Madagascar - Mangoky Agricultural Development Project

Docunmnt of The World Bank FOR OFFICIAL USE ONLY Report No. 8460-MAG PROJECT COMPLETION REPORT MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT (IDA CREDIT 881-MAG ANT IFAD LOAN 11-MG) MARCH 21, 1990 Agriculture Operations Division South-Central and Indian Ocean Department Africa Regional Office This docuuient bas a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otberwise be disclosed without World Bank authorization.  EXCHANG! RATES Currency Unit Malagasy France (FMG) Rate at-Appraisal (Feb. 1979) sMG 225 / US$. Intervening year averages 1980 FMG 211 / US$1 1981 FMG 271 1 US$1 1982 FMG 350 I US$1 6.983 : FHG 430 I US$1 1984 :MG 577 1 US$1 1985 : FMG 662 1 US$1 Completion year average 1986 : FMG 676 1 US$1 1987 FMG 1,069 1 US$1 1988 FMG 1407 / US$1 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS BTM Bankin' Ny Tantsahe Mpamokatra - National Rural Development Bank CFDT Compagnie Frangaise pour le D4veloppeaent des Fibres Textiles - French Textile Development Company FAC Fonds d'Aide et de Coop6ration -Aid and Cooperation Fund - (rench Government) FED Fonds Europeen de Ddveloppement - European Development Fund FOFIFA Agricultural Research Agency IFAD International Fund for Agricultural Development SAMANGOKY Socidte pour l'Amenagement et la Mise en Valeur de la Vallee du Bas Mangoky - Development and Production Agency for the Lower Mangoky Valley SINPA Societd d'Int4ret National pour la Production Agricole - National Agricultural Production Company SODEMO Societd pour le Developpement Economique de la Region de Morondava - Economic Development Agency for the Morondava Region. FISCAL YZAR Government t January 1 - December 31 SAKANGOKXY October 1 - September 30 (till 1983) January 1 - December 31 (after 1984)  FOR OMCIA USE ONLY THE WORLD: ANK Washington. O.C. 20433 U.S.A. O6k. 4M DvcwftwGcsw&1 March 21, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Madagascar Mangoky Agricultural Development Proiect 'Credit 881-MAG) Attached, for information, is a copy of a report entitled "Project Completion Report on Madagascar - Mangoky Agricultural Development Project (Credit 881-MAG)" prepared by the Africa Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment This document hw a testricted distribution and may be used by recipients only in the performance of their oflicbl duties. Its contents may not otherwlise be disclosed without World Bank authofition.  FOR OMCIL USE ONLY PROJECT COMPLETION REPORT HADAASCAR MANGOXY AGRICULTURAL DEVELOPMENT PROJECT (IDA Credit 881-HAG and SFAD Loan 11-MG) TABLE OF CONTENTS Page PREFACE i BASIC DATA SHET ii EVALUATION SUMOIARY iv PROJECT COMPLETION REPORT 1. Background ..1*.....0.....*. ... II. Project Objectives and Description.............. 3 III. Project Design and Organixation................ 3 IV. Project Implementation ......................... . 5 V. Project Results................................. 8 VI. Project SustaiDability......................... 10 VII. Bank Perform,ance............................... 10 VIII.- Borrower Perfooance ........................... 11 U. Conclusions and Lessons Learned................ 11 Statistical Annexes......................... 13 Map (13654 RI) rThis docurnent has a retricted distribution and rnay be used by recpients only in the performance | of tbeir ofllcisl dutios Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT COMPLETION REPORT MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT (IDA Credit 881-MAG and IFAD Loan 11-MG) P RE FA C E This is the Project Completion Report (PCR) for the Mangoky Agricultural Development Project in Madagascar, for which IDA Credit 0881- HAG in the amount of USS 12.0 millton and IFAD Loan 11-MG in the amount of SDR 5.0 million were approved in March 1979. The IDA Credit and IFAD Loan were closed in August 1986. The last disbursements were made in April 1987 and an unutilized balance of US$ 1,165,000 was cancelled in June 1987. The PCR was prepared by the Agriculture Operations Division of the Africa Region, South Central and Indian Ocean Department and is based on the Staff Appraisal Report, the Credit Agreement, internal Bank memoranda and the correspondence between the Bank and the Borrower. This PCR was read by the Operations Evaluation Department (OED). The draft PCR was sent to the Borrower on January 5, 1990, for comments by February 23, 1990, but none were received.  - ii - PROJECT COMPLETION REPORT MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT (IDA Credit 881-MAG and IFAD Loan 11-MG) BASIC DATA SHEET Key Proiect Data Appraisal Actual Actual as 2 Expectation of Appraisal Project Costs * IDA Credit Amount (US$ million) 12.0 10.835 90.3 * IFAD Loan Amount (SDR million) 5.0 4.997 99.9 Economic Rate of Return 18 I Negative Institutional Performance Poor Financial Performance Negative Technical Performance Poor Number of Beneficiaries (families) 8,200 3,900 47.6 Proiect Timetable Date Date Date planned revised actual Identification - - 1960s Preparation - 76177 Pre-appraisal - - 07177 Appraisal - - 03/78 Loan/credit negotiations - - 10178 Board approval - - 03/79 Loan/credit signature _ - 05/79 Loan/credit effectiveness 06/79 - 03/80 Loan/credit amendments . first 03/82 . second 07/83 Loan/credit closing 08/84 08/85 08/86 Loan/credit disbursement closing - - 04/87 Credit Disbursements Cumulative Estimated and Actual Disbursements IDA Credit 1980 1981 1982 1983 1984 1985 1986 1987 ................... US$ million .. Appraisal estimate 0.40 3.00 7.00 10.50 12.00 12.00 12.00 i2.00 Actual 3.04 7.31 8.37 8.50 9.41 9.81 10.24 10.83 ..... ..percentage of estiatei.. Actual 760 240 120 81 78 82 85 90  Use of lank Resources Wonth/ Number Days in Specialis. Perform. Rating Type of Year persons field represented status trend problems Post.Appraisal 02179 1 ll* u.a. n.a. n.&. n.a. Supervision I 09/79 3 13* E+A+F 1 2 II 01180 1 2 a 1 2 - III 02/80 2 23* A+I 1 2 - IV 08/80 2 12* 3.1 2 2 F,T v 01/81 3 22* 2A+1 2 3 F,M,T VI 07/81 2 29* +.A 2 2 F.X,T VII 01182 2 18* l+A 2 2 F,T VIII 06/82 3 21* A+e?4 2 1 F.M,T LX 01183 4 23* A+E+PM 3 2 F,M,T,P X 03/83 1 7 A - - - XI 07/84 3 2 A+29 3I F,T,? XII 02185 2 26* A+Z 3 1 F,T XIII 10/85 4 7 A,I*+2+ 3 - F.M.T. XIV 04/86 4 19 A+E4I9+F 3 . F,M.T XV 09/86 1 2 I 3 - F,M,T XVI 01187 2 9 I+F 3 - F,M,T, * Multi-project mission Specializations A - Agriculturalist; I - EconoList; t * Financial Analyst; I a Irrigation Engineer; M a Management Specialist. Performance Status: 1 - Problem-free or minor problemes 2 - Moderate problems; 3 - Major problem. Ratint Trends 1 - Improving; 2 - Stationary; 3 - Deteriorating TIne of Problems: F - Financial; T - Techbical; X - Managerial; P * Political Staff In-uts (Staff Weeks) 78 74 7 77 7 Is 0 e1 a a 55 a Os 57 TAL ProppeaomI - 25.5 14.9 2.5 S - - - - - - - - - 68.9 A_mesIl 0.1 - - 29.1 . - - - - - - - 50.7 NeIetlo. _ _ _ - - 11.2 -11. Supowvl.b. 1.4 - - - - 2.5 1S.2 t10. 