with Macartan Humphreys and Peter van der Windt — Democratic Republic of Congo
Summary: Despite growing emphasis on Community Driven Development as a model for development, there there has been no test of the mechanisms by which CDD promises to generate social impacts. Formally testing the mechanisms of participatory development is especially relevant for political science, since it is tantamount to testing the effects of elections and community contributions (the core aspects of CDD) on citizen’s participation, inclusion, and leaders’ legitimacy.
Design: We randomly assigned leaders’ selection methods and community contributions in a large scale community driven contribution program.
The community contribution variation failed at the implementation stage, but a large sample of communities were part of the randomization pool for the selection methods. In particular, we were able to randomly assign communities to selection processes in which communities were trained to, and implemented elections in order to elect the leaders of the development project from a pool of villagers’ candidates. In the control group, leaders’ selection followed a less intrusive process whereby existing village mechanisms identified a development committee without intervention from the partner organization.
This project is at the analysis phase.