Madagascar - Agricultural Extension Pilot Project

Document of The World Bank FOR OFFICIAL USE ONLY c4Z Z / SC) -r Report No. 8334-HAG STAFF APPRAISAL REPORT MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT APRIL 30, 1990 Africa Region South-Central and Indian Ocean Department Agricultural Sector Operations Division 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 disclowed witbout World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit Malagasy Francs (FMG) USS 1.00 = FMG 1620 rMG 1,000 US$ 0.62 WEIGHTS AND MEASURES 1 are (a) 0.0247 acres 1 hectare (ha) 2.47 acres 1 kilometer (km) G 0.62 miles 1 square kilometer (km2) 0.39 square mile 1 kilogram (kg) 2.20 pounds 1 liter (1) 0.26 US gallon 0.22 Imperial gallon 1 tonne (t) 2,204 pounds ABBREVIATIONS BAF c Administrative Bureau (Bureau Administratif et Financier) CCCE - French Development Fund (Caisse Centrale de Cooperation Economique) CS Field Supervisor (Chef de Secteur) CIRVA - Agricultural Extension District (Circonscription de la Vulgarisation Agricole) DVA - Directorate for Agricultural Extension (Direction L de la Vulgarisation Agricole) EIP - Production Intensification Agency (Entreprise de l'Intensification Agricole) FOFIFA = National Center for Applied Research for Rural Development. ICB International Competitive Bidding IFAD = International Fund for Agricultural Development LCB - Local Competitive Bidding MINAGRI - Ministry of Agriculture and Land Use (Ministere de la Production Agricole et du Patrimoine Foncier) MPAEF - Ministry of Livestock and Forestry (Ministere ee la Production Animale et des Eaux et Forets) MRSTD = Ministry of Scientific and Technologic l Research for Development (Ministere de la Recherche Scientifique et Technologique pour le Developpement) M&E Monitoring and Evaluation NARP National Agricultural Research PrJect NGO e Non-Governmental Organization PPF = Project Preparation Facility SOMALAC = Regional Development Authority for the Lac Alaotra Region SPVA = Provincial Agricultural Extension Service (Service Provincial de Vulgarisation Agricole) T&F = Subject Matter Specialist (Technicien Formateur) T&V Training and Visit VA ' Extension Agent (Vulgarisateur Agricole) Fiscal Year Government of Madagascar January 1 - December 31  FOR OFFICUIL USE ONLY MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT TABLE OF CONTENT CREDIT AND PROJECT SURIMARY ..... iii I. INTRODUCTION .. .................................................. 1 II. THE AGRICULTURAL SECTOR A. Main Features. ........................................ 1.= B. Government Policy for Agriculture ......................... . 2 C. Government Institutions in Agriculture . . 3 D. Bank Group Involvement in Agriculture . . 3 E. IDA Strategy for Agriculture . . 4 F. The Initial Pilot Activities ... 4 III. THE PROJECT A. General . . ............................................... 6 1. Project Objectives .................................. . 6 2. Summary Project Description . . . 6 B. Detailed Features . . ........................................ 7 1. Extension Services . . .................................. 7 2. Training .............................................. 8 3. Research .............................................. 9 4. Monitoring and Evaluation . . . 10 5. Financial Management System ... 10 6. Studies ... 11 7. Management Strengthening . . . 11 C. Project Costs and Financing .... 12 1. Cost Estimates . ...................................... 12 2. Financing Arrangements . . . 13 3. Impact on Government Budget ... 14 D. Procurement and Disbursement .... 15 1. Procurement . . ........................................ 15 2. Disbursements . . ...................................... 17 E. Accounts, Audits and Reporting Procedures . . 18 1. Budgeting and Reporting Procedures . . 18 2. Accounts and Audits . . .............. 18 This report is based on findings of an IDA mission which visited Madagascar in September 1989 comprising of Messrs. Charles Ameur (Mission Leader), Thomas Bredero, Dirk Prevoo and CW.is Ward (IDA); Alfredo Benites and Andre Rasolo (Consultants) This document has a restrictwd distribution and may be used by recipients only in the performance of their officiai duties. Its contents may not otherwise be disclosed without World Bank authorization.  - ii - F. Project Management ........ ........................... 18 IV. BENEFITS AND RISKS As Benefits.*** ....o ..*19 B. Risks ro.mental Imac...... ..... .0..... ... . 0 ....... 019 C. EnvironmentalImat2 D. Impact on Women. .......... . 20 E. Economic Rate of Return.... . ...... ........... a ........ 20 V AGREEMENTS REACHED WITH THE BORROWER, CONDITIONS AND RECOMMENDATION ................. . ..... 20 ANNEXES 1. Organization Chart of MINAGRI 2. Organization Chart DVA 3. Disbursement Profile 4. Economic Analysis 5. Economic Farmgate Prices 6. Documents in Project Files MAP: IBRD No. 21993, Madagascar: Agricultural Extension Pilot Project  MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT CREDIT AND PROJECT SUMMARY Botrower: Democratic Republic of Madagascar Amount: US$ 3.68 million Terms: Standard IDA terms, with 40 years maturity Beneficiaries: Ministry of Agricultural Production and Land Use; participating farmers Prolect Description: The project would constitute a three-year pilot phase of a long-term development effort to increase the efficiency of the agricultural extension services in Madagascar. It is part of a long-term strategy aiming at sustainable agricultural development through modernization of agricultural practices, environmental measures and population policies. The project would put particular