Science’s COVID-19 reporting is supported by the Pulitzer Center and the Heising-Simons Foundation.
Even before COVID-19 began to sweep through U.S. correctional facilities, Michael Daniels saw the storm coming. As the director of justice policy and programs for Franklin county in Ohio, Daniels knew the county’s two jails, with about 1950 inmates, wouldn’t allow for social distancing to control the coronavirus’ spread. So, back in March, he asked his team: How could they get as many people as possible out of there quickly?
In New York City, Elizabeth Glazer, director of the Mayor’s Office of Criminal Justice, was having similar conversations. The pandemic “distilled to its essence [how] we think about the use of jail,” she says. “Was it worth putting somebody in jail if you thought that they were at risk of getting COVID?”
As they feared, crowded jails and prisons have been deadly. By now 120,000 COVID-19 cases and 1000 deaths have been documented among people incarcerated in U.S. prisons alone. As cases surged, public health experts amplified a long-standing, unfulfilled demand of criminal justice reform advocates: Lock fewer people up. Because of the virus, such decarceration efforts suddenly made speedy progress. “Policy recommendations that we were unable to get traction on for 2 years—we were able to get them done in 3 weeks,” Daniels says.
Nationwide, jail populations plunged by about 25% between March and June, according to an analysis by the nonprofit Vera Institute of Justice. New York City and Franklin county both managed reductions of 30% in their jails, which primarily hold people charged with crimes but not yet convicted. Populations of prisons, which hold people serving sentences after a conviction, budged much less; an analysis by the Marshall Project and the Associated Press found an 8% decrease nationwide during that period.
The result is a major experiment in public health and criminal justice. Initial studies suggest decarceration has lowered infection rates in some jails. But overcrowding persists, and advocates urge further reductions. A committee convened by the U.S. National Academies of Sciences, Engineering, and Medicine (NASEM) is developing best practices for decarceration as a COVID-19 response, slated for publication in October. And scientists hope to study potential social consequences of population reductions, including changes in crime rates. “We’ve created … a society that has relied on incarceration as a solution to our social problems—and recently, that system was downgraded by like 30%,” says Vincent Schiraldi, a justice policy researcher at the Columbia School of Social Work. “Shame on us if we don’t study that in a sophisticated way.”
Prison and jail outbreaks heighten the inequality of COVID-19’s burden. People of color are incarcerated at higher rates than white people and tend to get longer sentences, and people who are incarcerated have higher rates of underlying health conditions that predispose them to severe COVID-19. Meanwhile, the safety of people in prisons is entangled with that of the surrounding community. The virus can travel back and forth with employees (23,000 infections have been documented among prison staff) and with people held for short jail stays or transferred between facilities. A June study in Health Affairs estimated that 15.7% of COVID-19 cases documented in Illinois by mid-April were associated with people moving through Chicago’s Cook County Jail.
“If we care about the community rates [of COVID-19], then we have to care about prisons and jails,” says Emily Wang, a physician at the Yale School of Medicine who heads its Health Justice Lab and co-chairs the NASEM committee on decarceration.
Jurisdictions have taken various tacks to reduce populations. New York City did it primarily by releasing two groups from jails: people being held for parole violations and those serving short sentences. The strategy in Franklin county included waiving some cash bail requirements, expanding the use of electronic monitoring to allow more people to await trial at home, and encouraging citations rather than arrests for certain misdemeanors.
Nationwide, the population drop in jails reflected a drop-off in arrests—likely because fewer crimes were committed during lockdowns and law enforcement officers aimed to avoid unnecessary physical contact, says Michael Jacobson, a sociologist at the City University of New York who has analyzed data on crime and policing in 50 cities.
To reduce prison populations, some states, including California, Oklahoma, Illinois, and Colorado, have halted the transfer of people who would normally move from jail to prison after sentencing. Governors have also commuted the sentences of inmates who were deemed medically vulnerable or were nearing the end of their sentences. And some states are trying to ramp up mental health care, addiction treatment, and other services that ultimately divert people from prisons. “The most successful [approach] is simply to not put people in to begin with,” Annette Chambers-Smith, director of the Ohio Department of Rehabilitation and Correction, told attendees in a 20 August NASEM webcast. “Turn the tap off.”
