Madagascar - Lac Alaotra Rice Intensification Project

Document of The World Bank FOR OmCLAL USE ONLY Report No. 11459 PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) DECEMBER 22, 1992 Agriculture Operations Division South-Central and Indian Ocean Department 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.  EXCHANGE RATES Currency Unit: Malagasy Franc (FMG) Year Malagasy Francs Malagasy Francs per SDR per US$ 1982 405.6 349.7 1983 515.3 430.4 1984 645.0 576.6 1985 698.4 662.5 1986 941.6 676.3 1987 1,751.0 1,069.2 1988 2,054.1 1,407.1 1989 2,014.0 1,603.4 1990 2,012.0 1,494.0 WEIGHTS AND MEASURES I are (a) = 0.0247 acres i hectare (ha) = 2.47 acres i kilometer (km) = 0.62 miles i square kilometer (km2) = 0.39 square mile i kilogram (kg) = 2.20 pounds I liter (1) = 0.26 US gallon 0.22 Imperial gallon I ton (t) = 2,204 pounds ACRONYMS ASAC Agricultural Sector Adjustment Credit BTM Agricultural Credit Bank CCCE Caisse Centrale de Cooperation Economique CIRIR Rural Irrigation District CIRVA Agricultural Extension District ERR Economic Rate of Return FAC (Fonds d'Aide et de Cooperation) FMG Malagasy Franc FOFIFA Center for Agricultural and Rural Development Research PCR Project Completion Report SOMALAC Development Agency for the Lac Alaotra Region WUA Water Users' Association ADMINISTRATIVE NOMENCLATURE Faritany Administrative equivalent of a province Fivondronona Sub-prefecture Firaisana Canton Fokontani Village level local government FISCAL YEAR January I - December 31  FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluatlon December 22, 1992 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Madagascar Lac Alaotra Rice Intensification Project (Credit 1337-MAG) Attached is a copy of the report entitled "Project Completion Report on Madagascar - Lac Alaotra Rice Intensification Project (Credit 1337-MAG)" prepared by the Africa Regional Office. While implementation took two an a half years longer than planned, the project achieved its physical objectives of rehabilitating modern irrigation schemes and improving water availability on traditional small schemes. Paddy yields and production have increased significantly and the project has been instrumental in getting farmers involved in operating irrigation works by establishing water user associations in the project area. As such, the project is rated as satisfactory. Project sustainability, however, is not assured because of uncertainty regarding maintenance arrangements for the civil works and political resistance to water charges. The issue of which Government agency should be given the responsibility for maintaining the irrigation schemes remains unresolved. Without adequate maintenance, the project may not generate the economic rate of return of 25 percent reestimated in the PCR. The Project Completion Report (PCR) is comprehensive and informative. No comments on Parts I and III were provided by the Borrower but a Project Completion Report written by the Borrower's implementing agency is available in the project files. An audit is planned jointly with two other agriculture investment projects and an agricultural sector adjustment operation (ASAC). Attachment 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.   FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) TABLE OF CONTENTS PREFACE ................................................... i EVALUATION SUMMARY . ........................................iii PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE ..................1 Identity of the Project ............ ............................ 1 Background ............................................... 1 Project Objectives and Description ................................ 2 Project Design and Organization .................................. 2 Implementation ............................................. 3 Results .................................................. 5 Economic Results . .......................................... 6 Project Sustainability . ........................................ 6 IDA Performance . .......................................... 7 Borrower Performance ............ ............................ 7 Consulting Services . .......................................... 7 Project Documentation and Data .................................. 8 PART II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE .... .......... 9 PART III: STATISTICAL INFORMATION .............. ................ 10 Table 1: Related Bank Credits .................................. 10 Table 2: Project Timetable .................................... 11 Table 3: Project Costs and Financing ............................. 12 Table 4: IDA Credit Disbursements .............................. 13 Table 5: Main Results ........... ............................ 15 Table 6: Compliance with Credit Condition ......................... 16 Table 7: Use of Bank Resources ................................ 17 ANNEX 1: ECONOMIC RATE OF RETURN ............................. 19 Map: IBRD 15665R 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.   i PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) PREFACE This is the Project Completion Report (PCR) for the Lac Alaotra Rice Intensification Project in Madagascar, for which Credit 1337-MAG in the amount of US$ 18.0 million equivalent (SDR 16.7 million) was approved on March 22, 1983. The credit was closed on September 30, 1991, two years and three months after the original closing date. The last disbursement was on February 20, 1992, at which time, the undisbursed balance of SDR 0.9 million was canceled. Parallel financing was received from the Fonds d'Aide et de Cooperation (FAC), equivalent to SDR 4.1 million, and from the Caisse Centrale de la Cooperation Economique (CCCE), equivalent to SDR 9.3 million. At the time of IDA credit closing, the actual FAC contribution had increased to SDR 7.9 million; from the CCCE loan, SDR 8.6 million had been used. 1/ The PCR was prepared by the Agriculture Operations Division of the South-Central and Indian Ocean Department, Africa Region (Preface, Evaluation Summary, Parts I and III). Preparation of this report was started during the final Bank supervision mission of the project in March of 1991, and is based, inter alia, on the Staff Appraisal Report, the Credit Agreement, supervision reports, correspondence between the Bank and the Borrower, internal Bank memoranda and a project completion report prepared by the implementing agency. The Borrower was sent a copy of the PCR with a request to prepare Part II. The Borrower did not prepare a Part II but a project completion report prepared by the implementing agency is available in the project files. 