emphasis on replicability and financial sustainability. More specifically, it would help: (i) reorganize and strengthen the extension services in 5 of the country's 23 CIRVAs (CIRVAs of Ambatondrazaka, Antananarivo, Miarinarivo, Moramanga and Amparafaravolo); (ii) provide systematic in-service training for project staff at all levels; (iii) focus adaptive research on validation and diffusion of economically and environmentally sound technology; (iv) strengthen the monitoring and evaluation capability at both the central and regional levels; (v) strengthen financial management; (vi) improve organizational and managerial capacity of the DVA; and (vii) make funds available for studies, an in-depth review of the project and alternative approaches to extension at the end of year 2, and the preparation of a follow-up project that would draw on the results of the project review. Part of the study program is intended to look expressly for ways to reduce the cost of extension to the Government through: (i) involving the private sector and NGOs more forcefully; (ii) seeking for ways to reduce costs at the DVA, e.g. by reducing excess personnel; and (iii) reviewing the possibility of cost recovery mechanisms. Also, both MINAGRI and IDA would carefully monitor the project's actual costs. Benefits and Risks: In the short-term the main benefits would be the achievement of improved efficiency and higher effectiveness of the DVA  - iv - extension service. Lriteria for measuring such achievements would be (a) increased collaboration between DVA &nd FOFIFA, ,b) a greater knowledge of farmers' demands and concerns by extension age%ts, and (c) the elaboration and implementation of measures to streamline the organization. In the long- term, the project would improve farmers' productivity, increase incomes for the rural population, increase food security, and would be a main tool in stemming environrmntal degradation. The risks include those usually associated with extension projects: (a) inadequate and/or untimely provisi&a of budgetary resources by Government; (b) staff of low compe- ance; (c) the possible inability of extension and research services to produce messages which farmers value in the short run; (d) slower than expected adoption rates by farmers of the recommended technical advice; and (e) poor inter-agency coordination. To minimize these risks emphasis will be put on strengthening financial and management capabilities at both the central and CIRVA levels; reinforcing links between extension and research; preparati( of annual work programs and related budgets; and an in-depth review of the project with emphasis on impact, sustainability and replicability at the end of year two. The adequate provision of counterpart funds will be reviewed annually in the framework of the Public Expenditure Program. In addition, a Bank staff member has been assigned to the Resident Mission to specifically advise on project implementation.  Estimated Project Costs: In current million US$ Equivalent Local Foreign Total Investment Costs: 2.66 0.99 3.65 Recurrent Costs: 0.94 0.01 0.95 Base Costs 3.60 1.00 4.60 Physical Contingencies 0.10 0.03 0.13 Price Contingencies 0.59 0.09 0.68 Total Project Costs 4.29 1.12 5.41 of which Taxes 0.81 Financing Plan Government 0.92 million 20% 1/ IDA 3.68 million 80% 2/ Total 4.60 million 100% Taxes 0.81 million Grand Total 5.41 million Estimated IDA Disbursements: FY89 FY90 !FY91 FY92 FY93 -US$-------------US$ million--------------- Annual 21 0.55 0.90 1.03 0.80 0.40 Cumulative 0.55 1.45 2.48 3.28 3.68 Economic Rate of Return: 17% Map: Madagascar: Agricultural Extension Pilot Project, IBRD No. 21993 Charts: Organigramme of the MINAGRI Organigramme of the DVA 1/ Excluding taxes and duties estimated at US$ 0.81 million. 2/ Including advances totalling US$1,000,000 from the Project Preparation Facility.  MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT STAFF APPRAISAL REPORT I. INTRODUCTION 1.01 The Government of Madagascar has requested IDA's assistance to improve the operation of its agricultural extension services. It is part of Government's long-term strategy to achieve sustainable agricultural development by addressing a nexus of inter-related problems (agricultural modernization, demographic pressures and environmental decline). A Government's analysis of the performance of the agricultural sector over the early 1980s revealed that production and productivity were declining steadily while population was growing at a 32 annual rate, leading to unsustainable farming practices and environmental degradation. Management weakness, lack of suitable messages to address farmers' problems and constraints, ineffective links with research, irrigated rice in the highlands receiving high priority in terms of research and inputs to the detriment of other crops, excess staff, lack of mobility, and inadequate operating funds characterized extension. As a result, the Ministry of Agriculture's extension personnel was poorly trained, unsupervised, ill- equipped, and insufficiently motivated and had lost most of its credibility with farmers. By contrast, a few region specific operations backed by outcide financing and technical assistance had in the past shown good results. Meanwhile, there had never been any particular effort by the Ministry of Agriculture and Land Use (MINAGRI) to harmonize and coordinate the various approaches to extension nationwide along a clearly defined strategy. 