Madagascar - Agricultural Extension Pilot Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15396 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (CREDIT 2150-MAG) MARCH 4, 1996 Agriculture and Environment Division Central Africa and Indian Ocean Department Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) CURRENCY EQUIVALENTS End of Period Figures Malagasy Francs (Fmg) Year per SDR per US$ 1985 698.4 635.8 1986 941.6 769.8 1987 1751.0 1234.3 1988 2054.1 1526.4 1989 2014.0 1532.5 1990 2085.4 1465.8 1991 2621.5 1832.7 1992 2626.5 1910.2 1993 2695.8 1962.7 1994 5651.2 3871.1 June 1995 6920.1 4411.2 Sources: International Financial Statistics, 1994 and February 1995 Note: the Fmg started floating on May ], 1994 WEIGHTS AND MEASURES Metric System GOVERNMENT OF MADAGASCAR FISCAL YEAR January I - December 31  FOR OFFICIAL USE ONLY ABBREVIATIONS AND ACRONYMS AMRT Monthly Technology Review Workshop (Atelier mensuel de revue de technologies) AUE Water Users Association (Association des Usagers de l'Eau) AVB Field Extension Agent (Agent de Vulgarisation de Base) CG Contact Group (Groupe de Contact) CS Field Supervisor (Chef de Secteur) CIRAGRI Agricultural District (Circonscription de l'Agriculture) DA Agriculture Directorate (Direction de l'Agriculture) DNVA National Directorate for Agricultural Extension (Direction Nationale de la Vulgarisation Agricole) DP Demonstration plot (Parcelle de Demonstration) DTI Import duties and taxes (Droits et taxes ai l'importation) EPVEF Livestock, Plant Protection, Water and Forestry Resources Directorate (Direction de l'Elevage, de la Protection des Vegetaux et des Eaux et For2ts). FAFIALA Environmental Development and Management Agency (Organisme de Developpement et de Gestion de l'Environnement) FIFAMANOR Malagasy-Norwegian Agricultural Development Agency (Organisme Malgache-Norvegien de Developpement Agricole) FOFIFA National Agricultural Research and Rural Development Center (Centre National de Recherche Agricole et du Developpement Rural) FQ Fortnightly training (Formation de Quinzaine) M&E Monitoring and Evaluation MINAGRI Ministry of Agriculture, Livestock, Water and Forestry (Ministere de l'Agriculture, de l'Elevage et des Eaux et Forets) NARP National Agricultural Research Project (Projet National de la Recherche Agricole--PNRA) PE Environmental Project (Projet d'Environnement) PIP Public Investment Program (Programme d'Investissements Publics) PNVA National Agricultural Extension Project (Projet National de la Vulgarisation Agricole) PPI Irrigation Schemes Rehabilitation Project (Projet de Rehabilitation des Perimetres Irriguis) PPVA Pilot Agricultural Extension Project (Projet Pilote de la Vulgarisation Agricole) PROVAGRI Agricultural Province (Province de l'Agriculture) PSE Livestock Sector Project (Projet Sectoriel Elevage) TMP Field Trials (Tests en Milieu Paysan) TTL Local Transactions Tax (Taxe sur les Transactions Locales) This document has a restricted distribution and may be uscd by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) TABLE OF CONTENTS PREFACE ...................................................................... EVALUATION SUMMARY .....................................................................; A. Project Objectives .................................................................... i B. Implementation Experiences and Results .................................................................... i C. Summary of Findings, Future Operations & Key Lessons Learned ................................................................... ii PART 1: PROJECT IMPLEMENTATION ASSESSMENT .................................................................... I A. General .................................................................... I B. Statement and Evaluation of Project Objectives .................................................................... 2 C. Achievement of Project Objectives .................................................................... 4 D. Implementation Record and Major Factors Affecting the Project .................................................................... 7 E. Project Sustainability ..................................................................... 9 F. IDA Performance .................................................................... 9 G. Borrower Performance .................................................................... 10 H. Economic Evaluation of the Project .................................................................... 10 L. Assessment of Project Outcome .....................................................................11 J Future Operations .................................................................... 12 K. Key Lessons Learned .................................................................... 12 PART I1: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE .................................................................... 14 A. Project Implementation .................................................................... 14 B. Project Results .................................................................... 15 C. Borrower Performance .................................................................... 16 D. Outlook .................................................................... 17 PART III: STATISTICAL ANNEXES .................................................................... 18 TABLE 1: SUMMARY OF ASSESSMENTS .................................................................... 18 A. Achievement of Objectives ......... ................................................................... 18 B. Project Sustainability ............................................................................ 18 C. Bank Performance ............................................................................. 19 D. Borrower Performance ............................................................................ 19 E. Assessment of Outcome ............................................................................. 19 TABLE 2: RELATED BANK LOANS/CREDITS .................................................................... 20 TABLE 3: PROJECT TIMETABLE .................................................................... 21 TABLE 4: ESTIMATED AND ACTUAL ANNUAL AND CUMULATIVE DISBURSEMENTS .................................................... 21 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION .................................................................... 22 TABLE 6: KEY INDICATORS FOR PROJECT OPERATION .................................................................... 22 TABLE 7: STUDIES .................................................................... 23 TABLE 8A: PROJECT COST .................................................................... 24 TABLE 8B: PROJECT FINANCING .................................................................... 25 TABLE 9: ECONOMIC COSTS AND BENEFITS .................................................................... 25 TABLE 10: COMPLIANCE WITH CREDIT CONDITIONS .................................................................... 26 TABLE 11: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS .................................................................... 27 TABLE 12: BANK RESOURCES -- STAFF INPUTS .................................................................... 27 TABLE 13: BANK RESOURCES-- MISSIONS .................................................................... 28 ANNEX A: SUMMARY OF AUDIT RECOMMENDATIONS .................................................................... 29 ANNEX B: CALCULATION OF ERR .................................................................... 31 MAP: IBRD NO. 21993R .................................................................... 37   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) PREFACE 1. This is the Implementation Completion Report (ICR) for the Agricultural Extension Pilot Project in Madagascar, for which Credit 2150-MAG in the amount of SDR 2.8 million (US$3.7 million) was approved by the Board of Directors on May 31, 1990 and made effective on March 22, 1991. The credit was closed on June 30,1995. Ninety percent of the credit was disbursed, and the undisbursed balance (SDR 238,164) was canceled on January 15, 1996. 2. The ICR was prepared by Ousmane Seck, Connie Luff and Christopher Trapman of the Agriculture and Environment Operations Division of the Central Africa and Indian Ocean Department of the Africa Regional Office (Preface, Evaluation Summary, Parts I and III) and reviewed by Nils Tcheyan, Division Chief AF3AE and Padmanabha Hari Prasad, project adviser, Central Africa and Indian Ocean Department. This report is based, inter alia, on the Staff Appraisal Report, the President's Report, the Development Credit Agreement, project supervision and progress reports, correspondence between the Bank and the Borrower, and internal Bank memoranda. The Borrower contributed to preparation of the ICR by preparing its own evaluation of the project's execution (Part II).   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) EVALUATION SUMMARY A. Project Objectives 1 . The main purpose of the project was to test a sustainable extension strategy in six pilot districts, covering the entire range of agricultural conditions found in Madagascar. The project was designed to focus on the reorganization and training of the extension service, and to establish closer links between extension and research. The objective was to define a cost- effective approach to improve the quality of field services to provide the effective transfer of agricultural technology to farmers, aimed at substantially and sustainably raising the production and yields of Madagascar's major crops. If successful, this system was to be replicated in the remainder of the country. 