With help from artist Kevin Pyle, we explain why very few people who apply for compassionate release are approved — even during a pandemic.

by Emily Widra and Wanda Bertram, May 29, 2020

With the coronavirus pandemic threatening to turn prison sentences into death sentences, many incarcerated people are seeking compassionate release — the release of people who are facing imminent death and who pose no threat to the public — to save them from dying of COVID-19 in prison. Unfortunately, what they, and the American public, are learning is that compassionate release is not a transparent and linear process, but an unpredictably ordered series of obstacles. As artist Kevin Pyle depicts here, these hurdles and delays kill the vast majority of compassionate release petitions:

Artwork by Kevin Pyle showing the complicated maze of steps required for incarcerated people to apply for compassionate release resulting in only 6% of applications being approved, while the other 94% applicants are left to die in prison. Analysis based on federal Bureau of Prisons data, no state data available.

Applying for compassionate release is a lengthy and cumbersome process. Given that those who apply are almost always terminally ill or profoundly incapacitated, the arbitrary nature of this process means many die before their cases are resolved.1

The compassionate release process varies tremendously between states (some states even give it a different name, like “medical parole,” “geriatric parole”, etc.),2 but the basic framework is the same: An incarcerated person is recommended3 for release on compassionate grounds to prison administrators, who then solicit a medical recommendation, and then administrators or members of the parole board approve or deny compassionate release. Some states allow only family and attorneys to recommend that someone be released on these grounds; others allow incarcerated individuals to apply on their own behalf, or allow prison personnel to do so.

Compassionate release programs are plagued by many shortcomings, including:

  • Requirements that a person be extremely close to death, or so incapacitated that they do not understand why they are being punished.4
  • Requiring medical professionals to attest that someone is within six months, or nine months, of death. Health professionals are reluctant to give such exact prognoses, which means prison officials will default to saying “it’s safer just to not let this person go.”5
  • Allowing the ultimate decision-makers6 to overrule recommendations from medical professionals and prison staff (e.g. by refuting or ignoring a medical prognosis).

The compassionate release process is frustratingly obscure not only for applicants, but for reporters, advocates, and others trying to understand the system. In their national survey, FAMM found that only three states are required to publish data on compassionate release grants, and eight other states publish some publicly available data, leaving most Americans in the dark about how often compassionate release is actually used. And despite that fact that FAMM has helpful memos for all fifty states and the District of Columbia detailing eligibility requirements for compassionate release, the application and referral process, the necessary documentation and assessments, and the decision-making criteria, the application process remains an arduous one.

There is plenty of room for states to improve their compassionate release processes. In our 2018 report Eight Keys to Mercy, we recommended that states:

  • Make compassionate release available to all incarcerated people, irrespective of the offenses for which they are incarcerated.
  • Streamline all compassionate release processes and set reachable deadlines so that petitioners don’t die due to bureaucratic bottlenecks before they are released.
  • Limit the ability of prison officials to overrule, on medical grounds, a recommendation of release by medical professionals.

But even when a compassionate release system operates efficiently and fairly, the majority of people in prison are still not eligible for it. As currently constituted, these programs exclude too many people and these systems were never designed for quick responses during a global pandemic. States need to look beyond compassionate release — including expedited parole, and mass commutations — to slow the spread of the pandemic and prevent a needless tragedy behind bars.

Footnotes

  1. According to the New York Times, between 2013 and 2017, the federal Bureau of Prisons approved only 6% of the 5,400 compassionate release applications received; meanwhile, 266 other applicants died in prison. Their analysis of federal prison data shows that it takes over six months, on average, for an incarcerated person to receive an answer on their compassionate release application from the BOP. In one tragic example, prison officials denied an application for someone because the BOP put aside prison doctors’ prognosis of less than six months and concluded that he had more than 18 months to live, despite. Two days after receiving the denial, he died.  ↩
  2. FAMM’s national survey found that not all states use the term “compassionate release” and instead refer to these programs as medical and geriatric parole, short- and long-term furloughs, suspension or reduction of sentences, and executive clemency on medical grounds.  ↩
  3. There are major differences between state policies, but for a more comprehensive view of state policies, go to FAMM’s detailed breakdown on state practices in their 2018 report.  ↩
  4. In Hawai’i, being “too ill or cognitively impaired to participate in rehabilitation and/or to be aware of punishment” can qualify someone for consideration for release on medical grounds.  ↩
  5. In a uniquely positive development, new guidelines for federal prisons issued by the U.S. Sentencing Commission do not require a short-term “terminal” prognosis, as an excellent article in Health Affairs points out.  ↩
  6. The “ultimate decision-makers” vary greatly by jurisdiction. For example, in the federal system, the prison warden makes a recommendation, but the final word on compassionate release comes from the Central Office of the Bureau of Prisons. In other jurisdictions, courts have the final say on compassionate release: in D.C., decisions are made in the superior court, and in Ohio the sentencing court rules on compassionate release. The Department of Corrections’ commissioner makes the decision in Massachusetts, Minnesota, and Mississippi (in Mississippi, the decision is made jointly with the Chief Medical Officer). In Arizona, Virginia, and West Virginia, the governor decides who is granted compassionate release after hearing a recommendation from the parole board. Many states rely on the parole board to make compassionate release decisions, including Alabama, California, and North Carolina. For the decision makers in every state, see FAMM’s state-specific memos.  ↩

Parole boards are granting parole contingent on participation in programs that are often not readily available for people behind bars, especially during the pandemic.

by Emily Widra and Wendy Sawyer, May 21, 2020

With public health officials and criminal justice reform advocates urging prisons to reduce their populations, people who have already been approved for release should be the first to return to their communities and families. Instead, thousands of them are waiting behind bars — where social distancing is impossible — as prisons across the country become the epicenters of the COVID-19 pandemic.

These are the people who have already been granted parole by the state parole boards, but have not yet taken a class or program that the parole board requires them to complete before they can go home. They are near enough to the end of their sentences to be parole-eligible, and the parole board has determined that they are “safe” to return to the community, but they cannot be released until they complete a program, often a drug and alcohol treatment program.

timeline showing that after being granted parole, people remain in prison waiting to participated in mandated programming Parole boards are granting parole contingent on participation in programs that are often not readily available for people behind bars. If states changed their policies to allow for these programs to occur in the community upon re-entry, they would see the prison population drop without making any other significant release policy changes. Studies show that, at least for the therapeutic community model used in many prison systems, there is nothing unique to the prison context that makes these programs more effective than when they are done in the community.