1.02 In October 1987, MINAGRI's extension service (DVA) with IDA's technical assistance initiated pilot activities in the three extension districts (CIRVAs) of Antananarivo, Miarinarivo, and Moramanga, to test ways to improve extension. Funds were provided by IDA through a US$750,000 PPF (P486-0). Despite delays during implementation, the pilot activities generated encouragiag results at the field level and helped verify some of the work methods and norms that could be replicated on a larger scale. The present project would therefore further test the main ongoing activities while including 2 additional CIRVAs, at Government's request. It would be executed over three years, at a cost of US$5.3 million, to be financed by Government and IDA. II. THE AGRICULTURAL SECTOR A. Main Features 2.01 Agriculture is the leading contributor to Madagascar's economy as it employs 85% of the population of 11 million and accounts for more than 80% of export earnings. Due to its climatic diversity both tropical and temperate crops can be grown. Most of the estimated 1.5 million farm  families have smallholdings rarely exceeding 5 ha of land, and are generally engaged in subsistence farming. More than half the total cultivated area is under rice, the country's main staple, which accounts for 30X of agricultural production value. Annual paddy production is estimated at 2.2 million tons with an average yield of 1.8 ton per ha. Most of the rice is under some form of water management. Otb.r important foodcrops are tubers (cassava, sweet potato and taro) and maize. Production from the beef cattle herd ranks second in value terms (16% of production value). Industrial crops (sugar and cotton) as well as edible oil crops (groundnut, coconut and oil palm) account for 7% of production value. A small group of export crops (coffee, cloves, vanilla and pe.per) account for 15% of production, but 60-7u% of export earnings. 2.02 Agriculture has generally performed poorly in the 1975-87 period with growth rates well below population growth, which is reflected in a continuous decline of per capita income. The main reason for this poor performance has been the Government's interventionist policies in pricing and marketing, as well as high export taxes, while at the same time allowing its services to the sector (extension, research and irrigation maintenance) to decline in quality and coverage. Thus, the country's main export, coffee, declined in both quality and quantity, ard Madagascar became a major importer of rice and edible oils during this period. Furthermaore, expansion of agricultural lands to compensate for stagnatir.g yields and population pressures has caused wide-spread deforestation and resulted in accelerated environmental degradation (erosion, siltation of irrigation systems and loss of bio-diversity). B. Government policy for Agriculture 2.03 Faced with this situation, thc Government has made a series c: policy changes in recent years that have much improved the macro-economic environment in which the sector operates. Under a series of sectoral and macro-edjustment operations, the Government realigned the exchange rate and liberalized foreign exchange allocations for imports, removed import controls, liberalized exports, and began the restructuring of the parastatal sector, including certain public enterprises engaged in agricultural production, processing and trade. Although this adjustment process is not yet completed, notably ir. parastatal restructuring, the beneficial effects of the measures taken are already becoming apparent. Rice production, buoyed by good weather and higher producer prices (due to increased competition) has increased to levels needing only marginal imports, and coffee exports have expanded. In 1989, agricultural production Is estimated to have grown well in excess of population growth for the first time in 15 years. 2.04 Parallel to efforts to reduce Government's role in agricultural production, marketing, pricing and trade, steps are being taken to improve the efficiency of its research and extension services. Similarly in irrigation, the objectives are to reduce the Government's concentration on large irrigation schemes through autonomous parastatal agencies, and to focus more on reinforcing re-habilitation and maintenance capabilities in smaller traditional irrigation areas, which comprise 90% of Madagascar's irrigated lands. An Environmental Action Plan (EAP) has beeu developed by the Government with assistance from IDA and other international donors, and  -3- the World Wildlife Fund. A credit for an Environment Program to fund the implementation of the first five years of the EAP was approved by the Executive Directors on April 17, 1990. There will be couslderable synergy between this Program and the proposed project in the areas of soil conservation and watershed management. C. Government Institutions in Agriculture 2.05 rhree ministries are involved in the agricultural sector: Agriculture and Land Use (MINAGRI), Livestock and Forestry (MPAEF), and Scientific Research (MRSTD). MINAGRI and MPAEF are both responsible for extension, each in their respective areas. The agricultural research institute (FOFIFA) is under MRSTD. Until the mid-70s, both research and extension benefitted from strong financial and tec$,nical bilateral French aid. When this support was reduced in 1974, combined with the subsequent erosion of Government's financial capacity to fund salaries, maintenance and other operational costs, the ability of these services to fulfill their roles declined sharply. Furthermore, their responsibilities were diluted by the creation of numerous autonomous project agencies, each with its own extension service and research components that were usually more crop- specific in orientation (e.g. in rice, coffee and cotton). Some of these autonomous agencies are presently in the process of, or are being studied for, reintegration into the Ministries. D. Bank Group Irvolvement in Agriculture 2.06 Since 1969, IDA has supported 20 projects in the agricultural sector, for a total of US$ 240.6 million. Most of the early projects were area specific development or enclave type operations and had large financial and management autonomy from the Government's own services. The projec . r.?letion reports and PPARs available for this generation of projects ':dicate that their overall performance was disappointing, primarily because of the interventionist policies of the Government. 