Sri Lanka - Project Completion Report

Docu_mt of The World Bank FOR OFIC IL USE ONY MICROFICHE COPY Rexwt No. 10191 Report No. 10191-CE Type: (PCR) DONALDSON,! X31680 / T9 032/ OEDDI. PROJECT COMPLETION REPORT SRI LANKA MAHAWELI GANGA DEVELOPMENT PROJECT IV (LOAN 2437-CE AND CREDIT 1494-CE) THIRD RURAL DEVELOPMENT PROJECT (CREDIT 1363-CE) DECEMBER 23, 1991 Agriculture Operations Division Country Department I Asia Regional Office This document has a restricted distribution and may be used bv recipients only in the performance of their official diuties. Its contents may not otherwise be disclosed without World Bank authorization.  :~~~~~~~~~~ ANDREVZATIONS AND ACROhNy AMDP - Accelerated Mahaweli Development Program CIDA - Canadian International Development Agency EtC - European Economic Community zCU - European Currency Unit ERR - Economic Rate of Return GOSL - Government of Sri Lanka IFAD - International Fund for Agricultural Development IRDP - Integrated Rural Development Project NDK - Nagolla Dee-ela Kadiunne-ela QECF - Japanese Overseas Economic Cooperation Fund SFD - Saudi Fund for Development  FORx OFFCIAL USE ONLY THE WORLD BANK Washington. D.C. 20433 U.S.A. Office of 0itectot-Cene' Opetatint Evak.MtUin December 23, 1991 MEMORANDUM TO TiHE EXECllTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Sri Lanka - Mahaweli Ganga Development Project IV (Ln. 2437/Cr. 1494-CE); and Third Rural Development Project (Cr. 1363-CE) Attached, for information, is a copy of a report entitled "Project Completion Report on Sri Lanka - Mahaweli Ganga Development Project IV (Ln. 2437/Cr. 1494-CE); and Third Rural Development Project (Cr. 1363-CE)" prepared by the Asie Regional Office with Part II of the Third Rural Development Project contributed by the Borrower. No audit of these projects have been made by the Operations Evaluation Department at this time. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. This~~ doum- ha..etitddsrbtonadmyb sdb eiiet nyi h erac  FOR OFFICIAL USE ONLY PROJECT COMPLETION REP SRI LANKA MAHAWELI GANGA DEVELOPMENT PROJECT IV (Loan 2437-CE; Credit 1494-CE) THIRD RURAL DEVELOPMENT PROJECT (Credit 1363-CE) Table of Contents Pate Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evaluation Summary . . . . . . . . . . . . . . . . . . . . . . . ii PART I - PROJECT REVIEW FORM BANK'S PERCEPTIVE A. INTRODUCTION ................ .. ... 1 B. POLITICAL AND ECONOMIC BACKGROUND Government and Bank Strategy . . . . . . . . . . . . . 2 Mahaweli . . . . . . . . . . . . . . . . . . . . . . . 2 Rural Development .... . . . . . . . . . . . . . . . 3 Economic Developments . . . . . . . . . . . . . . . . . 3 C. MAHAWELI GANGA DEVELOPMENT PROJECT IV Project Identity . . . . . . . . . . . . . . . . . . . 5 The Place of Mahaweli IV in the Mahaweli Program . . . . 5 Origins of Mahaweli*IV .... . . . . . . . . . . . . . 6 Project Features .... . . . . . . . . . . . . . . . . 6 Project Issues . . . . . . . . . . . . . . . . . . . . . 7 Cofiniancing Difficulties . . . . . . . . . . . . . . . . 8 Cancellation .... . . . . . . . . . . . . . . . . . . 9 Conclusion .... . . . . . . . . . . . . . . . . .. 9 D. THIRD RURAL DEVELOPMENT PROJECT Project Identity . . . . . . . . . . . . . . . . . . . . 11 The Project . . . . . . . . . . . . . . . . . . . . . . 11 Suspension of Disbursements and Cancellation . . . . . . 11 Issues During Suspension . . . . . . . . . . . . . . . . 12 PART II - PROJECT REVIEW FROM BORROWER'S PERSPECTIVE . . . . . . 13 PART III - STATISTICAL INFORMATION Mahaweli Ganga Development Project IV . . . . . . . . . . . . 15 Third Rural Development Project . . . . . . . . . . . . . . . 17 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.  PROORCT COMPLETION REPORT SRI LANKA MAHAWELI GANGA DEVELOPMENT PROJECT IV (Loani 2437-CE; Credit 1494-CE) THIRD RURAL DEVELOPMENT PROJECT (Credit 1363-CE) PREFACE This Project Completion Report is unusual in the sense that it concerns one project that was; not implemented and another that was only marginally implemented. - A Loan of US$12.1 million and a Credit of SDR 28.3 million (US$30.0 million equivalent) for the Mahaweli Ganga Development Project IV (Mahaweli IV) were approved by the Board in June 1984, but not declared effective until January 1988. Project implementation did not start, however, and Loan and Credit were canceled at the original closing Date of March 31, 1990, without disbursements having been made.V - A Credit of SDR 21.4 million (US$23.0 million equivalent) for the Third Rural Development Project (RD III) was approved by the Board in May 1983, and declared effective in November 1983, and implementation had started. Disbursements, however, were suspended in March 1986, and the undisbursed part of the Credit (about 89 percent) was canceled soon after the Closing Date of September 30, 1989. This PCR was prepared by staff of the Agriculture Operations Division, Country Department 1, Asia Region. It is is based, inter alia, on the basic documents for each project: President's Report, Staff Appraisal Report, Development Credit Agreement (and Loan Agreement for Mahaweli IV), supervision reports, project files and correspondence between the Bank and the Borrower. It was sent to the Borrower on March 13, 1991, and the Government's comments on RD III are attached. YExcept for an amount of US$30,175 representing the Bank's front-end fee for the Loan.  