Article: Silver, I. A., Burtch, E., Janda, K., Lattimore, P. K., & DeMichele, M. (2025). Slowing Down Reinvolvement in the System: A Multi-Site Examination of the Effects of COVID on Time Until Readmission into Jail. American Journal of Criminal Justice, 50(1), 58-74. https://doi.org/10.1007/s12103-024-09778-4
1. Background (PDF)
Each year, millions of Americans are admitted into local jails—over 7 million in 2022 alone—with roughly 1.5 million individuals re-booked multiple times within the same year. This cycle of arrest, release, and re-arrest creates a “revolving door” that disrupts employment, housing, and family stability for individuals involved in the system. Repeated re-admissions also increase the burden on police, courts, and jail facilities, contributing to jail crowding across the United States.
The COVID-19 pandemic disrupted this cycle. Beginning in March 2020, criminal justice agencies across the United States implemented emergency policies to limit COVID-19 transmission. These policies included reducing arrests for low-level offenses – through the increased use of citations – and using book-and-release procedures to lower jail populations. While these measures were designed as public-health interventions, they also represented an unprecedented, large-scale policy experiment in decarceration.
The current study sought to determine if the emergency policies implemented in response to the COVID-19 pandemic affected how quickly individuals cycled back into jail (were rearrested). By analyzing over 300,000 jail admissions across five U.S. jurisdictions from January 1, 2017 to October 18, 2022, this research provides the first multi-site empirical test of whether COVID-19 criminal legal system policies slowed how quickly individuals became reincarcerated in county jails.
2. Summary of Findings
The analytical sample for the current study included 322,117 jail admissions involving 176,648 unique individuals. Roughly, 46 percent of admissions resulted in a subsequent booking into jail between 2017 and 2022 (the study period). Average survival time in the community—the number of days between release and the next jail admission—was 687 days (~2-years), though this varied widely across people and jurisdictions. Approximately 3/4th of admissions occurred before the COVID-19 pandemic and 1/4th of admissions occurred after the pandemic (March 13, 2020).
On average, individuals booked after March 2020 remained in the community longer. In particular, being booked into jail after March 2020 was associated with a 73 percent increase in the number of days until readmission in comparison to those booked into jail before the start of the COVID 19 pandemic. The observed percentage increase equals approximately 1,400 days until readmission or 3-4 years. These effects were consistent across all five study sites, though their magnitude ranged from 36 percent to 304 percent longer survival times, highlighting policy differences in how a jurisdiction responded to the COVID-19 pandemic.
Individuals released from jail after the onset of COVID-19 were less likely to be re-admitted and, when they were, it took substantially longer for that to occur. The increase in community survival suggests that policies implemented by the criminal legal system in response to COVID-19 collectively slowed the rate at which individuals reentered jail. Notably, this effect persisted even after controlling for detailed offense types, prior criminal history, detention length, race, gender, and county fixed effects. This finding provides quantitative evidence that large-scale shifts in enforcement and detention policy can meaningfully interrupt the arrest, release, and re-arrest cycle of jail admissions.
3. Implications
Leverage Evidence-Based Decarceration Strategies: Rather than treating the policies implemented in response to COVID-19 as a one time effort, policymakers should view their immediate implementation as an opportunity to evaluate these approaches and assess their ability to reduce readmissions into jail. Streamlining arrest and booking procedures, prioritizing community-based supervision, and relying on technology (e.g., remote hearings, electronic monitoring) can preserve public safety while minimizing incarceration’s collateral harms. Additionally, criminal legal system leaders and legislative bodies could minimize the burden on the criminal legal system by learning about the beneficial effects of the policies implemented during the COVID-19 pandemic.
Identify and Retain the Most Effective Policy Components: Because the study relied on aggregate before-and-after comparisons, future analyses should isolate which specific COVID-19 policies produced the largest reductions in re-booking rates. This information could be used to revaluate the effects of these policies with consideration for long-term implementation. Jurisdiction-specific differences observed in the supplemental analyses suggest that local context and implementation fidelity strongly shape policy effectiveness.
4. Data and Methods
The study used administrative data collected through the Advancing Pretrial Policy and Research (APPR) initiative. Records captured all adult jail admissions across five diverse U.S. counties—from a large southern metropolitan area to smaller northwestern jurisdictions—between January 2017 and October 2022. The analytic sample comprised of 322,117 admissions representing 176,648 individuals. Admissions included new arrests, probation or parole violations, post-sentence confinements, transfers, and immigration detainers, enabling comprehensive assessment of local jail churn.
To measure the dependent variable, researchers calculated the number of days between each individual’s release and subsequent re-admission, creating both a binary indicator (whether any re-booking occurred) and a continuous time-to-event measure. The key independent variable distinguished admissions before versus after March 13, 2020, approximating the implementation of pandemic-related system reforms.
The models adjusted for 13 time-varying covariates (offense type, detention length, cumulative prior offenses by category) and seven time-invariant factors (race, sex, and county indicators). A mixed-effects parametric survival model using a log-normal distribution was estimated via the GAMLSS framework in R to capture both individual- and county-level variation. Supplemental analyses tested alternative pandemic onset dates and site-specific effects, with consistent direction and significance across replications.
5. Conclusion
This multi-site investigation demonstrated that the criminal legal system’s response to COVID-19 substantively slowed the rate at which individuals were re-booked into jail. After March 2020, people spent longer periods in the community and were less likely to experience new admissions, suggesting that pandemic-driven policy adaptations reduced reinvolvement and rebooking into jail.
While many of these policies were temporary, their unintended benefits highlight actionable pathways for reform. Scaling back unnecessary arrests, promoting cite-and-release practices, and adopting virtual court and supervision technologies could all help maintain smaller jail populations and improve community outcomes. These findings should not be interpreted as pandemic-specific anomalies but as evidence of structural flexibility within the criminal legal system. When policy deliberately prioritizes public health, appropriate enforcement, and efficient resource allocation, both individuals and institutions benefit.
The COVID-19 pandemic revealed that large-scale decarceration is possible without compromising system stability. By understanding which emergency-era practices slowed jail readmissions, policymakers can implement policies that minimize the burden on criminal legal system actors while also maintaining community safety.
Disclosure: This research brief was prepared by ChatGPT and reviewed/edited by Ian A. Silver.
ianasilver.com 