<p>How does the severity of incarceration—not just its occurrence—affect political participation? Past research has examined how interactions with criminal justice institutions feed back into the political marginalization of justice-involved individuals. However, this work, like most other policy feedback research, dichotomizes interactions with the state, overlooking important variation in these encounters. Leveraging discontinuities in the Pennsylvania Department of Correction’s custody level classification process as a (pre-registered) natural experiment, I study the effect of incarceration severity on post-release registration and turnout. I observe reasonably precise null effects up to three presidential elections following one’s custody assignment. These null effects are consistent across racial groups and other subpopulations. These results align with past work which has found small to negligible effects of prison sentences on turnout. Nonetheless, understanding heterogeneity in citizen-state interactions remains an important step for future research.</p>