Foreign Aid and Corruption: Clarifying Murky Empirical Conclusions

Abstract

This study considers foreign aid flow by sector in which the aid is directed and then estimates its impact on corruption in order to clarify the specific direction of aid flow that triggers (or does not trigger) corrupt practices. Data are from the Organisation for Economic Co-operation and Development database, Freedom House dataset, and the World Bank Governance Indicators. The dynamic system GMM and quantile regressions (QR) were estimated for robust estimation and correction of endogeneity issues. We found that aid flows for the development of economic infrastructure, multi-sector and programme assistance were consistently reducing corruption. This result stands for both the entire sample and for the African countries (especially for countries at the 25th, 50th and 75th quintiles). Aid flows to social infrastructure and debt relief significantly induce corrupt practices in the sampled countries. These forms of aid only spur rent-seeking behaviour for countries at the lower quintiles of corruption. Two robust checks were estimated, including: (a) using an alternate explained variable—the corruption measure by Transparency International; and (b) correcting for endogeneity in the QR estimation by instrumenting the independent variables of interest with their first-lags. For both checks, the signs and significant values of the variables were consistent with the earlier estimation. JEL Codes: B20, F35, F50, O10, O55