1. This is part of OED's six-country study of involuntary resettlement behind large dams, a follow up to a three-country study completed in 1992. OED added Indonesia to the new study, because it has one of the largest Bank-supported portfolios of projects involving involuntary resettlement and because OED has had occasion several times in the past few years to make observations on the Kedung Ombo project. The resettlement component has been among the most controversial of big dam projects and the only one with documented evidence of military intervention and coercion of households to move out from the reservoir area. The Bank and the government have been severely criticized for mismanaging their roles in resettlement at Kedung Ombo. An objective of this case study was to determine whether and how that management failure occurred.
2. The Kedung Ombo Multipurpose Dam and Irrigation Project was partly financed by Bank Loan 2543-IND, approved in May 1985 for US$156 million. The project was to finance construction of a dam and downstream weirs on the Serang River in the Jratunseluna Basin in Central Java (maps I and 2). The objectives were to supply irrigation to 87,000 farms on 59,000 hectares (ha) in four old and new schemes below the dam, control flooding from the Serang, supplement municipal water supplies for urban centers, and generate a small amount of hydroelectric power. The project was completed and the loan closed in December 1993, with cancellation of US$1.4 million. A project completion report (PCR) was issued in June 1995 with a comprehensive annex treating the resettlement component. PCR, Report No. 14636, June 19, 1995.
3. Government and the Bank claim the project was successful in reaching its objectives: in particular the irrigation component surpassed its physical targets and resulted in an overall economic rate of return (ERR) calculated for the PCR at 12.4 percent. If real paddy prices had remained at their 1984 level, the PCR says the project would have reached the ERR projected at appraisal of 17.4 percent. The dam has eliminated flooding from the Serang, although downstream tributaries continue to put the lower basin at risk. OED rated the project's overall outcome "marginally satisfactory," weighing these positive impacts against the unsatisfactory resettlement experience. However, other observers inside and outside the Bank challenged OED's relative weights of irrigation and resettlement. They asserted that the pain inflicted on the resettlers that suffered above the dam could not be offset by material advantages to populations downstream. The discussion provides some of the best-articulated material in Bank files supporting the primacy of resettlement policy. See especially Memorandum, A. Steer, ENVDR, to R. Picciotto, DGO, May 28,1995; and Memorandum, G. Davis, ENVSP, to R. Picciotto, May 31, 1995. An earlier Memorandum in the series, J.D. Clark to R. Picciotto, April 12, 1995, captures the spirit of their argument:
Later, M. Cemea, the Bank's senior sociologist, had his own words-criticizing the PCR's discussion of net costs and benefits-repeated back to the Bank by one of the NGOs concerned with Kedung Ombo: this kind of desk macro-accounting of costs and benefits, informed by crude statistical utilitarianism ... this] kind of spurious rationality ... detracts from seeking altemative approaches and solutions [to] the ill effects of such programs, which otherwise could be prevented or mitigated. (M. Cemea quoted in ELSAM, p. 102. See reference in para. 2.18).
The assumption that social costs to one group may be outweighed by benefits to another group is anathematic to the principles of protection and compensation enshrined in Bank ODs on resettlement, indigenous peoples etc. Costs and benefits must be aggregated for each vulnerable population group, not across all populations. If the latter were the case then there would be no point in having such ODs at all. Any resettlement could be ignored as long as more people benefited from the project.
4. This experience cannot be understood without referring to the geopolitics of the reservoir area. As shown in map 3, the Serang and other rivers enter the area from the southwest and southeast, respectively. Each passed through a stretch of relatively flat flood plain before moving into narrower valleys and hilly surroundings leading to the proposed dam. The reservoir occupied parts of three regencies: Grobogan, Sragen, and Boyolaii. Most of the occupied land, most of the people threatened by displacement, and all of the extensive south-western flood plain belonged to Kemusu district of Boyolali. Grobogan, Sragen, and Boyolaii are regencies (kabupaten), administrative units immediately below the provincial level. Kemusu is actually a sub-district (kecamatan), the next lower level. The "district" has effectively disappeared. It is used here for clarity and because that is what many other reports do.
Coincidentally, Kemusu plays a special role in the modem political history of Java as perhaps the most important center of repressed resistance by radical groups outlawed and driven underground by the military action of 1965 D. Butcher reported in 1988 the Boyolali Regency's assertion that there were 500 hard-core radicals (OT), most of them ex-communists, living in the reservoir area in Kemusu district (memorandum to G. Davis, ASTEN, June 16, 1988, p. 49)..
