Small but Growing Group Incarcerated For a Month or More Has Kept Jail Populations High

The COVID-19 pandemic has focused attention on the more than 700,000 people in jails across the United States because of the potential for spread of the virus to those working and confined there. Concentrated efforts to reduce admissions and increase releases from jails in recent weeks have resulted in significant population drops in many of these local facilities. According to a running census across 254 counties by the Public Safety Lab at New York University, the number of people in jail dropped 29% from March 1 to May 1.

This change represents a sharp contrast to pre-pandemic trends. A recent Pew analysis of Bureau of Justice Statistics (BJS) data showed virtually no change in jail populations from 2010 to 2017, although crimes, arrests, and jail admissions declined. Researchers found that the average time spent in jails nationwide increased about 22% between those years from approximately 21 to 26 days, a change that helps to explain the continued high populations.

To explore the dynamics behind these numbers, Pew analyzed BJS data from large jails. From 2010 to 2014, BJS asked these facilities—which generally housed over 500 people and represented about 20% of the total national jail population—for details on how long people had been in jail at the time of their release. Though not nationally representative, these jails experienced similar trends in admissions, length of stay, and average daily populations and therefore may reflect the complex dynamics found in many jails.

Those in jail more than a month occupy most jail space

Pew modeled the share of jail space occupied in 2014 by the share of people serving short, moderate, and long stays in the large jails for which data was available.

Jails are locally operated correctional facilities that hold diverse populations, including people who are awaiting trial, are serving sentences of a year or less for minor crimes, are being held pending transfer to another facility, or have violated probation or parole supervision. By contrast, state-run prisons typically hold people serving sentences of more than a year. Although they have only half the average daily population of prisons, jails see about 18 times more annual admissions.

The Pew analysis showed that the 21% of those released after serving a month or more took up almost 85% of jail space. (See Figure 1.) Those in jail between a week and a month (17% of those released) accounted for 11% of the population. And the 62% incarcerated for less than a week used just 4% of jail space. (See the methodology for detail on Pew’s model.)

To further illustrate the concentrated impact of longer stays, just 5% of people released from jail had stays more than six months, but they accounted for about 40% of jail space. Because the BJS survey’s longest grouping is “more than 180 days,” this figure could be even greater if a significant share of people in this group spent much longer in jail.

Recent data from the state and local levels also support the finding that the small share of long stays accounts for most of the jail space. For example, a January 2020 Michigan state analysis of 20 facilities of varying sizes found that people who spent a month or more in jail represented only 17% of admissions but accounted for 82% of space. Similarly, a Pew analysis of 2019 data from New York City jails found that the 14% of people who stayed more than 180 days took up 65% of the space.

Sizable decrease in short stays mostly offset by small increase in long ones

Pew then looked at changes in lengths of stay from 2010 to 2014. The study included only those large jails with reliable data for both years. These jails experienced a 5% increase in their average stay in that period. The analysis found that short jail terms of a week or less declined in both number and share of releases over that time, while longer terms of a month or more increased in both measures. Overall, the jails experienced an 8% decline in the number of stays, but only a 3% drop in population.

A person staying a month or more takes up at least as much space—known as “bed days”—as four people staying a week or less. Because of that dynamic, small increases in the number in jail for long periods can offset much larger decreases in the number of short stays.

This appears to be the story in Pew’s sample of America’s large jails: Those numbers show that 219 additional long stays mostly offset the impact of 2,606 fewer short stays. That helps explain why an 8% decline in admissions translated to just a 3% drop in population. (See Table 1.)

Table 1

Jail Population Dropped Only 3% Despite Larger Decline in Number of People Cycling Through

Decrease in those incarcerated a month or less mostly offset by increase in month-plus stays

Source: Pew analysis of large jail data, BJS survey of jails, 2010 and 2014

Understanding dynamics of jail length of stay is critically important now

Given the current spread of COVID-19 in some correctional facilities—two of the top 10 largest known outbreaks in the U.S. were in the Chicago and Houston jails—the drop in annual admissions by around 2 million people from 2010 to 2017 probably helped reduce virus exposure in jails. However, the reality that length of stay grew and jail populations remained high meant that a large number remained incarcerated as the pandemic took hold.

What accounts for the large and growing number of people spending weeks or months in jail—and what differentiates this group from those leaving in a matter of hours or days—cannot be determined from these data sources. Future work will analyze these dynamics to help policymakers as they seek ways to reduce jail populations safely. The bottom line, however, is clear: The success of these state and local efforts will depend on addressing not just the number of people coming in, but also how long they stay.

Jake Horowitz is a director and Tracy Velázquez is a research manager with The Pew Charitable Trusts’ public safety performance project.