Corruption, Financial Development and Economic Growth: Theory and Evidence From an Instrumental Variable Approach With Human Genetic Diversity

Because the predicted human genetic diversity created by Ashraf and Galor (2013a; American Economic Review 103: 1-46) can be considered strictly exogenous in cross-country growth regressions, we use it and its square as instrumental variables for corruption to investigate its impact on economic growth. Our empirical finding is that corruption has both a direct, negative impact on economic growth and an indirect, negative impact through financial development. Moreover, our evidence suggests that human genetic diversity is an origin of corruption, with a significant U-shaped effect on corruption: when human traits are more homogeneous or more heterogeneous, corruption occurs more frequently, and when the diversity of human traits is at an intermediate level, corruption occurs less frequently. Our theoretical framework demonstrates the endogeneity problem associated with corruption, clarifying causality from financial development to corruption, and it supports empirical specifications, clarifying the channels through which corruption negatively affects economic growth.