1/ At the time of the IDA credit closing on September 30, 1991, FAC and CCCE support was still on-going.   iii PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) EVALUATION SUMMARY Background 1. Agriculture is the most important economic activity in Madagascar, supporting over 80% of the population. It contributed about 40% to GNP and 80% to the export earnings of the economy at the time of project preparation in 1981. At that time, the growth of Madagascar's agricultural production was disappointing. As a result, Madagascar had to import about 175,000 tons of rice (staple food for the Malagasy population) in 1981, which represented an important use of foreign exchange resources. Government priorities were for the country to become self-sufficient in food production, increase the production of cash crops for export, expand basic agro-industries to process local raw materials, and increase rural employment. A previous irrigation project in the Lac Alaotra area (1970-1975) was cofinanced by the Bank (Credit 214-MAG) and other donors. This project was small and although it succeeded in expanding the irrigated area, it failed to strengthen the executing agency (Development agency for the Lac Alaotra region, SOMALAC); project results were therefore not sustainable. Project Description and Objectives 2. Project objectives were in line with the Government's agricultural objectives. The main project objective was to increase rice production and the marketed surplus in the Lac Alaotra basin, thereby saving foreign exchange expenditures on rice imports. Major project components were: a) rehabilitating 25,000 hectares of modern irrigation schemes and improving water availability on 3,000 hectares of traditional small schemes; b) reinforcing SOMALAC, the public enterprise responsible for irrigation in the Lac Alaotra region; c) strengthening the regional extension service (CIRVA) of the Ministry of Agriculture to assist farmers in small schemes; and d) strengthening research (FOFIFA) for seed breeding and multiplication. 3. The project was to be financed by the Caisse Centrale de Cooperation Economique (CCCE; mainly civil works, SDR 9.3m), the French Cooperation (FAC; mainly technical assistance, SDR 4. lm), Government (SDR 5. lm) and IDA (SDR 16.7m). The project was to be managed according to annual work plans, making it flexible in its execution. As part of project preparation, SOMALAC was reorganized in three semi-autonomous units: one for civil works and maintenance, one for production and extension, and one for rice milling and marketing, all under a General Directorate, with a view to progressively privatize some of its functions. Project Implementation 4. The main difficulties encountered during implementation have had to do with the functioning of the project authority, SOMALAC. During implementation, SOMALAC was one of the public enterprises to be restructured as part of the adjustment process. Extensive studies were carried out  iv (one of the main reasons for the delay in the project closing date of two years and three months) which led to (a) the elimination of SOMALAC's paddy purchase and milling monopoly and the opening up of paddy buying to private hullers and traders; (b) the transfer of SOMALAC's extension functions to the Ministry of Agriculture's regular extension services (CIRVA's); and (c) the transfer of SOMALAC's maintenance role to the Ministry of Agriculture's irrigation district (CIRIR). 5. The break-up of SOMALAC's paddy purchasing monopoly - in 1988 - has led to a considerable increase in competition between traders, leading to better farmgate prices and to a strong expansion of private hulling and milling. However, it also weakened SOMALAC's financial performance to the point that Government decided, in February 1990, to do away with SOMALAC and transfer its extension and maintenance functions to other services, as recommended in the studies. Extension activities in the area are presently supported by the Agricultural Extension Pilot Project (Cr. 2150-MAG). Government has yet to reinforce these services to assume their expanded role, as foreseen in the studies, and consequently, it is doubtful whether the project's achievements can be maintained. Proiect Results 6. The project has had considerable success in achieving its physical objectives. Some 31,000 hectares were rehabilitated (28,000 hectares estimated at appraisal) on schedule and without cost overruns (in SDRs; in local currency, the cost was considerably higher due to several devaluations during the implementation period). 7. The effect of rehabilitation has been an increase in paddy yields by just over 50%: from an average of 2.1 tons per hectare on the large schemes at the start of rehabilitation, to an average 3.2 tons per hectare in 1991, and from an average of 1.8 tons per hectare on the small schemes at the start of rehabilitation to an average 2.9 tons per hectare in 1991. This has entailed an increase in total annual paddy production from 57,900 tons in 1981 to 98,500 tons in 1991. 8. Results have also been generally positive in getting the farmers more involved and accountable for operating and maintaining irrigation works in their own areas. Water User Associations (WUAs) have been established covering the entire irrigated area, and WUAs have also increasingly been consulted on the design of rehabilitation works. However, payment of water charges, initially successful, has declined because of active resistance from local political interests. 