22996 Precis2 W O R L D B A N K OPERATIONS E V A L U A T I O NDEPARTMENT W I N T E R2 0 0 0 NUMBER 1 9 4 Involuntary Resettlement The Large Dam Experience T HE BANK'S RECENT STUDY OF INVOLUNTARY resettlement shows that if countries do not have both the capacity and the commitment to handle involuntary resettlement well, they should not embark on a large dam project. For years, the infrastructure and human sides of projects were utterly disconnected in most instances. That is now changing, as more countries voice their commitment to handle resettlement well. An even better indication of that commitment would be put- ting stronger monitoring and evaluation in place. The best signal will be when governments Capacity and Commitment treat resettlement not as a problem but as Are Both Essential an opportunity. One key element in this is Public agencies are typically limited in to mix land-based and diversified strate- their capacity to handle resettlement. In gies, not just to restore people's incomes most of the study's cases, public sector but to improve them. A second is to get agencies mishandled or ignored resettle- outside the traditional project cycle and ment. Frequently the public agencies instruments. That means planning for have a technical mandate, unrelated to income-generating opportunities well in resettlement. In India and Indonesia, the advance of resettlement and continuing to irrigation department was initially assess how the resettlers are faring after responsible for resettlement; in Thailand, the dam and relocation are complete. It Brazil, China, and Togo, it was the also means going beyond the project's power-generating authority (box 1). In funds to tap other sources. The third key addition, obtaining the cooperation of element is to work with NGOs, the other public agencies that are needed to private sector, government agencies, and make resettlement a success (agricultural external donors-to improve resettle- extension, health, education) can be ment's chance for success. difficult. 2 World Bank Operations Evaluation Department Figure 1: Displaced Persons, by Project Glenuine country commitment to doing resettlement resettlers needed, often the federal government did not or well is the key to success. In China, where the commit- could not allocate the funds needed to do so. This led to ment to restore settlers' incomes was a clear mandate, delays that further increased costs. In Indonesia, imple- the resettlement succeeded even in such remote areas as mentation proceeded without knowledge of the large Yantan. In India, the idea is beginning to gain accep- number of affected villagers who refused to move tance, but this has taken time to achieve. Above all, because local governments were not responsible for monitoring and evaluation have to he an integral part of reporting on their migration. Although central govern- planning and implementation-not a Bank-inspired ment's commitment was questionable in any case, the exercise that enters and exits with the Bank's presence. lack of follow-up at the local level aggravated the The borrowers' undisguised indifference must give way situation. At Nangbeto, too, responsibilities were not to their recognizing these tools as the essential basis for successfully transferred as planned to other ministries. improved management. Improving incomes is too big a challenge for imple- Governments are becoming more committed to menting agencies alone. Except in China-where decades good resettlement. In two of the cases here, borrowers of experience with a command economy, allocating jobs, exceeded Bank standards. There naturally are many and planning for incomes provided the necessary institu- levels of commitment. Sometimes there is commitment tional capacity and political will-government agencies at higher, more policy-oriented levels, but relatively less and bureaucracies lack the flexibilitv and grassroots expe- at the resettlement-officer level, where the resettlement rience to design income-generating options well-suited to posting is sometimes viewed as an undesirable two-year resettler capabilities and needs. NGOs should be called on career step. At other times dedicated individuals in the to make a much greater contribution. The private sector field are frustrated by indifference at higher levels, can also be drawn in. Looking ahead, governments should which robs them of the resources and other tools needed solicit cooperation and useful inputs from the earliest stage, to perform effectively. forging a wide range of public, private, community, and In Brazil, although the implementing agency institutional partnerships. intended to provide the infrastructure and services In all circumstances, adequate resource allocations are Pr&cis 194 3 essential both during and after the construction period. In The planners either failed to address the operating con- Brazil, Togo, India, and Indonesia, relocation preparations straints adequately or avoided the issue until it was too and resettlement activities lagged behind because too few late to implement a good strategy. They were forced resources had been dedicated to them. In Thailand, by into second-best solutions-or worse, they developed no contrast, ample resources enabled the implementing solutions. agency to achieve a satisfactory resettlement outcome Intelligent planning does not necessarily imply (resettlers are not necessarily satisfied, as NGOs urge them microplanning. A general structure of plausible income to complain and get more compensation). opportunities is enough to establish a basis for budget- ing and interactive implementation, after which the real The Devil Is in the Details lessons about feasible and popular options can be docu- Plannin /or U 0cnomniC Rcblhbilitatiol mented and fed back into the cycle. Ncc4ls to JImprovte Relying on the regional economy to take up the Economic rehabilitation of resettlers is the weakest slack is risky, especially where overall economic pros- aspect of resettlement planning. This was clearly the pects are uncertain. In none of the cases did the authori- case in Togo, where rehabilitation was not even an ties specifically rely on regional economic growth to afterthought. So, resettlers' incomes and living condi- ensure income restoration. But the absence of regional tions deteriorated for years without the implementing