PROJECT COMPLETION REPORT SRI AR MAHAWELI GANGA DEVELOPMENT PROZECT IV (Loan 2437-CE; Credit 1494-CE) THIRD RURAL DEVELOPMENT-PROJECT (Credit 1363-CE) EVALUATION SUMMARY 1. Obiectives. The Mahaweli IV Project was part of the Government's large multipurpose Mahaweli Ganga Development Program which was aimed at solving the country's persistent employment problems and severe food and power shortanes. In contrast, the Rural Development III Project was designed as a smaller, low cost, quick maturing progratn that would benefit large numbers of low income rural dwellers. 2. Imnlementation Experience. In both cases, implementation experience was cut short by the serious ethnic conflict which erupted in July 1983 between the Sinhalese and Tamil parts of the population. The conflict spread with increasing intensity from the northern part of the country, where the RD III Project area is located, to the east, where Mahaweli IV was to be implemented, eventually forcing tha Bank to cancel the loan/credits for both projects. 3. Results. Instead of focussing on narrow project-related topics and issues which usually form the core of project completion reports, the present report concentrates on the issues and considerations that played a role in the Bank's decisions: first, in the case of Mahaweli IV, to continue extending the effectiveness date; and finally, canceling the Loan/Credits for both projects. For obvious reasons, quantitative data on financial and physical implementation of the projects, and on their benefits and impact are not available. 4. Findings/Lessons Learned. In continuing to support the Mahaweli Project until the last moment, the Bank acted prudently, with the Government's best interests in mind and taking into account the broader political ramifications. In so doing, it managed to maintain good relations, both with the Government and the donors. The cancellation of the Credit for RD III was more straightforward than for Mahaweli IV. Early Bank supervision missions reported a generally satisfactory project start, but the security situation in the project area deteriorated steadily thereafter to the point where the Bank and the Government finally agreed to cancel the remaining amount of the Credit.  PROzCeT COMPLETION REPORT MANAWELI GANGA EVELOPMENT PROJECT IV (Loan 2437-CE; Credit 1494-CE) THIRD RURAL-DEVELOPMENT PROJECT (Credit 1363-CE) PART I - PROJECT REVIEW FROM BANRK' S PERCEPTIVE A. INTRODUCTION 1. This PCR is unusual in the sense that it concerns one project that was not implemented, and another that was only marginally implemented. In the case of Mahaweli IV, although the long delay between Board approval and effectiveness was caused by the withdrawal of one of the cofinanciers, the ultimate reason for the cancellation of Loan/Credits for both projects was the same: the security situation in the respective project areas. Serious ethnic conflict erupted in July 1983, between the Sinhalese and Tamil parts of the country's population. The conflict spread with increasing intensity from the northern part of the country, where the RD III project area is situated, to the east, where Mahaweli IV was to be implemented. A break in the hostilities appeared imminent in July 1987, when a Peace Accord was signed between the Governments of Sri Lanka and India by which India would maintain a peace- keeping force in the contested areas, and the Government of Sri Lanka would attempt to negotiate a settlement with the Tamil leaders. Prolonged negotiations, however, had no positive results, and the armed conflict flared up again. Eventually, the Xndian troops withdrew, and as of November 1991 there appears to be no immediate solution to the conflict in sight. 2. The causes of the delay in effectiveness of the Mahaweli IV Loan/Credit, the suspension of disbursements from the RD III Credit, and the eventual cancellation of both projects were, therefore, extraneous to the specifics of the projects. Apart from the estimated economic rate of return in the case of Mahaweli IV, more narrowly project-related topics and issues which usually form the core of project completion reports had no bearing on the Bank's decisions. For this reason, the present report deviates from the guidelines usually followed in preparing project completion reports. There was little implementation and there were no apparent benefits, and it would serve no useful purpose to speculate how preparation, design, Bank or Borrower performance would have affected project results had the projects gone forward. 