2.07 The first project to address an entire sub-sector on a national scale was the Cotton Development Project (Cr. 1433-MAG), approved in 1983. Around the same time, IDA started to support the agricultural institutions through the Agricultural Institutions Technical Assistance Project (Cr. 1249-MAG; 1982) and its follow-up, the Second Agricultural Institutions Development Project (Cr. 1709-MAG; 1986). Sub-sector investment lending continued with the 1985 Irrigation Rehabilitation Project (Cr. 1589-MAG), the 1987 Second Agricultural Credit Project (Cr. 1804-MAG), the 1988 Forests Management and Protection Project (Cr. 1878-MAG), and the 1989 National Agricultural Research Project (Cr. 2041-MAG). 2.08 The adjustment process in the sector was started through the Agricultural Sector Adjustment Credit (Cr. 1691-HAG), wh.ich supported the liberalization of rice marketing through the suppression of the monopsony of state rice marketing agencies and the liberalization of the edible oils and agricultural inputs markets. In 1988, IDA financed the Public Sector Adjustment Credit (Cr. 1941-MAG), which sought to remove administrative obstacles and controls over the export trade. including the elimination of state trading monopolies for traditional export crops, and the restructuring and/or elimination of sector public enterprises.  - 4- E. IDA Strategy for Agriculture 2.09 IDA's strategy in the sector has two main themes: (i) continuation of the macro-economic and sectoral adjustment proce(s; and (ii) a sectoral investment program fo-ussing on food security, llvestocs development, irrigation, agricultural extension and research, and environmental protection. Operations in aach of these areas are presentloy ongoing or under preparation. F. The Initial Pilot Activities 2.10 In October 1987, a Bank mission visited Madagascar at the invitation of the Governmerf to identify an agricultural extension project. Subsequently, a PPF in the amount of US$750,000 was approved in May of 1988 and, in September 1988 a staff member of the Bank was assigned to the World Bank's Regional Mission in Madagascar, to advise the Directorate for Agricultural Extension (DVA) on the implementation of the various activities. The latter included reorganizing DVA's extension services, training staff at all levels, strengthening the links with research and initiating appropriate monitoring and evaluation (M&E) routines. The strategy adopted takes Madagascar's specific conditions 4nto account by, inter alia: (i) dealing with groups of farmers instead of with contact farmers; (ii) gradually suppressing the supervisors' level (CS), as is already the case in the IFAD financed Highland Rice Project (IFAD Cr. 118- HAG) ; and (iii) taking advantage of the high literacy rate of 60-80Z in rural communities in part of the project area by investigating the possible use of pamphlets and radio broadcasts as a means of reaching more farmers while reduc.ng the number of Vi;StS. 2.11 During project preparation, it became apparent that a national project was premature in light of the organizational, managerial, technical and financial problems plagueing the DVA. Moreover, further testing was required of the adaptability, replicability and sustainability of the proposed extension strategy in the Malagasy context, which included several donor and IDA (Cr. 1337-MAG) financed projects each with its separate extension structures. It was therefore decided in consultation with the Government that the project would serve as a pilot operation to build up experience and improve the capability of the DVA. A follow-up project would only be considered if warranted by the project review at the end of year 2 (para 3.42). 2.12 Due to a variety of reasons, including extensive procurement delays of much needed equipment, cumbersome procedures in obtaining day- to-day operational funds, and late reussignment of staff, the pilot activities did not take off in earnest until April 1989. However, definite improvements are now visible in both work organization and management. Maps and programs of visits are currently used by most agents. Working with groups of farmers has become the dominant vehicle of extension as opposed to the earlier used method of visits to individuals. Ability tests were taken by all field staff in the three pilot CIRVAs in February 1989. Based on the results of these tests, bi-monthly training sessions of extension agents have been instituted and are adhered to. Monthly workshops for Subject Matter Specialists (Techniciens Formateurs or TF)  with the participation of research are becoming more regular. Area specific technical messages, covering land preparation, fertilization, planting material and planting methods, weeding and plant protection were extended to farmers for irrigated rice, cassava, maize, potatoes, groundnuts, wheat and a few vegetables. In order to continue progress, the CIRVAs stressed the need for additional technical training of extension staff in soil and water management, and in watershed pittection methods. They also bighlighted the importance of closer links with research, especially in areas of testing and plant protection. 