When Does Success Spread? Evidence from Project-Level Evaluations in Africa

This paper examines whether the success of prior aid projects affects the performance of subsequent ones based on a study of World Bank projects in Africa. While existing research has shown that aid effectiveness depends on macroeconomic conditions, institutional quality, and managerial capacity, far less is known about whether project success generates localized improvements that extend beyond its immediate boundaries. Using geocoded data and independent evaluation ratings of World Bank projects, this study finds that local success does not systematically translate into higher performance for subsequent projects. In fact, it can sometimes make it less likely for future projects to achieve top outcomes. I also found evidence of a negative effect in localities with dense exposure to highly satisfactory projects. Regions with a mix of highly rated and moderately rated projects tend to do better in the future than those with only top-rated projects. This means that while lessons have been learned from very successful projects, having too many of these in one place can have negative spillovers. Therefore, close monitoring is needed in regions with high concentrations of top-performing projects to ensure that early successes and lessons learned are managed appropriately, preventing them from undermining later performance.