Community Training, Information, and Demand for Accountability

We design a field experiment to test the effect of alleviating information and collective action constraints, or alleviating both of them together, on communities' demand for accountability in the context of a large anti-corruption intervention embedded in a community driven development (CDD) program in Uganda. We focus on demand for accountability, evaluation of government performance, and voting behavior as dependent variables. The first component of the intervention aimed at alleviating collective action problems by having beneficiaries select a small group of representatives, called a social action committee, and giving them the mandate to monitor service delivery. Committee members were trained to monitor the implementation of their local development projects and to address any issues with project implementers and government officials. The second component sought to address the information problem by providing citizens with a project scorecard, including a benchmark on how their local government performed relative to others in the same district. The idea is that the benchmark will provide citizens with a signal about the performance of the actors responsible for implementing their project. Scorecards were based on data from physical audits of all projects to construct simple indicators of the quality of project implementation, as compared to other projects in the same district.