18.1 1.7 e.9 5.5 20 0.7 10o.8 or- - 0.5 - - 0.1 - - - - 0., 1.4 26.5 14.9 2.5 37.1 25.7 10.2 10.1 18.2 12.7 e.9 e.g 20 o.7 24.6  - iv - PROJECT COMPLETION REPORT MADAGASCAR MANGORY AGRICULTURAL DEVELOPMENT PROJECT (IDA Credit 881-MAG AND IFAD Loan 11-HG) EVALUATION SUMHARY Origin i. The Mangoky Agricultural Development Project was the third Bank credit for agriculture in Madagascar, with a credit amount of US$ 12.0 million. IFAD participated jointly with the Bank in the funding of the Project with a loan amount of SDR 5.0 million. The Project was designed to improve and extend existing irrigation systems so as to increase rice and cotton production. This was intended to complete the first phase of the irrigation system under development since the early 1960s, financed by FED and PAC. as part of a long-term master plan aimed ultimately at bringing about 100,000 ha of land into production. Preparation of the project took place between 1976 and 1977. The credit wias approved in March 1979. Initial Components ii. The 5-year project, costing US$29.5 million, comprised the following components: a) development of primary and secondary irrigation and drainage networks serving 3,700 ha, and road infrastructure (US$5.1 million); b) land clearing and levelling, and development of tertiary and quaternary irrigation systems serving about 3,300 ha, including construction of concrete canals (US$9.5 million); c) detailed der.igns.. and supervision of civil works (US$0.5 million); d) incremental agricultural inputs (US$2.7 million); e) social infrastructure (six new primary schools) (US$0.5 million); f) applied research, training, monitoring and evaluation and studies to prepare future projects (US$3.4 million). Implementation iii. SAMANGOKY, a parastatal founded in 1961 to conduct the development of the Lower Mangoky valley, was responsible for the implementation and overall coordination of the Project. iv. Social and political problems, lack of local funds, inability of SAMANGOKY to execute works on force account and to coordinate the Project,hampered the project implementation from the outset and led to important delays. In 1983, through an amendment of the Joint Project Agreement, the initial target was scaled down and modified as follows : (a) development of 1,500 ha of land instead of 3,700 ha, (b) execution of urgent works for the rehabilitation of the existing irrigation  - V - infrastructure, and (c) provision of technical assistance, equipment and vehicles for SAMANGOKY. Corrective actions were also taken to concentrate SAMANGOKY's role and activities in the maintenance and operation of the scheme, extension services and input supply. A financial recovery programme of the agency was concurrently funded by the Government. These measures came too late to restore SAMANGOKY and the project to a healthier situation. Results v. Out of the additional 3,700 ha originally envisaged to be developed under the project, about 2,800 ha had been equipped by the closing date (1986). Total cultivated area fell from 4,800 ha without the project (1979-80) to an average of 3,500 ha during the 5-year implementation period (1981-85), with a record 5,700 ha cultivated in 1986. Cultivation intensity did not improve and was less than 80? during project execution, against 180? estimated at appraisal. Crop yields decreased steadily during project implementation, from pre-project (1971 to 1979) averages of 2.5 t/ha and 3.8 t/ha for cotton and paddy respectively, to about 1.2 tlha and 3.1 tlha between 1980 and 1987 while SAR estimates were 2.8 tJha and 3.5 t/ha (double cropping). At project completion, about 3,900 farm families were cultivating land under the irrigation scheme, almost half of the estimated appraisal target. Cropped plots per family ' are small, averaging about 0.8 ha cropped per year, and income levels, given the low yields and producer prices, are about US$190, almost 25Z of SAR estimates. SAMANGOKY's financial situation has remained precarious with substantial operating losses during project implementation despite a stringent reduction of staff and Government's periodic subsidies. vi. The Economic Rate of Return was estimated at 182 at appraisal, but, at the level of operation (5,000 ha) at project completion, the project is yielding a negative rate of return, since incremental benefits are non-existent and operation and maintenance costs are higher than estimated. Sustainability vii. The drastic lack of motivated farmers, qualified staff, adequate equipment prevailing in the region during project implementation were the major aspects which contributed to the project's failure. A highly sophisticated, capital intensive and technologically demanding project, requiring high operational and maintenance standards, an efficient production system, a sound political- economic environment cannot be sustainable and must not be replicable in isolated areas sitl.lar to the lower Hangoky region. The risks were obvious, well known, openly debated and alternatives considered on the technical, organizational and political fronts. The decision was to proceed nonetheless because the Bank was hopeful that its involvement would provide a vehicle for addressing the real sector problems that were intrinsic in all these risks. There were few other opportunities for the Bank's assistance and involvement in the sector. Mangoky did provide an important foothold at a critical moment in  - vi - the late 1970s when the objective of opening the dialogue on the agricultural sector was central and specific. The willingness to coneider Mangoky in the midst of difficult country relations and a clearly deteriorating sector and country situation did give the Bank an entree and some credibility deriving from direct field experience that proved beneficial. Findings and lessons vii. The main lessons learned from this experience are the following: (a) it should have been recognized at appraisal that the trend in country conditions raised serious questions about the suitability of the technical and managerial design of the existing and proposed Mangoky project. The irrigation construction featured a capital intensive and sophisticated technology calling for nigh 0 & M standards, an efficient production and marketing system and motivated and qualified farmers; it also required extremely good management skills, which in Madagascar had been already proven to be a rare commodity. Also, the project was based on the perception that the settlement process was stabilized, whereas this was not the cases economic and political conditions were precarious and incentives for farmers to move to the area hardly existed. (b) an enclave project has little chance for success in a situation where sector policies and the political environment are as difficult as it was in Madagascar. (c) supervision, although highlighting the main problems, failed to follow-up on crucial measures towards redressement of.the situation, pointing once more to severe staff and budget constraints compared to the needs to properly follow up such a difficult and complicated project. (d) Government policies should have been much more closely followed with a view to assess whether it was still useful to continue project execution with any chance of success. Notably, pricing policies and political interference in project's internal matters such as staff reduction, land allocation, cost recovery, had strong adverse effects on any chances to complete the project successfully. Drastic actions could then have been taken to stop or to comprehensively reappraise the project when it was still time. viii. Looking to the future, the main issues to be addressed to restore the project's sustainibility, are the following: (a) opening up the area by rehabilitating the Toliara-Tanandava and Tanadava-Morombe roads and the Morombe port;  - vii - (b) management approach by attracting the private sector and using foreign and local private capital as well as loanot (c) 0 & H system to be managed by private sector and Water Users Associations: (d) diversified crops as alternative to rice. In the context of t"eE ongoing Madagascar Irrigation Sub-sector Review, a thorough assessment of the future of the Mangoky Project will be done, with main focus on the above aspects.  PROJECT COMPLETION :'PORT MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT (IDA Credit 881-HAG and IFAD Loan 11-KG) I - BACKGROUND 1. Madagascar is predominantly an agricultural country, with about 852 of its population living in the rutal areas, the agricultural sector contributing about 402 to GDP and accounting for more than 802 of total export earnings. Following the political transition in the 1970s, growth in agricultural production failed to keep pace with demand and consequently imports of food, mainly rice and edible oils, increased dramatically. The country is nevertheless relatively well endowed with natural resources and the potential for agricultural development is considerable. With the introduction of appropriate policies and adequate external financial assistance. medium and long-term prospects for the economy were judged favourable at the time of loan approval (1979). 