9. The economic rate of return of this project is largely positive, estimated at 25% compared to 24% estimated at appraisal, but this is based on continued adequate maintenance of the irrigation system which under present conditions is not assured. Without adequate maintenance, production is likely to fall back to pre-project levels, which would imply a marginal rate of return. Sustainabil itv 10. Although the project was largely successful in achieving its objectives, its sustainability is in doubt (as was the case after the first project), largely because the institutional questions have still not been resolved. The main outstanding issue is that after dissolving SOMALAC, Government did not create an institution capable of assuring maintenance of the main civil works (storage dams, main canals and drains). The studies done on restructuring actually proposed one of two options to the Government: to create an autonomous maintenance organization (preferred by the French co- financiers) or to reinforce the existing CIRIR so as to enable it to take responsibility of maintenance  v (preferred by the Bank). No decision on either has yet been taken, which has potentially serious consequences for the security of storage dams and for a rapidly diminishing capacity of the distribution and drainage network (canals and drains need to be regularly cleaned because of on-going siltation due to erosion in watersheds). Unless such an organization is established (or the existing CIRIR reinforced), it will not be possible to sustain the results achieved. Lessons Learned 11. The major lessons to be drawn from this project are: - Flexible project execution through annual work plans, which helped to adjust activities during implementation and provided a mechanism for donor coordination during supervision missions, has been a main factor in the smooth implementation of the irrigation rehabilitation works. The difficult and protracted process of restructuring SOMALAC has been partly due to the fact that, at appraisal, there was no consensus on SOMALAC's ultimate role among the donors and between donors and Government; this situation delayed the decision to reinforce the Government's regular extension and irrigation services and to replace SOMALAC's functions. Pioneering work was done in establishing farmer-led WUAs and credit associations, with largely positive results, which are now being extended elsewhere in Madagascar. The project did include works on both large and small schemes. The results suggest that it is preferable to concentrate on the rehabilitation of small schemes: (i) the cost per hectare of rehabilitation is 56% lower on small schemes than on large ones (approximately SDR 200 per hectare versus SDR 450 per hectare, respectively); (ii) at present, paddy yields are only 10% lower on small schemes than on large ones (2.9 tons per hectare versus 3.2 tons per hectare); and (iii) maintenance on small schemes can be taken care of to a larger degree by farmers themselves, thus improving sustainability.   1 PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE 1. Identity of the Project Project name: Lac Alaotra Rice Intensification Project Credit number: 1337-MAG RVP Unit: Africa Regional Office Country: Madagascar Sector: Agriculture Subsector: Irrigation 2. Background 2.1 At the time of project preparation in 1981, agriculture was the most important economic activity in Madagascar and contributed about 40% of GNP, 80% of export earnings and supported over 80% of the population. However, because of a poor policy framework emphasizing state intervention in the economy at that time, the growth of Madagascar's agricultural production was disappointing. As a result, Madagascar, which was used to export 10,000 to 20,000 tons of rice through 1972, had to import about 175,000 tons of rice (staple food for the Malagasy population) in 1981, which represented a substantial use of foreign exchange resources. The Lac Alaotra basin, with an irrigated surface of 68,000 ha at the time of appraisal, has traditionally been the largest rice producing area in Madagascar. It provides the country with about one-third of total marketed surplus. 2.2 Faced with poor agricultural production performance, mounting food import bills and widespread disorganization in the provision of basic agricultural services, the Government defined new sectoral priorities including: a) self-sufficiency in food supply (particularly rice); b) increased production of cash crops for exports; c) expansion of basic agro-industries to process local raw materials; and d) increased rural employment. 2.3 IDA financed three rice irrigation projects in Madagascar, including one in the Lac Alaotra area. The Lac Alaotra irrigation project (Credit 214-MAG), was closed on June 1975. The Project Completion Report (PCR) was issued in December 1976 and estimated the ERR at 22%, although with a high likelihood of variation (instead of 11 % at appraisal), primarily the result of meeting the production targets combined with sharp increases in the world market price of rice. The project consisted of rehabilitating irrigation and drainage networks serving 4,000 ha, the construction of irrigation and drainage networks on 6,000 ha of peat marshlands and the construction of drainage works on 2,000 ha of peat marshlands. The project did not include any changes in the functioning of the existing project authority, the Development Agency for the Lac Alaotra Region (SOMALAC). It also did not pay any attention to social and land tenure aspects. The PCR criticized the project  2 design for not paying sufficient attention to the social aspects of the project and for not analyzing in sufficient detail the financial and managerial capabilities of SOMALAC. By the end of the project, Madagascar was in political turmoil and SOMALAC was in dire financial straits because payment for SOMALAC's services to farmers was delayed by Government, and farmers were unwilling to contribute to the maintenance of canal works. This situation led to a rapid deterioration of the irrigation works in the ensuing years, and consequently to a fall in rice production, which prompted the Government to request external financing for renewed rehabilitation. 3. Project Objectives and Description 3.1 Objectives. The objective of the project was to increase rice production for the market in the Lac Alaotra basin, thereby saving foreign exchange expenditures on rice imports. This would be mainly achieved through the rehabilitation of the existing irrigation infrastructure, the supply of farm inputs, and management and institutional technical assistance. 3.2 Components. The project, which was to be implemented over a period of five years (1983- 87), had the following components: (a) the rehabilitation of existing modern irrigation networks on about 25,000 ha; and the improvement of water availability on traditional schemes on about 3,000 ha; (b) the reorganization of SOMALAC, the regional development company responsible for agricultural development in the project area, through restructuring, material support, technical assistance and training; (c) the support of the regional extension service (CIRVA) of the Ministry of Agriculture, including the provision of vehicles and equipment, training and construction of a central warehouse; (d) the supply of farm inputs including paddy seed research and multiplication by FOFIFA (research service); (e) the establishment of a network of gauging stations; (f) pilot activities aimed at increasing agricultural production; (g) the establishment of a monitoring and evaluation unit; and (h) studies related to the project and for the preparation of other agricultural projects in Madagascar. 