Tennessee offers a striking example of this potentially devastating policy failure. Over 1,300 COVID-19 cases in Tennessee are connected to a single state prison — Trousdale Turner Correctional Center — making it the third largest source of COVID-19 cases in the country. As Nashville defense attorney David Raybin explained to NewsChannel5, over 1,000 people in Tennessee prisons have been approved for parole but are waiting to participate in the mandated programming, most often the Department of Correction’s therapeutic community program, which lasts 9-12 months. That means Tennessee could reduce its prison population by almost 4% by releasing just those who have already been approved for parole.1

Evidence shows that these programs are effective whether offered before or after release, but states have been reluctant to offer these programs in communities instead of in prisons. But of course, education and treatment programming across the nation’s prison systems have been interrupted by the virus, as volunteers and educators are no longer entering the prison system on a regular basis; in Tennessee, the Department of Correction released a statement that the virus is causing “some disruption in programming.” Even in the best of times, participation in these programs is limited and people wait behind bars for a space in the program before they can be released.

This is not a new problem for Tennessee. Before the pandemic, a taskforce commissioned by the governor found that 40% of people granted parole from 2015-2019 had not actually been released because they were still waiting to participate in pre-release programs mandated by the parole board. That means that over those four years, more than 6,000 people were parole-eligible, reviewed and approved for release by the parole board, and then remained in prison simply because the mandated program was not offered at their facility or the maximum number of participants had already been reached.

Where did this problem come from?

Every state is different, but in Tennessee and some other states, the parole board appears to decide the criteria for someone’s release, while the prison system runs the classes and decides who is eligible to take them. Governors and state legislators need to be aware that these two parts of the criminal justice system are working against each other and against public health.

Nor is this problem unique to Tennessee. A 2015 survey by the Robina Institute revealed that at least 40 states use “institutional program participation” as a factor in release decision-making for parole. In Texas, families have voiced their concern about loved ones who have been granted parole, but are still waiting to complete a pre-release program. Officials report that people often wait for months after being granted parole to begin these programs that provide education, life skills and employment training, substance abuse treatment, and other important re-entry supports. But waiting for programming that is on-hold or indefinitely postponed is no reason for people to remain in prison, especially when incarceration puts them at a heightened risk for contracting the virus.

If parole boards do not change this practice, for as long as the virus causes a “disruption in programming,” the number of people approved for parole but still in prison will continue to grow. The solution is obvious: Parole boards can waive the requirement or offer the therapeutic community programming after release. Especially given the current public health crisis, it makes sense for these programs — which, again, have been shown to be effective when offered after release — to be moved to the community setting when it is safe to do so. And in the meantime, people who have been approved for parole should be released as quickly as possible as part of the state’s efforts to protect incarcerated people and the larger community.

The Prison Policy Initiative is exploring doing a larger project evaluating prison programming, particularly the programming used to make parole decisions. If you happen to have copies of the curricula for any programs run in your state, please send a copy to virusresponse@prisonpolicy.org.

Footnotes

  1. Tennessee’s total prison population on March 31st, 2020 was 26,124, according to the Vera Institute of Justice’s recent report, People in Prison 2019, so releasing 1,000 people would be a 4% reduction.  ↩

Some correctional authorities - responding to bad guidance from the IRS - are intercepting and returning stimulus checks for incarcerated people. We explain why people in prison and jail are eligible for, and should be receiving, emergency aid.

by Stephen Raher, May 18, 2020

Updates

September 24, 2020: A judge in California issued a ruling requiring the IRS to process stimulus payments for incarcerated people. This isn’t quite the end of the story, though — the judge’s ruling could be stayed or reversed on appeal, or Congress could amend the law to prohibit payments to incarcerated people. Read more.

June 30, 2020: Since this article came out, other commentators have also raised questions about IRS’s lack of justification for withholding stimulus payments from incarcerated people who otherwise meet the statutory criteria. Some have even pointed out that the last time Congress enacted a program of individual stimulus payments (in 2009), some incarcerated people were excluded from eligibility. This illustrates that Congress knows how to exclude incarcerated people when it wants to, but it took no such action in the CARES Act.

As far as the prevalence of the issue, the Associated Press reports that “hundreds of thousands of dollars” in stimulus payments were sent to incarcerated recipients. The same article states that prison systems in Kansas, Idaho, Montana, Vermont, Mississippi, Pennsylvania, Arizona, California, and Oregon have all intercepted and returned payments that were mailed to people in state prisons. The General Accounting Office’s report on pandemic programs states that the IRS “worked with federal and state prison officials to assist in the return of payments made to incarcerated individuals.” This admission from IRS is particularly puzzling given that the agency is currently warning hospitals and nursing homes not to intercept stimulus payments sent to patients because the payments belong to the individual recipient, not the facility. The IRS has not given any explanation of why prisons are different from hospitals in this regard.

On March 27, Congress passed the Coronavirus Aid, Relief, and Economic Security Act, more commonly known as the “CARES Act.” One of the better-known aspects of the 883-page bill is the Treasury’s disbursement of one-time economic stimulus payments, which were designed with broad eligibility requirements to get financial relief into people’s pockets as quickly as possible.

Now, the IRS is claiming that incarcerated people do not “qualify” for stimulus payments and the agency is attempting to “claw back” badly needed funds from vulnerable people who may need it most. But this policy is contradicted by the unambiguous language of the CARES Act itself.

Does incarceration make people ineligible for stimulus payments?

In short: According to the CARES Act, no. The provision regarding the stimulus payments is fairly straightforward: the government is directed to distribute $1,200 to every “eligible individual.”1 An eligible individual is defined as “any individual” other than a nonresident immigrant, someone who is claimed as a dependent on another person’s tax return, a probate estate, or a trust.2 Other parts of the law reduce the size of payments to high-earning taxpayers3 and require eligible individuals to have tax ID numbers.4 These basic eligibility requirements appear in the law itself, and are repeated on the IRS’s webpage regarding stimulus payments. There is no language in the statute that directly or indirectly suggests that incarceration status affects eligibility.

Why have I heard that payments to incarcerated people should be returned?