3. Instead, the present report will concentrate on the issues and considerations that played a role in Bank decisions; first, in the case of Mahaweli IV, to continue extending the effectiveness date; and, finally, canceling the Loan/Credits for both projects. The present report will start with a brief discussion of the broader political and macroeconomic context in which both projects were conceived, and in which the Bank's assistance strategy and the Government's own development strategy fitted. This will be followed by two sections on the individual projects.  2 4. For obvious reasons, quantitative data on financial and physical implementation of the projects, and on their benefits and impact, are abeent. Therefore, the present completion report does not include the usual project completion data that, according to the guidelines should contain an accurate and comprehensive profile of the project at the time of completion. Some estimates of the manpower costs to the Bank of both projects and, ill the case of RD III, of the use made of the 11 percent of the Credit that was disbursed, will be made in appropriate places in the text. B. POLITICAL AND ECONOMIC BACKGROUND Government and Bank Strateav 5. The two projects under review had in common that they were parts rf several large development programs a newly elected Government had initiated upon taking office in late 1977. The 1977 elections marked a sharp turning point in the Government's economic policies and in its relations with the foreign donors who formed the Sri Lanka Aid Group./ The new Government had inherited a very difficult economic situation. The previous Government's system of pervasive state controls had caused economic stagnation, shortages and severe unemployment. The new Government stressed a major increase in employment through rapid growth and development of the country's natural resources. Its program included liberalization of the economy through greater reliance on market forces and a vast public investment program, of which the Accelerated Mahaweli Development Program (AMDP) was the largest, most capital- intensive element. The Aid Group members responded with unprecedented amounts of financial assistance to support the new economic prlicies. MahaweliV 6. The multipurpose Mahaweli Ganga Development Program has been an important part of the Government's development efforts for about thirty years. After some initial discussion and study, a Mahaweli Development Master Plan was prepared jointly by a UNDP/FAO team and Sri Lankan engineers in 1965-1968. It envisaged development in three phases over a 30-year period of about 365,000 ha of irrigable land, as well as about 500 MW of hydropower. Implementation of the Master Plan started in 1970, with the Bank participating in two projects (paras 13-15). 7. Late in 1977, however, the new Government announced its decision to speed up the implementation of the Master Plan. To help solve the country's !/The Sri Lanka Aid Group, composed of a number of multi- and bilateral-aid agencies, meets annually under chairmanship of the Bank to discuss assistance to Sri Lanka. YThe following two paragraphs provide a brief synopsis of the Mahaweli Program. For a more comprehensive discussion of the program, and the Bank's posture toward it, see OED Report No. 6074, dated February 24, 1986, "The World Bank and Sri Lanka -- A Review of a Relationship."  3 persistent employment problems and the severe food and power shortages more quickly, it had drawn up the AMDP, which included five major reservoirs and three hydroplants with a total power capacity of 500-600 MW, and downstream development of about 112,000 ha. The Government entered into direct contacts with several bilateral donors who were eager to support the relatively easy- to-implement large capital works involved that were also attractive to commercial interests in their home countries. 8. Late in 1977, the Government requested the Bank to support AMDP, followed, in early 1978, by a request to coordinate international support for it. These requests put the Bank in a somewhat awkward position. It had not been involved in the preparation of ANDP, which was, in fact, launched by the Government before full engineering and economic studies were available. The Bank decided to support AMDP in principle, agreed to the start-up of some components and advocated Aid Group support, after the Government had agreed to have an Implementation Strategy Study carried out to recommend a plan of implementation. The Bank also accepted its role as coordinator of the financial assistance rather hesitantly. When the request was made, bilateral negotiations were already far advanced, and the scope of the program was substantially decided. The Bank made it clear at the 1978 