4. Project Design and Organization 4.1 Preparation of the project was first discussed between Government and IDA in 1978 in the context of the Lac Alaotra Master Plan prepared in 1976. Pre-investment studies were financed by the French Fonds d'Aide et de Cooperation and an accounting and management audit of SOMALAC was carried out in 1981. Preparation took into account the experience drawn from the Lac Alaotra Irrigation Project (Credit 214-MAG). Total project costs in the amount of SDR 35.2 million were  3 to be financed by the French Caisse Centrale de Cooperation Economique (CCCE; mainly civil works, SDR 9.3m), FAC (mainly technical assistance, SDR 4.1m), Government (SDR 5.1m) and IDA (SDR 16.7m). 4.2 On the technical side, design of the main irrigation/drainage works was carried out by a consulting firm. Insufficient attention was paid to soil erosion on hilly watersheds even though this problem had already been identified at appraisal. During implementation, erosion of these watersheds created siltation of dams and canals that had been built and increased maintenance costs of the schemes. Modifications had to be made to minimize the effects of the erosion. This necessitated more detailed studies and caused delays during implementation. 4.3 In terms of organization, three agencies were involved: a) SOMALAC, which had been reorganized into three separate entities: one for engineering and maintenance, one for extension, and one for rice milling and marketing; b) the CIRVA, the local extension service of the Ministry of Agriculture responsible for areas outside SOMALAC's perimeter; and c) the Center for Agricultural and Rural Development Research (FOFIFA), for seed breeding and multiplication. It was envisaged that: a) the SOMALAC engineering and maintenance service would supervise execution of the physical works through contractors, but, once rehabilitation was completed, the service would decrease in size and eventually be privatized and Water Users' Associations (WUA) would take over maintenance; and b) the SOMALAC rice marketing service would loose its monopoly and would compete on an equal footing with other firms. This evolution of SOMALAC's organization was thought to have been well understood during appraisal but apparently was never fully accepted by Government, or by other donors. 5. Implementation 5.1 Effectiveness was delayed by about 6 months for lack of Government counterpart funds to recapitalize SOMALAC, which was a condition of effectiveness. Partly due to this delay, to changes in civil works design, to lack of timely counterpart fund availability during project execution (particularly in 1989), and to the studies on the restructuring of SOMALAC, the closing date had to be delayed three times from June 30, 1989, to September 30, 1991. 5.2 The design changes proved to be important as they focussed on erosion and siltation prevention, which have become a growing problem over the last fifteen years, as hillsides surrounding the lake area are increasingly grazed and cultivated. 5.3 Apart from the delayed start-up, no major problems in civil works execution were encountered: procurement, selection of contracting firms and execution have proceeded smoothly under the able supervision of SOMALAC, supported by technical assistance. The area on which rehabilitation works have been completed (31,000 ha) has exceeded the appraisal estimate by about 10%. Moreover, most of the civil works were completed within the time allowed and without exceeding appraisal estimates in SDR terms. 2/ Total project costs expressed in SDRs were 35.7 million (SDR 35.2 million estimated). However, in local currency (FMG), total cost has been about 2/ Increases in the cost of rehabilitating the large schemes as a result of the design changes, were more than offset by decreases in the cost of rehabilitating the small schemes. In SDR terms, the average rehabilitation cost dropped from SDR 475 per ha at appraisal to SDR 427 per ha during implementation.  4 210% higher, which has given rise to counterpart funding difficulties; allocations from the budget have been delayed - particularly in 1988/89 - which has contributed to delays in the final contracts (e.g. land planning). Actual Government financing was 60 percent of the amount estimated at appraisal (SDR 3.1 million actual versus SDR 5.1 million estimated). Financing by IDA was 95% of the amount estimated at appraisal (SDR 15.8 million actual versus SDR 16.7 million estimated); financing for studies and training was higher than estimated, while financing for inputs was less. Financing by the FAC was much higher than estimated (SDR 8.1 million actual versus SDR 4.1 million estimated); financing was increased for technical assistance and for the expansion of the seeds center. Financing by the CCCE was close to the appraisal estimate (SDR 8.7 million actual versus SDR 9.3 million estimated). 5.4 Pioneering work was done in getting the farmer more involved ,n water management and operation/maintenance of the irrigation network through the establishment of water-user associations (WUAs). None existed at the start of the project; at present, the entire area is covered by WUAs. The WUAs were set up by SOMALAC and received support from specialized SOMALAC agents. The WUAs are organizing water distribution effectively, and handling routine maintenance on the tertiary distribution system; they are also collecting water charges to contribute to the cost of maintenance of the main works. Collection of user fees, which was initially satisfactory (80-90%) has worsened to about 46% in 1991, mainly because of the appeal of local politicians, that maintenance has to be provided by the state. 