2. The project objectives were consistent with IDA's strategy for agricultural development that included support of the macro-economic and sectoral adjustment process, as well as a sectoral investment program that focused on food security, extension and research, livestock development, irrigation and environmental protection. The project preparation team correctly ascertained that a national project was premature given the political instability and the difficulties of the National Directorate for Agricultural Extension (DNVA), and redesigned the project as a pilot operation. Project objectives were well-defined, although only limited indicators were established in the appraisal report to evaluate the effectiveness of the extension service. Subsequently a systematic monitoring system to measure the level of farmer participation and the adoption rate of technical messages was perfected during project implementation, and was adopted for the follow-on National Agricultural Extension Program Support Project (Credit 2729-MAG). B. Implementation Experiences and Results 3. Project Results. Although the first 2 years of project implementation were very slow, the project achieved its main objectives. The links between the research and extension services were strengthened; the farmers' priority extension needs were identified and addressed more precisely than in the past, and the organization and efficiency of the extension system significantly improved. The project demonstrated that a large number of farmers could obtain good results at low cost with existing agricultural staff, systematic work programming, regular training and joint diagnosis of results by farmers and field staff. Major increases in production and yields were achieved without any cash investment, through use of techniques adopted directly by farmers. The proportion of farmers participating in organized fortnightly visits by front level staff (AVBs) rose substantially, in both percentage terms (from under 5 to over 20  ii percent) and absolute terms (from about 10,000 to about 40,000). Over 120,000 farmers are estimated to have benefited from the services provided, reflecting the impact of promoting the establishment of contact groups and improving the scheduling of farm visits. A total of 16,000 demonstration plots were set up in 1994/95 compared to 9,000 in 1992/93. The project achieved its objective of identifying a cost-effective system of agricultural extension, adapted to local conditions in Madagascar. 4. Project Implementation. Pilot extension activities were initiated in 1988 in three extension districts financed by a Project Preparation Facility of US$750,000. The Project was appraised in September 1989, the Credit (US$3.7 million) was approved in May 1990 and became effective in March 1991. The first two years of project implementation were seriously disrupted by political turmoil and civil service strikes. Numerous other constraints exacerbated this situation, including: (a) difficulties in providing counterpart funds to pay taxes and contractual staff salaries; (b) problems in selecting and placing competent and motivated staff to form the operational extension service; (c) slowness in procuring equipment and vehicles; and (d) lack of a commitment to implement an efficient extension management system. 5. However, major improvements in project performance were achieved in 1993 following the replacement of key staff to positions of responsibility at the central (DNVA) and local (CIRAGRI) levels and improved borrower commitment to the implementation of project objectives. In order to consolidate and properly evaluate the results of the pilot phase, the Bank agreed to extend the project closing date by one year until June 30, 1995. During the project period, an improved extension management system linked with tested principles of extension was introduced in six of the country's 28 districts, affecting a total farming population of 260,000. The experience and lessons of this project have served as the basis for the follow-on National Agricultural Extension Program Support Project (Credit 2729-MAG) that became effective in November 1995. C. Summary of Findings, Future Operations & Key Lessons Learned 6. Project Outcome. The results of the pilot project were assessed to be satisfactory. Although initially project implementation was slow, the project achieved its main objectives. The effectiveness of extension services has improved through fortnightly training of extension staff accompanied by regular visits with contact groups by both extension and research staff. Production and productivity of main season irrigated and rainfed rice and of interseason crops increased, with the introduction of relatively simple technologies, many of which required little or no cash investment. Net farm incomes using improved practices increased by almost US$400 per ha. Monitoring protocols and indicators were established to periodically measure the field impact of project activities. Through demonstration of simple, low-cost technologies on mini- plots, more and more farmers are becoming convinced of the validity of the recommendations presented and showing increased interest in the program. Smoother progress towards project achievements could have been obtained had there been a better programming of extension and  iii training activities in the early years and if other donors and associated projects in extension and research had been more closely involved in the implementation of the pilot project. 7. Future Operations. Based on the encouraging results posted during the pilot project after 1993, the Malagasy authorities decided to implement a first four-year phase of a longer-term program designed to consolidate and extend the pilot project experiment to the entire country. The Agricultural Extension Program Support Project (Credit 2729-MAG) will support this effort, and additional Bank funding for extension could be considered if the program is successful and remains a priority in the Country Assistance Strategy. The national agricultural extension program will be introduced to the country's 28 extension districts over a four-year period. 8. The principal lessons learned from this project can be summarized as follows: (a) The project demonstrated the feasibility of significantly raising the effectiveness of agricultural extension management in Madagascar through the combined application of basic management principles and the use of learning tools, such as regular training and individual in-field demonstration plots; (b) The project demonstrated that, for the improvements to materialize, the extension and training personnel must receive a minimum of material and financial support, as well as a uniform system of benefits and allowances for field staff. This support essentially consists of the provision of means of travel and operating equipment and supplies, particularly fuel, defrayal of training and field visit costs, and provision of office equipment and supplies. This implies the need for full Government commitment, not only to secure counterpart funding, but also to ensure the sustainablility of the project; (c) The feasibility of providing these resources on a regular basis depends, within the overall budgetary and treasury constraints, on the Government's ability to rationalize public investment in the sector and to coordinate donor assistance. To that end, the Government has developed a national strategy for extension services and is trying to integrate existing and future donor-funded projects into this strategy; (d) Finally, expansion of the pilot project experience was greatly facilitated by a clear Governmental policy on the provision of cost effective and relevant farmer extension services that depend on the coordination and harmonization of independent field approaches and the development of cooperation among the various entities involved (projects, NGOs, private sector). (e) The impact of intensive supervision from a full time specialist in the resident mission only really made itself felt with the appointment of a fully experienced  iv specialist in agricultural extension systems. The experience underlines the critical importance of the selection of the appropriate candidate for such a post.  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) PART 1: PROJECT IMPLEMENTATION ASSESSMENT A. General 1. Madagascar has considerable potential for agricultural growth due to its low labor costs, good soils and climate, an underutilized irrigation infrastructure, as well as the availability of diverse ecological zones. A variety of temperate and tropical crops may be grown for both domestic and export markets. Crop yields are uniformly low, creating strong potential to increase yields on the basis of known technologies. However, Madagascar remains one of the world's poorest nations. Inadequate economic policies since independence in 1960, coupled with population growth of about 3 percent per annum, have resulted in the persistence of low per capita income (US$230 in 1994). 2. Agriculture is the most important contributor to the national economy. Rice is the country's main staple and the major exports are coffee, vanilla, cloves and shellfish. In 1987, when the Agricultural Extension Pilot project was prepared, the sector employed more than 70 percent of the active population and accounted for nearly 35 percent of GDP. Despite favorable conditions for the development of most farm products, the sector's annual productivity growth has remained relatively low (under 2 percent from 1986-1991). As a result, and with a steadily growing population, Madagascar has been unable to produce enough to feed its 12 million people and to benefit significantly from its comparative advantages in order to increase exports. 3. Sustained sector growth has been severely hampered by the combined effects of a number of major constraints. These have included: (a) inappropriate macroeconomic and sectoral policies; (b) an unsuitable legal, regulatory and fiscal framework; (c) poor agricultural services (research and extension) for promoting the transfer of technology; (d) the lack of reliable transport and communication infrastructure linking farmers with markets; and (e) an agricultural credit system ill-suited to the constraints and requirements of the rural sector. 