On May 6, 2020, the IRS updated the frequently asked questions (“FAQs”) on its webpage to say that incarcerated people do not qualify for stimulus payments and should return any payments that they receive. The IRS cites no authority for this, and the only law mentioned in the FAQ is the statute that prohibits incarcerated people from receiving Social Security payments. But this is irrelevant since the stimulus payments are refundable tax credits5 having nothing to do with Social Security. Despite the fact that this new advice comes from an IRS FAQ page, and not from the CARES Act itself, it has been cited widely in publications like Forbes, leading to a lot of confusion.

How is it possible that the IRS website would give advice that’s not consistent with the law?

There is a well-defined process for the IRS to issue rules and regulations that supplement tax laws passed by Congress. The purported ban on stimulus payments to incarcerated people was not a result of this rulemaking process. Instead, it appears that IRS made up this “rule” out of whole cloth and announced it by posting it on a webpage.

It’s impossible to say why the IRS took this unusual approach, but here’s one theory: someone in the Treasury Department may have decided that giving money to incarcerated people is bad policy. Of course, the IRS is severely under-resourced, as a result of decades of attacks by grandstanding members of Congress, so the agency doesn’t have the time or staff to go after individual incarcerated people to claw back stimulus payments. And the IRS would likely lose in court if such an action were challenged in litigation. But, by placing an FAQ on the agency’s website saying that incarcerated people cannot receive the payments, some prison systems will probably do the IRS’s dirty work by using the FAQ as a justification to intercept payments or bring disciplinary action against people who take the steps to claim the money to which they are legally entitled. In fact, we’ve heard that at least one state prison system is already doing this.

It is entirely reasonable to give emergency financial aid to incarcerated people

Even though this issue is fundamentally about the rule of law (more about that in a minute), as a practical matter some people can’t fathom giving money to incarcerated people or others who are “dependent” on the government. But there are good reasons why people in jail or prison need emergency aid in these unprecedented times. First, many incarcerated people will be released soon (especially people in jail, where stays tend to be for short periods of time). Navigating the financial hurdles of post-incarceration life is difficult even in normal times. But to state the obvious, we are not in normal times: given the record-high unemployment rates, the well-documented challenges of finding work as a formerly incarcerated person are only going to get more formidable. It makes perfect sense for the government to provide monetary aid so that recently released people can obtain housing, clothing, and food. The CARES Act stimulus payments, while modest, can provide literally lifesaving assistance for people being released from incarceration.

It also makes sense to give money to people who won’t necessarily be released from custody soon. Prisons and jails have shifted more and more costs onto incarcerated people — costs for things like hygiene supplies, medical copayments, and communication with loved ones. Since incarcerated people have little ability to earn money, they tend to rely on money transfers from friends and family to pay for basic necessities. But as family members on the outside (who are often low-income to begin with) lose their jobs in the pandemic-induced economic collapse, families will be increasingly less able to send money to loved ones inside. Providing stimulus funds to incarcerated people helps protect the health and well-being of those behind bars and provides relief to their loved ones at home.

The implications of the IRS’s policy for our government and the rule of law

Beyond the immediate implications for incarcerated people and their families, the IRS’s errant attempt to prevent incarcerated people from receiving stimulus payments is troublesome because it upends our system of government, specifically the separation of powers. Executive-branch agencies (like the IRS) are charged with implementing the laws passed by Congress, not changing the law. But that seems to be exactly what’s happening here: Congress said to give everyone money, but then the Treasury Department thought that incarcerated people should have been excluded. As every first-year law student learns, it is settled law that unambiguous statutes are to be applied as written, even if that could lead to arguably unintended consequences.6

Furthermore, interfering with the administration of the federal tax system (which would presumably include interfering with someone’s ability to claim a valid tax refund) is a federal crime.7 But we live in a time when the national government operates under a philosophy that only some people (namely, people without the right connections) are obliged to obey the law. The IRS’s sudden about-face on stimulus payments provides a troubling illustration of this mindset: a government agency has ignored the clear-cut language of the governing law in an effort to impose additional punishment on people who are serving time for violating other laws.

Footnotes

  1. 26 U.S.C. § 6428(a) and (f).  ↩

  2. 26 U.S.C. § 6428(d).  ↩

  3. 26 U.S.C. § 6428(c).  ↩

  4. 26 U.S.C. § 6428(g)  ↩

  5. Even though the stimulus payments are tax credits, the CARES Act is very clear that people may claim the payments even if they have no taxable income. See Revenue Procedure 2020-28.  ↩

  6. See Magwood v. Patterson, 561 U.S. 320, 334 (2010) (“[Courts] cannot replace the actual text [of a statute] with speculation as to Congress’ intent.”); Henson v. Santander Consumer USA, 137 S.Ct. 1718, 1725 (2017) (“[I]t is never our job to rewrite a constitutionally valid statutory text under the banner of speculation about what Congress might have done had it faced a question that, on everyone’s account, it never faced.”).  ↩

  7. 26 U.S.C. § 7212(a).  ↩


Our updated analysis finds that jails are responding to the unprecedented public health crisis by rapidly dropping their populations. In contrast, state prisons have barely budged.

by Emily Widra and Peter Wagner, May 14, 2020

This article was updated on October 21st, 2021 with more recent jail and prison population data. That version should be used instead of this one.

In the last two months, local governments across the U.S. have drastically reduced their jail populations to slow the spread of the coronavirus. The typical jail has reduced its population by more than 30%. But state prisons — where social distancing is just as impossible as in jails, and correctional staff still move in and out every day — have been much slower to release incarcerated people: The typical prison system has reduced its population by only 5%. Below, we compare the population cuts in local jails to those in state prisons, discussing just how little states are doing to keep their residents (and the general public) safe. (And note, our use of the term “reduction” is a purposeful distinction from “release,” as we have found that there are multiple mechanisms impacting populations, of which releases are but one part.)

graph comparing jail population reductions to those of prisons in the time of coronavirus. While jails continue to make quick changes in the face of the pandemic, they house only 1/3rd of the incarcerated population, while the other two-thirds are held by state and federal authorities, who are moving far too slowly. After North Dakota, the six states with the largest reductions share an important quirk: they are all small state prison systems that serve as both prisons and jails. For that reason, much of their reduction could be the result of drops in the jail portion of their populations and it is possible that the reduction of their sentenced prison population may be much smaller. The one exception is Connecticut, which after the previous version of this report sent us data showing that their pretrial population decreased 10% and their sentenced population decreased by 11% from March 1st to April 29. (For detailed data on 131 large jails, see Table 1 below and for the data on more than 600 jails see our appendix, and for the smaller changes in 41 state prison systems and the federal Bureau of Prisons, see Table 2 below.)