5.5 The major problem during implementation has been institutional. Although during appraisal, SOMALAC was reorganized, the basic framework under which it operated, was left in place: SOMALAC's paddy purchasing, milling and marketing monopoly was maintained; it paid farmers below-market prices for paddy and out of the surplus it retained, it paid for irrigation maintenance and part of its operating costs (including the cost of running its own extension service). When in 1986, Government complied with conditionality under the Agricultural Sector Adjustment Credit (ASAC; Cr. 1691-MAG), and opened up the rice trade to private traders, SOMALAC was forced to pay considerably higher purchase prices for paddy, as it had to compete with other buyers. When the Government subsequently imported rice in order to reduce prices for urban consumers, SOMALAC incurred significant losses on stocks. These losses, combined with the fact that SOMALAC's functions in collection, marketing and milling were successfully taken over by the private sector, finally led to a Government decision, in February 1990, to do away with SOMALAC. 5.6 The decision to eliminate SOMALAC was taken after extensive studies on its restructuring were made. A consensus was reached on the studies' recommendations to sell off SOMALAC's rice mills and to transfer its extension service to the Agricultural Ministry's extension district (CIRVA) of Ambatondrazaka, and the newly created CIRVA of Amparafaravola. The Agricultural Extension Pilot Project (Cr. 2150-MAG) is presently supporting extension activities in these two CIRVAs. No consensus could be reached, however, on whether the Agricultural Ministry's irrigation district (CIRIR) could henceforth take care of all system maintenance, nor whether an autonomous organization in charge of maintenance of the main works was needed, nor on the number of personnel to be laid off. 5.7 To date, although SOMALAC nominally does not exist any more, no decisions on these matters have been taken, in spite of repeated proposals and requests by IDA and the French co- financiers to solve these issues urgently. As a result, adequate system maintenance is by no means ensured.  5 6. Results 6.1 The main objective of the project was to increase rice production, which has been achieved (incremental paddy production on rehabilitated schemes has reached 34,230 tons versus an estimated 34,300 tons at appraisal) due to a 50% increase in yields (from 2.1 tons/ha to 3.2 tons/ha) and an increase in irrigated land (31,100 ha versus 28,000 ha at appraisal). Yields are a little less than expected at full development (in year 8 after project completion) due to the unfinished land planning on some schemes. Land planning should be completed shortly. The extension services of the Ministry of Agriculture (CIRVA of Ambatondrazaka) have been strengthened and have established very good relations with farmers through on-farm demonstrations. SOMALAC's extension service has been successful in promoting urea use and chemicals for weed control, but has been less successful in introducing direct planting in straight lines. Two new rice varieties that are more productive and more flexible in terms of cultivation timing, were developed by FOFIFA and introduced to farmers. The seed multiplication center of FOFIFA was strengthened and is presently providing enough seeds for the region and is, in addition, shipping seeds to other parts of the country. The monitoring and evaluation activities carried out under the project yielded valuable information that was used during project supervision. 6.2 A second objective of the project was to trigger a participation process with the farmers to get them more involved in scheme management and maintenance. This objective has been partially achieved through the establishment of WUAs and Credit Associations. WUA's were successfully established on all schemes with strong assistance from the extension services of SOMALAC and the CIRVA in order to manage water use at the farm level and plan for maintenance works for which they would become eventually responsible. Rehabilitation of the last schemes was executed in collaboration with the WUAs. The WUA was a new concept for Madagascar and the project can be regarded as a good field experiment for setting up farmers' organizations. By 1991, twenty-six WUAs had been formed. Towards the end of the project, however, an increasing number of WUA's delayed paying maintenance dues because of interference by local politicians. 6.3 A third objective of the project was to strengthen SOMALAC while opening progressively its civil works and marketing services to competition. As a result of the poor planning of the restructuring, this was only partially successful. Four small local construction companies were created in the region for the execution of rehabilitation works on the micro-schemes. These companies were ready to take over part of SOMALAC's civil works activities. Unfortunately, as the implementation of the restructuring of SOMALAC was delayed, these construction companies moved to other regions to find other markets, or went bankrupt. The marketing of rice was quickly taken over by the private sector, after lifting SOMALAC's monopoly. 6.4 Although not planned at appraisal, the project made a special effort to create Credit Associations which could borrow funds to finance seeds, fertilizers, and equipment from the Agricultural Credit Bank (BTM) on an individual basis but with collective guarantees. The volume of credit distributed to farmers belonging to Credit Associations is growing and the rate of reimbursement to BTM is close to 100%. Yields of farmers in Credit Associations (primarily on the large schemes) are significantly higher than the average yield (4.5 t/ha versus 3.2 tlha) due to credit availability and to increased attention from the extension service of SOMALAC. Discussions are ongoing with BTM on the feasibility of increasing the number of farmers organized in Credit Associations by giving access to credit to farmers who do not own land. Farmers in Credit Associations account for about 25% of rice production in the Lac Alaotra region.  6 6.5 In order to protect farmers' incomes, the project attempted to intensify and diversify production in the Lac Alaotra basin. This action was unsuccessful. 3/ Research and extension promoted the introduction of other irrigated crops and inter-seasonal crops (rainfed rice, wheat, maize, beans, etc.), but farmers did not respond due to a preference among farmers for a rice variety that was well suited to the area, but that did not allow for multiple crops and to competition from other production activities in non-irrigated areas (cassava, groundnuts, livestock). 4/ No production figures are available for the non-irrigated areas. 