5. The villages and district headquarters in Kemusu formed the core of the controversy that was to arise over compensation rates for displacees, suitable substitute sites, refusal to evacuate even as the waters were rising, and ultimately, the insistence to stay put and cultivate the greenbelt as well as the fertile old flood plain as it reemerged every year in the dry season when water was released for irrigation (the drawdown area). Although there was resistance also-to the compensation packages and relocation plans-in the project-affected areas of two districts in Sragen, most of the households there took their packages anyway and moved out years ahead of the water. Households in Grobogan were never a factor of the controversy. They were few, and the valley lands they farmed were closest to the darn and destined for deep submergence Thus, the special conditions of Kemusu, with recoverable fertile lands and a leadership associated with an outlawed opposition, were to shape the resettlement experience of
Kedung Ombo. The dam and irrigation works were constructed by the central government's Directorate General for Water Resource Development (DGWRD). Other big power-generating dams on Java with Bank participation have been constructed by the State Electricity Corporation (PLN). Neither DGWRD nor PLN had experienced resettler resistance at this level of intensity in dam projects preceding Kedung Ombo. Afterwards, helped along by new central government regulations in place starting in 1993 that removed some of the arbitrariness over resettler rights of the earlier period, there has been no repeat of the Kedung Ombo violence, at least not on any Bank-supported project and nowhere at the level reported from Kedung Ombo. Thus, however helpful in considering and explaining an evolutionary resettlement record, this is indeed an unusual, watershed event in Indonesia.
Principal Findings: Performance
1. The Bank's appraisal team accepted Padjajaran University's projections that 75 to 80 percent of families in the reservoir would opt for transmigration. It raised the estimate to 90 percent, based on what it called "experience at similar dam sites in Central Java. 443 Staff Appraisal Report No, 5346a-IND, April 24,1995, p.33.3 This passed the problem of organizing for resettlement to the Ministry of Transmigration, which had long-established procedures for handling the influx from Kedung Ombo. Bank and outside critics were to claim later that the team misread the "experience," and should have anticipated a much greater resettlement load on local authorities.
2. The Bank's Sociology Advisor, who up to appraisal had played and was to continue to play the dominant role in formulating a policy for the Bank on involuntary resettlement, was paying close attention to developments at Kedung Ombo. He accepted the appraisal team's contention that most displacees would opt for transmigration. But he warned nevertheless in two memoranda dated July and December- 1984 that the local component was seriously deficient. Among other things he pointed out that (i) the institutional apparatus was underdefined and probably inadequate, (ii) the project provided no funds for whatever resettlement would be carried out in the catchment area 444 Sufficient transmigration funds were expected to be made available whatever the load, though they were not included in project costs , and (iii) provisions for resettling on productive land even 25 percent of the displacees under the crowded conditions of Central Java had not been spelled out 445 Memorandurn, M. Cemea, AGR, to J.C. Collins, AGR, "Comments on Issues Paper," July 16, 19845 . His concerns were ignored as appraisal passed from Issues Paper to Yellow Cover (draft) to Gray Cover (final). The appraisal report issued nine months after his first memorandum adds nothing on local arrangements except a generalized organizational chart and a promise that regency coordinating committees would prepare the remaining 10 percent of the families by giving them “preference for vocational training, work in local rural works programs, and in the project's civil works construction program 446 SAR, p. 33.6 .
3. The Bank may be excused for anticipating a large transmigration cohort, but the appraisal team's indifference to the warnings from the sociologist is inexcusable. The disconnect is all the more regrettable because the irrigation advisor to the team was a member of the Agriculture Department (AGR), not the Regional Department, and his sign-off memorandum to his Vice-President was silent on the fact-which he was obliged to report-that a colleague from AGR (in fact, his neighbor in the next office) was unsatisfied.
4. Later, given the warning, and the penalties for effort on this critical judgment, the Bank should have paid close attention during supervision to the processing of compensation and the organization of the transmigration program, starting with or even before the first recorded mission in May 1986. It paid no attention at all. Even if the government claimed that the compensation and transmigration plans were on schedule, the Bank would reasonably have been expected to so constitute its supervision team as to provide for visits to the reservoir area to observe the scheme unfold, and to the transmigration sites to investigate the infrastructure being prepared for the displacees. Again nothing. The task manager for supervision was the same engineer who had advised the appraisal team. He felt the Bank's attention ought to be directed to the technical problems at and below the dam, and government would be allowed to handle whatever problems appeared above it.