The strategies jails are using to reduce their populations vary by location, but they add up to big changes. In some counties, police are issuing citations in lieu of arrests, prosecutors are declining to charge people for “low-level offenses,” courts are reducing the amounts of cash bail, and jail administrators are releasing people detained pretrial or those serving short sentences for “nonviolent offenses.” (We’re tracking news stories and official announcements of the most important changes in the country on our virus response page.)

Table 1: Largest known population reductions in large local jails

Table 1. Most large jails have reduced their detained population by at least 25% in response to the pandemic, and many jails have gone much further. (And for jails of all sizes with available data, the median population reduction is 32%.) (This table is based on the daily populations of 607 jails collected by the NYU Public Safety Lab, and then filtered to show only 131 large jails — with a pre-pandemic population of at least 350 people — and those that had available population counts that pre-date the start of the pandemic. Our analysis excludes jails whose population counts were not collected prior to the pandemic because we did not want to under-report the scale of the population reductions in jails that took early decisive action. We excluded smaller jails from this table because small population variations in smaller jails can look more significant than they are. However, in the aggregate, smaller jails appear to be reducing their populations even more than larger jails because the median jail reduction for jails of all sizes is 32%. Subsequent versions of this briefing will experiment with calculating jail population reductions using a rolling 7-day average as a way to minimize the need to discuss large and small jails separately. For the data on all 607 jails with available data, see the appendix.
County jail State Percentage reduction Pre-COVID-19 jail population (large jails 350 or more people) Most recent jail population Pre-COVID date Most recent date
Clackamas OR 63% 403 148 1/27/20 5/12/20
Faulkner AR 58% 433 180 1/1/20 5/12/20
Bergen NJ 57% 573 248 1/31/20 5/12/20
Snohomish WA 55% 786 350 1/1/20 5/12/20
Scott IA 52% 464 224 2/11/20 5/11/20
Kenton KY 52% 722 350 1/29/20 5/11/20
Washington AR 49% 714 362 1/1/20 5/12/20
Pulaski KY 48% 371 192 1/29/20 4/30/20
Washington OR 48% 881 461 2/28/20 5/12/20
Jefferson CO 46% 1243 673 1/28/20 5/12/20
Rowan NC 46% 373 203 2/26/20 5/12/20
Yakima WA 46% 843 459 2/27/20 5/12/20
Cabarrus NC 45% 360 197 2/11/20 5/12/20
Yuba CA 43% 394 224 2/3/20 5/12/20
Polk IA 43% 876 503 1/1/20 5/12/20
Spalding GA 42% 409 236 2/26/20 4/29/20
Davidson NC 42% 368 215 1/7/20 5/12/20
Arapahoe CO 41% 1183 696 1/1/20 5/12/20
York SC 41% 421 250 2/18/20 5/12/20
San Juan NM 39% 458 278 1/1/20 5/12/20
Salt Lake UT 39% 2089 1268 1/31/20 5/12/20
Henderson KY 39% 439 268 2/11/20 5/12/20
Floyd GA 38% 678 418 1/29/20 4/14/20
McCracken KY 38% 567 350 2/11/20 5/11/20
Boulder CO 38% 602 372 1/1/20 5/12/20
Carroll GA 38% 464 287 2/6/20 4/20/20
Benton AR 38% 710 441 2/11/20 5/12/20
Clermont OH 37% 392 248 1/1/20 5/12/20
Lexington SC 37% 499 316 2/11/20 5/12/20
Putnam TN 37% 366 232 2/3/20 5/12/20
Bulloch GA 36% 376 240 2/21/20 5/8/20
Gaston NC 35% 631 407 1/30/20 5/12/20
Anderson SC 35% 410 265 2/27/20 5/11/20
Lafayette LA 35% 936 605 1/1/20 5/12/20
Hamilton OH 35% 1532 991 1/30/20 5/12/20
Berkeley SC 35% 511 332 1/1/20 5/12/20
Knox TN 35% 1415 920 1/28/20 5/12/20
Minnehaha SD 34% 504 332 1/1/20 5/12/20
Lafourche LA 33% 458 309 1/1/20 5/12/20
Daviess KY 32% 704 476 1/29/20 5/8/20
Shawnee KS 32% 530 360 1/28/20 5/12/20
Blount TN 32% 537 365 2/26/20 5/12/20
Baldwin AL 32% 559 380 2/28/20 5/12/20
Buncombe NC 32% 525 358 1/28/20 5/4/20
Kane IL 32% 489 334 1/21/20 5/12/20
St Joseph IN 32% 613 419 1/29/20 5/5/20
Racine WI 31% 753 517 2/28/20 5/12/20
Ellis TX 31% 410 282 1/25/20 5/12/20
Chatham NC 30% 1743 1213 2/2/20 5/6/20
Worcester MA 30% 753 529 2/11/20 4/28/20
Marion OR 29% 414 292 1/9/20 5/12/20
Galveston TX 29% 1002 707 1/28/20 5/12/20
Christian KY 29% 759 536 1/30/20 5/12/20
Houston AL 29% 361 257 1/23/20 5/12/20
Campbell KY 29% 604 430 2/11/20 5/11/20
Lancaster NE 27% 606 440 2/11/20 5/11/20
Tulare CA 27% 1548 1125 2/11/20 5/12/20
Cumberland ME 27% 354 258 1/1/20 5/12/20
Tippecanoe IN 27% 490 359 2/28/20 5/12/20
Franklin OH 26% 1923 1420 1/1/20 5/12/20
Monroe FL 26% 507 375 1/7/20 5/12/20
Spartanburg SC 26% 742 549 2/11/20 5/12/20
Bell TX 25% 857 639 1/1/20 5/12/20
Norfolk VA 25% 961 720 1/31/20 5/12/20
Bonneville ID 25% 376 282 1/1/20 5/12/20
Pamunkey VA 25% 361 271 2/11/20 5/12/20
New Hanover NC 24% 454 343 1/28/20 5/12/20
Terrebonne LA 24% 647 491 1/28/20 5/12/20
Milwaukee WI 24% 1890 1441 1/1/20 5/12/20
Guilford NC 24% 1060 809 2/11/20 4/29/20
Tangipahoa LA 23% 587 452 2/19/20 5/12/20
Boone KY 22% 427 331 1/1/20 5/12/20
Will IL 22% 739 573 1/27/20 5/12/20
Blue Ridge Lynchburg VA 22% 492 382 2/11/20 5/11/20
Warren KY 22% 684 532 2/29/20 5/12/20
Fulton KY 22% 497 387 1/29/20 5/11/20
Bernalillo NM 22% 1573 1227 1/1/20 5/12/20
Hopkins KY 22% 397 310 1/29/20 5/11/20
Tom Green TX 21% 438 344 1/1/20 5/12/20
Kenosha WI 21% 533 419 2/16/20 5/12/20
El Dorado CA 21% 389 306 1/21/20 5/12/20
Dauphin PA 21% 1121 882 1/1/20 5/12/20
Virginia Beach VA 20% 1486 1188 1/31/20 5/12/20
Ouachita LA 20% 1173 940 2/15/20 5/12/20
Walton FL 18% 471 385 1/1/20 5/12/20
Canyon ID 18% 420 345 1/1/20 5/6/20
Iberia LA 17% 409 338 1/28/20 5/12/20
Yavapai AZ 17% 473 391 1/1/20 5/12/20
Santa Rosa FL 17% 681 563 2/4/20 4/2/20
Avoyelles LA 17% 424 351 2/11/20 5/12/20
Sumter FL 17% 442 366 1/28/20 5/7/20
Franklin LA 17% 833 690 1/1/20 5/12/20
Richland LA 17% 755 626 1/29/20 5/12/20
Lancaster PA 17% 781 650 2/11/20 5/12/20
Monroe NY 17% 758 631 2/28/20 5/12/20
Shasta CA 17% 466 388 2/11/20 5/12/20
Stanislaus CA 17% 1305 1088 2/5/20 5/12/20
Riverside VA 17% 1368 1141 1/25/20 5/12/20
Middle River VA 17% 884 738 1/31/20 5/12/20
Prince Georges MD 16% 848 709 1/1/20 5/12/20
Aiken SC 16% 631 529 2/26/20 5/12/20
Shelby TN 16% 1819 1527 1/1/20 5/12/20
Wake NC 16% 1288 1082 2/11/20 5/12/20
Webster LA 16% 668 562 2/19/20 5/11/20
Claiborne LA 16% 581 489 1/1/20 5/12/20
Rapides LA 16% 875 737 1/31/20 5/12/20
Pike KY 16% 400 337 1/29/20 5/12/20
Escambia FL 14% 1450 1241 2/28/20 5/12/20
Kemper MS 14% 381 327 1/1/20 5/12/20
Brown WI 14% 721 619 1/31/20 5/11/20
St Charles LA 14% 469 403 1/28/20 5/12/20
Western Virginia VA 14% 880 757 1/25/20 5/12/20
Wayne MI 13% 2069 1800 1/1/20 5/12/20
Sarasota FL 13% 883 772 1/30/20 5/12/20
Alachua FL 12% 690 607 1/1/20 5/12/20
Jackson MO 12% 737 649 1/1/20 5/12/20
Morehouse LA 12% 484 427 1/29/20 5/12/20
Caldwell LA 11% 612 543 2/19/20 5/12/20
Randall TX 11% 389 347 2/22/20 5/12/20
Morgan AL 9% 600 547 2/26/20 5/12/20
Morgan TN 9% 600 547 2/26/20 5/12/20
Broward FL 8% 1685 1542 1/1/20 5/12/20
St Lucie FL 7% 1291 1196 1/30/20 5/12/20
Lubbock TX 7% 1243 1155 1/28/20 5/6/20
Meherrin River VA 7% 421 392 2/11/20 5/12/20
Comanche OK 4% 358 343 2/11/20 5/12/20
Clay FL 4% 397 381 1/30/20 5/12/20
Yazoo MS 3% 553 538 1/29/20 4/24/20
St Johns FL 1% 412 406 1/28/20 5/12/20
Ector TX 0% 592 592 2/21/20 5/12/20
Yuma AZ increased by 7% 356 381 1/1/20 5/12/20