2.13 According to surveys conducted by the M&E unit of the DVA in 1988 and during the first half of 1989, there were some 225,800 farmers and 397 extension agents (VA) in the three CIRVAs considered. Some 10 to 15% of the farmers were reached by extension agents (VA) with ratios varying between 25 and 195 farmers per VA. The latter spent 7 days per fortnight in the field visiting an average of two groups of farmers per day. These groups numbered on average 10 to 12 farmers, with attendance varying significantly during the cropping season with lows of 2 to 5 participants. It was estimated that some 50% of the farmers organized in groups and 35% of the not yet organized farmers were aware of the VAs work program. During the 1988/89 season, 6,194 small adoption plots were implemented by farmers. The main on-season crops for which technical recommendations were extended to farmers included rice, maize, groundnuts and in sowe areas cassava. The messages extended emphasized the need to produce more and better quality composts, the selection of seed and planting material, cultural practices and anti-erosion measures. The M&E unit was unable to record sufficient yield results to be meaningful, due to a lack of means. While in the field, in September 1989, the mission observed that where the technical messages had been properly implemented, marked improvements in winter season crops (wheat, potatoes and vegetables) were visible. 2.14 Field extension personnel has shown a high degree of commitment despite field allowances not having been paid for over a year. This was the result of a dispute between the DVA and the Ministry of Finance, over the level of field allowances. The latter were finally paid in November 1989 retroactively to April 1988. At negotiations, assurances were obtained that adequate field allowances (compatible with the PEP) would be paid for DVA staff needed for project execution 31 and that the corresponding budgetary allocations would be made. To smoothen the transition period it was also agreed that incremental field allowances would be financed by IDA on a decreasing basis (100% in 1990 and 50% in 1991), with Government providing full funding in 1992. A second factor affecting field staff performance is delayed payment of salaries to contractual extension workers 4/ during the first 2-3 mon .s of the year, just before the April planting season starts. As a result extension activities during this part of the year suffer. MINAGRI agreed at negotiations, to take the necessary steps to ensure that funds are 3/ Extension personnel outside of the project will not yet receive an increase in the level of flild allowances, because they have only very limited access to transportation and lack adequate training. 4! This affects over half of all the VAs.  - 6 - available, either as part of the government funded accounts, or as a revolving fund, from which these salaries can be paid during the first quarter of the year. III. THE PROJECT A. General 1. Proiect Obiectives 3.01 The project would serve as a pilot to test an extension strategy with particular emphasis on the specificity of the Malagasy context, sustainability in light of the country's current economical constraints, and possible future replicability nationwide. The project would therefore initiate the process of working out and implementing a viable long-term action plan for agricultural extension. The project would strengthen Government extension services, promote general principles of sound field management to ensure efficient and cost effective extension and harmonize the various approaches adopted by other dor.ors and non- governmental organizations. The strengthening of Government extension services would mainly entail improvir staff competence through regular training, gradually decreasing the overall number of personnel whenever warranted, keeping lean teams of competent staff at the MINAGRI and, involving the provincial levels (SPVA) in the development process. The harmonizing of the various approaches to extension would not be synonymous with standardisation. Rather, it should provide the opportunity to select the most cost effective approaches as results become avAilable, on the basis of an in-depth evaluation to be carried out at the end of year 2 (para 3.42). 2. Summary Project Description 3.02 The project would be implemented over a three year period and would include: (i) reorganizing and strengthening MINAGRI's agricultural extension services at the DVA level and in the CIRVAs of Ambatondrazaka, Antananarivo, Miarinarivo, I-loramanga and Amparafaravola; (ii) systematic in-service training of all project extension staff; (iii) improving the links between extension and research through adaptive research and on-farm trials, with the active participation of farmers; (iv) strengthening the M&E capacity at both the DVA and the CIRVA level; (v) introducing improved financial management for the DVA and the CIRVAs; (vi) strengthening the organization and management capacity of the DVA; and,  - 7 - (vii) providing for several specific studies, an in-depth review of the project at the end of year 2, and for the preparation of a follow-up project. B. Detailed Features 1. Extension Services 3.03 The project would continue to improve the organization and increase the efficiency of DVA's extension