Sri Lanka - Kurunegala Rural Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8273 PROJECT COMPLETION REPORT SRI LANKA KURUNEGALA RURAL DEVELOPMENT PROJECT (CREDIT 891-CE) DECEMBER 31, 1989 Agriculture Operations Division Country Department I Asia Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Appraisal (1979)s US$1.00 - Rs15.5 Last Year (1986); US$1.00 = Rs28.0 WEIGHTS AND MEASURES 1 long ton - 2.240 lb = 1.016 metric tons 1 hundredweight (cwt) = 50.8 kg = 112 lb 1 bushel (bu) of paddy = 45 lb 1 pint = 0.57 litres 1 acre (ac) - 0.405 hectare (ha) 1 mile (mi) = 1.609 kilometres (km) 1 square mile (sq mi) 8 640 ac (259 ha) 1 foot (ft) = 30.5 centimetres (cm) 1 chain = 66 ft PRINCIPAL ABBREVIATIONS ADA - Agricultural Development Authority ADB - Asian Development Bank ARTI - Agrarian Research and Training Institute BOC - Bank of Ceylon CCB - Coconut Cultivation Board DOA - Department of Agriculture DOI - Department of Irrigation DMEC - Department of Minor Export Crops GOSL - Government of Sri Lanka ICB - International Competitive Bidding KVS - Village Level Worker (Krushi Karma Viyaptha Sevaka) MPCS - Multi-Purpose Cooperative Societies O & M - Operation and Maintenance PB - People's Bark FISCAL YEAR January 1 to December 31  FOR OFFIAL USE ONLY THI WORLD SANK Washington. D.C. 20433 U.S.A. O0 of Owto-G.wai * ~~Opmatiam Ivahiatan December 31, 1989 HEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report - Sri Lanka Kurunegala Rural Develo2ment Project (Credit 891-CE) Attached, for information, is a copy of a report entitled "Project Completion Report - Sri Lanka Kurunegala Rural Development Project (Credit 891-CE)", prepared by the Asia Regional Office and the Food and Agriculture Organization of the United Nations. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment I This document has a restricted distibution and may be used by recipients only in the performance i of their official dutis. Its conents may not otherwise be disclosed witot World Bank authoriation.  FOR OMFCUL USE ONLY PROJECT COMPLETION REPORT SRI LANKA KURUNEGALA RURAL DEVELOPMENT PROJECT (Credit 891-CE) Table of Contents Page No. Preface .............................................. () Evaluation Summary ............................................ (ii) PART I - Project Review from Bank's Perceptive ................... 1 Project Identity ........................................ 1 Background ........................................., 1 Project Objectives and Description ...................... 2 Design and Organization ................................. 4 Project Implementation .................................. 5 Project Results .................................. 11 Project Sustainability .................................. 16 Bank Performance .................................. 16 Borrower Performance .................................. 17 Project Relationship .................................. 19 Consulting and Contractual Services ...... . ............. 20 Project Documentation and Data .......................... 20 PART II - Project Review from Borrower's Perspective .............. 22 PART III - Statistical Information ..... .......................... 23 Related Bank Loans and/or Credits . . . 23 Project Timetable . . ..................................... 24 Credit Disbursement .................................... . 25 Project Implementation .... 26 Project Costs and Financing . . . 28 Project Costs . . ...................................... 28 Project Financing .... 29 Project Results . . ....................................... 29 Direct Aenefits . . .................................... 29 Economic Impact . . .................................... 30 Financial Impact . . ................................... 30 Studies .. ............................................ 31 Status of Covenants . . ................................... 31 Use of Bank Resources .... 32 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.  Page No. TABLESs 1. Cropped Paddy Area and Cropping Intensity ............... 33 2. Credit Disbursement by Participating Banks .............. 34 3. Annual Credit Disbursement by Participating Banks ....... 35 4. Paddy Under Major Irrigation - Financial Budget ......... 36 5. Paddy Under Minor Irrigation - Financial Budget ......... 37 6. Paddy Under Rainfed Conditions - Financial Budget ....... 38 7. Paddy Production ..................... ............... . 39 8. Coconut Development Area Improvement .................... 40 9. Coconut Rehabilitation - Financial Budget ................ 41 10. Coconut (Mature Plantations) - Financial Budget ......... 42 11. Coconut Replanting/Underplanting - Financial Budget ..... 43 12. Economic Price Derivation for Paddy and Fertilizer ...... 44 13. Economic Price Derivation for Coconut . ....... 45 14. Economic Analysis ......... . ... .. ... .......... 46 ANNEX 1 -List of Reports/Studies .......................47  v i - PROJECT COMPLETION REPORT SRI LANKA KURUNEGALA RURAL DEVELOPMENT PROJECT (CREDIT 891-CE) PREFACE This is the Project Completion Report (PCR) for the Kurunegala Rural Development Projezt in Sri Lanka for which IDA Credit 891-CE for US$20.0 million was approved in March 1979. The credit was closed on June 30, 1986 after two one-year extensions. The final disbursement was made on January 21, 1987 and an undisbursed balance of US$2.9 million was cancelled at that time. The PCR was jointly prepared by staff of the Agriculture Operations Division, Country Department I, Asia Region, and staff of FAO/CP. The general findings of the PCR were discussed with Government of Sri Lanka officers and staff of the Project Office and the Agrarian Research and Training Institute following the PCR preparation mission, and the Preface, Evaluation Summary, Parts I and III were sent to the Borrower in June 1989 with a request to prepare Part II by October 30, 1989. None was received. Preparation of the PCR was carried out by an FAO/CP mission which visited Sri Lanka in May 1989 and is based, inter alia, upon a review of IDA records in Washington, discussions with staff, and documents and reports prepared by GOSL and the World Bank during project preparation, appraisal and implementation.  (ii) SRI LANKA KURUNEGALA RURAL DEVELOPMEN1T PROJECT (Cr. 891 CE) EVALUATION SUMMARY Obiectives 1. This project, the first of what was hoped would be a series of projects, aimed at assisting the Government of Sri Lanka (GOSL) to evolve a replicable model of regional development for raising productivity, employment, incomes and living standards of the rural pcpulation through strengthenJng of the development-oriented institutional services, accompanied by modest but critical complementary investments in physical infrastructure. More specifically, the project was to assist GOSL in helping the rural inhabitants of the Kurunegala District. Because of the predominant position of paddy and coconut in the district's economy as well as the scope of large production increases for both crops, the principal project thrust was to rehabilitate the production base of these crops and to improve the required supporting services. 2. To achieve these broad objectives the project provided fuinding for institutional and infristructural development th._ough sub-componesnts focussing on (a) improring irrigation and water management, (b) rehabilitating and replanting coconut plantations, (c) upgrading agricultural extension services, (d) improving agricultural input supplies for improving livestock management, (e) improving credit operation and resources, (f) improving livestock management, (g) increasing groundwater exploration. (h) improving rural roads, water supply and electricity, (i) increasing rural education facilities, and (j) establishing project coordination and monitoring capabilities. Implementation Experience 3. Project implementation was initially expected to be completed in five years (mid-1979 to mid-1984). However, implementation did not progress as anticipated particularly in the initial period and the project closing date was extended twice by one year in order to achieve most of the set targets. Actual disbursements were US$ 17.1 million (862) of the total Credit of US$ 20 million. The main reason for the slow start-up was a fourteen-month delay in the procurement of equipment and vehicles, which affected staff mobility ane execution of work. This was mostly due to the inexperience of Agencies' staff wfith the Bank's International Competitive Bidding (ICB) procedures. The other important reason for delay was a serious manpower shortage in the main implementing agency, i.e., the Department of Irrigation which slowed down identification of tanks and design of rehabilitation work. 4. Although in quantitative terms Staff Appraisal Report (SAR) targets were achieved for most components the absence of an effective institutional mechanism to absorb and react to farmers needs was a  (iii) constraint which reduced the effectiveness of the project. It was not until 1984 that the trained staff and equipment necessary for the project's success were available. In the area of irrigation and water management, implementation of the SAR model showed that it was best suited for adoption in the intermediate zone were rainfall was more regular, whilst in the dry areas of the district the model was less suitable. A larger sample with a longer time spread would be required before an adjusted model could be implemented throughout the district. So far, the attempts to introduce improved water management in the district have failed. In other areas successful implementation was influenced by prevailing practice, e.g., the purchase of agricultural inputs from private trader (who gave credit) rather than from the Multi-Purpose Cooperation Societies. 5. The envisaged implication of extension service did not materialize and the advisory services of the Coconut Cultivation Board, Department of Minor Export Crops as well as the Department of Animal Husbandry continued to render technical advice in their respective fields. However, the ratio of farm families per village '-evel worker was reduced from 1,200 in 1979 to 850 at project completion, which is slightly higher than the appraisal target of 600. 