5. When LBH provided the Bank a copy of its chronology of abuse in September 1987, the task manager considered the informal transmission irregular and did not warrant a response 447 The task manager wrote in the margin of one of the internal memoranda: "no further action needed. Attached has not been submitted to us officially so we need not take any notice of this. Also we know [of the] progress in resolution of the issue from [the Department of Irrigation]." Hand-written note dated October 14, 1987, on the Memorandum, M.C. Zemick to G. Fox, October 7, 1987 (available in ENV and central files).
6 . Even after the Bank's first resettlement expert visited the site in May 1988 (no Bank resettlement expert visited the site during preparation or appraisal), the resident mission was uncomfortable with his observations and felt his report was emotional and exaggerated. In fact, the headquarters division responsible for the project as well as the resident mission were uncomfortable even that his mission had been planned-by AGR and outside the regular supervision routine. At this point, the Bank's performance on the resettlement component would have to have been rated exceedingly unsatisfactory, especially since the Bank's new policy on involuntary resettlement was taking shape all around the responsible project staff.
7. The Bank was already mobilizing to correct for the lapse in attention. The resettlement component received more of it in the next two years than any other in the Bank's portfolio. Most missions were run out of the resident mission and not even registered in official files. The Bank was not only working to salvage a program that was already in deep trouble, but had to contend with mountains of angry letters arriving from NGOs and other critics around the globe. Whether or not the Bank was correct in concluding that it should not challenge the provincial government's compensation policies will remain a point of contention. But in all other respects the Bank tried to develop a package of supports and benefits that would ensure that the relocatees in the reservoir area had a good chance of reestablishing themselves whether they be self-directed settlers, residents in the new villages, or holdouts in the greenbelt. The Bank encouraged
the government to concentrate on establishing its three villages, while the universities, through their research, extension, and demonstrations, were expected to cover the whole of the reservoir area.
8. Thus, Bank activities in the field during phase two met all the standards of the new policy directives. That intensity of involvement continued until the end of the project. If one counts the months spanning the two phases, the second clearly dominates: (i) the loan became effective in November 1985; (ii) LBH sounded the alarm with a letter to the Bank in mid-1987 climaxed by handing over a version of its the detailed chronology in September that year; (iii) partly in response to that alarm 448 ELSAM points out that this mission was not devoted to Kedung Ombo, but part of a Bank-sponsored study of resettlement in four projects in Indonesia.
8 , the first resettlement expert was originally scheduled to visit the site as soon as possible and eventually made it there in May 1988; (iv) by March 1989 the Bank was mobilizing in strength to intervene, about 18 months after receiving the chronology; and (v) for the next five years the Bank closely attended to all the local resettlement operations.
9. In terms of the Bank's timely response to a rapidly developing crisis, however, 18 months was too long. In terms of the Bank's image as the world's new leader in resettlement policy, the delay was a disaster. The PCR says:
While supervision of resettlement should have been taking place on a regular basis, LBH's submission to the Bank was the first documentary evidence that there were major discrepancies between what the Bank thought was happening and what was taking place on the site. Furthermore, not only was an important opportunity to make early interventions lost, but the initial lack of Bank response made NGOs feel as if they had little choice other than to mount the bitter international campaign that ensued 449 PCR, op. cit., p. 64.
10. Unfortunately, this second phase was also darkened in its early years by the Regional Office's disingenuous efforts at damage control. At first the objective was to show the Bank had not been negligent either in planning or supervising the resettlement component. An accompanying theme was that the Bank had to accept government's reports that resettlement was under control. Later, the Bank was to agree that in hindsight errors may have been made. Several quotes give the flavor of that argument:
· On planning: "A detailed resettlement plan (including land acquisition, number of families to be resettled and compensation budgets-by each year 83/84 to 88/89) was prepared including plans for transmigration ... institutional mechanisms were also set up, i.e., a committee in Central Java...550 Supervision Report, D. Purcell, AS5RA, to T.R. Daves, AS5RA, Annex 4: "A Background Note on the Kedung Ombe Multipurpose Dam and Irrigation Project," April 14, 1989.