2. The broad objectives of the Government's long-term strategy for the period 1978-2000 were (i) to attain self-sufficiency in food and to develop basic industries, largely supplied by the agricultural sector; (ii) to decentralize economic activities, with the Government playing a more active and direct role in the economj, and (iii) to narrow the gap in income distribution and redress regional imbalances. As part of the establishment of a socialist regime, the Government intended to assume control of cultivated land on large farms and newly developed areas, through the creation of socialist cooperatives, and to monopolize agricultural marketing and processing activities. 3. Given the high pressure on the land in the developed cer-tral area of the country, future agricultural growth was naturally seen in the development of the isolated and sparsely populated lovwer Mangoky valley where hitherto cattle-raising had been the main activity. The climate in this region is favourable, the soils are relatively good and, with irrigation ihere necessary, suitable for most main crops. Irrigation water could be provided by tapping the Mangoky river which offers abundant hydaulic resources. With this excellent potential for agricultural development, the Mangoky delta was expected to permit low cost production of a number of crops and to encourage migration from the highlands. This vision conferred an image o' bread basket and population escape valve on the Mangoky project, with a huge impact on the whole southwestern region, if not on the entire country. These aspects made the project an interesting pilot undertaking and explained the continuing participation of various donors.  - 2 - 4. Development action in the lower Mangoky valley began in the 1950 with the creation of an experimental station and the preparation of a master plan for the development of over 100,000 ha of land. In 1961. the first phase of 10,000 ha of this development plan started, mainly for cotton. Eighteen years later, at the time of project appraisal, about 5,200 ha only had been developed, of which about 4,800 ha were cultivated, 3,200 ha devoted to the cultivation of cotton and the rest mainly to rice. Only 5,000 families were settled in the project area. 5. A parastatal company, Samangoky, was founded in 1961 as the development agency for the Mangoky Valley. with responsibility for (i) infrastructure and force account works, particularly land development; (ii) operation and management of the irrigation and drainage schemes, and (iii) operations related to cultivation of the developed area, together with the associated farmers. A special partnership contract between Samangoky and the associated farmers stated that the company should provide for mechanized land preparation, pesticide spraying, inputs, extension and marketing services, in exchange for which it would receive a portion of the production, collectable in kind for rice and as a price deduction for cotton. At appraisal, in the late 1970s, Samangoky was perceived as a very efficient agency, with good managerial capabilities and a sound financial position. In that respect, it was a "star case among Madagascar's troubled public institutions. 6. Up to that period, the project had received substantial financial support, initially from FAC (Fonds d'aide et de cooperation) of the French Government, and latterly from the European Development Fund (FED), which helped to sustain the momentum of progress in civil works up to 1981. The French Textile Company (CFDT) also played an important role in promoting cotton production in the area, providing extension, input supply and marketing and processing facilities. Once the third part of the 10,000 ha first phase development programme was nearing conclusion, FED announced the end of its commitment. The Government then applied for an IDA credit and an IFAD loan. 7. The Bank group had previously financed two irrigation projects in Madagascart the Lake Alaotra Irrigation Project approved in 1970 and completed in 1976, and the Morondava Irrigation and Rural Development Project, which was experiencing serious organizational, technical and financial problems at the time of the Mangoky project appraisal. Like the Mangoky project, these projects were designed to improve and extend existing irrigation systems and increase crop production. II - PROJECT OBJECTIVES AND DESCRIPTION 8. Objectives. The project aimed at developing about 3,700 ha in the Lower Mangoky Valley for production of rice on heavy soils and cotton on lighter soils, to be farmed by about 3,200 families. This was intended to complete the first phase of the irrigation system under development since the early 19609, as part of a long-term master plan aimed ultimately at bringing about 100,000 ha of land into production.  -3- 9. Components. The 5-year project, costing US$29.5 million, comprised the following components: i) development of primary and secondary irrigation and drainage networks serving 3,700 ha, and road infrastructure (US$S.l million); ii) land clearing and levelling, and development of tertiary and quaternary irrigation systems serving about 3,300 ha, including construction of concrete canals (US$9.5 million); iii) detailed designs and supervision of civil works (US$0.5 million); iv) incremental agricultural inputs (US$2.7 million); v) social infrastructure (six new primary schools) (US$0.5 million); vi) applied agricultural research, research and trials on operating problems, overseas study trips and training courses, monitorin, and evaluation and studies to prepare future projects (US$3.4 million). 10. In 1983, through an amendment of the Joint Project Agreement, the initial target was modified as follows : (i) development of 1,500 ha of land instead of 3,700 ha, (ii) execution of urgent works for the rehabilitation of the existing irrigation infrastructure, and (iii) provision of technical assistance, equipment and vehicles for Samangoky. III - PROJECT DESIGN AND DESCRIPTION 11. Although there was a clear conceptual foundation for the project, as a continuation of a reasonably successful development prograum, some design features were vulnerable to a high element of risk, particularly the capital intensive sophisticated technology, with gross costs of US$8,000 per ha (1). To be profitable, the project would have required high operational and maintenance standards and a very efficient production system. A sufficient number of capable managers and technicians with executive and planning abilities would have been necessary, as well as motivated and qualified farmers and the unrestricted iLccess to all required materials, inputs and other imported resources. These conditions, though only barely guaranteed at the time of the project's appraisal, became more difficult to fulfil soon after because of political disruptions, foreign exchange constraints, senior staff shortages and a dramatic deterioration of the infrastructure and equipment. (1) Featuring concrete lined and prefabricated canals, semi-automatized water level control systems, mechanized agricultural operations, aerial spraying, etc.  -4- 12. With the benefit of hindsight, it is now clear that the timing of project appraisal (early 1978) was unfortunate since only a few months later, significant political changes (para. 14 and 16) took place which considerably affected the project set-up. It was quite impossible, at appraisal, to clearly evaluate the political and sociological environment and to foresee the risks which were to occur to the project. It was then difficult for the Bank to cancel the credit or to suspend disbursement on this basis. 