6. Most of the project implementation difficulties could have been avoided through more detailed project preparation efforts. In particular, more attention should have been given to the strengthening of the Department of Irrigation and to ensuring better inter-agency coordination. The fact that the officials of the agencies involved in project implementation were not familiar with ICB procedures should have been recognized and technical assistance and/or training provided. 7. A more careful assessment of the agro-climatic conditions of the district would have helped to better formulate the coconut development component and to define the horticultural sub-component at an early stage. Also the downstream development works of minor irrigation schemes could have been identified at an earlier stage, if more detailed survey work would have been carried out. More accurate costings of civil works, based on detailed programs and specifications, would have avoided cost overruns and delays. 8. The project risks were rightly identified as being: (a) the ability to achieve the necessary coordination in implementing the various inter-related components; (b) the adoption of intercropping practices by coconut smallholders; and (c) the shortage of professional and technical staff, particularly design and construction engineers. However, insufficient provisions were made at appraisal to minimize these risks. 9. In retrospect, the decision to extend the project by two years proved useful. It enabled the agencies involved to use the experience gained during initial implementation to execute the remaining components quickly and efficiently and valuable experience was gained for similar projects in other districts. However, more decisive and concrete steps should have been taken to ensure a greater level of beneficiaries' participation in the identification process, in particular with regard to the irrigation rehabilitation and the rural water supply components.  (iv) Results 10. The project as a whole can be considered a success as it has achieved its main objectives, namely to increase paddy and coconut production, to create employment and improve income and living standards of the rural population in the district. It has also served as a model to be replicated in other districts and many of the lessons learned during implementation were incorporated in subsequent IRD projects. In addition, the project has helped to strengthen the social infrastructure in the area. It is, however, too early to assess the project's impact on the reduction of inter-regional disparities. Throughout project implementation, none of the project objectives were changed. 11. In quantitative terms incremental production of paddy is estimated at 56,000 tons compared with the appraisal projection of about 49,000 tons. Increased productivity of mature coconut plantatinns is expected to result in about 2,500 nuts/acre. The average yield of coconut arising from project activities is 2,350 nutslacre. Overall, incremental coconut production is expected to reach a maximum level of 107 million nuts in 2003 (i.e., 25 less than appraisal, with ten years delay) and production of MECs will reach about 40Z of anticipated production, partly compensated however by horticultural production which was not foreseen at appraisal. 12. The Economic Rate of Return (ERR) at appraisal was estimated at 321. Using the same assumptions, the reestimation of the ERR (at 1989 constant prices) is 382. The Financial Rate of Return (FRR) was estimated at 34Z. However, due to increases in fertilizer costs, stagnating prices for rice and depressed international demand for coconuts the estimated 1989 FRR was only 10X. 13. A positive trend was observed in income distribution. In FY79, before the commencement of the project, 192 or the population were below the poverty line which was recognized as Rs 3,600/year. According to GOSLs ex-post evaluation of the project in FY86, only 171 were below the poverty line of Rs 7,000/year. This achievement is not solely attributable to the project as the contribution of other national programs was also substant.al. However, the inflow of relatively larger allocations of funds under the project made a sizeable impact on the distribution of income, particularly in the dry zone of the district where project focus has been significant. It is assumed that more than 502 of the investments received by the dry zone during FY79/FY86 came from project provisions. In addition, the project created about 4 million man-days of direct employment through labor-intensive works and about 2.5 million man-days of field work under the coconut development program. Sustainability 14. All the facilities created under the productive components handed over to the respective agencies before or on closing and all irrigation schemes are still in full working order. Staff strength of the extension service has been maintained and slightly increased. However, due to the liberalizatior. of fertilizer marketing, the farm input stores created under the Department of Agrarian Services are underutilized. The state of repair of social infrastructure established under the project remained satisfactory.  (v) 15. The absence of water charges under the project for the major and minor irrigation schemes makes them entirely dependent upon scarce and fluctuating budgetary allocations for repair and maintenance resources which raises doubts about their long term sustainability. According to GOSL's ex-post evaluation report, the short and medium-term loan program introduced under the project shows satisfactory recovery rates compared with all island levels or pre-project levels. The overdue rate of crop loans has fluctuated between 22 and 4.8Z and in the case of equipment and machinery loans stabilized at 4Z in FY87. Together with improved procedures, the credit component appears sustainable. 16. The Government continues to support the investment undertaken by the project including the annual 0 & H costs for project works and services. Also, as proven by the conLinuously increasing cropping intensity, the farming community has generally accepted the improved crop husbandry techniques introduced by the project. The above are indications that the project will probably maintain an acceptable level of net benefits throughout its economic life. Lessons 17. The main lessons from project implementation are: (a) project design should focus more clearly on project organization and management especially when a multitude of implementing agenciacs are involved; (b) in the case of projects which are innovative in type and design, greater allowances for flexibility and a longer start-up period should be allowed for; (c) where new procedures are to be followed by the borrower (such as ICB), sufficient allowance for training and/or technical assistance should be made; ld) greater participation of beneficiaries in project design and implementation should be sought and encouraged to focus more clearly on their needs and ensure their acceptance; and (e) corrective action should be taken as soon as delays in component implementation occur to avoid slippage of the project as a whole.  SRI LANKA KURUNEGALA RURAL DEVELOPMENT PROJECT (Credit 891-CE) PROJECT COMPLETION REPORT PART I - PROJECT REVIEW FROM THE BANK'S PERSPECTIVE 1. Project Identity Project Name : Kurunegala Rural Development Project Credit No. s 891-CE RVP Unit : ASIAG - Ccuntry Department 1, Asia Region Country : Sri Lanka Sector : Agriculture Sub-sector : Rural Development 2. Background 2.01 Recognizing the pivotal role of agriculture in stimulating economic growth and rural development, the Government of Sri Lanka's (GOSL) medium-term plans and programs for agriculture concentrate on: (a) increasing agricultural employment as well as improving income levels, particularly for the economically disadvantaged groups; (b) increasing self-sufficiency in basic food items such as paddy, milk, sugar, fish and pulses so as to improve nutritional standards and to insulate the economy from excessive fluctuations in external markets; and (c) expanding export earnings both from traditional agricultural exports and 'non-traditional' items such as cocoa, coffee, pepper and spices. 2.02 The Government strategy for achieving these objectives had emphasized better utilization of the existing potential through institutional and policy improvements as well as rehabilitation of the production base, where necessary. Among policies and programs being given high priority by GOSL were establishment of appropriate price incentives for producers; strengthening of agricultural supporting services such as research and extension, credit, farm power supply and fertilizer distribution; improved irrigation water management; and stepped-up programs for rehabilitation of the plantation crops. 2.03 During FY76, the Government started to decentralize planning, monitoring and coordination by transferring more responsibilities and allowing greater autonomy to the district administration. To give an operational content to the decentralization programs, GOSL proposed to undertake a number of Integrated District Development Projects. These were to emphasize low-cost, quick-yielding, labor-intensive investments amid at better utilization of the existing potential by relaxing critical bottlenecks and constraints. 2.04 The Kutunegala Rural Development Project has been the first attempt to develop an entire district in an integrated manner. At appraisal, it had been expected that over the next decade all districts in  Sri Lanka would be covered by similar projects. This, however, did not fully materialize; by the end of 1988 only fifteen districts were covered (four IDA, Two IFAD and nine bilateral dor,ors), in three of which, activities had to be suspended due to civil disturbances. 