On the Political Economy of Felon Disenfranchisement

Abstract

Nearly 6.1 million U.S. citizens are politically disenfranchised because of a prior felony conviction, and these citizens tend disproportionately to be black. Specifically, more than 7.4% of the adult African-American population is disenfranchised compared to 1.8% of other Americans. This paper investigates the political consequences of this large racial disparity in disenfranchisement rates. To obtain the first causal estimates of the effects of felon disenfranchisement (FD), we build a new database that catalogs the annual state changes in disenfranchisement law. We show that these changes are driven by lengthy, uncertain, and complicated court cases which are outside of the control of individual state legislatures. We use a difference in difference strategy to analyze the impact of these changes in felon disenfranchisement laws. Our results suggest that FD legislation is associated with a 3 percentage point reduction in the likelihood of voting, allowing for a range of race-specific effects of demographic and geographic characteristics. This number is larger than would be implied purely by the mechanical effect due to the change in the number of eligible voters suggesting that FD also reduces turnout amongst those eligible to vote. Next, we show that relaxations in FD laws increase the number of Black U.S. Representatives. Finally, we show that relaxations also lead to an increase in state policy liberalism.