Abstract / Summary
Decarceration, the process of reducing incarceration rates, is increasingly proposed as a strategy to improve population health. Yet, its health effects remain understudied and may depend on how and where decarceration occurs. We developed a national decarceration "atlas" to systematically characterize local mechanisms of decarceration and examine their associations with population-level mortality across 2,870 counties in the United States between 1999-2019. Using longitudinal county-level data, we identified four operational types: reduced pretrial detention, reduced jail time, reduced prison admissions, and reduced prison time. We validated this classification against documented case studies of local and state decarceration. While most counties experienced at least one decarceration type during the period, declines were typically isolated to one part of one system and modest in magnitude, with median reductions of 19% to 38% ten years post-onset. Of the four types, reduced pretrial detention and reduced prison admissions - the two front-end strategies - were associated with 3.4% (95% CI, 0.7-6.1) and 1.9% (95% CI, 0.2-3.5) lower all-cause mortality, respectively, in large central metropolitan counties. By contrast, associations were consistent with the null for smaller urban and rural counties, and for mechanisms reducing the length of jail or prison incarceration. These findings indicate that decarceration is a heterogeneous phenomenon that is not universally health-promoting. Strategies to prevent incarceration may be more impactful for population health than those that shorten time served, but this pattern likely depends on local context. Further work is needed to understand the conditions under which decarceration advances public health.