3. Proiect Obiectives and Description 3.01 Objectives. The project aimed to assist the GOSL in evolving a replicable model of regional development for raising productivity, employment, incomes and living standards of the rural population through strengthening of the development-oriented institutional services, accompanied by modest but critical complementary investments in physical infrastructure. More specifically, the project was to assist the Government in helping the rural inhabitants of the Kurunegala District. 3.02 Because of the predominant position of paddy and coconut in the district's economy as well as the scope for large production increases for both crops, the principal project thrust was to rehabilitate the production base of these crops and to improve the required supporting services. 3.03 Components. At appraisal, the project was estimated to have an duration of five years with a total project cost of US$ 30 million equivalent and to comprise the following main components: (a) Irrigation and Water Management (US$ 5.50 million) i) improvement of nine major tank schemes; ii) rehabilitation of 500 village irrigation schemes; iii) improved water management for irrigation schemes. (b) Coconut Development (US$ 3.70 million) i) improved husbandry programs for 150,000 ac; ii) rehabilitation of 60,000 ac; iii) replanting/underplanting of 15,000 ac; iv) intercropping of 10,000 ac. (c) Agricultural Extension (US$ 1.00 million) i) provision for incremental staff and training; ii) provision of staff housing, vehicles and equipment. (d) Agricultural Input Supply (USS 1.05 million) i) establishment of a seed processing center; ii) construction of 80 fertilizer stores; iii) incremental staff salaries, vehicles and equipment. (e) Agricultural Credit (US$ 7.05 million) i) provision of short and medium-term loans; ii) improvement of credit operation.  -3- (f) Livestock (USS 0.15 million) i) construction of offices, clinics and staff quarters; ii) provision of equipment and vehicles. (g) Groundwater Exploration (USS 0.55 million) i) construction of 180 test wells; ii) provision for mapping, surveying, equipment and vehicles. (h) Rural Roads (USS 0.80 million) i) construction of 130 miles of high priority rural roads; ii) provision of construction equipment and vehicles. (i) Rural Water Supply (USS 0.60 million) i) installation of 400 dug wells. (j) Rural Electrification (USS 0.80 million) i) electrification of 16 rural areas; ii) provision of equipment and transport vehicles. (k) Health (US$ 0.35 million) i) construction of offices and staff quarters; ii) provision of ambulances, vehicles, generators and simple medical equipment. (1) Education (USS 0.70 million) i) construction of 140 science rooms, 40 teachers' quarters and upgrading of seven senior secondary schools; ii) provision of educational equipment. (m) Project Coordination and Investigation (US$ 1.05 million) i) vehicles, office equipment and additional staffing for project coordination and management both in Kurunegala and Colombo; ii) support for the preparation of a district statistical handbook, for the preparation of other development schemes in the district and the preparation of similar rural development projects in four other districts. 3.04 The irrigation and water management component which was almost entirely aimed at increasing rice production and the coconut development accounts for over 39Z of base costs. Together with the supporting credit, extension and input supply program, over 78Z of project base costs were almost entirely allocated to improvement in paddy and coconut production. In the following evaluation, major emphasis will therefore be given to crop production.  - 4 - 4. Proiect Design and Organization 4.01 Design. Project design was based on the Integrated Rural Development (IRD) concept which had already been adopted by GOSL. The main assumption underlying the project approach was that the existing economic and social infrastructure in the country and the district is substantially underutilized and prospects exist for a rapid and significant improvement through upgrading of institutional services and infrastructure as well as the provision of finance, especially for private sector investment in the production base. Although the project included 13 different components, it did not attempt to tackle all aspects of regional development. Instead the project design aimed at a pragmatic compromise between comprehensiveness and manageability. Due mainly to the diversity of its agro-ecological zones, Kurunegala was been selected as the representative district in Sri Lanka for a first IRD project to be implemented in the country and to serve as a model for replication in other districts. 4.02 Being the first of its kind, the project concept was innovative in its attempt to complement directly productive investments such as the rehabilitation of irrigation schemes (mainly for rice production) and coconut improvement with limited investments in transportation, health, education, etc.; to improve utilization and effectiveness of the existing facilities; to reduce intra-regional disparities and to develop the social infrastructure for growth in the long term. In the light of its objectives, the scope and scale of the project was appropriate. 4.03 Generally, the project was w,l prepared. However, with the benefit of hindsight, a longer project implementation period could have been foreseen. Also a better analysis of the agro-climatic conditions in the major coconut growing areas would have been useful to focus more sharply on the coconut intercropping component. Although the project stressed the need for strong political participation, it omitted to highlight the need for the involvement of the beneficiaries at an early stage, which adversely affected mainly the irrigation component. 4.04 The timing of the project was more appropriate as it came in the wake of the then recent liberalization of economic policy in the country and assisted further private investment in the production base. In addition, it supported the newly introduced decentralization program at a time when it was critical to demonstrate that this approach would lead to increased responsiveness of developmental efforts to local needs and priorities and would promote balanced regional development. 4.05 Organization. Due to its many facets, a total of around 20 agencies from 12 different ministries were involved in project implementation since the project was designed to utilize existing agencies rather than creating a new institutional set up. For the purpose of project coordination, a Project Steering Committee was established at national level and a Project Coordinating Committee at district level. A project office headed by a Project Director was in charge of monitoring, supervision and coordination of the work programs of the various agencies in the field.  4.06 To a certain extent, project pteparation suffered from a lack of clarity in defining objectives, roles and expected coordination for each sector and agency particularly where different agencies were called upon to work in the same sector. This was clearly observed in the coconut development sector which involved the participation of many agencies, like the Department of Agriculture (DOA) for extension, Department of Minor Export Crops (DMEC) for inter-cropping in coconut lands and the Coconut Cultivation Board (CCB) for the execution of the subsidy programs. The coordination among these agencies and their supporting roles were not sufficiently stressed at the time of the project appraisal. 4.07 The insufficiently defined roles in the implementation of the Unified Agricultural Extension System was another problem area identified at the project implementation stage. Under the proposed system, the DOA was assigned with the overall functions of agricultural extension including those crops coming under the purview of the DMEC and the CCB. As there was not general consensus among the involved agencies, the proposed roles of the DOA remained in ambiguity in these areas. Similarly, in the proposed component of groundwater exploration and rural water supply, the demarcation in respect of its implementation responsibilities was not sufficiently stressed at the planning stage. Subsequently, these problems had to be resolved both at national and district levels with the respective department and district heads, adding to delays in implementation. 4.08 Although the use of multiplicity of existing agencies may have contributed to a slower than anticipated project implementation, overall organizational experience has been positive and more cost effective than the establishment of yet another agency. 5. Prolect Implementation 5.01 The implementation experience of the various project components, with emphasis on the directly productive components and the most critical variances between planned and actual targets are discussed in the following paragraphs. 5.02 Project implementatior. was initially phased over a five-year period (mid-1979 to mid-1984). However, implementation did not progress as anticipated particularly in the initial period and therefore had to be extended twice by one year in order to achieve most of the set targets. Actual disbursements amounted to US$ 17.1 million or about 86? of the total IDA Credit of US$ 20 million. 