The Jail Incarceration Effects of Progressive Prosecutors

• In Event: Jails, Pretrial Detention, and Reentry: Evidence on Peer Effects, Prosecution, and Pathways to Work

Abstract

Do progressive prosecutors reduce incarceration? We address this question by linking district attorney elections to county-level incarceration outcomes and by improving the measurement of progressive prosecution in local elections. We define progressive candidates using direct coding of ten policy and organizational dimensions, including diversion, non-prosecution, police accountability, opposition to the death penalty, and reform-oriented affiliations.

Our primary design focuses on close elections in which the top two finalists are exactly one progressive-coded candidate and one non-progressive candidate. This produces a substantively clear treatment contrast while avoiding comparisons in which candidates differ only marginally or in ways that are difficult to interpret. We then merge these election episodes to county-level jail outcomes by expanding multi-county prosecutorial districts to their component counties and aggregating incarceration counts and denominators to the office level. We use annual outcomes based on the consistently observed Q2 county snapshot rather than constructing a pseudo-quarterly panel from sparse reporting in other quarters. We estimate stacked event-study difference-in-differences models centered on office entry in January following the election.

The main analytic sample contains 116 close-election episodes within a 10-point vote-margin bandwidth, including 69 progressive wins and 47 progressive losses. In this strict finalist sample, progressive victories are associated with lower total jail population, fewer jail admissions, lower sentenced jail populations, lower Black jail incarceration rates, and a smaller Black-White jail rate gap. These patterns suggest that progressive prosecutors reduce both aggregate jail exposure and some level-based racial disparities in incarceration.

We also examine broader ranking-based close-election designs that define the treatment as electing the more progressive finalist based on unified candidate scores. Those broader specifications generally continue to show declines in total jail population and sentenced jail populations, but estimates for admissions, pretrial detention, and racial disparity are more sensitive to how progressivism is scored and how races are selected into the comparison sample. In particular, score systems that rely more directly on raw progressive signals and their rarity produce results closer to the strict finalist design than score systems that introduce less transparent ranking differences.

Overall, the findings suggest that progressive prosecutors can reduce the jail population, but this inference depends heavily on carefully measuring substantive progressivism and using election comparisons with a clear treatment contrast. The paper contributes to debates over prosecutorial discretion, local democratic accountability, and the capacity of reform prosecutors to reshape incarceration outcomes.

Author

• Rebecca Sarah Goldstein, New York University
  Presenting Author