Agricultural returns and conflict

In 1994, the genocide in Rwanda caused massive waves of displacement. We use these migrations as a quasi-natural experiment to measure the cost of civil conflict. In 1997, the government implemented a resettlement and land redistribution policy to improve the conditions for returnees and to induce conflict resolution. We assess the asset effect of the programme on agricultural output, and use this intervention as an instrument to measure skill spill-over mechanisms across returnees and stayers. Time differentials in the implementation of the policy across villages are used to identify its impact. Evidence that the policy was successful in raising migrants' agricultural production by increasing access to land is found, as well as higher returns to on-farm labour among the returnees in all areas. However, evidence of lower returns to agricultural inputs for returnees in policy areas suggests that the policy induced a 'ghetto effect'. The policy implications of this study are that, in order to promote durable peace, and therefore sustainable development in the region, these sources of economic inequalities across groups need to be addressed.