5.03 The main reason for the slow start-up was the delay (up to 14 months) in the procurement of equipment and vehicles, which affected staff mobility and execution of work. This was mostly due to the inexperience of the agencies' staff with International Competitive Bidding (ICB) procedures. The other important reason for delay was a serious manpower shortage in the main implementing agency, i.e., the Department of Irrigation (DOI) which slowed down identification of tanks (only 130 were identified at appraisal) and design of rehabilitation work. 5.04 Due mostly to these delays, by mid-1984 only 50? of the IDA Credit funds were disbursed and less than a third of the directly productive  components Implemented. However, most of the social infrastructure components were completzd on time. 5.05 Irrigation and Water Management. The single most important component in achieving the rehabilitation and improvement of the paddy production base was the irrigation rehabilitation and impro*ement of water management. A total of nine major and 500 minor irrigation schemes were to be rehabilitated. By mid-1984, only three of the major and 263 of the minor irrigation schemes were completed due to various constraints (para. 5.03). However, by the end of the extended project period, all nine major schemes had been completed and 453 of the minor schemes. 5.Ot Although in quantitative terms appraisal targets were nearly achieved, the absence of an effective institutional mechanism to reflect farmers' needs with regard to design and implementing the rehabilitation work was a major deficiency and resulted, in several cases, in a reduced usefulness to farmers and a subsequent reduced commitment to repairs and maintenance. Another constraint was the lack of provision for the development of downstream works at minor irrigation schemes. This was rectified only in 1983 with the allocation of funds for the rehabilitation of 150 such works. 5.07 At appraisal, the need for the adoption of an improved package of agricultural and water management practices was stressed to optimize the benefits from the investments in irrigation improvement. The SAR specified that the 'Walagambahuwa' model, development at the Maha Illuppallama Research Station, be adopted. The principal hypothesis underlying this model water management program was that water saved during the main rainy (Maha) season can be used to cultivate a second crop in the less rainy (Yala) season. Dry sowing the first crop, rotational water issues in Maha season and crop diversification in the second season are some of the key elements in the efforts to improve the efficiency of water use. For this purpose, the appraisal report had emphasized the need for advancing cropping calendars, avoidance of staggered cultivation, use of short duration varieties and mobilization of group action through education and training. The implementation responsibilities were assigned to the Agrarian Services Department responsible for operation and maintenance of irrigation works. 5.08 According to the 1988 export evaluation review of the project carried out by the Agrarian Research and Training Institute (ARTI) the basic requirements for water management, both the trained staff and equipment, were only available in 1984 and a pilot program was initiated at 25 rehabilitated tanks in 1985. Experience gained there indicates that the model is best suited for adoption in the intermediate zone as rainfall is more regular, whilst for the dry areas of the district the model is less suitable. A larger sample with a longer time spread would be required before the introduction of an adjusted model throughout the district could be envisaged. So far, the attempt by the project to introduce improved water management in the district can be considered a failure. 5.09 Paddy production. The rehabilitation of major and minor irrigation schemes together with an overall increase in cropping intensity has resulted in an annual incremental production of 56,000 tons of paddy at  -7 - full development compared with the appraisal estimates of 59,000 tons. The next cultivated area was under-estimated but the total cropped area was over-estimated at appraisal thus resulting in a high cropping intensity (Table 1). Although a lower cropping intensity is assumed by the mission, based on the data provided by the district DOA responsible for the implementation of the project, the incremental cropping intensity (141) - mainly due to the provision of additional irrigation supply - is higher than the appraisal estimates (9X). 5.10 Input supplies. Supply of farm inputs for paddy production has improved considerably. The seed processing (2 tons/hour capacity) and storage complex was completed in 1983, one year before the original date of completion of the project. Although good quality seeds were produced, seeds of the reqsuired varieties were often not available at the time required. A jample survey by ARTI in 1985/86 indicated that only one-third of farmers obtained certified seeds from government agencies and others used seeds from previous harvests or from neighbors. Fifty-seven out of the revised target of sixty fertilizer stores were completed by 1986. However, only about 50X of the stores constructed for Multipurpose Cooperative Societies (MPCS) were fully utilized, partly due to the fact that farmers preferred to purchase fertilizers from private dealers from whom they could obtain credit readily and easily. 5.11 The provision of additional farm power for land preparation was a success. Although 200 units of 4-wheeled tractors were proposed at appraisal, only 103 units were provided and the program thereafter discontinued, partly because of poor financial resources of the average farmer and partly due to the operation of a similar program by another agancy and cases of misuse for non-agricultural purposes. The remaining balance and incremental funds were used for the purchase of 2,306 units of 2-wheeled tractors compared with the appraisal target of 500 units. 5.12 Coconut development. A comparison of appraisal targets and achievements is provided below: Appraisal Achievement Project Component Target 1984 1986 . . . . . . (acres) . . . . . . Rehabilitation 60,000 71,529 79,800 Replanting/underplanting 25,000 16,795 24,000 Increasing productivity 50,000 20,780 24,790 of matured plantation inter-cropping 10,000 4,100 4,415 5.13 Under the rehabilitation program, high priority was given by farmers to construction of contour drains for prevention of soil erosion, and provision of employment during the off-season. Only about 19Z of the farmers had obtained subsidies for removal of excess or unproductive palms because of the temporary reduction of family income. 5.14 The coconut replanting/underplanting sub-component did not progress as rapidly as anticipated at appraisal. Acreages completed by 1983 constituted only 612 of appraisal target although this increased to  - 8 - 902 at the revised project completion date. Development was slow and unsatisfactory in the first three years due to an insufficient supply of coconut seedlings, delays in identification of smallholdings and solving of land disputes. Other contributing factors included low subsidies and delay in subsidy payment to farmers. Sub;Ady rates were increased three times (1979, 1982, 1984) but were not commensurate with increasing production costs due to inflation. it is reported that only about 341 of the original participants of the scheme qualified for the second instalment; 172 for the third payment and only 12 was eligible to receive the fourth and final subsidy payment. Non-eligibility was due to the inability and reluctance of the farmers to complete specified works such as removal of senile or excessive palms and maintenance of underplanted palms. 5.15 In order to increase productivity of mature plantations, the SAR anticipated the provision of credit to enhance fertilizer use on 150,000 ac. However, only about 172 of the area benefitted from this program. Low usage of credit facilities for fertilizers was due to lack of farmers' awareness about the credit scheme and the high cost of fertilizers relative to farm-gate prices of nuts. Only about two-fifths of fertilizer users were smallholders owning 5-10 acres.1 5.16 The coconut intercropping sub-component was a partial failure due mainly to the selection of locations with inadequate rainfall; land with unsuitable soils; coconut holdings with excessive shade and inappropriate selection of crops. In addition, many farmers resisted this introduction as the technology of cultivation and processing was new to them, besides requiring a long gestation period of development and high costs. Also, the quantum of subsidy had been inadequate, particularly for smallholders. The system of providing subsidy payments in four equal annual instalments hap not been helpful to smallholders since the bulk of the development costs were incurred during the first two years. In view of the unsatisfactory progress of this scheme, the home garden and horticulture program(mainly bananas, mango and citrus) was introduced in 1983. 5.17 Eight new nurseries were established and two existing nurseries were improved for the production of coconut seedlings. Except in 1980 and 1983/84 (severe drought) sufficient seedlings were produced for supply to project participants. Nurseries were also established for the production of planting materials of minor export crops such as cocoa, coffee and pepper. Since the demand was less than expected, the nurseries were also used, commencing 1985, for the production of bananas, ginger, pineapple, etc. for supply to various home garden cultivation programs especially in the settlement schemes. The establishment of 1,450 home gardens was approved during revision in 1984 and 1,422 were established by project closing. 5.18 The ARricultural Extension Service was consolidated and reorganized to function along the lines of the training and visit (T & V) system, although certain aspects required improvement. The capacity to expand extension activities was improved. Closer linkage between extension and research was forged through the setting up a Research and Training I/ Sample survey by Agrarian Research and Training Institute.  9 - Group which met regulerly and the training of extension staff as well as farmers was strengthened. 