4. The Government has sought to address these constraints, in collaboration with the Bank and other international development partners, through economic policy reform and sectoral investment programs. The Bank's lending portfolio in the agricultural sector includes support for six major projects (Research, Extension, Livestock, Environment, Rural Finance and Irrigation Rehabilitation) that are designed to support increased agricultural productivity on a sustainable basis. The Government is also developing, with Bank support, strategies for strengthening rural infrastructure, focusing on feeder roads and market improvements. A key element of the Government's rural development strategy is the establishment and effective operation of a professional agricultural extension service as a means for increasing incomes of rural farmers and raising agricultural productivity.  2 Part I. Project Implementation Assessment B. Statement and Evaluation of Project Objectives 5. Project Objectives. The main purpose of the project was to test a sustainable extension strategy in six pilot districts (North Tana, South Tana, Miarinarivo, Antsiranana, Mahajanga and Amparafaravola), covering the entire range of agricultural conditions found in Madagascar. The project was designed to focus on the reorganization and training of the extension service, and to establish closer links between extension and research. The objective was to define a cost-effective approach to improve the quality of field services to provide the effective transfer of agricultural technology to farmers, aimed at substantially and sustainably raising the production and yields of Madagascar's major crops. If successful, this system was to be replicated in the remainder of the country. 6. Specifically, the project was to test the feasibility of significantly raising the effectiveness of agricultural extension management in Madagascar through the combined application of certain basic management principles and learning tools. These principles and tools included: a structured and transparent organization based on cost-effective farmer/extension agent and training ratios and clearly defined responsibilities at all levels of the hierarchy; a group approach, so as to reach and train a larger number of farmers; establishment of fixed work and farm visit timetables; systematic use of demonstrations to enable each farmer to judge the relevance of the technology offered and the suitability of adopting it; and close cooperation between researchers, farmers and extension agents, in diagnosing constraints and drafting technical recommendations incorporating solutions adapted to local conditions. 7. The project aimed to increase the ratio of Field Extension Agents (AVBs) to farmers from a range of 25-195 farmers per AVB to a range of 200 to 250 farmers per AVB. This ratio was to be area specific and adapted during project implementation to account for factors such as population densities, settlement patterns, types of farming, and road conditions. The overall percentage of farmers visited was to increase during project implementation from 15 percent to about 40 percent, without increasing the number of AVBs. In addition, the ratio of supervisors to AVBs was judged to be high, and the project aimed to reduce this ratio to one supervisor per 7-10 AVBs (SAR para. 3.04). 8. The project investments, based on tested principles of extension management organization, were to support (SAR para. 3.02): (a) reorganization and institutional strengthening of the extension service of the Ministry of Agriculture, Livestock, Water and Forestry (MINAGRI). This reorganization was to involve the delineation of a single line of command, with the AVB as the sole source of information to farmers, who otherwise often received contradictory messages;  Part I: Project Implementation Assessment 3 (b) systematic, on-going training of extension staff, monthly technology review workshops among researchers, farmers and technical specialists and fortnightly training sessions for AVBs; (c) development of close links between extension and research through adaptive research and on-farm trials, with the active participation of farmers, in order to diagnose constraints, establish priorities, formulate valid recommendations and evaluate results; (d) strengthening of monitoring and evaluation capacity at both the central (DNVA) and local (CIRAGRI) levels to provide timely information on the effectiveness of the extension service with emphasis on monitoring project progress. Feedback from farmers would serve as the basis to evaluate the performance of AVBs and the appropriateness of the technical messages; (e) provision of funds for studying ways to reduce the Government's costs of extension by more actively involving the private sector and the NGOs, exploring means to lower the costs of the National Directorate for Agricultural Extension (DNVA), for example by reducing excess personnel, and reviewing the possibility of cost recovery mechanisms; (f) strengthening of financial management; (g) improvement of the DNVA's organizational and managerial capacity. 9. Evaluation of Objectives. At the time of project preparation, a Government analysis of the performance of the agricultural sector during the 1980's revealed that production and productivity were declining while population was growing, leading to unsustainable farming practices and environmental degradation. The SAR attributed this poor performance to Government's interventionist policies in pricing and marketing, high export taxes and a decline in quality and coverage of its services to the sector (extension, research and irrigation), but recognized that the Government had made a series of policy changes to improve the macro- economic environment. The project objectives were consistent with IDA's strategy for agricultural development that included support of this macro-economic and sectoral adjustment process, as well as a sectoral investment program that focused on food security, extension and research, livestock development, irrigation and environmental protection. The project preparation team correctly ascertained that a national project was premature given the political instability and the difficulties of the DNVA, and redesigned the project as a pilot operation. Project objectives were well-defined, although only limited indicators were established in the appraisal report to evaluate the effectiveness of the extension service. A systematic monitoring system to measure the level of farmer participation and the adoption rate of technical messages was perfected during project implementation, and was adopted for the follow-on National Agricultural Extension Program Support Project (Credit 2729-MAG).  4 Part I: Project Implementation Assessment C. Achievement of Project Objectives 10. As a pilot operation to test a strategy for more effective agricultural extension based on the specific conditions of the Malagasy agro-economy, the project achieved its objectives. The links between the research and extension services have been strengthened through the adoption of a common approach to participative diagnosis of farmer problems and regular training of subject matter specialists by research staff. Farmers' priority extension needs were identified and addressed more precisely than in the past by regular programmed visits by extension staff, and the organization and effectiveness of the extension system significantly improved. This judgment is consistent with that of the Agricultural Extension Program Support Project Staff Appraisal Report of May 3, 1995: "... the pilot project has demonstrated that a structured and systematic approach to agricultural extension, using existing resources and simple, low-cost techniques within the direct reach of local technicians and farmers can have a rapid and substantial impact on productivity and income and thereby help to alleviate poverty." The pilot project has demonstrated that with available technology and a more effective extension service, it has been possible to substantially increase yields in demonstration plots (para. 12) and more broadly for about 45 percent of the total estimated farming population of 260,000 covered by the project area. For example, in addition to favorable rainfall conditions and improved water management from some irrigation improvements, extension services contributed significantly to the increased production of 45,000 tons of rice between 1993/1994 and 1994/95 by introducing early seedling transplanting and in-line planting techniques. 11. The project utilized a total staff of nearly 500 agents, almost all employed exclusively on field tasks. Depending on the number of farms in each project area and their individual characteristics, the range of farmers per AVB rose from 25-195 before the project to 500-600, and the AVB/supervisor ratio from 2-3 to an average of 8. In parallel, the proportion of farmers participating in organized AVB visits rose substantially, in both percentage terms (from under 5 to over 20 percent) and absolute terms (from about 10,000 to about 40,000). Over 120,000 farmers (against 125,000 estimated in the SAR) are estimated to have benefited from the services provided, reflecting the impact of promoting the establishment of contact groups and improving the scheduling of farm visits. A total of 16,000 demonstration plots were set up in 1994/95 compared to 9,000 in 1992/93, which was a significant increase, if lower than the objective of 25,000 plots set for 1994/95. The budgetary impact of the increased efficiency of extension coverage offers hope for longer term sustainability of the service by government. Unit costs of extension per farm family, which had been traditionally between twice and eighty times higher under previous or ongoing projects, were less than US$6 equivalent annually (cost includes salaries) (para. 30). 12. Extension Services. The pilot project demonstrated that a large number of farmers could obtain good results at low cost with existing agricultural staff, systematic work programming, regular training and joint diagnosis of results by farmers and field staff. For  Part I. Project Implementation Assessment 5 example, rice yields on some 1,700 demonstration plots in farmers' fields, with good water control, increased on average to nearly twice the yields on control plots using traditional techniques. On more than 1,350 separate demonstration plots, where the technical improvement was early transplanting of transplanting of seedlings (without improved water management), yields were on average 53 percent higher than on control plots. In both cases, these results were achieved without any cash investment, through use of techniques adopted directly by the farmers. The gross margins generated by application of these techniques amounted to an equivalent of approximately US$300 and US$125 per hectare, respectively. 