6.6 The land tenure situation on the large schemes is not clear. After rehabilitation, land holdings were to be consolidated and farmers would receive a valid title. At credit closing, half of the farmers still had no valid title and subdivision of land was increasing (in 1990, it was estimated that 21% of the reallocated parcels had been subdivided again). 7. Economic Results 7.1 The overall Economic Rate of Return (ERR) of the project is estimated at 25% (Annex I) as compared to 24% at appraisal. However, due to the unknown consequences of the political turmoil in 1991 and the uncertainty surrounding SOMALAC, the rate of return may turn out to be optimistic. Sensitivity analysis indicates that the rate of return remains above 10% even if net benefits decrease by as much as 25%, as a result of increased maintenance costs or decreased yields. It was assumed that irrigation works would continue to be maintained, either by the Government or the private sector, which may be overly optimistic. If no solution is found to the institutional problems that arose after the dissolution of SOMALAC, maintenance will suffer and production is likely to drop to pre-project levels; this would mean a rate of return close to zero. Separate rates of return for the SOMALAC (large) and CIRVA (small) irrigation schemes could not be calculated because the data collected during implementation were not detailed enough. However, a comparison of investment costs and yield increases between the large and smaller schemes suggests that the rate of return on smaller schemes may well have been higher. During implementation, it turned out that rice selling prices were higher than estimated during appraisal, but this was offset by higher production costs to farmers (mainly fertilizers, pesticides and small mechanization). 8. Proiect Sustainability 8.1 Although the project was largely successful in achieving its objectives, its sustainability is in doubt (as was the case after the first project), largely because the institutional questions have still not been resolved. 8.2 The main outstanding issue is that after dissolving SOMALAC, Government did not create an institution capable of assuring maintenance of the main civil works (storage dams, main canals and drains). The studies done on restructuring actually proposed one of two options to the Government: to create an autonomous maintenance organization (preferred by the French co-financiers) or to reinforce the existing CIRIR so as to enable it to take responsibility of maintenance (preferred by the 3/ The Staff Appraisal Report did not expect any significant impact of this activity on total output. 4/ Farmers owned land in both irrigated and non-irrigated areas.  7 Bank). No decision on either has yet been taken, which has potentially serious consequences for the security of storage dams and for a rapidly diminishing capacity of the distribution and drainage network (canals and drains need to be regularly cleaned because of on-going siltation due to erosion in watersheds). Unless such an organization is established (or the existing CIRIR reinforced), it will not be possible to sustain the results achieved. At present, the CCCE is reviewing the possibility of a follow-up project that would focus on resolving the present institutional impasse. Extension activities in the project area have been taken over by the Agricultural Extension Pilot Project (Cr. 2150-MAG). 9. IDA Performance 9.1 IDA supervision fell short during the design of the civil works, which did not properly take into account the existing erosion of the surrounding hills. Otherwise, the project was carefully and continuously supervised by the three co-financiers in good coordination with each other (CCCE, FAC and IDA). Supervision missions took place twice a year on average and received well prepared documents from SOMALAC that served as the basis of discussions between cofinanciers and Government. Because of the successful implementation of the physical components of the project, IDA decided to cancel the planned mid-term review. When problems arose later on, Government, IDA and CCCE met regularly in Paris. 9.2 SOMALAC suffered from a lack of timely counterpart funding, especially since 1987. In 1989, IDA was obliged to stop new procurement during a period of four months in order to press Government to provide counterpart funding. 9.3 During implementation, there were differences of opinion among the cofinanciers and between cofinanciers and Government on the timing of the liberalization of rice marketing and the restructuring of SOMALAC, which slowed down the decision-making process. 10. Borrower Performance 10.1 SOMALAC's management performed well in executing the project, but was unable to adjust to a changing environment. It failed to manage the necessary reduction in staff over time. This was true both for extension and for the marketing service where staff was kept even when SOMALAC's purchases of paddy declined by 80% after the break-up of its monopoly. 10.2 Due to lack of Government clarity and decisiveness, the studies on the future of SOMALAC took more time than needed and, as a result, demotivated SOMALAC personnel and diminished farmers' confidence in Government's capacity to deal with their problems. Some very good counterparts, trained by expatriate technical assistance left SOMALAC when they realized that the restructuring of SOMALAC was delayed amid interference by local and national politicians. This is also one of the main reasons why it is difficult at this moment to assess the sustainability of the project and of project impact on the development of the Lac Alaotra region. II. Consulting Services 11.1 The project used foreign technical assistance in different areas (extension, civil works, research, general management, etc.) and at different levels. In general, they performed well  8 throughout the project. However, due to a lack of long term planning on the part of SOMALAC management, and the scarcity of local trainable counterparts, some foreign experts left without having been able to train local staff. 12. Project Documentation and Data 12.1 The Staff Appraisal Report very often referred to policy orientation (liberalization, cost recovery, general organization, extension approach, etc.) and was helpful to the Borrower and the cofinanciers. For each supervision mission, SOMALAC management prepared in advance detailed reports on the relevant aspects of the project, which helped considerably in supervision discussions with SOMALAC and Government.  