This paper explores popular economic, political, and social conceptions of the relationshipbetween income inequality and corruption. Contemporary politics have become heavily divided on income inequality. Therefore, using current econometric methods such as panel Granger (non)-causality, dynamic panel models, and cross-lagged maximum likelihood estimation to understand better the underlying data generating process. The significance of this paper is threefold. First, multiple dynamic regressions are used to test the econometric assumptions identifying possible bidirectional causality. Second, the analysis finds that income inequality is not a statistically significant determinant of contemporaneous corruption. However, preceding values of the income share of the top 1% do correlate with future values of corruption in all dynamic models. Last, to test these assumptions against the new institutional hypothesis proposed by North (1990), a Panel Vector Autoregression tests whether specific political environments favor economic growth. The Panel Vector Autoregression results show that variations in per capita GDP can explain 8% of the variation, and 1% income share explains 13% of the variation in corruption perceptions. Further, robustness checks of the dynamic panel models include the use of two different periods. First, by using data from 2000 and 2002 through 2017 and because of the slow movement to change in corruption and inequality, I took the average over three-year periods from 2002 through 2017.