5.19 The envisaged unified extension service did not materialize. The advisory services of the CCB, DMEC as well as the Department of Animal Husbandry continued to render technical advice in their respective fields. However, the ratio of farm families per village level worker (KVS) vas reduced from 1,200 in 1979 to 850 at project completion, which is slightly higher than the appraisal target of 600. Fortnightly visits were made by KVSs, though not very regularly because they were still involved in work unrelated to extension. 5.20 Staff accommodation and mobility were improved under the project. However, about 501 of the quarters constructed for KVSs under the project were not occupied. Reasons for non-occupation included the lack of basic amenities such as drinking water and electricity and lack of medical and educational facilities in the localities of residence. Jeeps (12) and mini-buses (2) provided under the project have allowed agricultural officers to carry out supervisory and training activities also in remote areas of the district. Only about half of the total number of pedal cycles proposed (277) have been purchased by KVSs. 5.21 Agricultural credit. In order to facilitate farmers' access to inputs, the project had allocated US$ 7.05 million (including US$ 2.4 million and US$ 2.1 million IDA Credits for crop production and equipment and vehicle loans respectively) for the provision of short-term loans mainly for paddy cultivation and medium-term loans mainly to support the fertilizing program on 50,000 acres of coconut and the purchase of tractors. In addition, the credit operation was to be improved including the recovery of overdue loans. 5.22 As discussed above (para. 5.11), the restructured tractor credit program well exceeded its original target, the crop loans for rice cultivation and coconut fertilizer, however, showed a very poor disbursement performance (see Table 2). 5.23 Loans amounting to only Rs 84.3 million were provided for the latter category by the two participating banks (People's Bank - PB and Bank of Ceylon (BOC). The Development Credit Agreement, however, stipulated that Category 4 of the IDA Credit would only cover annual loan expenditures by the banks which exceed Rs 7 million (as the bank's contribution) and then only to cover 40X of the excess. As a result, only about 251 %US$ 0.6 million) of the IDA Credit target was disbursed (see Table 3). 5.24 It was mainly the non-availability of trained staff for the promotion of credit at farm level, slow and cumbersome credit procedures, increased transaction cost and more stringent rules in determining the credit worthiness of borrowers which affected crop loan disbursements. In uidition, generally declining net returns per acre cultivated as well as the impact of drought years on yields contributed to the poor performance in this component. 5.25 Another aim of the credit component was the recovery of some Rs 16 million in overdue debts (which were disbursed prior to the project) owed  - 10 - by an estimated 25,000 farmers in the district to the PB (Re 11.2 million) and the BOC (Rs 5.3 million). With regard to the BOC, recovery amounted to over 932, whilst only 35.7Z of the outstanding loans were paid back to the PB (as of December 31, 1986). 5.26 According to the ARTI Ex-Post Evaluation Study, the limited success of the latter could be attributed largely to the laxity on the part of the MPCS through which most of the agricultural credit had been disbursed before 1979. Besides the laxity in screening of individual loan applications, some of the societies had failed to maintain detailed records of loans disbursed, thus making the recovery of overdue loans difficult. Consequently, the PB in disbursing agriculture credit had withdrawn its links from 12 out of 14 MPCSs in the district. 5.27 Livestock. As foreseen, the number of veterinary ranges was increased from nWne to fourteen; the three clinics, offices and staff quarters were constructed in remote locations and the required vehicles provided. Although small, the sub-component was considered as very useful as it allowed vaccination programs to be carried out on time and pasture and fodder production increased. However, more attention should have been given to enhancing the facilities for artificial insemination. 5.28 Groundwater exploration. Of the 180 envisaged test wells, only 88 were constructed as this number was considered sufficient to provide the required data base. The remaining funds were used to construct 200 tube wells. Fiftv-four of the test wells were later converted into production wells whilst about twenty of the tube wells are now reported defunct. 5.29 Social infrastructure. The investment included the construction of 130 miles of rural roads, the provision of 400 dug wells, the electrification of 16 rural areas, improvement of the health care delivery system in 15 locations as well as the upgrading of educational facilities and equipment. With the exception of minor changes (see Part III/4), there were no major variances between planned and actual implementation of the various sub-components. 5.30 Proiect coordination and investigation. The project office was established in time and although different project managers led the project through implementation, there were no significant disruptions. It is, hotever, noted that although the project was to a large extent oriented towards agricultural production, no agricultural officer was assigned to the project office to assist with the implementation and monitoring and evaluation of agricultural operations. The office prepared the envisaged statistical handbook of the district and prepared similar IRDP projects for four other districts in the country which were taken up by IDA or bilateral donors. 5.31 Cost. In US$ terms, total project cost exceeded the appraisal estimate by about 122. Overall IDA Credit utilization, however, fell short by US$ 2.9 million (14Z) due mostly to the poor disbursement performance under the crop loan category. 5.32 In contrast, total project cost in local currency was estimated at RS 465 million with a GOSL contribution of Rs 80.5 million. Instead, total  - 11 - project cost increased to over Re 726 million with a reaulting Government contribution of over Rs 152 million which had to be covered by the GOSL Consolidated Fund. These cost overruns are due mostly tos (a) the increase in the average level of prices over the project period was much higher than what was provided for at the time of project appraisal; (b) additional items of work undertaken during the project period; and (c) depreciation of the Sri Lankan currency in relation to foreign currencies.2 From the above, it is quite evident that physical and price contingency provision allowed at the beginning of the project was inadequate. 5.33 Evaluation. Most of the above variances could have been avoided through more detailed project preparation efforts. In particular, more attention should have been paid to organizational matters including the need to strengthen the DOI and to ensure better inter-agency coordination. The fact that the officials of the agencies involved in project implementation were not familiar with ICB procedures should have been recognized and technical assistance and/or training provided. 5.34 A more careful assessment of the agro-climate conditions of the district would have helped to better formulate the coconut development component and to define the horticultural sub-component at an early stage. Also the downstream development works of minor irrigation schemes could have been identified at an earlier stage, if more detailed survey work would have been carried out. More accurate costings of civil works, based on detailed programs and specifications, would have avoided cost overruns and delays. 5.35 The project riots were rightly identified as beings (a) the ability to achieve the necessary coordination in implementing the various inter-related components; (b) the adoption of intercropping practices by coconut smallholders; and (c) the shortage of professional and technical staff, particularly design and construction engineers. However, insufficient provisions were made at appraisal to minimise these risks. 5.36 In retrospect, the decision to extend the project by another two years has proven to be useful. It enabled the agencies involved to use the experience gained during the first years to execute the remaining components in a fast and efficient way for the benefit of the rural population in the district and valuable experience was gained for similar: projects in other districts. At the same time, or even earlier, more decisive and concrete steps should have been taken to ensure a greater level of beneficiaries' participation in the identification process, in particular with regard to the irrigation rehabilitation and the rural water supply components. 6. Project Results 6.01 Project Objectives. The project as a whole can be considered a success as it achieved its main objectives, namely to increase paddy and coconut production, to create employment and improve income and living standards of the rural population in the district. It has also served as a 2t At appraisal, US$ 1 = Rs 15.5 at credit closing USS 1 - Rs 28.  - 12 - model to be replicated in other districts and many of the lessons learned during implementation were incorporated in subsequent IRD projects. In addition, the project has helped to strengthen the social infrastructure in the area. It is, however, too early to assess the project's impact on the reduction of inter-regional disparities. Throughout project implementation, none of the project objectives were changed. 6.02 Physical Results. At appraisal, it has been estimated that as a result of the project the area under irrigated rice would increase by 10,000 acres, that over 60,000 acres of coconuts would be rehabilitated and 15,000 ac replanted and 10,000 ac intercropped. Together with improved crop husbandry and increased fertilizer use, it has been estimated that incremental production would amount to 49,000 tons of paddy and 155 million coconuts at full development in 1986 and 1991 respectively. Results of other cropping activities such as horticultural crops and MECs were not specifically detailed but considered as a secondary but still important source of income. 