· On supervision: "Your contention that there was 'no monitoring or supervision of the project for three years (after loan approval)' is incorrect. Bank supervision missions have regularly supervised the project since its inception, and reviewed the progress in resettlement during these missions." 551 Letter, R.J. Cheetharn, AS5DR, to A. Hakiin and P. Kardoes, INGI, May 5, 1989. And: "Supervision missions from RSI regularly reviewed progress of resettlement. Supervision reports of March 1987, November 1987 and April 1988 reported satisfactory progress in acquiring land for the reservoir." 552 Memorandum, A. P. Cole, AS5AG, to R.J. Cheetham, AS5DR, February 3, 1989, p.22
· On confidence in movement’s reports: "These accusations of coercion were made in 1987. The Government has a clear policy of non-coercion in these matters. Nevertheless, in view of the allegations made, the Bank did raise this matter with Government and understands that Government has taken steps to curb such actions."553 Letter, R.J. Cheetam to A. Hakim and P. Kardoes, op. cit., p. 3.3
· On hindsight, and also to shift part of the blame: "We have concluded in hindsight that the Bank supervision of the resettlement component of this project could have been handled in a better way by allocating more resources during the earlier years of project implementation for the resettlement component, particularly by assigning staff with special qualifications. We treat this as an important lesson learnt from the Kedung Ombo experience. Supervision of the project initially concentrated more on the civil works as compared to resettlement, until resettlement issues surfaced in 1988. Under the project, monitoring of resettlement activities was to be carried by a consultant; this work did not start until early 1988 (due to a delay of two years in the employment of this consultant by Government).”554 Memorandum, I. Zincir, AS5DR, to Y. Yoshimura, Alternate Executive Director, October 24, 1990, p. 1.4
11. The references to hindsight, to lesson learning, and to resettlement issues surfacing in 1988, are unacceptable, given the senior sociologist's warnings four years earlier about the problems that could-and did--occur.
12. While senior management at headquarters was trying to control the damage, field staff ere working overtime to correct the damage. They also were trying to identify the appropriate lessons for the Bank to carry away from Kedung Ombo.555 Following the March 1989 supervision mission, two missions members dsrafted separate sets of lessons. Headquarters supported this work, although it warned against including lessons that had yet to be learned.5 They were supported by headquarters with successive revisions in the loan agreement and reallocation of US$7.1 million to finance resettlement, most of it near the reservoir. That is the correct image to remember from this experience, not the paper traffic at headquarters.
13. And yet RRDP was essentially a failure, and the Bank's attempts to rescue the situation were frustrated. The work of the universities was impractical, fragmented, and inconclusive. The components of RRDP were in any case unsustainable without ownership and follow-up by the provincial departments. An overall rating on Bank performance during implementation also has to be unsatisfactory. Nevertheless, the resettlers have progressed, and the ineffectual efforts to ensure that outcome do not appear to have mattered. The concluding chapter returns to this contradiction.
14. Movement’s performance also has to be rated unsatisfactory throughout the project period. It committed all the errors that the Bank now says are to be avoided while planning and executing a resettlement component, especially: no participation by the people living in the condemned area in planning for resettlement; failure to establish an institutional mechanism that can manage a large and painful exercise in social adjustment; setting compensation rates that do not acknowledge the project's impact On land prices and allow resettlers to buy substitute land assets of approximately equal earning capacity; and-highlighted by the NGO, student, and settler protests and the worst feature of the Kedung Ombo case-the physical violence, psychological harassment, and abuse of other civil rights that persisted throughout and after the acquisition period.
15. The judgment on government performance has to be harsh even while recognizing the mitigating circumstances: some of the Kemusu and Sragen heads of households and leaders were immovable; related to that, there was no easy solution to the residual compensation issue once most of the families had accepted payment: the protesters should have worked with government to find other solutions; the Supreme Court made a mistake is raising the ante and guaranteeing government intervention; and the worst errors have not been repeated, in fact they have been avoided in subsequent dam projects on Java and the other islands.
16. The government underfunded the program for Central Java, even after designing RRDP. Almost all of RRDP was financed by the Bank's loan. The provincial governments did not earmark extra funds for the reservoir during RRDP, and when the Bank withdrew, the resettlers lost whatever privileges they might have just enjoyed. There is an issue as to whether governments should accept the Bank's conditionality that involuntarily resettled families have to be treated deferentially in comparison with other poor rural families, even after a project closes. This case study will not address that issue.