growth was a complicating factor in the Itaparica area agency even realizing it. Nor were the strategies based of Brazil, and in India and Togo. on transmigration assumptions viable for many Planners have to present plans with land and non- resettlers at Karnataka, Itaparica, and Kedung Ombo. land options, flexible models, and fall-back positions in Box 1: A Summary ot Dam Projects THIS STUDY OF INVOLUNTARY RESETTLEMENT In Mabarashtra, two much smaller dams displaced associated with big dams selected projects from six countries- about 40,000 people. Resettlers were moved into the down- all appraised well after the Bank's resettlement guidelines were stream area irrigated by the reservoirs displacing them. Some first issued and while they were being strengthened. Two resettlers have done well, but many have not received irriga- projects each were included from India and China, to assess the tion, or have received it sporadically and in limited quanti- representativeness of the primary selections from the two ties, insufficient to compensate for the reduction in land- countries that dominate the portfolio. holdings. A follow-up plan to make up for these deficiencies The projects range widely in size. On average they had not done much more than improve community infra- displace about four times as many people as other Bank structure in its first year. projects with dams. The dams were not significantly more China-Shuikou and Yantan. At Shuikou, 67,000 expensive or larger in reservoir surface area, but they were in people were relocated from the valley floor, and another more densely populated areas. 17,000 from Nanping City, at the upstream end of the reser- A higher percentage of the projects had resettlement voir, to make way for embankments that avoided the need to problems than the portfolio as a whole, largely because relocate more of Nanping's 200,000 inhabitants. All reloca- more recently approved projects have generally had fewer tion was completed by 1992. Although the original resettle- problems. Three are among the most controversial in the ment plan called for using traditional agriculture to rehabili- Bank's portfolio. tate 74 percent of those displaced, in actuality 75 percent India-Upper Krishna. In Karnaraka, two dams are dis- were rehabilitated by other means. Local government offic- placing a total of 40,000 households, or about 240,000 ers aggressively developed the reservoir fishery, oyster beds, people, the largest resettlement operation in the Bank's fruit and timber trees, and township and village enterprises, history. This does not include another 150,000 to 200,000 and even recruited foreign investors to establish factories to people who will be displaced when the dam is raised to its employ resettlers. The incomes of displaced people recovered ultimate height, not part of the Bank-financed project. Imple- to premove levels by 1994 and increased 44 percent by mentation of the resettlement plan ran into problems almost 1996, almost doubling the increase in non-resettler income immediately, and the Bank suspended the project twice, in 1992 growth. Treating resettlement as a development opportunity and 1995. Resettlement has finally improved. Compensation led to the most successful resettlement outcomes among the rates have approached market levels; a consent award system case study projects. has almost eliminated the need for resettlers to sue for increased Yantan displaced 43,000 people and affected the compensation; house construction, land purchase, and income- incomes of another 19,000. Located in a much more remote generation grants have helped supplement compensation; and and isolated region, Yantan did not benefit from a rapidly nongovernmental organizations (NGOs) have begun income- growing coastal economy. Nonetheless, resettler incornes generating activities. Resettler incomes have started to increase increased, and were supplemented by a grain ration until after a long period of decline. they reached the target level. Furthermore, the government 4 World Bank Operations Evaluation Department case the regional economy cannot quickly absorb the mentation, this assumption has not held up for involun- displacees. Planners should aim for a viable package, tary resettlement. As a subsidiary operation, resettlement adequately managed and funded, that can establish the continues to receive inadequate attention during imple- conditions for restoring full employment-even when the mentation. In none of the six cases did implementation economy is stagnant. follow plans, either because implementation failed to keep up with the timetable, or because plans made incor- Results, Not l'lanss, Are the Appropriate Touchstonc rect assumptions about resettler needs and resource /or Quality Managatement endowments. Countries that monitored results and had Although intelligent planning is a prerequisite to sound the flexibility to alter plans when warranted did the best. implementation, excessive reliance on "paper plans" is China, which altered its land-based income restoration dangerous. Reality almost always differs from plans. When strategy when farm employment did not reach high it came time to move in Indonesia, despite surveys of enough levels, is the best example of this. villager desires before planning the Kedung Ombo resettle- ment, far fewer people were willing to enter the transmi- (ComPensa,etion Al MVsi Be Adc1uatc and Tinecly gration program than originally said they would go. In Land compensation-the most difficult part of the dam- China, plans estimated that larger numbers of people could related compensation package to get right-needs to be find farm-related employment than proved to be the case. handled early, but even this does not guarantee success. In both instances, adjustments had to be made. Pak Mun's original land compensation rates proved too This is the downside of the progress with planning. low-land prices increased when the reservoir rose, and Although better planning usually means better imple- successive compensation increases only pushed land ere still waiting for their plantings 12 years after theyv sugaraestates and another state farm in other parts of the moved. Ther was Xvery poor follow-through in monitoring province. Ave=rae yincomesamong thlose resettlers have resettlerconditions to see that the plans were implemented in T<