Aid Group meeting that it would act as a coordinator only after it was understood and agreed by all parties that the Bank would not interfere with individual bilateral discussions, but would only attempt to ensure that in those discussions no elements of the program were left uncovered. Rural Development 9. The Government's Integrated Rural Development Program (IRDP) was much less controversial than ANDP. It was initiated in 1979 to counterbalance to some extent the large, centrally controlled investments, particularly the Mahaweli scheme. In contrast to the large investment programs, the rural development program was designed as a smaller, low-cost and quick-maturing program that would benefit large numbers of low-income rural dwellers. The program also follows from the Government's decision to decentralize development efforts. It encourages "grass roots" institutional development at district and village levels. The program consists of a number of individual district projects, implemented by the district administrations under general guidance and support of the Central Government. The individual projects were geared to the specific needKs of the districts, and to an extent to the wishes of the sponsoring donors. Economic Developments 10. The response of Sri Lanka's economy to the Government's 1977 economic program was rapid. Except for the tree crop sector, virtually the entire economy grew. Exports, employment and real incomes advanced. In a remarkably short time, the major capital components of the AMDP were well under way. As a result of the second oil shock, however, the international economic tide started turning against Sri Lanka in 1979. Rampant international inflation caused substantial cost increases in Sri Lanka's investment program. Also, because of relatively low export prices, the Government found it difficult to generate higher levels of domestic savings.  4 By 1980, the economy showed serious signs of .verheating, with unsustainable budget and balance of payments deficits. 11. Initialli, the Government attempted to cover the deficits by short- term borrowing. But when the longer-term nature of the economic problems became more evident in the early 1980s, the Government and the Aid Group began to respond. The emphasis in Bank and IMF reports, and in discussions at the Aid Group meetings, shifted from the Government's expansionary new economic policies to what were essentially stabilization issues. The Government had to take some painful decisions. Tighter discipline was imposed on Government spending and commercial borrowing, and fewer new projects were started. Cuts in the Mahaweli Program, however, were limited, as the Government felt that because of the political, financial and contractual commitments already made, the program could not really be rephased or altered to any signific-.nt degree. 12. Just as the stabilization efforts were becoming effective in 1983, the ethnic conflict broke out with far-reaching negative consequences for the economy, including delays in the execution of the public investment program. In spite of continued financial shortages, however, the Government remained for a long time committed to carrying out the Mahaweli Program. Particularly after the Peace Accord of July 1987, in its negotiations with Tamil leaders, it attached great political significance to continuing development in the contested areas. When the Peace Accord fell through, however, the Government began to recognize that given the financial stringency, an overstretched capability to execute new projects, and deteriorating existing assets, a new economic course was called for. It now recognizes that projects that rehabilitate and intensify the use of existing assets, and have high rates of return, should have highest priority in the investment programs.  C. MH AWELI GANGA DEVELOPMENT PROJECT IV Prole t Identity Project Name: Mahaweli Ganga Development Project IV Loan/Credit No.: Loan 2437-CE of US$12.1 million Credit 1494-CE of SDR 28.3 million (US$30.0 million equivalent) Vice Presidential Unit: Asia Regional Office Country: Sri Lanka Sector: Agriculture Subsector: Irrigation TheaPlace of Mahaweli IV in the Mahaweli Program 13. At the time of inception of Mahaweli IV, towards the end of 1982, the Bank had made four investments in the Mahaweli Program: two before the Government accelerated the program in late 1977, and two thereafter. The first project was approved in 1970 and completed successfully in 1978. It provided water supply to about 53,000 ha of existing irrigated land, and to about 29,000 ha of unirrigated land to be developed later. Parallel financing provided for a 40 MW hydropower plant. The second project was approved in April 1977 and completed in 1985. With cofinancing from Canada, the Nether!