13. Also, production and financial results of the 1978 and 1979 campaigns revealed serious problems (para. 19) which, if knovn, would have certainly led to a different project design and scope. Moreover, some aspects of project design were inapproprate, such as ti) drainage to control salinization was not adequate in some areas, (ii) agricultural machinery and maintenance equipment for the extended area were not included, (iii) the proposed social infrastructure was insufficient, and (iv) there were no provisions for technical assistance. More specific provision for these matters would have given the project a greater chance of success, since lack of them led to serious problems during implementation. 14. At appraisal, it was difficult to impose permanent land tenure rights as a prerequisite for proceeding with the Project. Therefore, arrangements were agreed with Goveroment, defining roles and responsibilities of Samangoky in this matter. But fundamental changes were to occur later which could not have been foreseen. Samangoky, which was expected to be in charge of execution of most of the project's activities, had full authority in allocating land to 'associated farmers' and imposing discipline in cropping activities. However, during the first project year, the impact of national decentralization policies on the Samangoky area increased markedly, shifting the land allocation decision from Samangoky to political-administrative authorities elected in the region ('Fokonolonas'). As a result, Samangoky was deprived of one of its most important instruments in imposing discipline on farmers to ensure appropriate cropping and systems' maintenance. Although, in principle, the democratization measure appeared sound, it did not in practice improve the equity in the land allocation process. A recent survey on farmers' reactions revealed general dissatisfaction with land allocation procedures which adversely affected the project's productivity and cost recovery because farmrs who were not fulfilling their obligations could no longer be penalized. 15. Until 1985, Samangoky executed all mechanized works for land preparation and infrastucture maintenance, provided inputs and water, ensured the marketing of agricultural products, and assumed applied research and extension activities. Divesting these activities to farmers, private sector or national services would alleviate Samangoky's role, but Madagascar was not ripe enough - nor the Bank - for restructuring its parasatals (1). Farmers performed manual activities such as weeding and harvesting. The rationale on which was founded the partnership arrangement (1) The Public Sector Adjustment Credit under which the Bank started the restructuring of Malagasy parastatals became effective only in 1988.  -5 between Samangoky and its associated farmers became undermined since the latter had virtually assumed the status of tenants vithout any inheritance rig ts to their plots, and were therefore not interested in the proper conservation or maintenance of the quaternary canals and drains. Moreover, the cost recovery system offered no incentives to increase productivity in cotton and was not adjustable in relation to the quality or timeliness of services provided by Samangoky (1). Once Samangoky had lost its disciplinary role and was unable to provide adequate agricultural services, productivity fell dramatically, and so did cost recovery, leading Samangoky into its subsequent precarious financial situation. 16. During the first project year (1980), political and social unrest among Samangoky's own personnel led to an extended strike and eventually to the resignation of the General Manager. Since the enterprise's efficiency wq founded in large part in his managerial capabilities and dynamism, and recruitment of capable senior staff to work in the project area was very difficult, Samangoky's future performance was seriously affected. 17. In retrospect, it seems that a capital intensive and technologically demanding project was not manageable at that time, given the country's political-economic conditions and the existing human resources. Overestimation of Samangoky's strength and capabilities and insufficient attention paid to productivity trends (para. 27), production organization and operation and maintenance problems were also major aspects which contributed to the project's eventual failure. The risks were obvious, well known, openly debated and alternatives considered on the technical, organizational and political fronts. The decision was to proceed nonetheless because the Bank was hopeful that Mangoky's isolation might allow it to continue to function as an enclave, and that the Bank's involvement would provide a vehicle for addressing the real sector problems that were intrinsic in all these risks. There were few other opportunities for the Bank's assistance and involvement in the sector: Mangoky did provide an important foothold at a critical moment. The dialogue on the project was a major element in the opening of the dialogue on rice marketing and pricing and on the cotton sector in the late 1970s. It was the first occasion on which the Bank was able to join an effective dialogue on these topics, which eventually helped to complement the work under the Agricultural Institutions Technical Assistance Project, the Agricultural Sector Adjustment Credit, the Public Sector Adjustment Credit, etc... The objective of opening the dialogue was central and specific. The (1) Global charges, comrrising all services, were fixed uniformly and collected from users as a price deduction for cotton (from 50 X of the price in 1979 to 66S in 1984) and as in-kind payment for paddy (1 t/ha, when all services were provided, 0.4 t/ha when land preparation was done manually). However, given the declining yields and unfavourable product input prices' evolution, charges fell below Samangoky's costs and collection rates were too low (between 50 to 70Z). In 1985, individual accounts were introduced, services and inputs were charged at cost and a fixed water charge was established.  -6- willingness to consider Mangoky in the midst of difficult country relations and a clearly deteriorating sector and country situation did give the Bank an entree and some credibility deriving from direct field experience that proved beneficial. IV - PROJECT IMPLEMENTATION 18. Despite social and political problems which hampered project implementation from the outset, and delays in construction works due to lack of local funds and Government slowness in fulfilling fonmalities for loan effectiveness, construction of the main irrigation and drainage networks contracted to a private enterprise was nevertheless concluded four months ahead of schedule. Conversely, development works (land clearing, levelling and quaternary drains and canals) on force account with Samangoky fell behind schedule owing to shortages of spare parts and cement and the hopeless state of the canal factory (producing concrete canal sections), partly attributable to foreign exchange constraints, and partly to weaknesses in Samangoky's management. When cost overruns became excessive and Samangoky was incapable of completing development works, the contract was cancelled, scaled down and eventually awarded to a local enterprise under the project reformulation agreement (para. 21). Despite their importance for project success, some other components, such as applied ' agricultural research and construction of social infrastructures were never fully executed because of ill-defined institutional roles and responsibilities. 