2. The compensation process was satisfactory overall; the main shortcoming regarded lost trees. Even on that count, CEB allowed resettlers to take their final harvest and offered seed1ings.47 This offer was inappropriate considering that the areas to which the resettlers were moving were unsuitable for oil palm. This raises the question of how to compensate an asset that cannot be replaced. At the very least, CEB should have compensated for the cash value of the lost trees, and CEB now agrees. It is unfortunate that this rather small shortcoming marred an otherwise good compensation process.

3. The relocation process was well planned and implemented. The sites were well chosen, the process was generally participatory (and worked better where it was), the disturbance was minimal, and in many ways it was a model. The entire problem was what came next, or rather, what did not come next. Nangbeto is a case Study in resettlement without rehabilitation. There was no income restoration strategy beyond re-creating the previous farm economy, nothing done for the host communities, no definitive resolution of the land tenure question, no baseline surveys or monitoring. no clear delegation of responsibility for ongoing services, no follow-up, and no way to really know what was happening to resettlers. R&R (resettlement and rehabilitation) effectively ended with relocation, final compensation payments, and cessation of food assistance. Ironically, at the time that appeared satisfactory ( except to Bank resettlement specialists ) and no one objected. At most they expressed concern about the future and the need for ongoing monitoring.

4. Despite compensation and relocation being generally well handled there was frustration due to poor communication, confusion, and misunderstanding during the resettlement process. CEB inventoried trees, prepared to pay compensation, offered seedlings, and then dropped the matter. Resettlers thought that CEB would clear the land for them, then that they would be paid for clearing new land, but finally they did the work but were not paid. They thought CEB was replacing their old houses, and that the CEB houses were meant to be permanent, not temporary . The compensation payments dragged on for years. The food assistance was for an indefinite period, and then cut off without warning. CEB maintained the pumps for years, but then quit. Fishing was prohibited, then banned, and now lies somewhere in between. These shortcomings are not all CEB's fault, but the net effect of this vacillation is that resettlers have a sense of distrust or even betrayal.

5. While the agricultural decline due to population pressure and insecure land tenure was not exclusively the fault of the project, certainly more could have been done to avert its worst impacts, especially as these problems were anticipated from the outset. The Bank's resettlement experts warned, even before relocation, that the extensive agricultural system in practice required more land than was being provided in the resettlement zones. The CEB resettlement officer foresaw a shortage in the future. The ambiguous land tenure situation was noted but not resolved. The potential for poor relations with the host communities was pointed out. The PCR' s warnings that the situation was deteriorating as forecast were ignored. Basically, everyone abdicated responsibility and let the problem unfold. If it were not for the Adjarala project in preparation and the OED mission, it is unclear when, if ever , anything would have been done to ameliorate the situation.

6. The difficult issue for evaluation is that many of these problems would have unfolded eventually regardless of what actions were taken. People displaced by Nangbeto cannot be sheltered from making the shift from extensive to intensive agriculture. What can be done is that they be assisted during this transition. They desperately need affordable credit and agricultural . technician services, but then so does much of the country . CEB is very wary of creating a privileged class of resettlers. Nonetheless, the Bank's stated policy of restoring pre-project incomes in an environment where most incomes have declined would do just that, protect resettlers from declines affecting most other people. This suggests the need to clarify the Bank's resettlem~t policy in situations where incomes are declining. Presumably the Bank's policy of do no harm should be interpreted to mean that resettlers should be no worse off in relative terms (relative to nearby, similar, unaffected people), not in absolute terms.