6.03 For paddy, crop yields were very much underestimated at appraisal for both the with and without project situations. Modest yield increases are assumed with the project and comparison of yields at full development with appraisal estimates are summarized below: Maior Irrigation Minor Irrigation Rainfed Average Schemes Schemes Without With Without With Without With Without With -............ (tons/acre) .........0...... SAR 1.24 1.41 0.96 1.08 0.90 1.02 1.00 1.12 PCR 1.73 1.93 1.40 1.49 1.06 10.6 1.36 1.44 Yield increases with the project are due to more assured irrigation and small increases in the rate of fertilizer application. Revised crop budgets per acre are based on data provided by the DOA. Details are given in Tables 4 to 6. At full development, annual incremental paddy production is estimated at 56,000 tons compared with the appraisal projection of about 49,000 tons (Table 7). 6.04 Performance of the coconut sector and the minor export crop sub- sector (coconut intercropping) was significantly affected by the drought years. Although the civil works program was completed as scheduled and all the nurseries established and fully productive, the field works program of intercropping and replanting/underplanting suffered a setback. 6.05 Based on project data, it is estimated that y.elds will average 1,700 nets/ac at full development in replanted/underplanted areas and 2,300 nuts/ac in rehabilitated areas. Increased productivity of mature plantations is expected to result in about 2,500 nuts/acre. The average yield of coconuts arising from project activities is 2,350 nuts/acre. 6.06 It is estimated that the project resulted In an incremental yield of 880 nuts/acre compared with the appraisal estimate of 400 nuts/acre. Contrary to appraisal anticipation of yield increase from 1,700 nuts/acre  - 13 - pre-project to 2,000 nuts/acre without the project, it is assumed that there would be a decline in yield (evident presently) to 1,470 nuts/acre without project intervention, due to farmers retention of old and unproductive palms and deteriorating husbandry practices. 6.07 Appraisal estimates assumed that the entire coconut area (370,000 ac) in the district would benefit from coconut development given an annual increzental production of 144 million nuts at full development in 1991. However, project benefits only attributed to areas affected by the project (see Tile 8) in order to reflect the actual situation. The incremental production is summarized below: Without (.. ...... .Without Project .............) Proiect Annual Production Incremental Production 1992 2003 1992 2003 ....................... .(million nuts).. Replanting/under- 14.4 25.7 40.8 11.3 26.4 planting Rehabilitation 127.7 183.5 183.5 55.8 55.8 Increasing produc- 37.3 62.6 62.2 24.9 24.9 tivity of matured plantations TOTAL 179.4 271.4 286.5 92.0 107.1 6.08 Full development is expected to be attained in 1990 in the case of rehabilitation and increasing productivity of matured plantations but only by 2003 under replanting/underplanting because of delayed and slow underplanting of palms. The estimated net income per acre for the different coconut schemes at full development is given in Tables 9 to 11. 6.09 Although, at appraisal it was assumed that about 200,000 acres of coconut lands would be suitable for intercropping and only 52 was proposed to be tackled by the project, implementation experience haR shown that even this modest target was over-optimistic. Agro-climatic, soil and topographical conditions combined with manpower constraints prevented a further expansion of the intercropped area. Also the cumbersome subsidy procedures in connection with the minor export crops program and the sometimes poor quality of planting material, further affected the program. 6.10 As a result of the above, the incremental coconut production is expected to reach a maximum level of 107 million nuts in 2003 (i.e., 252 less than appraisal, with ten years delay) and production of MECs will reach only about 40? of its anticipated production level, partly compensated however by horticultural production which was not foreseen at appraisal. 6.11 Economic Rate of Return. The economic rate of return (ERR) at appraisal was estimated at 322. This ERR was derived by including all investment costs (except price contingencies and agricultura. credit) and post-project costs for the completion of coconut development works started under the project, as well as imputed cost of family labor used in coconut development works. Also included were incremental crop production costs as  - 14 - well as operation and maintenance (O & M) costs during project implementation and subsequent 0 & M costs for project works and services. On the benefit side, the analysis was based only on incremental production of paddy and coconut. 6.12 Using the same assumptions, the re-estimation of the ERR has been carried out in 1989 constant prices. Past expenditures excluding taxes and duties were adjusted to 1989 constant prices by using GDP deflator factors.3 All local costs, except for labor, were converted into border prices by using the SCF of 0.75. For labor, a SF of 0.80 was applied. For the projected 1995 equivalent paddy prices, values of Rs 6,139/ton were used (Rs 3,185.0 at appraisal) and for coconuts Rs2.79/nut (Rs 1.10 at appraisal) were calculated based on World Bank Commodity Prices and Price Projections as were fertilizer prices (see Tables 12 and 13). 6.13 The re-estimated ERR of the project is 38Z. Factors contributing to the increase above the SAR estimate of 32? are the higher incremental price production and a steep increase in the economic value of paddy and coconuts (see Table 14). 6.14 Financial performance. At appraisal, it was estimated that net farm incomes would increase by between 15? to 342 but actually the increase amounted to only about 102 or even less. This lower than expected increase can be mainly attributed to a considerable increase in fertilizer cost (due to the phasing out of subsidies), a rather stagnating price for paddy and in the case of coconuts a depressed international demand. Nevertheless, the incremental income (after labor cost) for paddy under major and minor irrigation schemes with about Rs 500 and Rs 200 per acre respectively is still attractive enough to induce farmers to adopt improved technology. Similar conditions apply for coconut improvement where income increases between Rs 1,300 and Rs 1,500 per acre for replanting, rehabilitation and improvement of mature plantations respectively are calculated. 6.15 It has also been anticipated that about Rs 28 million (in 1979 prices) per annum would be received as taxes and duties on incremental coconut exports assuming that 602 of the project related increase would be exported. Due to the partial failure of this component and a reduction in export levies, revenue generated by the export of 602 of the incremental production is expected to reach only Rs 24 million by 1991 and is expected to slowly increase to Rs 31 million by the year 2003. 6.16 It was further assumed that about Rs 20 million per annum would be saved due to procurement of one-third of the incremental production of paddy for ration at below world market price. During the project period, Government paddy purchases from producers in the district fluctuated between 1I and 4X with an annual average of 1.8?. Based on an incremental production of about 56,000 tons (at full development) and total Government purchases of about 1,000 tons, total savings would amount to US$ 40,000 (Rs 1.32 million). 3/ Wholesale Price Index during 1979 to mid-1989 as follows: 188.0, 229.5, 268.6, 283.3, 354.1, 444.7, 377.1, 366.0, 414.0, 488.7 and 525.6.  - 15 - 6.17 Incremental 0 & M cost of project facilities and services vere expected to reach Rs 4 million at full development. The fact that not all incremental staff had been employed resulted in a reduction in expenditure on salaries, and a parallel reduction in vehicles, vehicle operating and replacement cost. However, 0 & M costs for irrigation works and road maintenance are expected to increase and total 0 & M costs are now re- estimated at Re 8 million (Re 6.4 million in economic terms). 6.18 Impact of Proiect. The project had a significant impact on the agricultural sector of the district, particularly in the areas of paddy production and to a lesser extent on coconut development. Although the project accounted for nearly 50S of the Annual District Capital Budget, it was still a rather small investment (with an average annual loan disbursement of US$ 2.8 million) considering that it was meant to affect over 270,000 ac (sown area) under paddy and 370,000 ac of coconuts. Besides the project's impact on the grow'th of the agricultural sector, it also influenced the overall policy approach to rural development by providing and testing a replicable model of decentralized accelerated development with a major thrust on the rehabilitation of the production base and minor investments in infrastmActure and social services. 6.19 By utilizing and strengthening the existing agencies, the project had a substantial impact on institution building especially by improving the extension services and enhancing the monitoring and coordinating capacities in the overall district institutional setup. As foreseen at appraisal, field and supervisory staff of the extension service were increased, including the provision of adequate coverage in remote areas and the T and V system was partly introduced. During the project period, the Project Officer successfully managed the coordination of largely autonomous government agencies, which in itself was already a considerable achievement. At the time of termination, the Project Office and the District Planning Unit were amalgamated to work functionally as a more broad-based District Planning Office with a view to undertaking district planning and monitoring with the experience and capacities developed during project implementation is a further example of the institution building impact of the project. 