Meanwhile, state Departments of Correction have been announcing plans to reduce their prison populations — by halting new admissions from county jails, increasing commutations, and releasing people who are medically fragile, elderly, or nearing the end of their sentences — but our analysis finds that the resulting population changes have been small.

Table 2: Most state prison systems show only very modest population reductions (showing 41 states — and the Federal Bureau of Prisons — where the data was readily available)

Table 2. The Vera Institute of Justice has collected and made available for this report the pre-pandemic population counts (as of December 31st, 2019) and current (as of late April/early May) counts for 41 state prison systems and the federal Bureau of Prisons. For information about the most important policy changes announced in the states that made these small reductions possible, see our COVID-19 response tracker. *Importantly, there are six states with small state prison systems that serve as both prisons and jails.
State Percentage reduction Pre-COVID-19 prison population Most recent prison population
North Dakota 19% 1,794 1,461
Hawaii* 18% 5,179 4,260
Vermont* 15% 1,608 1,369
Rhode Island* 13% 2,740 2,395
Alaska* 11% 4,475 3,985
Connecticut* 11% 12,293 10,973
Delaware* 11% 5,692 5,081
Utah 10% 6,731 6,064
Oregon 9% 15,755 14,355
Kentucky 9% 23,436 21,397
New York 8% 44,284 40,956
Colorado 7% 19,714 18,419
Nevada 6% 12,942 12,127
Louisiana 6% 31,609 29,682
New Jersey 6% 18,613 17,519
Wisconsin 5% 23,956 22,681
Massachusetts 5% 8,205 7,778
North Carolina 5% 34,510 32,795
California 5% 125,507 119,327
Texas 5% 158,820 151,126
Mississippi 5% 19,469 18,553
Pennsylvania 4% 45,875 43,852
Idaho 4% 9,437 9,028
Kansas 4% 10,177 9,740
New Hampshire 4% 2,622 2,513
Iowa 4% 9,282 8,899
Alabama 4% 28,266 27,164
Maine 4% 2,205 2,123
West Virginia 4% 6,800 6,550
Florida 4% 96,009 92,574
Missouri 3% 26,044 25,133
Georgia 3% 55,556 53,648
BOP 3% 175,116 169,426
Oklahoma 3% 25,712 24,947
Michigan 3% 38,053 36,980
Ohio 3% 49,762 48,453
Arizona 2% 42,441 41,386
Arkansas 2% 17,759 17,331
South Carolina 2% 18,608 18,160
Indiana 2% 27,268 26,707
Nebraska 2% 5,651 5,537
Wyoming 1% 2,479 2,465

Some states’ prison population cuts are even less significant than they initially appear, because the states achieved those cuts partially by refusing to admit people from county jails. (At least Colorado, Illinois, California, and Oklahoma are doing this.) While refusing to admit people from jails does reduce prison density, it means that the people who would normally be admitted are still being held in different correctional facilities.