13. A representative selection of the 80 different technical messages outreached in the pilot area, and the associated impact they had on farmers' yields in 1993-1994, is shown in Table 1. The significant differences between yields on control (traditional methods used) and demonstration plots (recommended methods), illustrate the potential for improving yields in a variety of crops. The significant differences in yields between demonstration plots and control plots resulting solely from the introduction of simple, low-cost technologies, are convincing increasing numbers of farmers to try the improved techniques. Table 1: Impact of Improved Techniques Recorded in Six Pilot Districts Crop and Recommendation Yield Increase Difference between Control Variation (tons and Demonstration Plots* per hectare)* Rice - intensive rice system (transplanting at B days) 2.24 - 4.4 69% -258% Rice - early transplanting of plants (15-20 days) 0.60 - 3.0 43% - 251% Rice - rainfed in-line planting and manure 0.5 - 1.0 55% - 111% Potatoes - sowing density with manure 5.0 - 10.6 100% - 273% Maize - improved density and weeding 0.6 - 2.3 75% -276% Cassava - improved density and manure 8.0 - 10.0 114% - 200% Beans - density and manure 0.2 - 1.3 400% - 216% Ground Nuts 0.3 - 0.6 16% - 72% Pasture Crops 5.0 Not Practiced Previously Rice - Compost 0.3 - 0.5 Not Practiced Previously * Ranges reflect the highest and lowest impacts observed in the demonstration plots across the 6 pilot districts. 14. Training. Regular training sessions were not implemented under the project until 1993, although sporadic workshops were delivered. Starting in 1993, a systematic training program was introduced that followed a timetable set at the beginning of each season. The monthly technology review workshops were led mainly by FOFIFA, FIFAMANOR and FAFIALA research staff. The first day was dedicated to observation and analysis of the various farming techniques in use, in order to identify exogenous solutions that would serve as a basis for  6 Part I: Project Implementation Assessment recommendations. The second day was devoted to practical applications in farmers' fields. At the end of each workshop, a data sheet of all the lessons learned in this process was prepared. This information was used by the extension agents as a permanent technical reference source to support field activities. 15. Research. On-farm research activities were systematically carried out in the various regions. More than 700 test plots, covering some 10 different crops, were regularly established during the season. These trials, which covered an average of about 60 topics per season, were conducted under the direct responsibility of the subject matter specialists. Joint monitoring and supervision missions, comprising research staff, extension agents and farmers were carried out periodically during the crop year. Their purpose was to ensure the proper execution of field operations and to gather the opinions and comments of participants, especially beneficiary farmers, to accurately register their concerns as a basis for effective recommendations. At the end of the crop year, a review meeting was held with the farmers and extension agents in order to validate the results. 16. Although there has not been an independent assessment of adoption rates, the increased application of the technical recommendations relating to many of the farmers' major concerns is readily apparent on the ground and is borne out in the fortnightly reports of the field extension staff. The topics of greatest interest to farmers included the transplanting of young rice plants (under 45 days), intensive rice-growing, stock selection, seed treatment, drill-hole sowing on upland rice, choice of better-yield varieties, rice/fish-farming, rapid propagation of slips and cuttings (cassava), seed densities, nursery management and fruit-tree maintenance, compost and manure production techniques, animal habitat improvement, production and establishment of fodder reserves, and soil protection and water management. 17. Monitoring and Evaluation. Monitoring indicators were defined by the national coordination unit to facilitate coordination of activities, and to track the impact and performance of extension in the field. The advance programming of work at all levels of the hierarchy has resulted in regular attendance by farmers at the regular demonstration visits conducted by the AVBs, and in a steady improvement in the quality of supervision, reflected in greater interest and involvement on the part of farners. A detailed evaluation of extension effectiveness was conducted every 3 months and served as a major input in the design of the national program. 18. Studies. The mid-term review decided to concentrate the efforts of the project on getting extension staff effectively operating in the field and to reallocate funds to meet this objective. It was therefore decided to reduce the studies component to assistance in the preparation of the follow-up project. The training activities included under this component were maintained, but were to focus on means of reaching women more effectively. 19. Financial Management System. During project appraisal it was recognized that the Administrative Bureau (BAF) would need to be reorganized and that special assistance would need to be given to both the BAF and the project CIRAGRIs in defining accounting and  Part I. Project Implementation Assessment 7 procurement procedures and in strengthening local capacity. In June 1994, the BAF was reorganized and a head of the financial section was nominated. The delay in implementing these measures led to poor accounting and budgetary management in the early years of project implementation. Further improvements have been introduced, although belatedly, including a decentralized financial management system and training of BAF and CIRAGRI staff. D. Implementation Record and Major Factors Affecting the Project 20. Pilot extension activities were initiated in 1988 in three extension districts financed by a Project Preparation Facility of US$750,000. The Project was appraised in September 1989, the Credit (US$3.7 million) was approved in May 1990 and became effective in March 1991. The first two years of project implementation were seriously disrupted by political turmoil and strikes in the civil service that preceded the advent of the Third Republic in 1992. Numerous other constraints exacerbated this situation, including: (a) difficulties in providing counterpart funds to pay taxes and contractual staff salaries; (b) problems in selecting and placing competent and motivated staff to form the operational extension service; (c) slowness in procuring equipment and vehicles, and (d) lack of a commitment to implement the detailed stages of an extension management system. 21. However, major improvements in project performance were achieved in 1993 following the replacement of key staff to positions of responsibility at the central (DNVA) and local (CIRAGRI) levels and improved borrower commitment to the implementation of project objectives. A notable characteristic of the new nominations was the number of women appointed to positions of responsibility. By 1995, 50 percent of the central PNVA team were women. In order to consolidate and properly evaluate the results of the pilot phase, the Bank agreed to extend the project closing date by one year until June 30, 1995. During the project period, an improved extension management system linked with tested principles of extension was introduced in six of the country's 28 districts, affecting a farming population of 260,000. The experience and lessons of this project have served as the basis for the National Agricultural Extension Program Support Project (Credit 2729-MAG) that became effective in November 1995. 22. Projectfinancing. The project had to contend regularly with a serious dearth of counterpart funds, owing in large measure to the wide discrepancy between the available local resources and the demands of the many ongoing projects. This situation led to substantial delays in the financing of expenditures, notably local travel for extension agents, and operating expenses. The May 1992 mid-term review concluded that substantial delays and irregularity in paying field allowances were the most important implementation difficulty confronting the project. At project appraisal this risk had been foreseen and a decision was made to support field allowance payments with IDA financing on a declining basis (100 percent in year 1, 50 percent in year 2, and 0 percent in year 3). At the mid-term review, the Bank decided that this decline was too rapid for the local budget to support and agreed to increase the financing percentage for field allowances to 1 00 percent, on the condition that the level of field allowances be reduced.  8 Part I: Project Implementation Assessment 23. Delays in Procurement. Procurement for project motorcycles was substantially delayed due to the Borrower's lack of compliance with IDA procurement regulations. Due to this delay, AVBs were obliged to walk to farmers' fields and villages creating serious problems for effective project implementation. 24. Technical and organizational aspects. The project began to function better from 1993, as a consequence of (a) major structural changes made in the extension service; (b) a more systematic process of technology generation and diffusion; (c) the training provided to the extension agents and farmers; (d) the development-- though still weak--of collaboration with other extension agencies, both public and private, operating in the same geographic areas as the project; and (e) coordination of the various field activities and performance oversight. 