19. By far the most important factor adversely affecting project implementation was Samangoky's managerial weakness and its accumulated financial losses. One of the main reasons f6r this was that for many years the highest technical positions in the works and agriculture departments were not filled due to lack of qualified candidates interested in working in such an isolated location, and senior managers spent most of their time in the central office in the capital. The accounting and auditing services were deficient and planning for operation and maintunance was insufficient. Samangoky began to experience financial losses in 1978/79, and by April 1981, they had already accumulated to about FMG610 million (US$2.3 million). All reserves were used up and the enterprise's liquidity situation became precarious, the company relying for its operations mostly on bank overdrafts and Government advances. By that time, construction equipment and agricultural machinery were old and insufficient to execute the normally required works, and the factory fabricating canals was practically unutilizable. Consequently, Samangoky had no operational capacities to provide for the scheme's operation and maintenance or to carry out land preparation. As the production system was no longer operative, yields decreased and consequently farmers failed to pay their service charges, thus worsening still further Samangoky's financial situation. Although Samangoky's management capabilities were seen at appraisal as the main source of risks, no provisions were made to improve its operational and planning systems nor to revise its accounting procedures.  -7- 20. Project risks appear to have been underestimated at the time of appraisal, particularly those associated with yield decreases. Starting in 1976, there were already signs of a declining trend in cotton yields which, combined with increased insect attacks and deficiency in cropping practices, led to the early deterioration of cotton productivity (para. 27) and thence to the project's economic and financial debacle. 21. An attempt to redress the project's disappointing perfomance was made in 1983 when, under an amendment to the Joint Project Agreement, the following corrective actions were proposeds i) Samangoky should undergo reorganization, abandoning its construction activities and reducing the scope of agricultural services provided to farmers, and concentrating on the scheme's maintenance and operation, extension services and input supply. The agency would also have to improve its operational and managerial capacities and reduce its redundant personnel. To achieve these objectives, Samangoky would benefit from technical assistance to be provided under the project. Concurrently, the Government would fund a financial recovery programmee; ii) unfinished works in the extension areas should be scaled down and contracts tendered to private contractors; iii) urgent rehabilitation works to the existing scheme should be executed under the project; and iv) financing be provided for the replacement of part of Samangoky's equipment and machinery. 22. Other actions/decisions eventually taken to restore project' production levels included: (i) modification of the cost-recovery system, introducing individual accounts for inputs and services provided to farmers; (i') more flexibility in cropping patterns, a decision which entailed a rapid conversion to rice and the introduction of other crops; (iii) allocation of medium-size plots to private farmers and enterprises to encourage higher land occupancy. 23. Most of the decisions and actions taken during implementation were well conceived but directed to solving only specific problems. However, successive supervision missions (para. 36) failed to comprehend the growing severity of the project's diminishing viability. The 1983 project reformulation was a concerted attempt to limit losses, as was the subsequent inclusion of a Mangoky component in the Irrigation Rehabilitation Project (para. 35). These measures came too late and neither of them had the necessary impact to restore Samangoky to a healthier financial position and give the project a chance of sustainability. In retrospect, it appears that it would have been wiser to stop the project right after the 1980 crisis, when Samangoky's financial losses became critical. At that stage, perhaps the situation could have been salvaged had the project been redesigned. reappraised and, if its eventual profitability could have been demonstrated and the Government commltment to support Samangoky's rehabilitation secured, additional resources poured intc it.  24. A major lesson learned from the experience of this project is that when projects perform poorly within a deteriorating environment they should be stopped and comprehensively reappraised at the earliest possible moment or instance, as soon as one element of the risk analysis has reached a threshold level or if an unforeseen factor appears to affect project economic or financial viability. (1) V - PROJECT RESULTS 25. The proj6ct has not only failed to achieve the objectives established at appraisal but also those of the rehabilitation plan proposed in the 1983 amendment of the financing agreement. Out of the additional 3,700 ha originally envisaged to be developed under the project, about 2,800 ha had been equipped by the closing date (1986), although some 320 ha of these were never fully completed. The total equipped area in the perimeter had, by that time, reached 7,000 ha, of which about 2,000, however, were classified as unexploitable due to soil restrictions (salinity, or sandy soils inappropriate for rice cultivation) or drainage and levelling problems. 26. Total cultivated area fell from 4,800 ha without the project (1979-80) to an average of 3,500 ha during the 5-year implementation period (1981-85); a record 5,700 ha was cultivated in 1986, only to fall again th6 following year to 4,900 ha. Cultivation intensity did not improve and was less than 0S during project execution. against 1802 estimated at appraisal (2). 27. Crop yields decreased steadily during project implementation, from pre-project (1971 to 1979) averages of 2.5 t/ha and 3.8 tlha for cotton and paddy respectively, to about 1.2 tlha and 3.1 t/ha between 1980 and 1987 (1). SAR estimates were 2.8 t/ha and 3.5 t/ha (double cropping) for cotton and paddy respectively. Among the factors contributing to decreasing cotton yields weres (i) Samangoky's incapacity to provide timely land preparation and to plan and supervise agricultural operations; (ii) the absence of incentives to obtain higher yields, since farmers had to pay the same portion of their production regardless of the amount and quality of services received; (iii) farmers' outlook discouraged by poor income prospects, farmers neglected their plots or even abandoned them; (iv) inadequate land levelling and deterioration of irrigation conditions, and (iv) inadequate extension services and input supply to cope with technological problems, particularly insect infestation. (1) In this particular case, when cotton yields fell below 2.0 t/ha Samangoky's losses became beyond control. A declining trend in cotton yields was already evident before project appraisal, being 2.8, 2.4, 2. 0 and 1.7 t/ha in 1971, 1976, 1978 and 1979 respectively. (2) The cropping pattern shifted from predominantly single cropping cotton cultivation to double cropping rice, in early 1987 after the sharp increase in paddy price when liberalization of rice pricing and marketing took place. There is now a great demand for land for rice cultivation from the High Plateau individual farmers, mostly the Betsileos from the Fianarantsoa province.  28. At project completion (1986), about 3,900 farm families were cultivating land under the irrigation scheme, fewer than before the project (about 5,000 in 1980) and almost half of the estimated appraisal target (1). Cropped plots per family are small, averaging about 0.8 ha cropped per year, and income levels, given the low yields and producer prices, are about US$190, almost 252 of SAR estimates. 29. In an attempt to increase land occupation, three large companies were allocated land in 1984 to grow cotton, but all of them went into liquidation after the 1984-85 catastrophic campaign, when their cotton yields averaged 0.9 t/ha (202 less than those of small farmers). A number of medium-scale farmers started cultivating under the irrigation scheme in 1985; presently there are about 140 of them, each cropping an average of 15 ha. While they are supposed to be more efficient producers, yields recorded up to now are often lower than those obtained by peasant producers, hence the future prospects of their entrepreneurial activities are not bright. Most of these medium-scale farmers are local traders, public officials or Samangoky employees, lacking practical experience, capital and depending highly on the scarce peasant labour. 