A spatial analysis of local corruption on foreign direct investment: Evidence from Chinese cities

This paper examines local corruption's direct and spillover effects on city-level foreign direct investment (FDI) flows in China. The empirical analysis is based on a new dataset on local corruption at various disaggregation levels. Using a convex combination structure in spatial econometrics, we decompose the overall effect of local corruption into a direct and an indirect effect and derive six novel empirical results. First, corruption has an economically and statistically significant negative effect on home city FDI inflows (the direct effect). Second, home city corruption has a significant and positive spillover effect on FDI inflows to competing cities (the indirect effect). Third, the direct effect is 16 times greater than the indirect effect, which leads to an overall negative effect of corruption on FDI. Fourth, the effect of corruption on FDI is heterogeneous across different FDI quantiles with the most pronounced effects being present in the lowest 80th percentile. Fifth, FDI inflows to the inland and lower-income cities are more sensitive to local corruption. Sixth, the indirect effect is more pronounced for cities within the same province.