Sri Lanka : Mahaweli Ganga development

NUMBER 86 ED Precis Operations Evaluation Department April 1995 Sri Lanka: Mahaweli Ganga Development Development of the Mahaweli Ganga new more market-oriented gov- families, also with homestead has been a centerpiece of Sri Lanka's ement elected in 1977 acceler- plots. The project also provided development thinking for 35 years. ated the development of the for fuelwood (3,000 ha) and Formal work on the Mahaweli Ganga Mahaweli Ganga Development cashew plantations (2,000 ha). Development Program started in 1958 Program. The Mahaweli Authority provided and a 30-year master plan was formu- support services such as extension, lated in 1964-68. The program has been Project goals marketed inputs and outputs, expected to provide, and in large and operated and maintained the measure has provided, employment in To assist the accelerated pro- irrigation system. It installed construction, land for the landless, food, gram, the Bank-assisted Third roads, urban centers, and social and electric power, together with an Mahaweli Ganga Development infrastructure including schools. appealing focus for foreign assistance. Project was approved in 1981; a credit of $90 million supported a Implementation The World Bank extended IDA total project cost of $200 million. credits for a series of projects that This was an archetypal integrated Implementation took twice as supported the Mahaweli Authority, rural development project, which long as originally planned. The which managed the program. Though started from raw jungle to provide cost of Rs. 8.45 billion ($263 the Bank provided relatively little productive investments and all million) was about 30 percent money (6.2 percent of the total funding social infrastructure and services higher, in real terms, than origi- for Mahaweli by 1990) it played an for a population projected to reach nally budgeted. active role in advising on the size and about 180,000. The project was composition of the Mahaweli program expected to produce 186,000 The proposed project size and and in coordinating external assistance. additional tons of paddy, with phasing were too optimistic, given significant volumes of other crops. the experience with earlier phases OED recently audited the third of Most importantly, it would of the scheme and the capacity of the Bank-assisted projects, which was provide land to the landless, and the implementing agency. The executed in 1982-92.* It finds that the relieve population pressure. schedule at appraisal called for project was soundly conceived and jungle clearing at around 5,300 well, if slowly, executed, assisting large Before development most of the ha/year, on-farm development at numbers of poor families. But environ- project area was virtually 5,700 ha/year, and settlement of mental aspects were neglected. And unpopulated, but population 5,300 farm families/year. The economic returns were much lower pressure from the wet zone than expected, due partly to a long (southwest and central highlands) Performance Audit Report. delay in implementation and partly was leading to increased immigra- Sri Lanka: Rehabilitation and to a dramatic fall in the price of rice to tion, with associated shifting producers. The audit draws lessons. cultivation. The project involved jungle clearing, land leveling, and No. 13835, Decemnber 1994. During the 1970s Sri Lanka's on-farm development over a gross Performance audit reports are annual growth of agricultural area of about 31,000 ha, and available to Bank executive production fell behind that of settlement of a projected 18,500 directors and staff fron the population. Real per capita GDP families each with 1 ha of irrigated Interial Documents Unit and declined by 1.3 percent a year and land and a 0.2 ha homestead plot, r Regional Information open unemployment increased. The as well as 5,500 nonfarming Service Center,. Land sales to settlers were de- World Bank (IDA) assistance to Sri Lanka, to June 1994 ferred. Settlers did not receive title to their land. The project design $million Percent provided that the cost of land Irrigation 244 13 improvements was to be recov- Mahaweli 158 9 ered by sale of the farm and house Other 86 5 plot given to each settler, and that Other agriculture 267 14 settlers would receive clear title to Total agriculture 511 27 their land. But no payments have been made by settlers and owner- All sectors 1,863 100 ship rests with the government. ___________________________________________________________ Farmers remain dependent on the government. Most farmers, and implementing agency was Results especially those in the most already overstretched, and the recently settled areas, still depend project posed new technical and The project achieved close to its heavily on government initiatives organizational issues. Many agency targets for physical works, numbers of and services. staff were newly recruited and beneficiaries, and rice production. It were working with contractors provided irrigated farms to 16,136 Maintenance has been mad- from a recently emancipated settler families (87 percent of the equately managed and funded. private sector. Civil unrest number estimated at appraisal) at a Operation and maintenance caused disruptions. total cost of $17,350 per irrigated arrangements are deficient and farm. This gross cost includes forest the irrigation facilities built under Delays in providing water to clearing, irrigation works, roads, the project face premature deterio- cleared land led to problems with social infrastructure, supporting ration. Though it actively pro- spontaneous settlement and to towns, and more than 2,000 fami- vided support for farmer organi- jungle regrowth. As designed, lies who were settled without zations, the project had limited the project gave priority to farmers irrigation. Jobs were created for success in transferring manage- displaced from areas flooded by 8,000 landless families. Training ment responsibility to farmers other parts of the Mahaweli under the project upgraded the for planning water use and scheme. It did not foresee that technical competence of project cooperating to maintain the these involuntary resettlers- staff and farmers. distribution system. and also refugees from civil disorder-would arrive before The project significantly ex- Water charges were collected enough irrigated land was ready panded the land available for from farmers at first, covering for distribution. Some would-be paddy production. By 1993, allow- about half the cost of operating settlers had to be accommodated ing for double cropping, the project and maintaining the irrigation for two years before they could be was contributing 122,000 hectares facilities. But cost recovery was assigned land; the World Food of paddy from about 100,000 ha of abandoned as farm incomes Program played a major role in new arable land formerly under turned out to be lower than feeding them. forest. Additional rice production is expected. Collections fell from estimated at 145,000 tons, or 95 90 percent of the amounts due The relationship between percent of the appraisal estimate. in 