval, Effectiveness, Extensions, Completion, Closure 4.01 The Board approved an IDA credit of SDR 8.7 mlllion on September 9, 1980. The Credit became effective on August 17, 1981. During the period of the Credit three amendments were signed which: (1) reallocated the lion's share of credit funds to finance imports of inputs and tractors, the counterpart passing to the BTM for its original purpose of funding agricultural lending; (2) greatly enlarged the institution building component, mainly for consultancy and computer equipment; and (3) extended the closing date progressively from June 30, 1984 to December 31, 1986. The last disbursement was made in July 1987. Disbursements totalled SDR 8.27 million, 95% of the total. SDR 426,000 were cancelled. B. Actual Implementation v. Proiect Design 4.02 The quite substantial changes to design of this project during implementation are shown in the table overleaf:  HADAGASCAR: FIRST AGRICULTURAL CREDIT PROJECT - PROJECT SUNNARY THE PROJECT AT EFFECTIVENESS IN 1980 ! THE PROJECT At COMPlETION IN 1987 !--------------------------------------------------------------------------- !-----------------------------------------------------------------------! !LOANS FOR ShALLHOLDERS CATEGORY I SDR 7,500,000 !LOANS FOR SNALLHOLDERS CATEGORY I SDR 141,817 !Export crop loans to 25,000 farsers for FnG 678 million (S3.23 million) !Disbursement ade for incremental lending to smaliholders in 1981. In !Sigarcane loans to 1,200 farmers for FRG 253 million (51.20 sillion) !addition the equivalent of FRG 950 million (:SDR 1,470,643) lending to !Seasonal foodcrop loans to 6,000 farmers for FRG 414 million (S1.97 million)!smallholders 1982-6 is eligible under this category and has been !Term foodcrop loans to 6,000 farcers for fnG 860 million ($4.0 million) !financed from the counterpart fund from sale of imports under Categories !V and VI. !------------------------------- --------------------------------------------- !---------------------------- --------------------------------------------! !EXPERIMENTAL CREDIT CATEGORY II SDR 350,000 !PILOT PROJECT CAtEGORY 1I SDR 177,908 !Loans to test viability of lending for vineyards, steer fattening, !Vehicles and operating costs for a test of intensified smallholder !scallholder dairying, pig fattening, and loans to cooperatives. credit sanagement in,6 pilot branches. !---------------------------------------------------------- ------------------! ----- --------------------- ------------------------------------------ ---! !CONSUL7ANCY AND TRAINING CATEGORY III SDR 350,000 !CONSULTANCY AND TRAINING CATEGORY III SDR 1,346,588 '.eieloptent of an in-th;>se training program using 24 months consultants. !Development of an in-house training progras (36 months technical 'Training of coz,uter staff. !assistance). 1Feasibility study for a data-base management system (3 sonths !Developoent of a computer master plan (42 months technical assistance). consultants). !Participation of an international firm in audits of accounts for years !1982-6. !Consultancy from same firm in portfolio analysis, organisation etc. '