As populations dropped, some researchers tried to track the effects on disease spread. Wang and her colleagues estimated the reproduction number of the virus—how many people are infected by each newly infected person—over 83 days at a large urban jail, which they did not identify publicly. As the jail reduced its population by 25% and moved about two-thirds of residents into private cells, that number dropped from 8.25 to 1.72, they reported in a June preprint on medRxiv. (It later dipped below one, indicating the outbreak was in check, after the jail set up widespread testing of asymptomatic people.)
In another study, published 21 August in JAMA, Harvard University epidemiologist Monik Jiménez and colleagues found that among 13 county jails in Massachusetts, those with greater reductions in population from early April to early July also had lower rates of COVID-19 infections. Jiménez notes, however, that limited and inconsistent testing data make it hard to sort out exactly how much decarceration helped prevent infections.
Lauren Brinkley-Rubinstein, a community psychologist at the University of North Carolina, Chapel Hill, aims to better predict such health effects. Through the COVID Prison Project, her team pulls daily case counts from state prison reports. She collaborated with researchers at Stanford University and the University of Miami to group 103 Texas prisons based on rates of COVID-19 cases and deaths. Prisons classified as “low outbreak” were at 85% capacity, the team reported in a preprint last week on medRxiv, proposing that number as a “benchmark” for reducing infections.
With Wang’s team and researchers at Stanford, Brinkley-Rubinstein hopes to combine case numbers with publicly available data about the layout of different facilities and how inmates are housed. That might help them forecast how changes in a given facility’s population will influence its risk of COVID-19, she says. “I can say all day long, ‘Reduce your population,’ [but] a department of corrections might come back to me and say, ‘OK, but how many? Who should I target? How many should I release?’ That precision is very important.”
Other researchers aim to document the effects of the speedy decarceration on public safety. Decades of criminology research suggest many inmates can be released with minimal risk of recidivism, Jacobson says. But the fear of releasing even one person who might commit a crime helps explain why researchers have had little opportunity to study the effects of rapid, large-scale decarceration before the pandemic. Even now, political calculations explain why jails—most of whose inmates have not been convicted—shrank more than prison populations during the pandemic, says Sharon Dolovich, a law professor at the University of California, Los Angeles, and head of the Covid-19 Behind Bars Data Project, which tracks efforts to improve conditions and reduce populations in jails and prisons.
There’s no evidence so far that pandemic-inspired releases have raised crime rates. A July analysis of 29 U.S. locations by the American Civil Liberties Union found no relationship between reductions in jail populations and crime trends between March and May. Both Glazer’s and Daniels’s teams have thus far found very few reoffenses among the people released early from the New York City and Franklin county jails. Criminologist Daniel Nagin and statistician Amelia Haviland at Carnegie Mellon University plan to document the impact of the pandemic on jail populations and explore how population changes in U.S. jails relate to crime rates.
A potential downside of the pandemic’s speedy decarceration, says Matthew Akiyama, a clinician and public health researcher at the Albert Einstein College of Medicine, is that “discharge planning wasn’t as rigorous as it might have been.” People released from prison already struggle to access medical care, addiction treatment, and other supports for re-entry into society, he notes, and the new releases “left people floating in the wind, to a certain extent.” But the threat of COVID-19 has also inspired new forms of support. On 27 August, California Governor Gavin Newsom announced a joint effort with philanthropic groups to provide $30 million to organizations that offer transportation, quarantine housing, health care, and other services to people released from prison. California and New York City have set up hotel stays for people leaving jail, allowing them to quarantine and avoid crowded homeless shelters. Such initial stability may help them thrive long term, Glazer says.
But many jail systems, including Glazer’s in New York City and Daniels’s in Franklin county, have seen upticks in their populations since the rapid plunges earlier in the pandemic—likely at least in part because rates of arrest rebounded.
Local officials are trying to hang on to the recent progress, Daniels says. Franklin county’s municipal court has made the issuance of citations standard for some offenses and downgraded failure to appear in court from a jailable offense. Now that he’s confident the county can quickly shrink its jails without risking public safety, there’s no reason they shouldn’t stay that way, he says. “Not if I can help it.”
With reporting by Eli Cahan.