1985 to 25 percent in 1989 and the borrower and the three 5 percent in 1990. Farmers' organi- cofinancing agencies-Japan's Incomes were much lower than zations have willingly contributed Overseas Economic Cooperation expected. The project turned out to labor for cleaning and for minor Fund, the Kuwait Fund, and IDA- be unsatisfactory in economic maintenance of distributor and was generally satisfactory. Project terms, less because of poor design, field canals, but have not supplied managers responded well to or poor implementation, but even small amounts of cash for advice provided by Bank supervi- because the terms of trade shifted needed hardware. (The same sion missions, which was relevant against rice production. pattern was observed in the and realistic. The good relation- recently audited Village Irrigation ships were especially helpful in The present income of farming Rehabilitation Project in Sri maintaining activities through settler families is about twice their Lanka.) As noted by the project periods of civil unrest in the estimated "without project" in- completion report on this opera- project area, and in agreeing come. This is much lower than tion, government maintenance the four extensions of the the five or sixfold increase pro- budgets have been used mainly credit closing date. jected at appraisal. for salaries. April 1995 Issues and lessons Sri Lanka: Amount of paddy needed to cover production costs, Speed of execution. The acceler- 1980 and 1993 ated program was an idea that (Kilograms of paddy; production costs per hectare) energized all parties. Recruitment of some of the "best and the 1980 1993 brightest" from national line Hired labor 425 985 agencies was meant to speed up Fertilizer 219 523 implementation. But in practice Agrochemicals 89 256 the staff appraisal report was Farm power 492 572 overoptimistic in projecting what Source: Cost series maintained by World Bank Colombo office. could be achieved by an over- stretched organization facing new technical issues. Elsewhere in Sri Lanka, as the project completion report pointed out, budgetary cleared. But conditions on the analysts. In part, it came from pressures forced the national Bank's credit called for an environ- the increased supply of small extension service to employ fewer, mental action plan to be submitted grains that resulted from the better qualified staff, but the for review and concurrence by the Green Revolution, the Common Mahaweli Authority retained a end of 1981 and to be implemented Agricultural Policy of the large number of low level, poorly by March 1982. In fact a very European Community, and qualified field assistants. Staffing unsatisfactory document was farming subsidies in other difficulties made it necessary to submitted in March 1983. Concerns OECD countries. hand over responsibility for public expressed by the Bank's Environ- safety, education, and health to mental Advisor and Agriculture For Sri Lankan rice producers, the relevant line ministries. Department were not addressed. yields were reflecting most of the benefits of the Green Revolu- Ultimately the rate of farm Today, increased population tion by 1980. Yields rose little in development in the project area pressure as a result of the settlement the ensuing decade, while world seems to have been determined as of nonfarming families and the paddy prices fell, but production much by factors outside the children of original settler families costs per hectare rose steeply. control of project staff as by those is leading to encroachment on The opening of the economy led to within their competence. Civil remaining forest areas for shifting an increase in real wages. Fertil- strife, credit problems, poor cultivation or for fuelwood. Thus izer prices rose, following the availability of suitable varieties of much more attention is needed to abolition of subsidies. Using paddy, and lack of area-specific the management and protection of constant production technology, fertilizer recommendations are all the remaining forest within the in 1980 paddy production costs nationwide concerns. project area. accounted for 40 percent of the revenue from paddy production; Farmers' organizations. Farmers Cropping in unirrigated areas. by 1993 production costs ac- in the project area will need to The appraisal seems to have counted for 80 percent of the take more responsibility for their misjudged the suitability of revenue (Table 2). In practice, own affairs and play a greater role unirrigated land for cashews. farmers have been changing in managing the irrigation system. Even the highly permeable soils in technology in response to the Project design and government the area get very waterlogged in changes in prices. administration of services can help the long rainy season. Possibilities make farmers' organizations more demonstrated by an NGO for In the light of the changed effective, but village dynamics and changing to mixed cropping, price ratios, it may now be the relations among the members taking advantage of the different difficult to design any new have a more important influence soil moisture regimes on the ridges, investment or national project (see Pricis No. 85--ed.). slopes, and gully bottoms, should that would yield a 10 percent be taken seriously, economic rate of return on the Environment. Environmental basis of increased irrigation water issues in this project were largely Economic viability. By the early for paddy production. Indepen- subordinated to the task of build- 1990s rice production was much less dent studies have warned that ing the irrigation infrastructure. attractive than when this project Sri Lanka cannot afford to provide Mitigatory measures can have was approved in 1980. The halving much additional employment only a partial impact in a situation of the world price of rice was not through schemes as capital where so much land is to be foreseen by the Bank's commodities intensive as the Mahaweli. OED Pricis Other possible alternatives for tural and Rural Development * Better education and assistance assisting small and poor farmers are: Project on Crop Diversification, for entrepreneurial young farmers. * Agricultural research (local or funded by USAID, has not yet Education, at least to the diploma imported) to increase the genetic shown the successes first hoped level, of two young farmers per merit of rice plants. for, this direction of public sector village would seem to hold the * Diversification into higher value research should continue to highest promise for increasing the crops. While the Mahaweli Agricul- be supported. dynamism of smallholder agriculture. OED PrPcis is produced by the Operations Evaluation Department of the World Bank to help disseminate recent evaluation findings to development professionals within and outside the World Bank. The views here are those of the Operations Evaluation staff and should not be attributed to the World Bank or its affiliated organizations. Please address comments and enquiries to the managing editor, Rachel Weaving, G-7137, World Bank, telephone 473-1719. Internet: rweaving@world bank.org April 1995