Wirtschaftswissenschaft im Dienste der Armen
The question of whether development aid is effective in fighting poverty has been subject of controversial debates, often based on ideological assumptions. Searching for a general answer is similarly futile than the question whether government expenditures are good or bad: it will depend on how the money is spent. It is therefore important to know which development projects are most effective in fighting poverty.
In recent years, the methods with which such evaluations can be conducted have seen important improvements, through the use of randomized studies, which have been developed and promoted in large part by researchers of MIT’s Poverty Action Lab. This article describes the randomized evaluation methods, illustrated by studies that allowed detecting large differences in cost-effectiveness between a-priory similarly plausible educational projects and provides a discussion of methodological limits (external validity, etc) and ways to mitigate them. It concludes that if aid has had mixed results, the reaction should not be resignation, but a reorientation of aid to a more evidence based selection of interventions. If the new approaches were more widely used and had a direct impact on the distribution of the money, the quality of development aid could be increased substantially.