25. Concerning organization of the extension service, key personnel have been appointed at the strategic command echelons. The entire project coordination team was replaced in 1993 by personnel with greater authority and managerial ability. The basic organization, comprising the extension agents and their supervisors, was radically overhauled and its staff downsized to eliminate a large number of redundant personnel, who generally lacked the skills needed to improve services to farmers. A special effort was also mounted to select subject matter specialists with the ability to identify and resolve, in close liaison with research, the various problems underlying farmer concerns and to ensure proper training of the AVBs. 26. Links with the research service. Joint teams of researchers and extension agents were established, at both the national and regional levels, to perform diagnostic work, prepare and implement annual work programs, including field extension and research (trials and tests), and develop training activities for subject matter specialists, AVBs and farmers. Regular meetings between research and extension officials, at both the national and regional levels, assessed the status of program execution and made necessary adjustments. 27. Training of agents andfarmers. Clearly defined training programs were drawn up annually, together with an execution schedule based on the major crop cultivation timetables. Researchers, subject matter specialists and farmers met regularly in monthly technology review workshops to exchange information on their experience with particular technologies. Workshops were held to prepare and update technical protocols or data sheets that provided practical guidance for the entire extension service staff. The knowledge so gained was then transferred by the subject matter specialists to the AVBs at the fortnightly training sessions. 28. Collaboration with other institutions. Cooperation with other donors was often hampered by differences of approach and strategy. Notable improvements were made, however, in developing synergy with certain projects, development agencies and NGOs following the successful experience with start-up workshops in late 1994 in three new districts (Tulear, Mahajanga and Diego Suarez), started under the new National Extension Program support project. These workshops brought together representatives of all the projects, agencies and technical services of a given district involved in extension, research or agricultural training  Part I: Project Implementation Assessment 9 activities, and afforded all participants an opportunity to share their respective objectives and strategies. Given the success of these workshops, similar workshops will be held regularly and gradually extended to all districts. E. Project Sustainability 29. Based on the results described above, the project is sustainable. The extension service is staffed entirely by former staff of the agricultural agencies who had responsibility for the preparation and implementation of project activities. This base is being progressively consolidated and secured through the institutional strengthening of the Agriculture Directorate and the DNVA in the areas of coordination, monitoring and evaluation of project activities and collaboration with non-governmental organizations. 30. The unit costs of extension services per farm family under the project were estimated to be FMG 24,000 (US$6.00). This estimate varies from between one-half and one- eightieth of that of ongoing agricultural projects, according to a February 1994 study ("Evaluation du systeme d'enseignement de recyclage et de formation agricoles" BEST/CINAM for PREFTEC/MEADR). Under the national agricultural extension policy, which was inspired by the positive results of the pilot project experience, all extension activities undertaken by other director-tes would gradually be integrated under the national agricultural extension program and the elimination of duplicate activities should result in significant cost savings. At full development, the annual costs of extension would be US$7.5 million per annum, of which UMS3.85 in recurrent costs (salaries, field allowances, maintenance and operating expenses). This represents about 8 percent of total public expenditure for the agricultural sector in 1994, and about US$3 million less than the government is currently paying for dispersed extension activities, a factor which should considerably improve the Government's ability to meet its share of the costs. 31. The project's sustainability is strengthened by the many extension activities that were implemented to introduce efficient tools and methods for upgrading soil fertility and water conservation management. These activities, based on endogenous strategies already developed elsewhere, produced durable positive changes in cultivation practices. Their countrywide application should reverse the current depletion of agricultural potential currently underway and lead to better uses of relevant technologies in line with farmers' needs. F. IDA Performance 32. Project Preparation. The Bank's performance during project preparation was satisfactory. The project design was consistent with the Government's long-term strategy to achieve sustainable agricultural growth. The project was planned to complement IDA's strategy for agricultural development that included support of the macro-economic and sectoral adjustment process and a sectoral investment program (focused on food security, extension and research, livestock development, irrigation and environmental protection). When it became  10 Part I. Project Implementation Assessment apparent during project preparation that a national project was premature in light of the organizational, managerial, technical and financial difficulties of the DNVA, the Bank supported the Government's decision to use the project as a pilot operation to acquire experience and improve the capability of the DNVA, and the project was redesigned appropriately. 33. Project Supervision. Most of the regular supervision of the project was assured by the resident extension specialist in the Bank's Mission in Antananarivo. He also provided advice to the Director of Extension. However, it was only after the appointment of the third incumbent to the post of extension specialist that the quality of this intensive supervision became fully cost-effective. Turnaround of project impact and results followed the April 1993 mission of the AF3AE Division Chief and the Bank's senior extension training specialist, which achieved, among other things, agreement of the Minister of Agriculture to the replacement of key members of the project coordination team and an action plan for intensified field supervision and regular training. The permanent in-country assignment explains the higher than anticipated number of staff weeks spent on supervision (Table 12 - Part III). Headquarters staff participated in 7 supervision missions. In December 1990, the Bank's agricultural extension unit (AFTAG) assisted the Government in organizing a week-long training seminar to familiarize Malagasy staff with the key concepts of extension policy. The Bank's disbursement department (LOAAF) organized two training missions for Project Implementing Agencies in January 1991 and in December 1994, to improve Madagascar's general procurement and disbursement performance. G. Borrower Performance 34. Project Preparation. The relatively low level of borrower participation in project preparation may have contributed to the lack of borrower commitment during the first years of project implementation. This was corrected in the preparation of the National Agricultural Extension Program Support Project (Credit 2729-MAG) where Government and Bank staff prepared the project jointly and worked together to ensure that beneficiaries and stakeholders were closely associated in the process. 35. Project Supervision. The first two years of project implementation were seriously disrupted by: (a) lack of counterpart funds to pay taxes and contractual staff salaries; (b) delays in selecting competent and motivated staff both in project management and to form the operational extension service; (c) slowness in procuring equipment and vehicles, and (d) lack of a clearly delineated chain of command and the consequent dilution of responsibilities as well as insufficient monitoring and evaluation of the services delivered to farmers. Aware of these shortcomings, Government commitment improved from 1993 onward with the nomination of a new project management team. H. Economic Evaluation of the Project 36. The project shows an economic rate of return (ERR) of 24 percent, against 17 percent estimated at appraisal. The calculation is based on: (i) comparisons of production under  Part I. Project Implementation Assessment 11 with and without project scenarios; (ii) the incremental benefits derived from implementation of the project, and (iii) assumptions of the adoption rate of new technologies and their related costs. A sensitivity analysis based on a possible decline in adoption rates of 5 percent from 1996 onward and an equal variation in costs and benefits, lowered the estimated ERR to about 19 percent (Annex B). 1. Assessment of Project Outcome 37. The results of the pilot project were assessed to be satisfactory. Although initially project implementation was slow, the project achieved its main objectives. The effectiveness of extension services improved. Production and productivity increased, and more efficient technologies - including technologies that are environmentally sound and that often do not require an important cash investment -- are increasingly being adopted by farmers. Monitoring protocols and indicators were established to periodically measure the field impact of project activities. Through demonstration of simple, low-cost technologies on mini-plots, more and more farmers are becoming convinced of the validity of the recommendations presented and showing increased interest in the program. Attendance rates at AVB demonstrations are high; the number of demonstration plots almost doubled from 9,000 to 16,000, and the rate of participation in the services delivered grew fourfold, from 5 percent to nearly 20 percent. Exceptionally strong rice cultivation in comparison with prior years particularly attests to the impact of extension activities, even after allowing for a favorable rainy season. The cumulative impact of these activities, and those undertaken to improve water managernznt in the rehabilitated irrigation schemes (PPIs), is estimated at an additional 45,000 tons of paddy between 1993/94 and 1994/95 alone. Smoother progress towards project achievements could have been obtained had there been a better programming of extension and training activities and if other donors and associated projects in extension and research had been more closely involved from the start in the implementation of the pilot project. 