Can a market-assisted land redistribution program improve the lives of the poor ? evidence from Malawi

This paper uses a rural household survey dataset collected in 2006 and 2008 to investigate the impact of a market-based land resettlement project in southern Malawi. The program provided a conditional cash and land transfer to poor families to relocate to larger plots of farm land. The average treatment effect of the program is estimated using a difference-in-difference matching technique based on propensity score matching; qualitative information complement the analysis to ensure unobservable characteristics do not bias the findings. As expected, the results show a significant effect on landholdings and agricultural production, with land size increasing and maize production increasing by more than 100 kilograms relative to the control. However, the impacts on food security and asset holdings were mixed. Households that relocated great distances had systematically lower impacts than those households that stayed within their district of origin because they had to adapt to unfamiliar agro-ecological, cultural, and market environments. Impacts also varied across gender of the household head; female-headed beneficiary households increased their productive and consumption assets significantly, while male-headed households increased their asset holdings less so.