30. The Samangoky audited accounts for 1985 show an operating loss of PMG664 million (US$0.9 million equivalent) in itu overall Income Statements, while losses of FMG114 million and FKG410 million are reported for 1986 and 1987 respectively. Despite Government subsidies (FMG1,198 million support for the financial balance of the company in 1985) and periodic contributions to maintenance works, Samangokys' financial situation remains critical. Even if cost recovery were to improve from its disappointing 502 to 602 level, and if the agency were to adopt more stringent austerity policies (its staff has already been reduced from 850 in 1979 to 300 in 1986), its financial future would continue to be precarious as long as land occupation in the irrigation scheme remains at the current low level. 31. The Economic Rate of Return was -ttimated at 18? at appraisal, but at the present level of operation (5,000 ha) the project is yielding a negative rate of return, since incremental benefits are non-existent and operation and maintenance costs are higheJr. It was difficult to determine project benefits, because despite the expansion onto new land under the project, the cultivated area remains almost the same as it was before the project. Iloreover, the degradation of the irrigation infrastructure is such that unless considerable rehabilitation works are undertaken the future of the scheme is bleak. For the time being, operation and maintenance costs will continue to be very high and exceed Samangoky's financial capacity. (1) Between 1982 and 1985, about 2,200 producers left the project after Samangoky decided not to allocate my more land to cotton growers with yields under 1.5 t/ha or to rice producers with outstanding debts.  - 10 - 32. SAR estimated profitability thresholds for cotton and rice cultivation (under mechanized cultivation) are yields of about 2.0 t and 2.3 t per hectare for cotton and paddy respectively. However, average yields during implementation were significantly lower for cotton (1.2 t/ha) and in respect of paddy cultivation, which has become the main crop in 1987. the labour intensive technology adopted by peasant farmers resulted in a yield of only about 2.7 t/ha, which represents a net return to labour of little more than US$1.0 per day (more or less the equivalent of the official minimum wage). Production on the average 0.8 ha of peasant farm is estimated at 2.2 t of paddy per year, which just about cover family consumption, thus generating a meagre cash income. The financial viability of these farms is therefore uncertain. 33. The financial viability of medium-scale farms is even worse, since their yield levels are no better than those of the peasant farms, and operational costs for mechanized land preparation, inputs and hired labour are significantly higher. 34. Declining international prices for cotton and the instability of institutional prices for paddy have certainly contributed to the low production performance, along with the deteriorating physical condition of the hydraulic infrastructure and the road to Toliara (1) and the lack of efficient marketing channels. Though the increased official prices for paddy in 1986 were instrumental in boosting production that year, the very' low 1987 prices were decisive in undermining the following harvests. VI - PROJECT SUSTAINABILITY 35. With the drastic lack of qualified staff, motivated farmers and adequate equipment prevailing in the region during project implementation, there is little hope that the project could improve. The whole scheme soon started suffering from water shortages and bad drainage because of improper management and insufficient maintenance. Soils are salinized due to poor drainage in some areas; farmers abandoned the region. Highly sophisticated hydraulic infrastructure, especially with canals carried above ground cannot be sustainable and must not be replicable in areas similar to the lower Hangoky region. VII - BANK PERFORMANCE 36. Since the project was confronted with serious problems from the outset, it required higher than usual supervision inputs. However, the Bank's supervision efforts were not sufficient: supervision missions were fielded twice a year, but they often had to deal with several projects at the same time and there was little staff continuity between missions. During the critical period when project reformulation was to be implemented (1983-84), missions were curtailed because of staff-constraints. The project was given a lower priority at a time when major attention and iitensive efforts were concentrated in addressing sector policy and management issues with the Government. (1) The 210 km road connecting to Toliara, the nearest important city. When practicable, the journey requires 6 hours.  - 11 - 37. The lank failed to take decisive actions aimed at correcting emerging problems, particularly at the early stages of project implementation. The amendment to the Joint Project Agreement was made without an exhaustive analysis of the project's viability. VIII - BORROWER PERFORMANCE 38. Borrower's performance during project implementation was less than satisfactory. The lack of managerial capabilities, the shortage of qualified and motivated staff and the declining commitment of the borrower towards the project hampered the achievement of the project's development goals. 39. Project reporting was poor during the first years but improved notably afterwards, owing mainly to technical assistance which appeared necessary due to insufficient local qualified staff. Three positions were filled with expatriates which provided support of mixed quality in planning and organizing Samangoky's agricultural services, workshop and maintenance works. The achievements of locally recruited assistance to improve the accounting system were not totally satisfactory since there are still serious gaps in cost accounting procedures. A monitoring and evaluation system, which was supposed to be designed and established during the first years of the project, was never implemented. Finally the project experienced substantial non compliances with credit covenants (see Table 9) . IX - CONCLUSIONS AND LESSONS LEARNED 40. The main lessons learned from this experience are the following: ii) it should have been recognized at appraisal that the trend in country conditions raised serious questions about the suitability of the technical and managerial design of the existing and proposed Mangoky project. The irrigation construction featured a capital intensive and sophisticated technology calling for high 0 & X standards, an efficient production and marketing system and motivated and qualified farmers; it also required extremely good management skills, which in Madagascar had been already proven to be a rare commodity. Also, the project was based on the perception that the settlement process was stabilized, whereas this was not the cases economic and political conditions were precarious and incentives for farmers to move to the area hardly existed. (ii) an enclave project has little chance for success in a situation where sector policies and the political environment are as difficult as it was in Madagascar. (iii) supervision, although highlighting the main problems, failed to - follow-up on crucial measures towards redressement of the situation, pointing once more to severe staff and budget constraints compared to the needs to properly follow up such a difficult and complicated project.  __12 . (iv) Government policies should have been much more closely followed with a view to assess whether it was still useful to continue project oxecutiot with any chance of success. Notably, pricing policies and politlcal interference in project's internal matters such as staff reduction, land allocation, cost recovery, had strong adverse effects on any chances to complete the project successfully. Drastic actions could then have been taken to stop or to comprehensively reappraise the project when it was still time. 41. Looking to the future, the main issues to be addressed to restore project sustainability, are the followings ti) opening up the area by rehabilitating the ToliaLa-Tanandava road and the Morombd port; (ii) managemont approach by atti.acting th, private sector and using foreign and local private capital as well as loans; (iii) 0 & H system to be managed by private sector and Water Users Associations; (iv) diversified crops as alternative to rice. In the context of the ongoing Madagascar Irrigation Sub-sector Review, a thorough assessment of the future of the Mangoky Project will be done, with main focus on the above aspects.  Table 1 _ 13 - MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Basic Data Related Bank Loans and/or Credits Credit Title and Purpose Total Cost Year of Status Approval 322-MAG Morondava Irrigation and US$15.3 06/1975 completed Rural Development Project million al Rehabilitate and extend existing irrigation systems (9,300 ha) for rice, tobacco and cotton production 1337-MAG Lac Alaotra Agricultural US$18.9 03/1983 on-going Intensification Project million Reinforcement of agric. development institutions; rehabilation of irrigated schemes; supply of inputs; organisation of water users associations 1589-MAG Irrigation Rehabilitation US$29.6 09/1985 on-going Project million b/ Strenghtening of existing institutions. rehabilitation of 16 medium irrigation schemes (17,100 ha) and rehabilitation of the large-scale Mangoky scheme a/ Serious management and financial problems, production is poorly organized and yields are very low. The project was revised substantially and reduced in 1976. Seo PPAR dated December, 1984 (OED Report No 5403) b/ Financing of rehabilitation works for the Mangoky project was cancelled due to misprocurement; equipment component is currently also suspended.  1 14 - Table 2 MADAWSCAR IANGOKY AGRICULTURAL DEVELOPMENT PROJECT Project Timetable Date Date Date plamned revisea actual Identification a/ - - 1960s Preparataion bl - - 76177 Pre-appraisal - - 07/77 Appraisal c/ - - 03/78 Loan/credit negotiations dl - - 10/78 Board approval - - 03/79 Loan/credit signature - 05/79 Loan/credit effectiveness 06/79 - 03189 Loan/credit amendments el * first 03/82 * second 07/83 Loanicredit closing 08184 08/85 08/86 Loan/credit disbursement closing - 2187 a/ The project was identified by the Government in the 1960s, as the first phase of a long-term masterplan for the development of the Lower Mangoky Valley. Developments efforts started in the 1950Os and at the time of appraisal some 5,200 ha had been developed with financial support from FAC and FED. b/ Project preparation was dcoe by the Government, Sauangoky and consulting engineers (SOGREAR/SOMEAH). c/ Main issues raised at appraisal and board presentation related to disappointing experiences and impleme..tations difficulties with on-going agricultural projects in Madagascar- fears of consequences of low producer prices and pricing policies, the cost recovery issue and its relatior. to land tenure arruzgements. Technical issues related to drainage and water managements were also raised. d/ The project was co-financed by IFAD (Joint Project Agreement dated 17 May 1987 as its first operations in Madagascar (Loan HAG-11). IFAD negotiations, which had to be preceeded by the Government's applications to IFAD membership, delayed board presentation. el A first amendment concerned Schedule 2 of the Joint Project Agreement. to increase disbursement percentages for foreisn expenditures (Cat. 2 and 3). The second amendment scaled down land development objectives to 1,500 ha and modified the disbursement schedule, introducing new categories for rehabilitation works, technical assistance and equipment and vehicles for Samangoky.  Table 3 - 15 - MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Credit Disbursements Cumulative Estimated and Actual Disbursements 1/ IDA Credit 1980 1981 1982 1983 1984 1985 1986 1987 .............. ..... US$ million ......... Appraisal 0.40 3.00 7.00 10.50 12.00 12.00 12.00 12.00 estimate Actual 3.04 7.31 8.37 8.50 9.41 9.81 10.24 10.83 *. . .... . 0.... percentage ........... Actual Ps X of 760 240 120 81 78 82 85 90 estimate Date of Final April disbursement 1987 11 Calendar years, ending December.  - 16 - Table 4 MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT ProJect Implementation KEY INDICATORS Unity Appraisal Revised 11 Act4al Estimate Major Civil Works (Area to be served) ha 3,700 1,500 1,800 Extension of the main canal km a - 9 Secondary canals km 2; - 14 Drains km 73 - 59 Roads km 121 - 101 Development Works Bush clearing ha 3,400 1.500 1,800 Land levelling ha 3,300 1,500 1,600 Tertiary & quatern.canals km 158 - 93 Drains km 60 - 166 Roads km 198 - 89 Area to be served by ha 3,700 - 3,700 main civil works 21 Area developed ha 3,300 1,500 1,800 3/ Area potL'ially cultivable ha 3,300 1,500 1,470 41 of whichs cotton area ha 700 - 150 rice area ha 2,600 - 1,320 Potential cropped area 41 ha 5,900 - 2,700 1/ The amendment to the Joint Project Agreement states only the total area to be developed (1,500 ha). 2/ Main civil works comprised the extension of the main canal, secondary canals and drains, tertiary drains and roads. These works were concluded before the project revision. 3/ Of which 200 ha were only clearea and 120 ha were levelled but not equipped. 4/ After conversion to rice cultivation. S/ Assuming appraisal estimates on cropping intensity (two crops of rice and one crop of cotton).  - 17 . Table 5 Page 1 MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENS PROJECT A. Project Costs Appraisal Estimates Actual 11 FMG US$ US$ million ... .miilion . Major civil works 2/ 1,145 5,089 7,000 Land development 2,130 9,466 9,500 Design studies & supervision 126 560 1.100 Incremental agricultural production costs 614 2,728 n.a. Applied agriculture research 3 13 n.a. Social infrastructure 25 11l 0 Technical services 3/ 757 3,364 1.860 Base costs 4,800 21,330 n.a. Contingencies 1.825 8,111 Total costs 6,625 29,441 n.a. NOTE: Due to changes in project scope and financing, inadequate or missing according records and the overlapping with the Irrigation Rehabilitation Project, it was not possible to accurately establish actual project costs. n.a. - data not available I/ Estimates based on disbursements and financing percentages. 21 Including urgent rehabilitation works. 3/ Including Mangoky Phase II Project Preparation, applied investigations, training and technical assistance.  . 18 - Table S Page 2 MADAGASCAR MANGOXY AGRICULTURAL DEVELOPMENT PROJECT B. Project External Financ9pg Planned 21 Revised 3/ Final 41 Cat. No. Category IDA IFAD IDA IFAD IDA IFAD 1/ USS SDR US$ SDR USS SDR ............... '0000 1 Civil works in extension areas 4,400 5,600 5,577 2 *Civil works for rehabilitation 1,000 332 3 Land cleaning and levelling 3,605 3,000 2,802 4 Manufacture of canals and 4,400 3,700 3.277 installation of networks 5 *Technical assistance 300 55 6 Design studies and supervision 400 400 879 7 *Equipment and vehicles 1,000 981 8 Applied agriculture research 100 100 48 and investigation, research on operating problems 9 Project preparation studies 1,300 1,300 320 training 10 Studies for preparation of 600 600 611 agricultural projects Unallocated 1,500 700 TOTAL 12,000 5,000 12,000 5,000 10.83S 4,997 Unwithdrawn amount 1,165 3 1/ Category number corresponding to the amendment of the Joint Project Agreement; asterik indicated new categories introduced at that time. 21 As established in the Joint Project Agreement (1979). 3/ Amendment of Joint Project Agreement (1983). 4/ As disbursed at closing date (1987). NOTE: Some discrepancies appeared in category allocations, especially with categories 3 and 4. Totals may not add due to rounding.  - 19 - Table 6 Page 1 ADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT trojecI_Results Cultivated Areas, Yields and Production Year Cultivated Areas b.yrp (ha) SAR Actual Cotton Ricell Total Cotton Rice Other Total Crops 1978179 - 3,140 1,660 - 4,800 1980 - - 3,390 1,390 - 4,780 1981 3,160 2,220 5,380 3,210 1,170 - 4,380 1982 3,260 3,020 6,280 2,100 1,150 - 3,250 1983 3,390 3,960 7,350 2,590 1,100 - 3,690 1984 3,540 5,060 8,600 2,150 900 - 3,050 1985 3,690 6,160 9,850 1,890 1,340 180 3,410 1986 3,840 7,260 11,100 1,660 3,240 840 5,740 1987 3,840 7,260 11,100 110 4,600 170 4,940 Year Yields 2/ Incremental Production tIha) (t) 31 SAR Actual SAR Actual Cotton Paddy Cotton Paddy Cotton Paddy Cotton Paddy 1978179 - - 1.7 3.1 - - - - 1980 - - 1.3 3.8 - - (1,220) (110) 1981 2.4 3.0 1.5 3.3 SO 720 (700) (1,260) 1982 2.5 3.2 0.8 3.3 290 3,230 (4,090) (1,400) 1983 2.6 3.5 1.0 3.2 570 6,420 (3,200) (1,640) 1984 2.8 3.5 1.3 3.3 1,000 10,420 (2,900) (2,230) 1985 2.8 3.5 1.0 2.7 1,410 14,520 (3,840) (1,580) 1986 2.8 3.5 0.7 3.0 1,820 18,740 (4,500) (4,480) 1987 2.8 3.5 1.2 2.5 1,890 19,670 (5,560) (6,240) 1/ Additional land taken into production with double cropping for rice, except for the Mongolovolo area (400 ha). 2/ Yields in the extension areas, except Hongolovolo. 3/ cased on situation without project.  - 20 - Table 6 Page 2 MADAGASCAR MANeOKY AGRICULTURAL DEVELOPMENT PROJECT Proiect Results Without Appraisal PCR prolect estimate estimate 1979 Project Beneficiaries (Number) Peasant families 5,000 8,200 3,900 Medium-size farms - _ Average Cropped Area (ha) Peasant farms 0.9 2.0 0.5 Medium-size farms - - 40 Net Benefits (US$) 1 ha cotton 560 (negative) 1 ha rice (mechanized) - 386 30 1 ha rice (manual) - - 230 Estimated Farm Family ?acome (USS) 130 800 110  - 21 - Table 7 MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Studies Title Consultinag Firm Date Irrigation Rehabilitation/ ZEDR Hamokatra/SOMEAH-SOGREAR 02/81 Petits pdrimetres start Livestock production in the SATEC 09181 Middle West Rehabilitation of citrus farm GERSAR 09/81 Bezeika Bas Mangoky Phase II SOGREABISOIWAR 03/82 NOTE: With the exception of the Bas Mangoky Phase IX study, no references vere found for studies executed under the project.  -2t- Page 1 MADAGASCAR MANGOKY AGRICUJLTUJRAL DE~VELOPMENIT PROJECT 3.02 Coloyeent of *aIn @rtfg Complied with consul tants 3.04(ol(b) turnishin9 of informtio n Coapi e with keepin,g Ot eOrS 3.04 (c) CoUclepio0 report 6 months after Coemiance sot Partially valted closing date entirely 2 years after satisfactOry closing date 3.04 (8) Propsol on meonitorla 31 O e. 19)t not comuliOe Io. "St system and eveluettn witth 3.04 () oaterly reports N ot copliee with 3.05 Acquisition of lead Complied with 3.06 Yearly repots and I Oct. of CoMpiQeS with Doloyed during work prograef_s each year the first years 3.07 tOR for Pesmorcb prcaMIe - Comoliena "st No research programe anE studio& an future entirely satist. 3.08 Proposals on recowery of 30 Juoe 1960 Not Complied with Country wide investment costs ropeosal to te 4.01 CO~n" of accounting Comolieg with records & suOsissolo of yearly finsacial reports 4.02 Accounting of SnangoykY Complied with Sesw of auditeo & subuuSst.n of awUtv accounts requested accounts several times 4.05 (a) Quolity of SaMengoky C-mplian e ot mnanaenent entirety satitf. 4.OS (0) Consultstion with 104 an Not complied with IDA was not Consulted *esointnent of Dlrector of Sanangoky's Works Dee.  _ 23 - Table 8 Page 2 MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Compliance with Credit Conditions Section Content Deadline Compliance 4.06 Autonomy of Works Dept. - Complied with & information on land development costs 4.07 Consultations with IDA on - Compliance not change in organization of entirely agricultural production satisfactory methods of Samangoky 4.08 Consultation with IDA on - Complied with in cotton & rice pricing and the context of marketing policy other operations (ASAC, Cotton Sect.) 4.09 Schistomiasis control in - Not complied project area with 4.10 Financing of cotton - Compliance not and marketing of cotton entirely satisfactory  - 24 - Table 9 MADAGiASCAR MANKOGY AGRICULTURAL DEVELOPMENT PROJECT Use of Bank Resources Missions Mgnth/ Number Das in Snecializ Perfor. Raling Tvoe of Yea gersons field renrgsentedf status trend PCoblems 1/ 2/ Z/ i/ at 5l W Post-Appraisal 02/79 1 11 * n.a. n.a. n.a. n.a. Supervision I 09/79 3 13 * E+A+F 1 2 - II 01/80 1 2 E 1 2 - III 02/80 a/ 2 23 * A+I 1 2 - IV 08/80 2 12 * E+I 2 2 F,T V 01/81 3 22 * A+A+I 2 3 F,M,T VI 07/81 2 29 * E+A 2 2 F,M,T VII 01/82 2 18 * E+A 2 2 F,T VIII 06/82 3 21 * A+F+M 2 1 F,M,T IX 01/83 4 23 * A+E+F+M 3 2 F,M,T,P X 03/83 1 7 A - - - XI 07/84 3 2 A+2E 3 1 F,T.P XII 02/85 2 26 * A+E 3 1 F,T 7/ XIII 10/85 4 7 A+I+E+F 3 - - XIV 04/86 4 19 * A+E+I+F 4 - - XV 09/86 E/ 1 2 * I 3 - - XVI 01/87 2 9 I+F 3 A/ Loan effectiveness, March 1980. 2/ Final disbursement completed on April 1987. JJ Date of the mission report. 2/ Multiple purpcse missions marqued (*) Xl A = Agronomist or agriculturalist 4/ 1 = Problem-free or minor problems E = Economist 2 = Moderate problems F = Financial Analyst 3 = Major problems I = Irrigation Engineer M = Management Specialist j/ 1 - Improving 2 = Stationary 3 = Deteriorating f/ F = Financial T = Technical M = Managerial P = Political Z/ New supervision summary format from this date.  -25 - ANNEXI Table 1 QADAJ&= MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Crop Budgets: Rieg 11 ha) Su I/ E Psg2 (mechanized) (manual) Seed 1,200 14,000 12.000 Fertilizers 5,060 48,000 33,000 Agro-chemicals 1,350 10,400 10,400 Samangoky services and water charges 27,390 50,000 50,000 Hired labour 500 173,000 - Mechanized services .-. 5/ 67.QO - Total cost 35,500 362,400 105,400 Family labour requirements (man/days) 150 - 4/ 203 Estimated unit cost per man/day 300 - 950 Yields (kg/ha) 3,500 2,500 2,700 Price (FMG/kg) 35 165 165 Value of production (FMG) 122,500 412,500 445,500 Net benefit 87.000 50,100 340,000 Return to labour (FMG/day) 580 - 1,670 (USS/day) 2.60 - - 1.11 Profitability threshold i/ (kg/ha) 2,300 2,200 1,800 1/ From Model 1 (1 ha rice double cropping). Z/ First campaign 1988. 1/ Total costs, including labour, divided by product price. A/ Commercial farms. j/ Figures included in Samangoky services.  - 26 - AEX 1 Table 2 M8AGASCAR MANGQKY AGRICULIURAL DEVELOPMENT PROJECT Crop Budaets: Cotton ft ha) 5i8 I/ Actual 1982183Z Seed 500 500 Fertilizers 15,180 32,500 Agro-chemicals 74,460 86,300 Sacks l 2,500 Aerial Spraying / 34,600 Mechanized activities A/ 67,900 Hired labour A/ 58,300 Samangoky's services and water charges 21_a60 86.500 Il Total cost 112,000 369,100 Family labour requirements (man/days) 180 90 Estimated unit cost per man/day 300 650 Yields (kg/ha) 2,800 960 Price (FMG/kg) as 177 Value of production 238,000 169,920 Net benefit 126,000 (199,180) Return to labour (FMG/day) 724 neg. (US$/day) 3.2 neg. Profitability threshold (kg/ha) 1,950 2,400 A/ Figures included in other items. 1; From Model 3. Z/ From Charancy/Estur, CFDT mission 1983. After this period cotton production declined steadily J/ Estimated to cover full costs of Samangoky's operations. 4/ Total costs including labour, diovided by price.  ANNEX I -27 - Table 3 MADAGASCAR MANGOKY AGBICULTURAL DEVELOJPMENT PBDJECT Samangokv K(ey Indicators Oge ating Personnel ZI Rates of actual Results.I/ arges collected .I FMG million No.employee % X 1977 (41.4) - n.a. 1978 (45.7) - - n.a. 1979 (246.0) 846 - n.a. 1980 (211.7) 791 - 79 1981 (876.0) 772 - 63 1982 (718.8) 768 - 58 1983 (1,185.8) 743 - 56 1984 (716.0) 411 60 1985 (664.2) 393 68 54 1986 (114.7) 298 40 68 1987 (410.0) 284 58 55 1/ *Risultats dexploitation generale de 1'exercisew, Accounting Services, Samangoky. 1976/77 is the last year with a positive result. aV Total number of employees. Samangoky. J/ Until 1983/84 charges for cotton were recovered directly through price deductions. n.a. = not available.  ANNEX 1 -28 Table 4 MADAGASCAR MANGOKY AGRICULTURAL DEVELOPMENT PROJECT Price Indicators Producer Prices 1/ Exchange Price Index i/ R*