38. The regular provision of several types of training activities, including monthly technology review workshops and fortnightly training sessions, have contributed to sustainability. These activities provided refresher training to the entire service in the understanding and skilled use of the principles and tools basic to the relevant methodological approach, and enhanced the agents' technical and organizational skills, boosting their confidence in talking to the farmers. At the same time, progressive redeployment of research staff in the various agro-ecological regions and phased transfer of responsibility to the decentralized units have helped develop close links between researchers, farmers and extension agents, and better tailor programs to major farmer concerns. The mechanisms of collaboration between research and extension services, and their mutual responsibilities in generating and disseminating results, have been clarified. 39. Periodic consultative meetings helped identify and develop bridges with various projects, agencies and technical services and sharply reduced the sometimes negative competition among various activities. One result of these discussions, was that project execution  12 Part I: Project Implementation Assessment was widened to include other agencies, including FAFIALA (bottomland slope development and protection), FIFAMANOR (tubers and fodder crop development), and RAMILAMINA (better rice field fertility management). As the program expands, similar agreements will be negotiated in various regions. J. Future Operations 40. In the past, attempts to redress the weaknesses of poor government services in the agricultural sector by external donors resulted in the creation of a plethora of projects (more than seventy in the 1994 PIP), exacerbating problems of coordination and sustainability. Subsequent attempts to bring together all those involved in research, training and extension have provided much improved local support and collaboration. Based on the encouraging results posted during the pilot project after 1993, the Malagasy authorities decided to implement a first four-year phase of a longer-term program designed to consolidate and extend the pilot project experiment to the entire country. The Agricultural Extension Program Support Project (Credit 2729-MAG) will support this effort, and additional Bank funding for extension could be considered if the program is successful and remains a priority in the CAS. 41. The national agricultural extension program will be introduced to the country's 28 extension districts over a four-year period. In addition to the 6 pilot extension districts, 3 new districts became operational in 1995 and a further 7 districts will be operational in 1996 and 1997, respectively, with the remaining 5 districts to be launched in 1998. Almost 3,000 field extension staff from both public and private sector sources will be involved by the fourth year of the project with the objective of making extension services available to the country's approximately 1.5 million farm families. To ensure smooth implementation, it is planned to decentralize the management of the project to the provinces and agricultural districts of each project area. Special emphasis will be placed on strengthening the producers' organizations and developing relationships with the NGOs and the private sector, bearing in mind the comparative advantages offered by each partner. K. Key Lessons Learned 42. The principal lessons learned from this project can be summarized as follows: (a) The project demonstrated the feasibility of significantly raising the effectiveness of agricultural extension management in Madagascar through the combined application of basic principles and tools; (b) The project demonstrated that, for the improvements to materialize, the extension and training personnel must receive a minimum of material and financial support, as well as a uniform system of benefits and allowances for field staff. This support essentially consists of the provision of means of travel and operating equipment and supplies, particularly fuel, defrayal of training and field visit costs, and provision of  Part l. Project Implementation Assessment 13 office equipment and supplies. This implies the need for full Government commitment, not only to secure counterpart funding, but also to ensure the sustainability of the project; (c) The feasibility of providing these resources on a regular basis depends, within the overall budgetary and treasury constraints, on the Government's ability to rationalize public investment in the sector and to coordinate donor assistance. To that end, the Government is focusing its strategy around a "prograrn approach" in order to reduce the number of ongoing or future projects, to enhance their efficiency in the field and to avoid unnecessary duplication of effort; (d) Expansion of the pilot project experience was greatly facilitated by a clear Governmental policy on the provision of cost effective and relevant farmer extension services that depend on the coordination and harmonization of independent field approaches and the development of cooperation among the various entities involved (projects, NGOs, private sector). (e) The impact of intensive supervision from a full time specialist in the resident mission only really made itself felt with the appointment of a fully experienced specialist in agricultural extension systems. The experience underlines the critical importance of the selection of the appropriate candidate for such a post.  'Is  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) PART II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE A. Project Implementation 1. The project comprised extension of a first pilot project financed by the World Bank in October 1987 and executed in the CIRAGRIs (agricultural districts) of Antananarivo, Miarinarivo and Moramanga with IDA technical assistance. At the end of this pilot phase, the project covers a total of nine CIRAGRIs: those of Tana Nord, Tana Sud (because of its huge size, Antananarivo was split into two districts), Miarinarivo, Moramanga, Ambatondrazaka, Amparafaravola and, more recently, Antsiranana, Mahajanga and Toliara. The project got off to a difficult start owing to the socioeconomic situation that prevailed in Madagascar in 1991 and 1992. The Agricultural Extension Directorate was reorganized at the end of the first pilot project to make it more efficient and fit it to meet the future demands of transition to the national phase of the current project. The successive internal reorganizations of MINAGRI did not take these PPVA concerns into account. 2. Regarding financial management, the project officials were greatly hampered by the delay in setting up the special account: the expense allowances were not paid promptly, settlement of suppliers' invoices was delayed somewhat (the field agents are no longer motivated to continue, and the project has trouble in finding suppliers willing to accept its order notes). 3. In July 1993 the MINAGRI project officials were compelled to change the entire coordination team because of the unsatisfactory results being obtained. Since then the results of the 1993 off-season, the 1993/94 season, the 1994 off-season and the 1994/95 season have been substantiated in figures. The various extension ratios were revised and adjusted in the interest of greater efficiency. Collaboration agreements were signed with the various project partners. 4. The most recent results posted by the PPVA led the World Bank to agree to the Malagasy Government's request that the project move on to its national phase. The first expansion of the project took place in the CIRAGRIs of Tulear, Majunga and Antsiranana. Start- up workshops have been conducted to determine if all the ongoing development projects in these regions and the PPVA are complementary. The expected results are very encouraging. 5. Concerning equipment, the PPVA suffered substantial delays: the 182 motor cycles for which a call for bids was issued in late December 1992 were not received until June 1995, on the eve of project closure.  Part II: Project Review From Borrower's Perspective 15 B. Project Results 6. The activities carried out during this pilot phase mainly comprised: (a) instituting and developing links with research (FOFIFA or other agencies involved in research); (b) setting up a more cost-efficient structure and organization (determination of standards and ratios); (c) training of the extension agents to improve their grasp both of the system generally and of extension techniques and methods; (d) setting up an M&E system. as the basis for retargeting the extension activities and measuring their impact. 7. Technical results. Despite the initial difficulties, technical gains were posted in the field. The technical activities focused mainly on: (a) setting up demonstration and test plots, essential tools for persuading the farmers to adopt the various messages. The pilot phase successfully demonstrated that the application of simple messages by a number of farmers, with the consequent multiplier effect, can yield very cogent results (an incremental 13,000 tons of paddy in the season 1993/94 and 35,000 tons in 1994/95); (b) supervision of the extension agents to help them perform their work properly. An efficient system of supervision at each echelon was set up, making it possible to correct certain errors in time and prevent them from spreading elsewhere; (c) monitoring field performance and measuring message adoption rates. (Tables of results appended). 