Punished Communities: Felon Disenfranchisement, Political Participation and Public Good Provision

Abstract

This paper explores the consequences of political participation on public good provision. Answering this question has proven difficult to due to endogeneity issues, but I provide an instrumental approach and show that an extra percentage point in turnout is associated with an increase of 2 to 3% in per capita spending. This result is the opposite of what would be found in a simple OLS setting. The identification strategy relies on different voting right regimes for current and former felons across states in the United States. In states that disenfranchise all felons for life, I find that political participation in communities with a high proportion of former felons has decreased compared to other states between the 1970s and the 1990s as the United States shifted to a system of mass incarceration. I show that a similar pattern can be observed for public good provision and by linking the two outcomes, I am able to estimate the value of a vote. In the absence of data on the size of the disenfranchised population at the county-level, I use the proportion of African Americans and the poverty rate as geographic markers for the proportion of former felons given that these communities are especially vulnerable to imprisonment and support this approach in several ways. This paper therefore sheds light on mechanisms of redistribution and on political processes. It also calls to attention the potential formation of vicious circles of poverty, crime and low education, as I find that the decrease in public spending could be as high as 18% or $640 per year and per child forgone in educational investment in disenfranchising states, with effects concentrated in poorer communities.