&..is, the UK, the US and the EEC, it provided for construction work and social infrastructure on the 29,000 ha of newly irrigated land provided by Mahaweli I. The project was audited by OED, and the audit report noted that "the Mahaweli II project should be considered successful in regard to attaining major objectives ...," but that "the project ERR has been recalculated at -1% (21% at appraisal), mainly due to much lower actual and forecast prices for rice than those that prevailed at the time of appraisal." 14. The first Bank project under the Government's AMDP was a technical assistance (TA) credit in 1980 (with EEC cofinancing) for consultants' studies of the transbasin water diversion and other studies related to proposed future projects. The TA project made a useful contribution to the implementation of AMDP and provided useful on-the-job training by the consultants of Sri Lankan staff. The second Bank project under AMDP (Mahaweli III), with cofinancing from Japan and the Kuwait Fund, was approved in 1981 and scheduled to be completed in 1987. It is still not completed, however, partly due to the difficulties described in paras 11 and 12 above. 15. The purpose of Mahaweli III is to provide full irrigation supply for about 24,000 ha of land newly irrigated with water that had become available from headworks financed by bilateral agencies under the AMDP umbrella. This set a pattern, agreed at Aid Group meetings, by which bilateral agencies generally would finance the large capital works, including hydropower plants, and the Bank would stand ready to serve as a lender of last YSee OED Report No. 7336, Project Performance Audit Report, Sri Lanka, dated June 30, 1988.  6 resort for down-stream irrtgation development. The Mahaweli IV project fits entirely into this patte"t Orkins of Mahaweli IV 16. Towards the end of 1982, one of the major headworks of the AMDP, the Maduru Oya Project, was nearing completion. That project, financed by Canada, consisted of a dam and reservoir on the Maduru Oya River, together with works to divert Mahaweli Ganga water into the reservoir. Since the project had only limited hydropower potential, its main justification was providing water for downstream irrigation development in an area known as System B. Feasibility studies and detailed designs had been carried out by Canadian and U.S financed consulting firms. At that time, downstream development on the left bank of the Maduru Oya had already started with external financing from USAID, the EEC, the Saudi Fund for Development, Australia and the OPEC Fund. But there were no concrete plans yet to start development on the right banrs. 17. Completion of the Maduru Oya was somewhat ahead of schedule. The Canadian authorities, who were planning to become involved in the downstream development on the ri&.it bank, were faced with a dilemma: to demobilize the Canadian contractor working on the dam, or to convince the Government to advance right bank development so that costs could be saved by retaining the contractor. By October 1982, the Canadians had found the Saudi Fund prepared to participate, but there was still a considerable financing gap. The Canadian authorities then approached the Bank, which agreed tentatively to participate after it was understood that the Bank would be "lead agency," and that there was no objection on the Bank's part to continued work by the Canadian contractor on the Nagolla Dee-ela Kadiunne-ela (NDK) dam and feed. canal before formal Bank Board approval. The Government of Sri Lanka made a formal request for Bank financing in March 1983. The project was subsequently appraised by a joint Bank/CIDA mission in June/July 1983, and approved by the Bank's Board in June 1984. Proiect Features 18. Conceptually, the Mahaweli IV Project was very similar to the Bank- financed Mahaweli downstream development projects that had preceded it. It would provide facilities for new irrigation and settlement on about 14,000 ha; improvements on an additional 1,800 ha of existing cultivated land; and some development of conservation forests, fuelwood, cashew plantations and fisheries. It would also include: management facilities, staff quarters and vehicles; social infrastructure and settlement assistance; monitoring, training and the development of crop cultivation programs. 