Evaluating and Optimizing Mass Decarceration as a Pandemic Mitigation Strategy

The COVID-19 pandemic exposed the vulnerabilities of carceral systems as high-risk environments for disease transmission. In response, jurisdictions across the United States implemented mass decarceration policies aimed at reducing jail populations. This paper investigates such interventions and their optimal planning, leveraging system-theoretic insights. First, we empirically demonstrate the influence of jail population changes on COVID-19 infections by applying convergent cross mapping, transfer entropy, and linear regression to real-world data. We find that the jail population dynamics were associated with the evolution of the pandemic. Building on this empirical finding, we present a compartmental epidemiological model that captures the coupled dynamics of disease spread in jails and their surrounding communities, with fluxes modeled as time-dependent control inputs. Using optimal control theory, we derive an optimal design for these fluxes to balance public health and public safety. Our findings provide both empirical and model-informed evidence that mass decarceration was not only a reactive policy choice but also a justifiable intervention for pandemic mitigation. We emphasize that our optimization is carried out within the space of admission and release policies and does not rank decarceration against other public-health measures.