Other states are indeed transferring people in prison to outside the system, either to parole or to home confinement, but these releases have not amounted to significant population reductions. For example, the Iowa Department of Corrections has released over 800 people nearing the end of their sentences since March 1st, but the overall net change in Iowa’s incarcerated population has only been about 4%. Kentucky Governor Andy Beshear commuted the sentences of almost 200 people convicted of felonies in early April, and the state also planned to release 743 people within 6 months of completing their sentences. Since December 2019, the Kentucky prison population has only decreased by a net 9%, while more than 85% of the jails we analyzed had dropped their populations by 10% or more.

Of the states we analyzed, those with smaller pre-pandemic prison populations appeared to have reduced their populations the most drastically. The prison population has dropped by 19% in North Dakota, the same state that we found to have the most comprehensive and realistic COVID-19 mitigation plan in our April 2020 survey. North Dakota has done more to reduce its state prison population than any other state, but even that state has done less than the typical jail in the country which has reduced its population by more than 30%.

States clearly need to do more to reduce the density of state prisons. For the most part, states are not even taking the simplest and least controversial steps, like refusing admissions for technical violations of probation and parole rules, and to release those that are already in confinement for those same technical violations. (In 2016, 60,000 people were returned to state prison for behaviors that, for someone not on probation or parole, would not be a crime.) Similarly, other obvious places to start are releasing people nearing the end of their sentence, those who are in minimum security facilities and on work-release, and those who are medically fragile or older.

If the leadership and success of local jails in reducing their populations isn’t enough of an example for state level officials, they may find some inspiration in the comparative success of other countries:

Table 3: Countries reducing their incarcerated populations in the face of the pandemic (showing 13 countries where current population data was readily available)

Table 3. The United States incarcerates more people than any other country, and all U.S. states incarcerate at higher rates than most countries. Countries around the world are recognizing that public safety includes protecting society from the unnecessary spread of COVID-19, and are reducing their prison populations in order to meet that goal. (Release counts collected by Prison Policy Initiative from news stories covering international prison and jail releases. Percentage of reductions calculated by the Prison Policy Initiative based on pre-pandemic populations — including pretrial and remand detainees — from the World Prison Brief.)
Country Percentage reduction Pre-COVID-19 prison population Number released Pre-COVID date Date of releases
Afghanistan 33% 30,748 10,000 2018 3/26/20
Turkey 31% 286,000 90,000 2019 4/14/20
Iran 29% 240,000 70,000 2018 3/17/20
Myanmar 26% 92,000 24,000 2018 4/17/20
South Sudan 20% 7,000 1,400 2019 4/20/20
The Gambia 17% 691 115 2019 4/26/20
Indonesia 14% 270,387 38,000 3/31/20 4/20/20
France 14% 72,000 10,000 3/2020 4/15/20
Ireland 13% 3,893 503 2018 4/22/20
Italy 11% 61,230 6,500 2/29/20 4/26/20
Kenya 9% 51,130 4,500 2018 4/17/20
Colombia 8% 122,085 10,000 2/29/20 3/31/20
Britain 5% 83,189 4,000 3/27/20 4/4/20

Prisons and jails are notoriously dangerous places during a viral outbreak, and public health professionals, corrections officials, and criminal justice reform advocates agree that decarceration will help protect both incarcerated people and the larger communities in which they live. It’s past time for U.S. prison systems to meaningfully address the crisis at hand and reduce the number of people behind bars.

This article updates one published on May 1st with a larger dataset of state prison population reductions collected by the Vera Institute of Justice and released alongside their report Prisoners in 2019, and with updated jail reduction figures collected by the NYU Public Safety Lab.


Our table shows that more than 10% of people incarcerated in state prisons are 55 or older - and in some states, like Montana, the percentage is much higher.

by Emily Widra, May 11, 2020

This briefing has been updated to a new version with data through 2021.

Prisons and jails have become the epicenter of the COVID-19 pandemic, with seven of the ten largest hotspots identified as state prisons and local jails. With the CDC having warned that older adults are at heightened risk for severe complications and death from COVID-19, readers have asked us: Just how many people in state prisons are older adults? We’ve answered this question — state by state — in a handy table below.

To prepare our table, we drew on the most recent age data from the National Corrections Reporting Program, 1991-2015. Age data for state prisons is broken down into categories, and older adults fall into the category of “55 and older.” Although outside of correctional facilities, the term “older adults” often refers to people 65 and older, incarceration itself shortens life expectancy and hastens physiological aging. So for the purposes of addressing how vulnerable different groups are to the coronavirus, it makes sense to consider adults 55 and older behind bars as “older adults.”

We found that, on average, more than 10% of people in state prisons are over the age of 55. Some state prison systems have much higher percentages of older adults, like in Montana, where over 17% of the state prison population is 55 years or older.

Regardless of their preexisting health conditions, all older adults are at greater risk for complications from COVID-19. As this virus threatens to turn their prison sentences into death sentences, states should use all possible strategies to release them to the care of their families.