19. At appraisal, total project cost was estimated at about US$160.1 million. Cofinanciers would finance about US$132.1 million (Bank/IDA US$42.1 million, the Canadian International Development Agency (CIDA) US$40 million equivalent, and the Saudi Fund for Development US$50.1 million equivalent); and the Government the remaining US$28 million. Broadly, the Bank/IDA contribution would finance the irrigation distribution and drainage systems, roads and land development, as well as settlement assistance. The Canadian  7 contribution would be for the NDK dam and feeder caral, social infrastructure and training and monitoring, and consultants' services. The Saudi Fund's contribution would be for main and branch canal construction. The cofinancing arrangements would be in the nature of parallel financing by which the components financed by each participant would be procured in accordance with their respective procedures. Proiect Issues 20. The major issues that came up between identification and Board approval were threefold: macroeconomic considerations, cost recovery, and the economic rate of return. The macroeconomic conslderations centered on the question whether it was appropriate to start a new large capital investment project in Sri Lanka at a time when the economy was seriously overheated, budget cuts had to be made, and the public investment program had to be delayed. High-level discussions on this subject took place within the Bank just before and after project appraisal in June/July 1983. In the end, it was decided to go ahead with the project on the basis of the argument that if the Bank withdrew, it would risk unnecessary damage to the relations with the Government and donors who were already proceeding with implementation of components of Mahaweli IV. It was also thought that questions of size and composition of public investment programs could be addressed more appropriately in the framework of structural adjustment lending, which was being considered at the time. This, however, never materialized. 21. As in many other countries, cost recovery and the imposition of an adequate level of land development and water charges had been a perennial problem in Sri Lanka. Attempts had been made in previous Bank irrigation projects to address the probilema, but so far without much success. Once again, relevant covenants were included in the Loan/Development Credit Agreements of this project to increase recovery of operation and maintenance costs of irrigation projects and to implement a program of transfer of land rights to beneficiaries coupled with fees to recover land improvement costs. Similar covenants were again included in the Bank's next project in the irrigation sector: the Major Irrigation Rehabilitation Project (Credit 1537-CE), approved in December 1984. 22. The economic rate of return (ERR) was estimated at 12 percent at appraisal. For that reason it was considered a marginal p:oject, but Bank management was persuaded to go ahead on the basis of a number of arguments in favor: the Bank would maintain its influence in policy issues, such as cost recovery; the project would make productive use of large investments already made; secondary employment effects; soft funds were available from donors, for which there were no readily available alternative uses.  8 23. During the long period of delay between Board approval in June 1984 and Effectiveness in January 1988, when the originally agreed cofinancing was being rearranged, it became clear that a combination of cost increases and lower forecasts of rice prices would adversely affect the economic rate of return. In fact, in 1986 and 1988 the ERR was tentatively re-estimated at about three percent. But even then the Bank remained committed to financing its share of the project, first because its support was key in the negotiations between the Government and potential cofinanciers, and later because the project had gained important political significance, as it would mainly serve Tamil settlers, would address Tamil grievances and would be of help in the Government's peace negotiations. Cofinancing Difficulties 24. At the time of Board approval, the Bank had received firm declarations of intent from its prospective cofinanciers, Canadian International Development Agency (CIDA) and the Saudi Fund for Development (SFD). Effectiveness of the Bank's Loan/Credit was dependent on effectiveness of their contributions. Soon after Board approval, however, communications ceased between SFD and the Government. Verbal enquiries solicited evasive answers, and written correspondence remained unanswered. 25. This was the beginning of a long period of uncertainty in which the Government probed other potential cofinanciers about taking the place of SFD. But others were hesitant to enter the scene before the Sri Lankan Government had formally declared that it was no longer seeking Saudi financial participation. The Government was slow in doing so. It was not until the June 1986 Aid Group meeting that a solution started presenting itself. The SFD had no objection to others taking its place, and the EEC expressed its willingness to finance part of the project component that would have been financed by SFD. Aided by the Bank, negotiations between the EEC and the Government started and on October 30, 1987 an EEC grant of ECU 25 million (US$28.5 million equivalent) was approved and declared effective. At that time, the Government had entered into negotiation with the Japanese Overseas Economic Cooperation Fund (OECF) about financing the remaining parts of the original Saudi component. 