Post Conflict Community Driven Development

with Macartan Humphreys and Peter van der Windt — Democratic Republic of Congo

Paper: Social Engineering in the tropics: Evidence from the Congo

 Funders: DFID, International Initiative for Impact Evaluation (3ie)

Final Report: Final Report Tuungane I.

Partners: International Rescue Committee, DFID

In press: one of the world’s largest ever randomized trials”;  Financial Times;  World Bank Development BlogTim Hartford’s Blog.

Ex-ante analysis plan: We registered the analysis plan in february 2011 prior to data collection. The Outcomes and Data Sources that we designed are here: ODS. In addition, we wrote a mock report using scrambled treatment data as an example of a particularly hard form of registration of analysis plan. The mock report is here: Comprehensive registration

To our knowledge, this was among the first studies registered in economics or political science (in another study assessing the impact of a similar intervention, Katherine Casey, Rachel Glennerster and Edward Miguel pre-registered, in 2009: Reshaping Institutions: Evidence on Aid Impacts Using a Pre-Analysis Plan).

Based on this registration, we explored advantages and limitations of such an extreme form of registration. We demonstrate the room for fishing when no registration is present and discuss the plausible impacts and adverse incentive effects for different models of registration of analysis plans here: Fishing (Political Anlysis, 2013).