A new Bureau of Justice Statistics study, conducted in collaboration with Abt Global, on the impact of COVID-19 on state and federal prison systems found a sharp drop in the number of prisoners from January 2020 to February 2021. The decline of 215,800 prisoners—16 percent of the 1.3 million persons incarcerated— occurred in every state, in federal Bureau of Prisons (BOP) facilities, and in privately operated prisons under contract with a state or the federal government.

The report also documented nearly 2,500 prisoner and 196 correctional staff deaths from COVID-19-related causes from March 2020 to February 2021, with 83 percent of those who died age 55 or older. Prisons administered 4,816,400 viral tests from March 2020 through February 2021, with 8.2 percent of the tests positive. Infection rates were 219 per 1,000 state prisoners, 298 per 1,000 federal prisoners, 269 per 1,000 state prison staff, and 188 per 1,000 for BOP staff.

The data collected from 49 states varied widely because states used different procedures to address the pandemic. Some tested everyone at a site, while others tested only those with symptoms. The study did not compare prison data with the pandemic in the general population because of the state variations. Instead, the study assumed the risk of contracting COVID-19 was equal across all months and states, while among the general population, researchers documented waves of increases and decreases.

Beyond testing differences, states varied in their approaches to early releases and reducing admissions to minimize transmission. The drop in the U.S. prison population was primarily because of fewer admissions, rather than a rise in people released. Only six percent of releases were expedited.

The difference in treatment of violent and non-violent offenses was notable. In one state, for example, violent-offender admissions began to increase after the initial drop, but non-violent admissions kept going down. In another state, both violent and non-violent admissions started to increase again in mid-late 2020, but at a faster rate for non-violent offenders.

The study did not determine which state approach worked best. Nor did it address why admissions declined. Future research could explore possible explanations such as courts shutting down, people being held in jails for longer periods, and fewer parole revocations.