services in the three CIRVAs covered by the pilot actions started in 1988. The CIRVA of Ambatondrazaka (Lac Alaotra area), which has already made progress in improving its extension services alonZ similar principles of extension as the above three CIRVAs through funding from other sources, would be included in the project. At the request of Government, a yet to be created CIRVA (Amparafaravola), would be added to the project. The Government plans to create this CIRVA within a year as part of the restructuring of the SOMALAC parastatal. The present staff of the extension services of SOMALAC would be transferred to the new CIRVA. A number of issues would need to be resolved prior to the inclusion of this CIRVA in the project, especially as to personnel levels, the readjustment of salaries and allowances, and task allocations. Specific studies have been commissioned to address these problems and IDA and CCCE are working in close collaboration in this matter. It would be a condition of disbursement that (i) the studies have been concluded; (ii) Government and IDA have agreed on the recommendations; (iii) the CIRVA of Amparafaravola has been created; and the Government funded account has been established (para 3.25). 3.04 The present coverage ranges from 25-195 farmers per VA. An indicative ratio of 1 VA per 200 to 250 farmers would be aimed at from project inception. The ratio would be area specific and adapted during the project taking into account such parameters as population densities, settlement patterns, types of farming, road conditions, and communication facilities. The overall percentage of farmers visited would increase during the project from 15% to about 40%, without modifying the current number of VAs. It is thus expected that the total number of farmers directly re&ched by extension would be about 125,000, and an undetermined number through spill-over effects. As to the supervisors (CS) the present ratio of 1 CS per 2 or 3 VAs in some areas is unacceptably high and would be reduced to at least 1 CS per 7 to 10 VAs. In areas where it would not be possible to modify this ratio because of geog-:aphic conditions (low population density, poor road network, etc.), the CS would also perform part-time extension activities. Finally, there are currently some 30 Subject Matter Specialists (TF) mainly in the fields of general agriculture and plant protection, in the five CIRVAs considered. The total number of TFs which is oni the high side would not be increased during this three- year phase. Instead, there is a definite need to up-grade their levels cf competence and to specialize them in fields of priority to farmers, i.e. vegetable and tree crops production, and anti-erosion and watershed management. They would also be used in the on-farm trials program which require close supervision by technical personnel. As the TFs become more competent, their average number -Br CIRVA would need to decrease in subsequent project phases. This could be achieved through transfers to other CIRVAs, that are in need of trained TFs.  - 8 - 3.05 Sixteen low-cost guesthouses would be build in the CIRVAs for the purpose of training (para 3.08) and to facilitate supervision *y headquarters and CIRVA staff. They will relieve the burden caused by inadequate lodging facilities in rural areas. 3.06 Unqualified field staff is a major problem affecting the cost- effectiveness and credibility of the extension service. Following the second competency test (December 1989), a plan for the gradual retirement or redeployment of unqualified staff is being developed by Government for submission to IDA. The submission would also include proposals on filling vacancies resulting from this exercise with qualified and competent staff. Agreement between Government and IDA on this plan would b;. a condition of credit effectiveness. 2. Training 3.07 Insufficient training z.i staff remains one of extension's main weaknesses. This is a major issue which can be traced back to the country's inadequate agricultural education system currently under review. It is essenzial that extension staff be better qualified. The project would therefore provide training to all staff. Training for field personnel would be essentially practical and have strong emphasis on technical knowledge, communications skills, feedback responsibilities and the ability to diagnose farmers problems, constraints and aspirations. The various training needs would be determined on the basis of updated job descriptions, profiles, performance and ability tests. Provision has been made for short-term consultancies by a local Human Resources Specialist. The latter would contribute to the updating of staff job descriptions, and advise on career development and incentives and, together with the DVA training department, would check the content of the various training sessions and refresher courses as well as the objectives of seminars and study tours, to ensure they properly address the staff training requirements. The training department would also examine the potential for multi-media support to the extension service. 3.08 Training would consist mainly of: (a) bimonthly one-day- sessions for the VAs conducted by the TFs; (b) monthly workshops for the TFs led by Research and/or any source of knowledge including the University and the private sector; and (c) refresher courses and special training sessions for field staff at all levels. To minimize costs, seminars would be reduced in number and duration and, visits overseas would be the exception. The project would also provide for the training of the accountants, monitoring and evaluation agents, and staff involved in adaptive research. Details of the training component are in Working Paper 1. The project would finance: (i) the rehabilitation and/or construction of three training centers to be equipped with audio-visual and office equipment; (ii) training rooms as part of small sector offices; (iii) guest houses; and (iv) incremental operating costs.  -9- 3. Research 3.09 The project would put particular emphasis on strengthening the current informal and weak links between extension and research. This comes at a time when a first seven year phase National Agricultural Research Project (NARP), co-financed by IDA, is being launched. Besides addressing the specific aspects of Research, NARP would improve the linkages with agricultural extension through: (a) joint field diagnosis; (b) participation of researchers in the training of TFs during monthly workshops; and (c) implementation of on-farm trials jointly with extension. ;.10 Adaptive research programs would be prepared and discussed annually by the regional research-extension committees to be set-up under NARP. The specifics of these programs would be derived from joint field diagnosis by research, extension and farmers. Adaptive research would be confined to on-faLm trials and designed to adjust general recommendations from applied research to different farm situations including: (a) cropping sequences and mixes; (b) level of farmers' skills, and (c) farmers' resources and land tenure status. The main contribution of the adaptive research programs would be to provide farmers with alternatives to choose from. 3.11 The proposed project would provide funding for short-term adaptive research programs by FOFIFA in collaboration with the extension services in the project area. The programs would concentrate on specific areas of concern to farmers and where a definite impact on production and yields can be expected. In addition, particular attention would be gi7en to the loss of soil fertility, soil degradation and to watershed management. The project would have a major effect through better trained staff regularly advising the rural communities on environmental issues. In this respect there is a variety of aspects for which techniques would be developed including: anti-erosion measures such as alley cropping, contour ploughing and planting, zero tilling, crop rotations, the use of compost and of manure, cultural practices for tanety 5/ and for bottomland, and the use of agricultural tools and equipment to enhance labor efficiency. Basic animal husbandry and agro-forestry would also be sources of recommendations on the assumption that extension would have a multi-disciplinary approach. This important aspect would require careful attention during the pilot so as to reach a common understanding with the MPAEF. Details of the research component are in Working Paper No. 2. 3.12 The recommendations to be extended in 1990 have been selected with extensive participation of farmers and technical support by FOFIFA. They are closely related to farmers requests for advice on watershed management, erosion control, the growing of off-season crops, and to demands for training on Farmers Associations type of activities including seed production and marketing. 3.13 Funding for the short-term adaptive research program would, under contract with FOFIFA, be limited to the purchase of farm inputs since the researchers salaries, allowances and vehic7es are provided for by the 5/ Hillside lands, usually of poor quality, surrounding bottomlands.  - 10 - National Agricultural Research Project. The DVA and FOFIFA entering into an agreement for short-term adaptive research, acceptable to IDA, would be a condition of credit effectiveness. Given the fact that FOFIFA does not have the staff capacity to fully participate in the adaptive research component, implementation as well as the collecting of most data would be the responsability of the extension service. FOFIFA would limit itself to preparing protocols and, together with the DVA, interpret and analyze resulcs. The efficiency of this arrangement would be assessed as part of the annual renewal of the agreement. Similarly, the efficiency and the appropriateness of the adaptive research program and the on-farm trials would be assessed during the in-depth review at the end of year 2 (para 3.42). 4. Monitoring and Evaluation 3.14 The monitoring and evaluation division of the DVA would be strengthened to provide the director of the DVA and the heads of the CIRVAs with timely information on the effectiveness of the extension service with emphasis on monitoring project progress. The present M&E system comprises a central unit within the DVA and regional units at the CIRVA level. The project would provide for 6 man-months of technical assistance and training of key staff in order to assist the central unit in standardizing concepts, harmonizing actions, supporting the steff of the regional units, and supervising and analyzing the final results. It would be key to determining the project's impact and the future role of extension. In this respect, the feedback from farmers would be an important factor in determining the performance of field extension agents and the appropriateness of the technical messages. From project inception, the central unit, with the assistance of the consultant, would define the type, amount, purpose and timing of data and information to be collected. It would also organize -raining for staff in the regional M&E units and provide support to these units. Furthermore, it would be responsible for continuity in the collection of data series requiring extended time periods. Details of the M&E component are in Technical Paper No. 3. 