Guided by the Rusche and Kirchheimer thesis, this study examines variation in incarceration rates across states. Time-series regression analysis is applied to 30 years of state-level data to examine how economic factors interact with aggregate measures of race/ethnicity in predicting rates of incarceration. The analysis indicates that income inequality, not unemployment, is the most salient predictor of incarceration rates. That is, state-level measures of income inequality exert a strong, positive effect on state-level incarceration rates, and this effect is particularly salient in the presence of higher percentages of African-Americans.