Foreign Aid and Corruption: Clarifying Murky Empirical Conclusions

This note reconciles an on-going debate on the effect of foreign aid on corruption by introducing a previously missing heterogeneity dimension of aid. The relationship was estimated using dynamic system GMM and quantile regressions (QR). Results show that both narratives in the debate are correct, contingent on the type of development assistance. The QR results are robust to endogeneity when the independent variables of interest are instrumented with their first-lags.