8. Training provided. Despite the lack of a training plan, several types of training were provided during this pilot phase: (a) extension service initiation training for the entire team; (b) monthly technology review workshops (AMRTs) and fortnightly training sessions (FQs). These AMRTs are more or less firmly established as periodic and continuing training activities. The workshops are beginning to be held regularly with the participation of various resource persons (research service, NGOs or other specialist services), depending on the training topics suggested by the constraints identification studies. The main topics of these periodic training sessions relate essentially to the upgrading of production techniques. (c) specific training sessions, on topics vital to the proper conduct of the extension activities that could not be covered in an AMRT program; (d) training sessions on teaching techniques and methodology, to enable the extension agents to transmit the messages effectively;  16 Part II: Project Review From Borrower's Perspective (e) study trips and regional workshops and seminars (both in Madagascar and abroad) to allow the extension service officials to compare extension methods and introduce improvements into the project. 9. Collaboration agreements. The project used the services of specialist agencies to conduct training sessions for its extension staff and carry out specific extension activities. To that end, collaboration agreements were concluded with: (a) FAFIALA, concerning tanety development and protection (in year 3); (b) the NGO RAMILAMINA, concerning the use of azolla (year 2); (c) FIFAMANOR, concerning the cultivation of potatoes, sweet potatoes and fodder crops (year 2). 10. Collaboration with the Ministry's other directorates and technical agencies was also beginning to be effective: the staff of these directorates were acting as resource persons during the monthly workshops and participating in the diagnosis and monitoring of field results. 11. Financial results. The delay in setting up the special account (three years after project start-up) seriously hampered the financial management of the project. Project operating expenditures were limited by the inadequacy of the counterpart funds available to pay the various duties and taxes. C. Borrower Performance 12. It was agreed that the PPVA staff would undergo continuing training and testing following which agents deemed to be incapable of improvement would be taken out of the project. The expected impact of this measure was limited by a number of administrative, political and social factors. 13. There is a shortage of personnel in the field but an over-abundance in the large cities. For the same reasons as mentioned above, personnel redeployment took place on only a very small scale (assignment of supervisory staff to a Firaisana lacking in social infrastructure (hospital, school, etc.) is often a difficult decision. Most of the basic extension agents are contractual staff and therefore not assignable). 14. The counterpart funds were never adequate or released on time. For example, the motor cycle spare parts accepted in June 1995 are still being held in customs because of inadequate funding. The annual allocations to the project barely suffice to pay salaries and a tiny fraction of taxes and duties. Purchases of operating supplies and spare parts by National Coordination was limited by lack of sufficient counterpart funds to pay the taxes and duties on them.  Part II. Project Review From Borrower's Perspective 17 15. The opening of a special account at a primary bank greatly helped the project settle its invoices and pay off its staff travel allowance arrears. A weakness noted, on the other hand, was the failure to decentralize the Special Advance Accounts (Caisses d'Avance). D. Outlook 16. Experience during this pilot phase yielded a number of lessons concerning ways of improving the various components of the national phase of the project in order to achieve its objective of enhancing the quality of life in the rural areas. They point to the need to: 'a) implement decentralization so as to confer accountability on the regional services (the PROVAGRIs and CIRAGRIs); (b) improve coordination with the other ongoing projects and the other agencies active in rural development (NGOs and public services); (c) upgrade the project's budgetary and financial accounting, taking into account the external audit recommendations; (d) furnish the various categories of personnel with material support; (e) strengthen the diagnosis process in order to refine the extension messages and thereby tailor them more closely to the farmers' real needs; (f) step up the results to be obtained in order to remunerate the resources deployed; (g) consolidate the program M&E system and information circuit; (h) monitor the impact of the extension activities by measuring the rates of adoption of the messages by the farmers; (i) consolidate clear supervision responsibilities and organization at each level; G) ensure regular and continuing training of the extension agents.  IMPLEMENTATION COMPLETION REPORT REPUBLIC OF MADAGASCAR AGRICULTURAL EXTENSION PILOT PROJECT (Credit 2150-MAG) PART III: STATISTICAL ANNEXES TABLE 1: SUMMARY OF ASSESSMENTS A. Achievement of Objectives Objectives Substantial Partial Negligible Not Applicable Macro Policies X Sector Policies X Financial X Objectives Institutional X Development Physical X Objectives . | Poverty X Reduction Gender Issues X Other Social X Objectives Environmental X Objectives l Public Sector X Management l Private Sector X Development . B. Project Sustainability Project Likely Unlikely Uncertain Sustainability x  Part III: Statistical Annexes 19 C Bank Performance Performance Highly Satisfactory Deficient Satisfactory Identification X Preparation X Assistance Appraisal X Supervision X D. Borrower Performance Performance Highly Satisfactory Deficient Satisfactory Preparation X Implementation X Covenant X Compliance l Operation (if X applicable) _ _ E. Assessment of Outcome Performance Highly Satisfactory Satisfactory Unsatisfactory x  20 Part HII. Statistical Annexes TABLE 2: RELATED BANK LOANS/CREDITS Loan/Credit Title Amount Purpose Year of Status (US$ million) Approval l National Agricultural 24.0 To strengthen agricultural 1989 Active Research Project research and make it more (Credit 2042-Mag) responsive to the needs of l __________________ producers. Environment Project 26.0 To protect and preserve the 1990 Active (Credit 2125-Mag) environment on a sustainable basis by managing and protecting the ecosystems and establishing anti-erosion systems. Livestock Project 19.8 To increase livestock 1991 Active (Credit 2243-Mag) output and to promote exports. Rural Finance Project 21.2 To assist rural groups 1993 Active (Credit 2459-Mag) develop savings and loan associations. Second Irrigation 3.7 To promote growth by 1995 Active Rehabilitation Project increasing agricultural (Credit 2644-Mag) output through better water management and assessment of irrigation rehabilitation.  Part III: Statistical Annexes 21 TABLE 3: PROJECT TIMETABLE Identification December 1989 December 1989 Preparation July - August 1989 July - August 1989 Appraisal September 1989 September 1989 Negotiations February 1990 February 1990 Board Approval March 19, 1990 May 31, 1990 Credit Signature July 18, 1990 July 18, 1990 Credit Effectiveness July 1990 March 22, 1991 Midterm Review March 1992 March 17-30, 1992 Credit Closing June 1994 May 5, 1995 TABLE 4: ESTIMATED AND ACTUAL ANNUAL AND CUMULATIVE DISBURSEMENTS Appraisal Estimate Actual Disbursement Cumulative Actual Year US$ 000 US$ 000 as % of Cumulative Estimated Disbursement Annual Cumulative Annual Cumulative 1988 -- -- 208 208 -- 1989 550 550 213 421 76.54 1990 900 1,450 471 892 61.52 1991 1,030 2,480 304 1,196 48.22 1992 800 3,280 158 1,354 41.28 1993 400 3,680 400 1,754 47.66 1994 -- I 669 2,423 65.84 1995 -- I 1,028 3,451 93.78  22 Part IlI: Statistical Annexes TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION INITIAL * INDICATOR OBJECTIVES ACTUAL PERFORMANCE NOTES No. of AVBs 305 298 98 No. of CGs 4,500 3,950 88 No. of DPs 25,000 16,000 64 No. of TMPs 1,200 780 65 Participation rate * 100% 70-90% Incremental rice -- 45,000 production (tons) 1994/95 ** * Participation Rate: measures the rate by which the contact group member farmers participate in the visits of the extension agents and indicates the degree of farmer interest and the relevance of the technical messages demonstrated. ** Figures for incremental rice production only available for 1994/95 campaign. Extension staff obtained figures from farmers practicing improved techniques. TABLE 6: KEY INDICATORS FOR PROJECT OPERATION No key indicators for project operation were provided in the SAR. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  Part III: Statistical Annexes 23 TABLE 7: STUDIES Studies Planned Studies Completed Studies of phase of transfer of extension Gradual transfer of responsibilities was responsibilities from the regional parastatal undertaken as a result of the work plan authorities and autonomous projects to the developed jointly between the parastatals and Agricultural Extension Directorate. the PNVA. Future powers and responsibilities of the Roles and responsibilities defined in SNVA Provincial Agriculture Service. document. Influence of land tenure (type of utilization) on Roles and responsibilities defined in SNVA disseminated technologies adoption rates. document. Role of women in agriculture, and No study undertaken but special efforts were improvement of their participation in the made to adapt technical messages to the needs agricultural sector. of women farmers. Special studies on environmental protection No special studies conducted. and conservation.  