6.20 A positive trend was observed in income distribution. In FY79, before the commencement of the project, 19Z of the population were below the poverty line of Rs 3,600/year. According to the ARTI ex-post evaluation in FY86, only 17Z were below the poverty line of Rs 7,000/year.4 This achievement, as the ARTI study concludes, is not solely attributable to the project as the contribution of other national programs was also substantial. However, the inflow of relatively larger allocations of funds under the project has brought a sizeable impact on the distribution of income, particularly in the dry zone of the district where project focus has been significant. It is assumed that more than about 502 of the investments received by the dry zone during the period 1979/86 came from project provisions. In addition, the project created about 4 million man- days of direct employment through labor-intensive works and about 2.5 million man-days of field work under the coconut development program. 4/ Adjusted upwards to reflect increase in cost of living.  - 16 - 6.21 The physical environment was little affected by the project since it was mostly geared to rehabilitating existing facilities. however, the construction of contour drains on about 20,000 acres contributed substantially to improving land quality through reduced soil erosion. Besides the electrification of 17 selected rural areas, the project did not have any major impact on the technological environment and relied on the proven technology which was already available within the district. 7. Proiect Sustainability 7.01 Over three years have passed since Credit closure (June 1986) and all facilities created *mder the directly productive components (see Part III/5) were handed over to the respective agencies before or on that date and all irrigation schemes are still in full working order. Staff strength of the extension service has been maintained and even slightly increased. However, due to the liberalization of fertilizer marketing, the fanm input stores created under the Department of Agrarian Services are underutilized. The state of repair of social infrastructure established under the project remained satisfactory. 7.02 However, the absence of any water charges for the major minor irrigation schemes makes them entirely dependent upon scarce and fluctuating budgetary allocations for repair and maintenance which raises doubts about their sustainability in the long run. According to the ARTI ex-post evaluation report, the short and medium-tenm loan program introduced under the project shows satisfactory recovery rates compared with all island levels or pre-project levels. The overdue rate of crop loans has been f;luctuating between 2Z and 4.8Z afte: the project implementation period and in the case of equipment and machinery loans stabilized at around 42 during the last project year. Together with improved procedures, the credit component appears sustainable. 7.03 GOSL continues to support the investment undertak.n by the project including the annual 0 & M costs for project works and services. Also, as proven by the continuously increasing cropping intensity, the farLing community has generally accepted the improved crop husbandry techniques introduced by the project. The above are indications that the project will probably maintain an acceptable level of net benefits throughout its economic life. 8. Bank Performance 8.01 Throughout the project cycle, the Bank's involvement has been quite intensive. Being the first IDA-funded Integrated Rural Development Project in Sri Lanka, initial preparation of sector proposals and subsequent appraisals extended over a considerable period. The initial project proposal prepared by the line agencies had to be reviewed by five missions before it was finalized. This, together with further delays due to the June 1977 general election and subsequent change in Government, resulted in a rather long preparation period (Part III/2). 8.02 In general, the project was well conceived but displayed some deficiencies which were rectified in collaboration with subsequent supervision missions. A total of 13 supervision missions were fielded over  - 17 - the seven-year period (Part III/8). Although most of these missions were multipurpose missions, supervising other projects or providing preparatory assistance, all were considered beneficial to the project and properly timed. 8.03 Realizing, however, the difficulties implementing agencies had with ICB procedures, the supervision missions should have taken rectifying actions earlier. The same applies to the irrigation and water management components where lack of sufficient technical staff in the DOI prevented a timely implementation. The supervision missions should have advocated the immediate use of the consultancy funds available for that purpose. 8.04 The main lessons learned from project implementation that may be relevant to other Bank-financed projects are: i) project design should focus more clearly on project organization and management especially when a multitude of implementing agencies are involved; ii) in the case of projects which are innovative in type and design, greater allowances for flexibility and a longer start- up period should be allowed for; iii) where new procedures are to be followed by the borrower (such as ICB), sufficient allowance for training and/or technical assistance should be made; iv) greater participation of beneficiaries in project design and implementation should be sought and encouraged to focus more clearly on their needs and ensure their acceptance; v) the project's impact on beneficiaries should be measured through clearly defined evaluation measures; and vi) corrective actions should be taken as soon as delays in component implementation occur to avoid slippage of the project as a whole. 9. Borrower Perfor..ance 9.01 Generally, the borrower performed very well throughout the whole project cycle. All ministries and agencies collaborated fully with the UNDP and World Bank missions during the preparation stage and were supportive to the supervision missions and open to suggestions for improvement. 9.02 The borrower adopted all the agreed covenants (Part III/7), with submission of audited accounts being the only area erperiencing considerable delays. The main weaknesses of the borrower were their unfamiliarity with ICB, leading to considerable delays in the procurement of important items and its initial failure to attract sufficient technical staff for the implementation of the irrigation and wate-r management components.  - 18 - 9.03 Regarding project management, the borrower performed particularly well. Even though three different project directors headed the project during its implementation period, the difficult task of coordinating the numerous agencies towards the common project goal was achieved. Also financial management and monitoring of accounts was satisfactory throughout. 9.04 Staff resources were adequately deployed and trained. However, the borrower's failure to employ consultants for the irrigation and water management component, as foreseen at appraisal, has contributed to the delays in implementation. 9.05 Experience gained by the borrower was that: General (a) The model (blue print) approach to standardize replicable IRD projects has the great advantage of better withstanding local pressure. This rigidity may, however, hamper the diversion of resources into areas and/or sub-components found to be more feasible than those suggested at appraisal. (b) Adequately trained staff, in sufficient numbers. have to be recruited at an early stage of the project to provide the required implementation capacity. (c) The Government should not hesitate in deploying local and foreign consultants if expertise is not available within its own institutions, to avoid project slippage with all its adverse economical and financial implications. Paddy production (d) In minor irrigation systems, the total number of tanks to be rehabilitated should not exceed manpower availability. Problems of water management practices by farmers would be reduced if there were early participation of agencies such as the Agrarian Service and the Agricultural Extension Services and farmers' groups, starting from the selection of tanks for rehabilitation. Sustainability of these practices would be in doubt unless there is strong leadership and high motivation of farmers. (e) In agricultural extension, contact farmers selected should be progressive and dynamic and should have the ability to communicate readily with farmers for whom they are responsible. In areas where fortnightly visits or meetings have not become fruitful because of lack of introduction of new technology or extension messages, it would be more beneficial for extension workers to conduct adaptive trials or demonstration plots on water management practices, etc.  - 19 - (f) More storage facilities, distribution or sale centers for the supply of good quality seeds of the required varieties and fertilizers would have to be provided by government agencies in the more remote areas (i.e., dry and semi-dry zones) rather than in areas likely to be served by private dealers. Coconut development (g) In the replanting/underplanting scheme, farmers carried out underplanting without removing senile and unproductive palms, resulting in inappropriate spacing and excessive planting density. Removal of unproductive palms should be regarded as a prerequisite prior to the provision of other project inputs. Farmers would be encouraged to remove these palms during the off-season if sufficiently renumerated for their labor, and also given planting materials and fertilizers for the cultivation of quick return cash crops to compensate for the temporary loss of income. (h) The provision of fertilizers (in kind) in an adequate amount in the first five years for the underplanted or new palms, is considered necessary for high nut yields. These proposed incentives are also applicable to coconut rehabilitation, although cash crop cultivation should only be encouraged if there is no overshading from existing palms. i) The construction of contour drains was popular in that it gave the farmers additional labor wages during the off-. season. However, the need for such drains should be more closely and objectively reviewed. (j) In the intercropping of coconuts, the introduction of crops (cocoa, pepper and coffee) which are new and have long gestation periods, has not been accepted by farmers, particularly at locations with inadequate rainfall, on smallholdings of less than two acres, without orocessing facilities and know-how, and marketing outlets. Alternative crops proposed include mango, pineapple, banana and papaya, provided existinb palms do not provide overshading. (k) There is a need for more effective streamlining of coordination amongst agencies fin the production of planting materials, provision of credit for fertilizer purchase, subsidy schemes and extension activities. 10. Project Relationship 10.01 The Bank-Borrower relationship was satisfactory throughout the project period. A fruitful dialogue was maintained between the national government, as principal borrower, and the implementing agencies on the one  - 20 - side and the Bank on the other which cot,siderably helped to achieve the project objective. 11. Consulting and Contractual Services 11.01 The inital failure to recruit consultants to supplement the limited manpower resources of the Irrigation Department contributed to delays in project implementation. Subsequently (in 1981), this problem was overcome by obtaining the services of planning consultants appointed with the provision made available under IRDP II. These consultants were to assist the department in the surveying, preparation of plans, designing and elaboration of cost estimates for the irrigation schemes. The services of these consultants were useful to the department by improving the quality of designs and estimates and by reducing the department's workload in this area. 11.02 In addition, also under IRDP II, consultants were provided to assist the project management in the supervision of engineerilng works and in accounting and management functions. With regard to the supervision of engineering works, the consultants' services were considered useful, whilst with regard to accounting the management, the assistance was negligible as sufficient skills in these filds were available. 11.03 At appraisal, Rs 1.5 million were included for contracting local research institutions and firms to undertake monitoring and evaluation studies which were subsequently contracted to ARTI. A total of 11 studies (annex 1) were carried out, mainly based ou field investigations and sample surveys. Generally, the performance of ARTI was satisfactory. However, the inexperience of staff with IRD projects, on occasions let to quality problems. Also the required interaction between project staff and ARTI was not always maintained. 11.04 The execution of earth and construction works was the responsibility of the relevant implementing agencies and carried out on a force account basis by using direc'A labour and machinery or on a contract basis. In general, design and quality specifications were adheared to. Only in the case of a few local contractors executing irrigation works, sub-standard results were observed which had to be rectified later. The reason for their poor performance was due mainly to a lack of experience with irrigation works and especially the need for adequate compaction, as well as their inadequate understanding of the English contract documents. Subsequently, detailed contractor briefing and a translation of documents in local languages brought considerable improvement. 12. Project Documentation and Data 12.01 As noted in Part III, the borrower adhered to practically all legal agreements. It was, however, observed that the prosecution of defaulters as stipulated with the framwork of the agricultural credit component was not implemented with the necessary vigour. This led to poor recovery, especially for long overdue loans, which were given prior to the project, in particular those obtained through MPCSs, and did not help to improve repayment morale. Supervision missions should have paid greater attention to this adpect.  - 21 - 12.02 The SAR provided a useful framework for the Bank and the Borrower to implement the project and monitor its progress against the targets set. However, the two successive one-year extensions and the high depreciation of the Sri Lankan rupee reduced the SAR's usefulness especially in providing a financial guideline. The usefulness of the document could have been further increased if project organization and management had been formulated with greater care9 more attention had been given to project design and sufficient base-line studies carried out to focus the agricultural development components more sharply. 12.03 In spite of the fact that the Project Officer maintained satisfactory project implementation records and that ARTI carried out a number of evaluation studies, not all data relevant to the preparation of the PCR were readily available. This was mainly due to the fact that the data were not collected and processed to serve as a basis for a broad ex- post evaluation but rather for short-term monitoring of project results against targets. Furthermore, it appears that the Borrower has not been sufficiently well briefed on the format and required techniques to be used for PCR preparation. 12.04 To facilitate PCR preparation, the Borrower should be provided with very detailed instructions including tailor-made outlines of tables to be prepared.  - 22 - PART II PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE The Asia Regional Office sent Parts I and III of this PCR to GOSL in June 1989 with the request to prepare Part II by Oc -.ber 30, 1989, but none was received.  - 23 - PART III - STATISTTEAL NPFNRMATTlM 1. Related Bank Loans and/or Credit Year of CreditlTitle Pu!L #mm& Status Cents Cr. 66G-CE To ipwove irrigation 1977 Copleted No major probles were Tank Irrigation water suply. reduce encountered. Iodernization seepage losses, provide Inputs, strengthen agri- cultural extension ser- vices and water manage- ment In five major tank sch_mes. Cr. 819-CE To diversify about 9.500 1978 Cupleted Encountered mjor imple- Tree Crops ha of arginal tea lad mentation problems due to Diversificea- into minor export crops inadequate budgetary allo- tiov. I (pepper. cardamom. cloves. cations. administrative etc.). establish forests and political opposition. for fuelwaod and provision of SW, dairy cows. Cr. 1079-CE To iwrove productivity. 1981 Extended Progressi.4 SAtisfac- Rural Develop- incme and living stan- by one year torily. oent II dards in Natale ad to 6/90. Puttalsm districts through provision of inputs and services and stimulating planting of coconuts and sinor export crops. IFAD 111-SR To raise Inome and living 1962 Cwpleted world Bank as executing Rural Develop- standards in the Sadulla agency. ment Badulla district by iqwoving sup- District ply of agric. inputs and services and stimlating cultivation of minor export Crops. Cr. 1363-E To increase inac and 1983 Extended lnpleaentation seriously Rural Develop- living standards of the by one year affected since Decenher ment II1 rural population in Panr to 12/84 1984 by poor security and Vavuniya districts by situation. iqpoving supply of agric. inputs and services. ensure timely irrigation water dis- tribution and enhace soCial infrastructure (education. health and water supply).  * - 24 - IFAD Ln.179 SR Increase incomes and 1986 Progress- World Bank as executing Rural Develop- employuent for rural pwor lug without agency. ment Kegalle In Kegalle District through major problems. integrated act of investments Expected comple- in agriculture (sinor export tion date 12 /92. crops and tea) soal Indus- tries, rural infrastructure and social services. Cr. 1160-CE Rehabilitation of 1.200 1961 Extended for Iaplementation In several Village Irriga- village irrigation schemes, third time by districts suspended due to tion Rehabilit. modernization of 500 schemes. one year to poor security situation. Project Introduction of systematic 12/89. water management progranes. 2. ProJect Timetable Item Date Planned Date Revised Date Actual Identification October 1976 December 1976 Preparation / August 1977 Pre-appraisal October 1977 April 1978 Appraisal mission August 1978 Project Brief October 1978 Loan/Credit February 1979 February 1949 Negotiations Board Approval March 1979 Credit Signature April 1979 Credit Effectiveness July 1979 August 1979 y Loan/Credit Closing Deceer 1983 December 1964 December 1985 Loan/Credit Dis- June 1984 June 1985 June 1986 burse ent Closing ) Prepared by 6OSL with assistance from IJP and World Sank missions. p/ With retroactive finance starting January 1979.  - 25 - Comments: World Bank identification mission did not include 'social and economic overheads' in the project or small rural Industries as initially reported by GOSL. This, however, was later resolved by including components on rural education and health, water supply, electrification and social services. No other main issues arose during the project cycle other than those .escribed in Part I reflecting the implementation history of the project. 3. Credit Disbursement Cumulative Estimated and Actual Disbtrsements (USS millocn) 1988 1981 1982 1983 1984 1985 1986 Appraisal estimate 3.00 8.00 12.00 16.00 20.00 - - Actual 1.57 3.38 5.00 7.60 9.70 11.60 17.IC I/ Actual as % of estimate 52 42 42 48 49 58 56 :/ Including payments up to final disbursoent date. Date of final disburseent - January 21. 1987.  4. *Prnaet Inhplementation Cnmaonanat Unit SAR Tarcet Achievement Revised wu