26. When the EEC grant was declared effective, however, the Bank could still not declare its Loan/Credit effective because the arrangements between Sri Lanka and Canada had not yet been completed. The Canadian authorities had initially approved financing and completion of the NDK dam and feeder canal (completed in March 1986), but the Canadian Treasury Board had withheld approval of a CIDA grant for the other components to be financed by Canada. The main outstanding issue was an adequate representation of Tamil-speaking farmers among the settlers. The CIDA grant (US$43 million equivalent) was finally approved in January 1988, whereupon the Bank declared its Loan/Credit effective. Later, in July 1988, an agreement was concluded between the Government of Sri Lanka and OECF on a soft loan of US$67 million equivalent to finance the remaining parts originally to be financed by SFD.  9 Cancllaion 27. When the Bank Loan/Credit were declared effective, however, the security situation in the project area had deteriorated so much that no start could be made. Periods of optimism and pessimism followed the ups and downs of the peace negotiations. With the Closing Date approaching (March 31, 1990) and no quick solution to the ethnic strife in sight, the Bank was faced with the decision to cancel its Loan/Credit or to extend the Closing Date. Apart from the seemingly insoluble political problems, economic considerations, such as the low ERR and the low priority of the project in the country's new development strategy (para 12) added to the Bank's concern. The Government, however, continued to attach high political priority to the project. 28. In October 1989, the Bank suggested to the Government that it establish a small task force to review all relevant aspects of the project, on the basis of which decisions could be made by all participants on the future of the project. The task force was to concentrate particularly on prospects for reducing project costs, increasing benefits, increasing commercial activities, and shortening the project implementation period. 29. The Task Force report was reviewed by the Bank and representatives of all three cofinanciers in February 1990. The conclusion reached by the Bank was that some improvements in costs and benefits seemed possible, but that the revised estimates appeared overly optimistic so that the expected ERR would remain low. At the recommendation of the Bank's final mission, Bank management subsequently decided to cancel the Loan/Credit at the Closing Date of March 31, 1990 on the grounds of: * the project's marginal rate of return; * the Government's weakened implementation capabilities as witnessed by the large delays in the implementation of Mahaweli III; * the shift of priorities in the Government's investment program towards rehabilitation; and * the continued bad security situation. Subsequently, all cofinanciers followed suit. Conclusio 30. The extraneous circumstances leading up to the cancellation of Mahaweli IV are so unique that it is difficult to distill lessons of wider interest to the Bank, other than perhaps to state that it is difficult to change the course of a project of such size and complexity once it gets under way. But even with the benefit of hindsight, it is difficult to argue that the Bank should have acted differently. In conclusion, it can be said that in continuing to support the project until the last moment, the Bank has acted prudently, with the Government's best interests in mind, and taking into account the broader political ramifications. In doing so, it has managed to maintain good relations, both with the Government and the other donors.  10 31. On supervision missions and at headquarters, the technical staff involved has always been active in taking all possible preparatory steps to ensure a rapid start-up once the circumstances would permit it. This involved, in 1985, financing from remaining funds under the Bank's own Mahaweli TA Credit, consultants' studies and the preparation of tender documents that would have been financed by the Saudi participation. Later, at the time of renewed negotiations with cofinanciers, it made sure that all project elements remained covered and, after effectiveness, that consultant services were lined up to ensure a speedy start. The Bank's activities created a good deal of goodwill with the Government, and the costs to the Bank in terms of manpower were not high. Supervision of Mahaweli IV was always undertaken in conjunction with supervision of other Mahaweli projects in execution. The project, which had cost 98 staffweeks up to Board approval, cost another 29 staffweeks on supervision over the entire six-year period FY85-FY90.  