The percentage and count of state prison population that is 55 and older, as well as total prison population, by state for the most recent year possible. Compiled by Prison Policy Initiative from National Corrections Reporting Program, 1991-2015 using year-end populations. For the number and percent of state prison populations 55 and older over time (1999-2015), see our spreadsheet, Percent of prison populations 55 and older, by state, 2013-2015 [xlsx].
State Percent 55 and older Count 55 and older Total population Year
Ala. 12.33% 3,266 26,487 2015
Alaska 10.48% 397 3,597 2014
Ariz. 9.38% 3,971 42,352 2015
Ark. 11.08% 1,734 15,647 2015
Calif. 13.07% 16,826 128,717 2015
Colo. 11.30% 2,191 19,394 2015
Conn. 7.91% 898 11,359 2015
D.C. 10.53% 641 5,414 2014
Del. 10.90% 568 5,210 2015
Fla. 12.91% 12,848 99,532 2015
Ga. 10.95% 5,717 52,188 2015
Hawaii 11.31% 669 5,917 2015
Idaho 10.74% 781 7,270 2015
Ill. 7.76% 4,041 48,159 2014
Ind. 8.72% 2,401 27,535 2015
Iowa 10.06% 944 9,388 2015
Kans. 10.92% 1,070 9,795 2015
Ky. 7.90% 1,729 21,877 2015
La. 11.51% 4,079 35,434 2015
Maine 10.81% 241 2,230 2015
Mass. 15.23% 1,391 9,134 2015
Md. 9.21% 1,850 20,095 2015
Mich. 11.66% 5,032 43,171 2013
Minn. 7.83% 795 10,153 2015
Miss. 9.43% 1,758 18,648 2015
Mo. 10.89% 3,447 31,666 2015
Mont. 17.34% 439 2,531 2015
N. Dak. 7.33% 126 1,720 2014
N. Mex. 9.51% 684 7,195 2015
N.C. 10.81% 3,947 36,524 2015
N.H. 14.86% 402 2,705 2015
N.J. 9.34% 2,021 21,638 2015
N.Y. 10.37% 5,289 50,992 2015
Nebr. 10.12% 535 5,289 2015
Nev. 12.36% 1,644 13,299 2015
Ohio 10.96% 5,969 54,455 2015
Okla. 10.68% 3,115 29,156 2015
Oreg. 12.11% 1,757 14,503 2014
Pa. 12.10% 6,049 50,005 2015
R.I. 9.45% 253 2,678 2015
S. Dak. 10.03% 345 3,441 2015
S.C. 10.29% 2,190 21,288 2015
Tenn. 8.87% 2,749 30,978 2015
Tex. 11.59% 17,456 150,627 2015
Utah 10.24% 647 6,318 2015
Va. 11.21% 4,106 36,631 2015
Vt. 10.61% 177 1,668 2015
W. Va. 11.98% 777 6,487 2015
Wash. 11.20% 1,967 17,560 2015
Wis. 10.48% 2,379 22,695 2015
Wyo. 12.81% 309 2,413 2015

Our fact sheet for advocates shows how rapidly the coronavirus can spread through correctional facilities, and how high infection rates in prisons and jails already are.

by Wendy Sawyer, May 8, 2020

To help advocates argue for more aggressive decarceration as COVID-19 spreads rapidly through the nation’s prisons and jails, we’ve created a one page PDF fact sheet.

Factsheet thumbnail.

The fact sheet includes new analysis of recent COVID-19 data, largely gathered by the UCLA School of Law COVID-19 Behind Bars Data Project, put into context using other government data sources. As a result, we were able to make a series of tables and charts to show:

  • Prevalence rates of the virus in facilities that have conducted widespread testing;
  • How many asymptomatic people test positive in facilities with universal testing — indicating that in places where only the few people with symptoms are being tested, many more untested people are spreading the virus;
  • The largest outbreaks in jails and prisons where facilities are testing incarcerated people; and
  • The rapid spread of the virus over time in the few places that publish historical data, such as the Cook County (Chicago) jail and Arkansas prison system.

We also discuss the problems with prison and jail COVID-19 data; namely, that the data we have only reflect test results, and most places still are not testing widely. What we’ve learned from the places testing everyone — not just people showing symptoms — is that the virus is rampant among incarcerated people and correctional staff, which means that some prisons and jails are acting as “spreaders” of the virus in local communities.

The rapid spread of the virus among incarcerated people is unsurprising, since social distancing is impossible in the close quarters of prisons and jails. And incarcerated people, who disproportionately suffer from chronic illnesses that make them more vulnerable to the virus, are at incredible risk. As our new fact sheet shows, federal, state, and local authorities must freeze admissions and release more people now to prevent further spread of the virus through incarceration.


We review how federal courts are modifying their procedures in the face of the COVID-19 pandemic.

by Wendy Sawyer, May 6, 2020

As we have argued recently, in order to prevent more unnecessary COVID-19 related deaths among incarcerated people, authorities must minimize the use of pretrial incarceration. In the case of local jails, “authorities” means a complex web of jurisdictions and officials who have the authority to release people from jail. The federal system is a different story. Unlike local jails, federal pretrial detention is governed by a single set of laws, but that doesn’t simplify matters much when the actual decision-makers are spread out among 94 judicial districts.

So what are federal courts doing to reduce pretrial incarceration? We looked at some of the orders that federal courts have issued to deal with pretrial populations during the COVID-19 pandemic, and identified some important issues, promising approaches, and places where more dramatic action is needed.

Judges are traditionally hesitant to involve themselves with carceral operations, given the central role of separation-of-powers in American governance. But pretrial detention is a special situation, since people who are not convicted are in the custody of the court, and judges must decide whether or not defendants should be incarcerated pending trial.

Every federal court in the country has issued general orders modifying their operations during the current pandemic. Dozens of these orders address the treatment of pretrial detainees, but most of these provisions are disappointing in their narrow scope: Many orders simply require screening of symptoms, or encourage the use of video appearances, but do nothing to actually address the public health crisis posed by incarceration during a viral pandemic. Most notably, many orders instruct the US Marshals Service (which operates federal pretrial detention) to develop procedures for monitoring the health of incarcerated people, despite the Marshals’ well-documented indifference to the health of the people in its custody.

However, a handful of court orders — ranging from generalized statements of goals to specific policy changes — stand out as actually addressing the problem, and are worthy of discussion:1

Statements of policy. On the more generalized end of the spectrum, Minnesota’s district court has entered an order directing its office of pretrial services to “reassess whether alternatives to detention exist that in its judgment will reasonably assure the appearance of the defendant and the safety of the community.” This approach is commendable, but without careful attention to the details, it runs the risk of widening the use of troublesome “alternatives” like electronic monitoring.