24 Part III: Statistical Annexes TABLE 8A: PROJECT COST Appraisal Estimates Actual Expenditure Category (JUS$ million) (US$ million) Local Foreign Local Foreign currency exchange Total currency exchange Total Civil eng. works 695 174 869 0 0 0 Equipment 511 365 876 755 603 1,358 Consultants and studies 876 0 68 76 144 Training and documentation 168 171 339 606 233 839 Unspecified investment 414 276 690 10 22 32 Total investment cost 2,664 986 3,650 1,439 934 2,373 Recurrent costs 937 14 951 1,077 0 1,077 Total basic costs 3,601 1,000 4,601 2,516 934 3,450 Technical contingencies 104 26 130 0 0 0 Price contingencies 586 92 678 0 0 0 Total Cost 4,291 1,118 5,409 2,516 934 3,450  Part III. Statistical Annexes 25 TABLE 8B: PROJECT FINANCING Appraisal Estimate Actual Financing Source (US$ million) (US$ million) Local Foreign Total Local Foreign Total Currency Exchange Currency Exchange IDA 3,680 2,516 934 3,450 Government 920 1,312 Taxes 810 Total 5,410 3,828 934 4,762 TABLE 9: ECONOMIC COSTS AND BENEFITS Annex B: Calculation of ERR  26 Part lII: Statistical Annexes TABLE 10: COMPLIANCE WITH CREDIT CONDITIONS Clause Subject Status 2.02 b Opening of the Special Account Opened at BMOI. l ____________ ______________________________ Operational July 1993 2.03 Closure date: June 30, 1994 Actual closure: June 1995 3.01 a The Government shall provide the Clause partially complied with: funds, installations, services and other funds allocated inadequate and resources required by the project as and released late when the need arises 3.01 b Opening of the Special Advance Clause partially complied with: Account at DIRAGRI and CIRAGRI only the DIRAGRI benefited. Delays in appointing the new CIRAGRI chiefs as managers 3.03 Presentation of the PTA by September Clause complied with 30 of each year 3.05 Payment of indemnities to the Clause complied with despite delay extension agents in payment timetable 4.01 b Financial Audit Clause complied with  Part III: Statistical Annexes 27 TABLE 11: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS There was compliance with the applicable Bank Operation Manual Statements. TABLE 12: BANK RESOURCES -- STAFF INPUTS Planned Actual Actual Stage of Project Cycle Staffweeks Staffweeks US$ (000) Preparation 70.0 75.3 177.4 Appraisal 54.0 60.7 165.8 Board-Effectiveness 26.0 25.6 72.2 Supervision 64.0 251.2 779.0 Completion 4.0 9.0 25.4 218.0 421.8 1,219.8 Total  28 Part III: Statistical Annexes TABLE 13: BANK RESOURCES -- MISSIONS Stage of Date No. of Specialization Implementation Development Types of Project persons Skills Status Objective Problems Cycle Identification 10-87 2 AES Preparation 8-89 2 AES Appraisal 9-89 4 AES, FA Supervision 8-90 2 AES 2 2 LC Supervision 3-91 1 AES 2 1 LC Supervision 5-92 2 AE, AES 3 3 LC, P Supervision 6-93 3 AE,AES, 3 3 LC, M FA F, FP Supervision 1-94 2 AE,AES 2 2 P,F,FP Supervision 7-94 4 DC,AE, S HS NI Completion 9-95 1 1 AES S S Key to specialized staff skills: DC = Division Chief; FA = Financial Analyst; AE = Agro-Economist; AES = Agricultural Extension Specialist; Key to types of problems: LC = Legal Covenant Compliance; M = Management; P = Procurement; F= Funding; I = Institutional; P = Technical; FP = Financial Performance NI = None Identified  ANNEX A: SUMMARY OF AUDIT RECOMMENDATIONS 1. The audit reports on the various National Agricultural Extension Program (PNVA) credits (PPF-486, MAG-2150, FNDE 99-46-G, 91-01, Malagasy Counterpart Fund) were unqualified by the auditors but provided the following recommendations, most of which were acted on by the project coordination unit: (a) systematically post all fixed asset acquisitions to the accounts (with date of receipt) and carry their value (value net of tax per invoice) to the fixed assets ledger, and note therein all transfers to users; (b) post to the ledger all requests for reimbursement against the special account as recorded in the certified statement of expenditures; (c) prepare the statement of financial condition for Credit 2150(a) MAG as of December 31, 1994; (d) keep accounts of the Counterpart Funds at the PNVA central office so as to allow regular monitoring of the Malagasy Government's contribution per CIRAGRI (accounting charge, etc.) [and] transfers from the Counterpart Funds; (e) obtain from the Treasury's External Relations Service (SREET) the serial numbers of transfer orders and requests for withdrawal of funds; (f) obtain the World Bank statements from the Treasury (Ministry of Finance); (g) obtain a copy of the statement of the corresponding bank from the Central Bank of Madagascar (BCM); (h) maintain a continuous statement for Direct Drawings and the Special Account; (i) prepare a bank reconciliation statement at the end of each month; (i) have the expenditure charge categories checked (by the official in charge) before the expenditures are posted; (k) number each operation in step with its registration in the books; (1) record all payment references in the books (payment orders and BRs); (m) attach all pertinent vouchers to each request; (n) conclude service arrangements between the DALF chief and the comptable depositaire whenever necessary; (o) prepare an inventory of the assets acquired under the project; (p) find suitable premises for installation of the data processing equipment in order to avoid its under-utilization and otherwise make sure that it is operational (training of agents, etc.); (q) return all unused equipment to the central store, drawing up minutes of proceedings; (r) locate the equipment acquired under the project; (s) draw up minutes of proceedings for all transfers of equipment by the holders and the comptable depositaire; (t) have the inventory listing signed and dated by the comptable depositaire; (u) draw up an official declaration of loss for all equipment lost and send a copy to the National Coordinator of the project;  30 Annex A (v) repair broken equipment that is the subject of spare parts acquisitions; (w) maintain and keep current a fixed asset monitoring register and post the corresponding operations when movements take place; (x) post requests for reimbursement against the Special Account and carry them to the certified statement of expenditures; (y) utilize suitable local resources and management tools.  RECAPI I ULAITIF DlES COU [S EI AVAN IAGES D)U PROJE ANNEE SUPERFICIE TAUX DE PART' BENEF ADD /IIA TOT.BENEF ADD. COUT DU PROJET COUT ADD /HA TOT COUT ADD. TOTAL BENEF. 25% 24% 1990 230000 1 160 368,000 998,852.60 143 1.328,631.95 -960,632 0.8 -768.505.56 0.806 -774,269.3S 1991 230000 3 160 1,104.000 457,853.40 143 1,447,190.84 -343,191 0.64 -219,642.14 0.65 -223.074 05 1992 230000 5 160 1,840,000 157,321.67 143 1,806,217.41 33,783 0 512 17,296 69 0 524 17,702.08 1993 230000 10 160 3,680,000 397,705.44 143 3,695,496.91 -15,497 0.41 -6,35373 0 423 -6.555.19 1.994 230000 15 160 5,520,000 555,828 66 143 5,502.515.87 17,484 0 328 5,734.80 0.341 5.962.09 1995 230000 20 160 7,360,000 878,119 34 143 7,473,702.28 -113,702 0.262 -29,790 00 0 275 -31.266.13 1996 230000 25 160 9,200,000 2,448.35 143 8,246,927.03 953,073 0.21 200,145.32 0222 211,58220 1997 230000 25 180 9.200,000 686 00 143 8,245,164 68 954,835 0.168 160.412 33 0.179 170,915 52 1998 230000 25 160 9,200,000 68600 143 6,245,164.68 954,835 0 134 127,947.93 0 144 137.49629 1999 230000 25 160 9,200.000 68600 143 8.245,16468 954,835 0.107 102,16738 0.116 110,760 90 2000 230000 25 160 9,200,000 686.00 143 8,245,164.68 954,835 0.086 82,11584 0.094 89.754.52 2001 230000 25 160 9,200,000 686.00 143 8,245,164.68 954,835 0 069 65,883.64 0.076 72,567.48 2002 230000 25 160 9,200,000 686.00 143 8.245.164 68 954,835 0 055 52.515 94 0.061 58,244.95 2003 230000 25 160 9,200,000 686.00 143 8,245,164.68 954,835 0.044 42,012.75 0049 46,786.93 2004 230000 25 160 9.200,000 686.00 143 8,245.164.68 954,835 0.035 33,419.24 0.04 38,193.41 2005 230000 25 160 9,200,000 686.00 143 8,245.164.68 954,835 0.028 26,735 39 0.032 30,554.73 2006 230000 25 160 9,200,000 686.00 143 8,245,164.68 954,835 0.022 21,006.38 0 025 23,870.88 2007 230000 25 160 9,200,000 686.00 143 8,245,164.68 954,835 0.018 17,187 04 0.021 20,051.54 2008 230000 25 160 9,200,000 686.00 143 8.245,164 68 954,835 0.014 13,367.69 0 017 16,232.20 2009 230000 25 160 9,200,000 686.00 143 8,245.164 68 954,835 0.011 10,503.19 0.013 12,412.86 45,839.88 27,921.87 TAUX DE CHAN $1= 4,260.00 FMG :z 0 .i  RECAPITULATIF DES COUTS ET AVANTAGES DU PROJET ANNEE SUPERFICIE TAUX rAdop BENEF ADD /HA TOTBENEF ADD VaCoUtsts5% COUTDUPROJET COUTAAIAOTCOUTADD TOTALBENEF Var 5Bn-5% 18% I 19% 1990 230000 1 160 388,000 998,853 998852.80 143 1.328.63195 9 -980632 -960,632 1 -980,631.95 1 -90,631 95 1991 230000 3 160 1.104.000 457,853 457.85340 143 1.447,19084 -343,191 -343.191 1 -343,19084 1 -343,190.84 1992 230O0O 5 180 1.840.000 157,322 157.321.67 143 1.808.217 41 33.783 33,783 1 33.782 59 1 33,782.59 1993 230000 10 180 3,680.000 397,705 397,705.44 143 3,6895496 91 -15,497 -15,497 1 -15,496 91 1 -15,498 91 1 994 230000 1 5 1 0 5.520,000 555,829 555.828.66 143 5.502,515 87 17,484 17,484 I 17.484 13 1 17.484.13 1995 230000 20 160 7.360,000 878,119 878. 119 34 143 7,473,702 28 -113.702 -113,702 1 -113.702 28 1 -113,702 28 1996 230000 20 160 7.380.000 2,571 2.44a 35 143 6,598,031.29 761,969 723,870 0 314 239,258.17 0 296 225.542 74 1997 230000 20 160 7,380,000 720 68800 143 6,596,268.95 763,731 725,845 0.286 203,152.46 0 249 190,169 03 1996 230000 20 1 e 7.380.000 720 8 88o 143 6,596.268 95 763,731 725,545 0.225 171.839 49 0.209 159,819.79 1999 230000 20 160 7.360,000 720 686 oo 143 6,598,268 95 763,731 725,545 0 191 145.872 63 0176 134,416 67 2000 230000 20 160 7,360.000 720 68600 143 6,596.268 95 763,731 725,545 0 162 123,724 43 0o 147 112.268 46 2001 230000 20 160 7,360,000 720 88 00 143 6,596,268 95 763.731 725,545 0 137 104,631 15 0.124 94,702.65 2002 230000 20 180 7.360,000 720 686 00 143 6,598,26895 763,731 725,545 0 116 88.592 80 0104 79,428.03 2003 230000 20 180 7,360,000 720 88 00 143 6,598,268.95 783,731 725.545 0.098 74.845 64 0.087 66.444 60 2004 230000 20 160 7.360.000 720 688.00 143 6.596.268.95 763.731 725.545 0 083 63,389.68 0 073 55,752.37 2005 2300O0 20 160 7,360,000 720 868 o8 143 6,596.288 95 763,731 725.545 0 071 54,224.90 0.062 47.351.33 2008 230000 20 160 7,360,000 720 888.00 143 6.59,256 95 763,731 725,545 0.08 45,823.86 0.052 39.714 01 2007 23000O 20 180 7,360.000 720 686.00 143 6,596,268 95 763,731 725,545 0 051 38,950.28 0.04 33.604 17 2008 230000 20 180 7,360.000 720 686.00 143 6,596,268.95 763,731 725,545 0 043 32.840.44 0.037 28,258.05 2009 230000 20 160 7.350.000 720 6ee.o 143 6,596.2e895 763,731 725,545 0036 27.494 32 0.031 23.876.68 3,454,735 8,774,194 32.885 01 -90,8.07.70 1.05 0 95 lb  Annex B 33 REVENU ADD.PAR EXPLOITATION |_______ |_________ _ Saison Contre saison Revenu add/Ha | r Revenu add/expl Riz irrigue 2,208,384 0.40 0.20 1,325,030 Riz pluvial 490,752 0.10 0.00 49,075 Mais 750,000 0.10 0.00 75,000 Arachide 1,440,000 0.10 0.00 144,000 Manioc 3,200,000 0.30 0.00 960,000 TOTAL 8,089,136 2,553,106 T O T A L US $ 599 /Ha 2,127,588 499 COUT ET PRODUCTION ADD.PAR EXPLOITATION Saison Contre saison Cout add/Ha Cout add/expl