9 PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) PART II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE The Borrower did not submit a Part II. However, a project completion report was prepared by the implementing agency, which is available in the project files.  10 PROJECT COMPLETION REPORT MADAGASCAR LAC ALAOTRA RICE INTENSIFICATION PROJECT (CREDIT 1337-MAG) PART III: STATISTICAL INFORMATION Table 1: Related Bank Credits Loan/Credit Title Year of Purpose Amount of Status Approval Loan/Credit (Millions of SDR) Cr. 121 1-MAG 1982 Second Village Livestock and Rural 15.0 Completed Development Project June 1989 Cr. 1433-MAG 1983 Cotton Development Project 7.5 Completed June 1990 SF-008-MAG 1983 Cotton Development Project 9.4 Completed June 1990 Cr. 1589-MAG 1985 Irrigation rehabilitation Project 12.6 On-going Cr. 1709-MAG 1986 Technical Assistance to Agricultural 8.7 Completed Institutions Project Dec. 1990 Cr. 1691-MAG 1986 Agricultural Sector Adjustment 50.0 Completed Credit June 1990 Cr. 1804-MAG 1987 Second Agricultural Credit Project 8.0 On-going Cr. 1878-MAG 1988 Forestry Management and 5.1 On-going Protection Project Cr. 2041-MAG 1989 National Agricultural Research 18.6 On-going project Cr. 2125-MAG 1990 Environment Project 19.8 On-going Cr. 2150-MAG 1990 Agricultural Extension Pilot Project 2.8 On-going  11 Table 2: Project Timetab Project Timetable Date Planned Date Revised Actual Date Identification 1/ - 1978 Preparation 2/ - 1979/81 Appraisal - 09/81 Post Appraisal - 12/81 Credit Negotiation - 12/82 Board Approval - 03/83 Credit Signature - 05/83 Credit Effectiveness l/ 01/83 - 03/84 Completion 01/89 06/90,03/91 09/91 Credit Closing 06/89 01/91,09/91 09/30/91 Last Disbursement 12/89 06/91,01/92 02/20/92 1/ Identification was done by Government in 1978 in line with the master plan of the Lac Alaotra region prepared in 1976. 2/ Technical and organizational preparation were done by two consulting firms over a period of two years and an accounting and management audit of SOMALAC was carried out by a consulting firm in 1981. 3/ Delays due to lack of counterpart funds and delays in putting technical assistance in place.  12 Table 3: Proiect Costs and Financing A. Proiect Costs Appraisal Estimate Actual Costs Expressed in SDRc; Actual us % of M'n FMG '000 SDR M'n FMG '000 SDR Estimate Buildings 293 700 1,090 695 99 Irrigation Rehabilitation 5,390 13,300 17,656 13,270 100 Vehicles & Equipment 4,022 9,800 6,436 6,169 63 Inputs, Research & Seeds 1.325 3,200 7,196 6,482 203 Technical Assistance, Studies, Training 3,226 8,200 11,994 9,078 111 Total 14,256 35,200 44,372 35,694 101 Comments: Actual cost in FMG are much higher than estimated at appraisal because of several devaluations of the currency. Actual costs in SDR are close to those estimated at appraisal, although less equipment (farmer mechanization) has been bought due to delays in rehabilitation, more studies and technical assistance have been provided, and the seeds center was expanded. B. Project Financing Estimate Actual (estimated) Financing Expressed in SDRs; Actual as % of M'n FMG '000 SDR M'n FMG '000 SDR Estimate IDA Buildings 206 507 895 489 96 Irrigation Rehabilitation 2,774 6,842 9,053 5,551 S1 Vehicles & Equipment 2,141 5,281 5,905 5,380 102 Inputs, Research & Seeds 562 1,385 306 226 16 Technical Assistance, Studies, Training 1,089 2,685 5,009 3,420 127 Total 6,772 16,700 21,168 15,805 95 FAC 1,640 4,091 10,123 8,066 197 CCCE 3,763 9,27S 9,155 8,731 94 Government 2,081 5,131 3,926 3,092 60 Total 14,256 35,200 44,372 35,694 101 Comments: FAC financing was increased for technical assistance and for the expansion of the seeds center. For IDA, the financing of input supply was stopped in 1987, but more studies and training were financed.  13 Table 4: IDA Credit Disbursements A. Cumulative Estimated and Actual Disbursements (million SDRs) Fiscal Year 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Appraisal Estimate 0.9 3.3 6.5 10.4 13.8 16.0 16.7 _ __ _ ActualJ 0.8 1.6 4.8 6.8 10.9 13.2 14.2 15.2 15.8 Actual as % of Appraisal 0 24 25 46 49 68 79 85 91 95 Estimate IlI_ I I_____ Date of last disbursement: February 20, 1992. Amount cancelled: SDR 895,013.28.  14 Table 4: Credit Disbursements B. Disbursements by Credit Category (million SDRs) = Category Original Actual I. Equipment not included in Category 4 below, vehicles and spares for: (a) The Borrower (b) SOMALAC (excluding vehicles and workshop equipment 0.12 0.25 (c) FOFIFA 4.13 4.86 0.28 0.26 2. Civil works for building construction and rehabilitation (a) Under Part E.3 and H.2 of the Project 0.11 0.05 (b) Under Part C of the Project 0.19 0.17 (c) FOFIFA 0.14 0.27 3. Civil works for irrigation rehabilitation (a) Under Parts A & B of the Project 5.85 5.42 (other than on PC 15 and the Andranobe perimeter) (b) Under Part G. of the Project 0.02 0.13 4. (a) Incremental agricultural inputs and equipment, excluding ploughs, for CVA 0.49 0.00 under Part H. I of the Project (b) Incremental agricultural inputs and equipment, excluding ploughs, for 0.69 0.23 SOMALAC under Part H.2 of the Project 5. Consultants' services (a) Under Part F of the Project 0.37 0.30 (b) Under Part G. of the Project 0.46 0.18 6. Training (a) Under Parts E.l & E.2 of the Project 0.23 0.83 (b) Under Parts EA4 & G0 of the Project 0.23 0.04 7. Services under Part D of the Project 0.23 0.44 8. Studies (a) Under Part 1.5 & 1.6 of the Project 0.49 0.32 (b) Under Parts 1.1 through 1.4 of the Project 0.28 1.29 9. Refunding of Project Preparation Advance 0.93 0.74 10. Unallocated 1.46 0.00 Total 16.70 15.80  15 Table 5: Main Results Appraisa Fstimate Actual Actual as a % of Large Small Total Large Small Total Appraisal Schemes Schemes Schemes Schemes Estimate Rehabilitated surfaces (ha) 25,000 3,000 28,000 27,569 3,546 31,115 111 Paddy yields (tlha) . prior to project 2.10 1.80 2.07 2.10 1.80 2.07 - . at full development (year 8) 3.40 2.40 3.29 3.20 2.90 3.17 96 . credit associations n.a. n.a. n.a. 4.50 2.90 4.32 n.a. Incremental Yields 1.30 0.60 1.23 1.10 1.10 1.10 89 Production (t) . prior to project 52,500 5,400 57,900 57,895 6,383 64,278 - . at full development (year 8) 85,000 7,200 92,200 88,221 10,283 98,504 - Incremental Production 32,500 1,800 34,300 30,326 3,901 34,227 100 Comments: On large schemes, yields are a little lower than expected because land surfacing is not yet completed. It is expected that, when surfacing is finished, yields will be higher than originally estimated. It is also expected that the number of farmers in Credit Associations will increase as a result of changes in BTM's requirements for providing credit to farmers (no longer necessary to be a land owner). The main reasons for the Water Users' Associations to delay paying their dues for maintenance are: a) all land surfacing is not yet terminated and farmers are reluctant to pay dues for unfinished works; and b) pressure from local politicians who claim that water should be free; SOMALAC and the Ministry of Agriculture are now taking disciplinary actions to force payments.  