"There is some evidence indicating that voting behavior and criminal behavior are related. In a longitudinal survey of 1,000 young adults, Uggen and Manza (2004) found that only around 5% of the voters had been arrested or incarcerated compared to the non-voters of whom 16% had been arrested and 12% had been incarcerated. And, among former arrestees, approximately 27% of the non-voters had been rearrested compared to 12% of the voters. Similar patterns emerged when examining self-reported crime such that voters reported significantly less crime than non-voters. Thus, voting appears to be part of a 'package of prosocial behavior.'"
Dhami, Mandeep K. "Prisoner Disenfranchisement Policy: A Threat to Democracy?" Analyses of Social Issues and Public Policy (Washington, DC: The Society for the Psychological Study of Social Issues, December 2005) Vol. 5, Issue 1, p. 9.