11 D. THIRD RURAL DEVELOPMENT PROJECT Proiect Identity Project Name: Third Rural Development Project Credit No.: Credit 1363-CE of SDR 21.4 million (US$23.0 million equivalent) Vice Presidential Unit: Asia Country: Sri Lanka Sector: Agriculture Subsector: Rural Development The Proiect 32. The cancellation of the Credit for RD III was much less complicated than that of Mahaweli IV. As mentioned earlier (para 9), the Government's Integrated Rural Development Program, initiated in 1979, consists of individual district projects implemeiited by district administrations under general guidance and support of the Central Government. RD III was one of those projects. It aimed at increasing productivity, employment, income and living standards in the districts of Mannar and Vavuniya. It included rehabilitation of irrigation works; forestry and horticultural development; provision of inputs and services; and the development of economic and social infrastructure. 33. At appraisal, in November 1982, the project was judged to face no major risks. It was patterned after similar ongoing projects that were being implemented successfully, some of which were financed or supervised by the Bank. No major issues came -up during project or Credit preparation. The Credit was approved by the Board in May 1983 and declared effective in November thereafter. SusDension of Disbursements and Cancellation 34. By the time the Credit was declared effective the ethnic conflict had broken out. Early Bank supervision reports speak of a generally satisfactory project start. In the course of 1985, however, the conflict spread to the project area; it became impossible for Bank staff to visit the area and for most local staff and contractors to continue project execution. 35. As of January 13, 1986, the Government agreed to an informal stoppage of disbursements under which it would not submit any disbursement requests for expenses incurred after January 13, 1986. When the security situation did not improve, this was replaced by a formal suspension of disbursements as of March 13, 1986, except for expenses incurred before January 13, 1986. The security situation remained poor thereafter, and the  12 suspension was never lifced. After the Closing Date of September 30, 1989, the Government and the Bank agreed to cancel the remaining amount of the Credit as of January 15, 1990. By that time about 11 percent of the Credit had been disbursed: Original Amount SDR 21,400,000 Disbursed SDR 2.406,226 Canceled SDR 18,993,774 Issues During Susnension 36. During the time disbursements were suspended, two issues came up that required the Bank's attention. First, some twenty newly ordered project vehicles had arrived in the port of Colombo and could not be delivered to the project area. Second, the Government was in default with respect to submitting audited project accounts. 37. The Bank was concerned that the project vehicles that had arrived should be used for development purposes. After weighing some alternatives, the Government, the Bank and IFAD agreed that the vehicles would be taken over by an IFAD-financed, but Bank-executed new rural development project, the Kegalle Rural Development Project, implementation of which had started in 1987. 38. The only consolidated audited project accounts ever received were for 1983. Some unaudited accounts were received from individual project offices for later years, which stated that all project activities were stopped March 31, 1986, and that all financial accounts submitted after that date were confined to the settlement of liabilities incurred prior to March 31, 1986. The Bank kept pressing for audited accounts, but none were received. The matter was deemed important, because the Bank did not want the Government to be in default in the event of a lifting of the suspension. After that became unlikely, the matter was not pursued any further. No explanation was ever given for the failure to submit audited accounts, but it probably had to do with the auditor's own inability to visit the project area. 39. According to the Bank's own accounts, 51 percent of the disbursed amount was for civil works; 42 percent for equipment, vehicles, livestock and furniture; and 7 percent for consultancies and training. Without the possibility to visit the project area, it is impossible to ascertain what has become of the goods and services financed by the disbursements. The total cost to the Bank in terms of manpower was 75 staffweeks in project and Credit preparation, and 28 staffweeks on supervision. Of the latter, about 7 or 8 staffweeks were incurred after the suspension of disbursements.  13 PART II - PROJECT REVIEW FROM BORROWER'S PERSPECTIVE The Asia Regional Office sent Parts I and III of this PCR to Government of Sri Lanka for completion of Part II. Comments from the Ministry of Policy Planning and Implementation for the Third Rural Development Project are attached. However, no comments were received for the Mahaweli Ganga Development Project IV.  14~~~~~~~~~~~~~*. U4Urdfl'r J.S42S35 1 4 * I 0 J1 56 3,0 1 . s i 'I 9X2 ei