16 Table 6: Compliance with Credit Condition Section Obiect 3.01 (b) (i) Price of paddy seeds, rice and inputs. Mostly complied with. Since 1987, paddy price to farmers is on a free-market basis. Since 1988, SOMALAC does not sell inputs. 3.01 (b) (ii) Counterpart funding. Often late, particularly in 1989 when IDA suspended new procurement financing for 4 months. 3.01 (c) Supplement to Subsidiary agreement with Complied with. SOMALAC. 3.01 (d) Proceeds of credit available to FOFIFA. Complied with. 3.02 Employment of consultants. Complied with. 3.03 Equipment insurance. Complied with. 3.04 (a) Preparation reports on certain parts of the Not consistently on time. project. 3.04 (b) Quarterly reports on certain parts of the Complied with. project. 3.04 (c) Inform on nationality of suppliers. Complied with. 3.04 (d) Reports on project execution after Not yet due. completion. 3.05 Annual work plans. Complied with. 3.06 Funds available for purchasing paddy and Not always complied with, particularly in 1986. marketing rice. 3.07 Transportation plan by June 1984. Complied with. 4.01 Audit reports and separate accounting. Complied with. 4.02 Inform IDA of SOMALAC and CVA status Complied with. change. 4.03 Proper maintenance of equipment by CVA. Complied with. 4.04 Land titling to farmers on irrigated land. Not complied with. 4.05 Adequate head of CVA and sound CVA Complied with. administration. 4.06 Foreign exchange allocation. Complied with. 4.07 Adequate SOMALAC salary structure. Complied with. 4.08 Water charges adequate at project Not complied with. completion. 4.09 Credit available to farmers. Hardly complied with (only to Credit Association and big farmers).  17 Table 7: Use of Bank Resources A. Staff Inputs (staff weeks) Stage in Project FiscaL Year Cycle- -- - - - - - - - - - - ____Cycle ____ 79 80 81 82 83 84 85 86 87 88 89 90 91 92 Total Preparation and 2.7 5.8 9.4 5.2 23.1 Identification Appraisal 46.7 7.5 54.2 Negotiations = =_= = 11.5 = = = 11.5 -Spuervision _ _ _ _ 1.3 29.9 12.6 30 19.5 15 13.1 9.4 8.5 4.9 144.2 PCR I___ 8 8 TotaL 2.7 5.8 9.4 51.9 20.3 29.9 12.6 30 19.5 15 13.1 9.4 8.5 12.9 241 B. Supervision Mission Information Supervision Month/ Days in Number SpeciaLization Performance Rating Type of Mission Year Field of Represented Status Trend Problems ____ ___ ____ ____ Persons __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ 1 Dec-83 15 4 Financial Analyst 2 2 Sr. Agriculturalist F,T,M Irrigation Engineer _______ ________ Training SpeciaList | __ __ __ 2 Feb-84 21 1 Financidl Anelyst 2 2 F,T,M 3 Jul-84 20 2 Sr. Agriculturalst F,T,M I______ I__ I__I_ FinanciaL Analst yst I_ 11 4 Nov-84 26 1 Sr. Agriculturalist 2 2 F,T.M 5 N.A. N.A. N.A. NA. 2 2 F,M 6 Dec-85 9 4 Sr. Agriculturalist 2 F,M Irrigation Engineer | Training Specialist 7 Apr-86 26 4 Economisat 2 F Irrigation Engineer _______| Training Specialist | 8 JuL-86 N.A. 1 Irrigation Engineer 2 ______ F 9 Mar-87 9 4 Sr. AgricuLturalist 2 F,M Irrigation Engineer | Financial Analyst 10 Oct-87 15 1 Irrigation Engineer 2 _ | 11 _ 11 Jun-88 8 2 Irrigatlon Engineer 2 M _______ ____ |___| FinanciaL Analyst | _____ 12 | Feb-90 8 1 Financial Analyst 2 | | F,4T | 13 Jul-90L 14 1 Financial Anaylt s2 7 I F.4T | 14 Apr91 20 1 Fnancial Analyst 2 T Note: N.A. means data not avaiLable. Key to types of probLem F: Financial M: Managerial T: Technical   Lac Alaotra Rice Intensification Project Economic Rate of Return p..j-I I Y 2 2 4 5 a 1 a a 10 II I2 13 14 I a II IA 1 20l 2 1 22 22 24 21 Cd.A. V.0 244 US I'M 10417 Is" late In le ss 0 *2 1*81 11,4 1494 less ls) lose 13*8 2000 2001 2002 2002 2004 200 a2, 20 2 44.44 Wm 444.70 M. 412.20 72840 uon0.40 1003.101) 1122.10 1280A0 484 -* lob 100.SO 11.OL 1246.4 144.14 184.44 01.24 3221.01 4SS47 264.47 214.1 27 S4.4)7 21.167 254.47 28.817 214.41 284.41 248.41 284.41 21441 2a 27 441 2*.41 214.17 244.4) 23*_7 _yM4 D_4.42 I OO C0 4 70.4* "Al 441.43 4.1 4.481 44.44 34.44 28.44 24.44 24.44 24.* 34.44* 4.s4s 204 2.4 4 2 382 . 28 44 2S." 34.44 28.11 34.44 2348 IIL.... 61-4 I*OUALAcI S.4.d ". 0 14401 21444 22414 21742 24284 2)848 2116 21144 21141 214413379 27144 2)4 27141 2,141 27444 2)541 27141 21184 21141 2144s 2114g3 2124 2)141 21613 _.d4 P -. .. 2_ 87 2.17 2 .7 3 82 ** Z 3 2. ) 2.17 2.11 3 .1 1 2.1) 2l2 2.1 I 11 211 3.17 217 21 2 I1) 2.1) 211 2.11 0.00 4oo )41 4444 761114 41220 7434* 30428 0)94 41214 83124 913)14 *3)24 47124 47124 874124 )3114 12734 41214 0)314 )3284 42144 8344 41214 41234 43124 s20>. 0.00 24114 404I 4 a1110 40271 4 401 41440 *4424 1124 18424 14424 41424 11424 64420 1442379 61424 14414 24428 S1421 14424 12423 1142S 11428 12428 41424 p... ........ 114 140 o71 41 IsI 42 41 1a 24 4 III 122 *5 644 441 11 441 "*1 47 41 447 So0 187 101 7170 u. F caG 410 *"44 1384 1)342 24224 21*5 3414 31142 2144 20212 21444 32 42 341 32,41 32441 34641 21441 22041 91441 31*4 42012 42012 47012 42012 . I ..k. 1444 424 12 9440 4404 12024 12440 14121 10S 11426 11760 12224 11321 16321 IS,12s 24221 116 11322 11211 16326 63211 14141 14241 16342 IS241 23 4410 Uft- OCVAN a- _.a 42 2210 2844 2144 2148 *44 2 *4* 2644 2144 2244 2644 2644 2144 2444 2144 244 2144 246 21544 3244 21s44 pTMy s..I. 1I. 24 2 )*0 2.40 3.01 2.0 2.24 32.1 3.70 3 20 220 la 2.70 3 10 3.70 3 10 2 l .10 320 3 21 3 1O 2.10 - _ p.44 2243f *127 10242 10114 31221 1104 I1200 12126 13120 12120 11210 13120 13120 19210 12120 12120 12120 11210 17210 11120 12120 t. o0l 1404 0)1 332 1242 701 40S S0 424 43224412 *122 3211 4322 0422 84122 422 4922 3322 4122 4222 4123 *d.p.. 6.0.GAI I ISO 271 431 .I 44s 41 24 443 sit 422 441 S*t &II 447 447 441 as) ") 447 202 701 0 07 0) 0M) :sA 441. 2 2114 2131 4144 40* 4043 4347 41*4 1811 1341 S161 4411 1111 1*41 4411 521 1 40) 4201 4201 4201 VA5M .h. 1844 824 24*4 Zile 2441 201 l)4 1440 142 201 2203 201l 2201 2203 201 1201 201 2m0l 2423 2434 1243 2424 P80OlmI03i COSTS T0 fAINAM" .*. 04 --,u 210.00 224A0 240.Ao 401.00 414.00 601A0 414.00 20.00