5. Financial Management System 3.15 The Administrative Bureau (BAF) within the DVA is responsible for financial and personnel management. The section responsible for financial management, however, is staffed with employees that functionally belong to other divisions within the DVA and work only part-time for the BAF. Furthermore, in the absence of a head of the fir.ancial section, the Director of the DVA must also act as head of the BAF. Under the project the BAF would be reorganized and a full-time head with staff would be assigned. The latter would be a condition of effectiveness. Local technical assistance would be provided by a management specialist to aid in the definition of the role and management of the central BAF at the DVA and the BAFs at the CIRVA level. Further technical assistance would be provided periodically by a financial specialist, who would assist the central BAF and the project CIRVAs in defining accounting and procurement procedures and strengthening local capabilities for increased financial autonomy in the five project CIRVAs. The financial specialist would also participate in the quarterly supervision of the BAFs at the project CIRVAs.  - 11 - 3.16 The BAF at the DVA would be equipped with a micro computer and receive several man-months of training each year in the use of this computer and standard software. The computer would be used to aitomate accounting procedures and to allow for improved management information. 6. Studies 3.17 The project would provide for a study fund (US$ 400,000) to finance several studies to be carried out. in conjunction with the preparation of a follow-up phase. Without presetting the exact studies to be launched, the following topics could be envisaged: (i) the phased transfer of extension responsibilities from parastatal regional authorities and autonomous projects to the DVA; (ii) future role of the Provincial Agricultural Extension Services (SPVA); (iii) the effect of land tenure on the adoption rate of technology by farmers; (iv) women's role in agriculture and how to increase their participation to the development of the agricultural sector; (v) farmer groups in the sociological context of Madagascar and their possible involvement in various support services to agriculture; and (vi) farmer awareness of pesticides risks and origin of supplies. In addition, one or more studies would be undertaken to assess the feasibility of alternative means of financing (including cost recovery mechanisms) and providing extension (private sector and/or NGOs). Specifically environmental studies would be carried out under the Environment I Project. 3.18 Following the studies on alternative means of financing and providing extension, the project would provide up to US$100,000 for a test program. 3.19 Assurances were obtained at negotiations that terms of Reference, cost estimates and arrangements for executing the studies would be prepared by DVA and submitted to IDA for approval. 7. Management Strengthening 3.20 Strengthening management within the DVA is a necessity. DVA's director has several functions and many tasks to attend to, and would need a deputy, with a clear definition of responsibilities, to oversee the operation of the extension system in the 5 project CIRVAs. A project coordinator has been selected and agreement was reached during negotiations on his terms of reference. There is also a need to arrive at a better delineation of responsibilities between the central divisions and a clearer tasl. description for each of them. The Human Resources Specialist (para 3.07) would assist the Director in this task. 3.21 Similarly, strengthening of management at the CIRVA of Antananarivo is vital, if extension service quality is to be improved. This CIRVA covers a vast area, which cannot be adequately managed by a CIRVA head alone. A deputy to the head of the CIRVA would therefore be nominated who would supervise the implementation of the project in outlying areas.  - 12 - 3.22 With the above arrangements, and with the identification and action program on unqualified staff (para 3.06), DVA management would be able to carry out the pilot project. However, as part of the preparation of the follow-up project on National Extension, othez organizational issues will have to be considered, such as the decentralization of the DVA, the role of the SPVAs, the role of regional authorities with separate extension responsibilities, and the issue of redundant personnel. These issues cannot be adequately addressed now as their solutions depend to a large extent on the results of the pilot project. Studies to examine these issues are, however, funded under the proposed project (para 3.17) C. Proiect Costs and Financing 1. Cost Estimates 3.23 The costs total US$5.41 million over a three year period and include taxes and contingencies. The foreign exchange component is 46% and taxes and duties are 16% of project cost. Base costs were calculated using prices and exchange rates in effect at appraisal in September 1989. Taxes and duties have been applied on all costs at prevailing rates. Physical contingencies have been applied at the following rates: Civil works 10% Agricultural 10% and office supplies Price contingencies have been applied at the following rates: 1989 1990 1991 1992 Local 9.5 8.5 8.0 7.0 Foreign 3.6 4.3 4.3 4.3 Costs are summarized in Table 3.1 below. The full detailed cost tables are in Working Paper 4.  - 13 - Table 3.1: Proiect Costs 6/ (USS '666) X Total -------