Protecting health one case at a time. In the middle of the spectrum of judicial responses are courts that have decided to make case-by-case decisions on pretrial detention, but have provided new procedures for speeding up the process. The federal court for Alaska has created an expedited procedure to rule on requests from people seeking release from custody prior to trial or sentencing, which includes a procedure for defendants to obtain their own medical records for use as evidence. The courts in Massachusetts and the Eastern District of Michigan have entered similar orders, with the Michigan court specifically noting the need to “reduc[e] population density in BOP and detention facilities.”

Anticipating how facilities may undermine justice. The federal court in Montana has recognized that social distancing measures will inevitably lead to more defendants being forced to communicate with their lawyers via phone or video. Accordingly, that court’s general order specifies that when that happens, “the attorney-client privilege…is not waived by the presence of third parties or the existence of monitoring.” Importantly, these protections apply whether or not facilities or phone companies advise callers that their communications are being monitored.

The court for the Eastern District of New York has taken the most comprehensive approach, by emphasizing the need for meaningful information on the conditions in facilities where people are held. The court has identified the four facilities that hold most pretrial defendants in the Eastern District, and has ordered the wardens of those facilities to provide twice-weekly reports on mitigation measures and test results. The court order goes on to direct that such reports be posted on the court’s public website.

As one can see from reading the reports submitted to the New York Court, they are sparse on details and use the bureaucratic jargon common among correctional administrators. Nonetheless, this reporting requirement is an important first step in combatting one of the pandemic’s most important drivers of fear: the lack of information.

As a currently incarcerated author noted in an article published last week, “The only way to significantly reduce inevitable deaths from an outbreak inside is to reduce the number of people inside.” While none of the court orders discussed here directly release anyone from custody, they highlight some of the issues that we will have to address when pushing for large-scale reductions in incarceration: facilitating judicial decision-making about releases, preserving privacy in an age of digital communications, and prying factual information out of correctional bureaucracies that are generally hostile to transparency.

Footnotes

  1. This is not to say that every court that hasn’t issued a general order on pretrial detention is failing to do something. Some courts may be addressing the issue informally, through robust case-by-case determinations, or through amendments to local rules. This briefing focuses on general orders because they are comparatively easy to locate.  ↩


New BJS reports show that jail and prison populations remain stubbornly high despite decreasing crime rates, and point to the shifting demographics of correctional populations.

by Alexi Jones, May 5, 2020

The COVID-19 crisis is illustrating yet another danger of our overreliance on incarceration, as jails and prisons are rapidly becoming coronavirus hotspots. As correctional facilities around the country grapple with the crisis, two new Bureau of Justice Statistics (BJS) reports, Jail Inmates in 2018 and Prisoners in 2018, provide crucial details about our nation’s correctional populations. The reports highlight the slow pace of decarceration over the past decade, the persistence of pretrial detention despite calls for reform, and the changing demographics of prisons and especially of jails.

Of course, because of COVID-19, jail incarceration in particular has changed dramatically in just a matter of weeks. It remains to be seen whether jail populations will bounce back up after the pandemic subsides, or whether decarceration will become the “new normal” – or at least a more politically acceptable strategy in places that have been reluctant to reduce jail populations. Either way, the pre-pandemic data in the BJS reports are helpful in understanding how we got to the point where so many people are needlessly jailed, and they reveal important demographic shifts in jail populations that merit further action.

Incarceration rates are dropping far too slowly, largely due to pretrial detention

Both of the new BJS reports boast of declining correctional populations, but a closer look at the data reveals the pace of decarceration is still far too slow. Prisoners in 2018 reports that prison populations decreased 9% between 2008 and 2018, meaning prison populations, on average, declined by less than 1% each year. As the nation with the highest incarceration rate in the world, such small declines represent a national failure.

The rate of decarceration in jails is similarly slow, and jail populations have even ticked up in recent years. Although Jail Inmates in 2018 and its press release boast that the “jail incarceration rate decreased 12% from 2008 to 2018,” most of that drop happened over five years ago; the jail population barely budged between 2015 and 2018. There were actually over 18,000 more people in jail on an average day in 2018 than in 2015 – despite the fact that the overall crime rate declined 11% over the same period.

Even worse, the growth of jail populations over those years can largely be attributed to an increase in the number of people held pretrial. The vast majority of people in jails have not been convicted and are simply stuck in jail waiting for their day in court, and their number has increased by 6% since 2015, while the number of people in jail who were convicted declined by 9%. That means pretrial detention has continued to drive all of the net jail growth in recent years, despite the fact that counties around the country are reforming their bail systems to reduce pretrial incarceration. Clearly, these measures have not gone far enough.

Gender, race, and geographic shifts in prison and jail populations

Another key takeaway from the recent reports: There have been striking demographic shifts in jail populations and, to a lesser extent, in prison populations. The number of women incarcerated in jails has increased, and while the women’s prison population is slowly falling, the decarceration of men in prisons continues to outpace that of women. Racial disparities remain persistent, but have actually narrowed in both prisons and jails. Finally, we see that rural jails have grown while urban jail populations have taken more significant steps toward decarceration.

Decarceration efforts seem to impact men’s populations more than women’s

graph women's jail populations increased 15% from 2008 to 2018 while men's decreased 9%.

In the ten-year period from 2008 to 2018, the number of women in local jails grew by 15%, while the number of men in jails fell by 9%, mirroring trends seen in many state prison systems and in arrests over the last decade. Our 2018 report found that, since 2009, women’s populations have fared worse than men’s populations in 35 state prison systems. Prisoners in 2018 reveals that the rate of decarceration for men still far outpaces that for women: The male prison population declined 9.3% from 2008 to 2018, but the female population only declined by 3.3%.

While not discussed in these reports, arrest trends may provide some insight into why women’s correctional populations aren’t dropping the way men’s are. Between 2008 and 2018, total arrests declined by 26.5% for men but only declined 13.6% for women. Worse, over these years, which were marked by the opioid crisis, the number of women arrested for drug offenses increased by 34%, while men’s drug arrests fell by almost 8%.

As we explain in Women’s Mass Incarceration: The Whole Pie, it is important to focus on women in local jails, because jails hold almost half of incarcerated women, and they can be especially harmful for women, who suffer from health problems at greater rates than men and more often the caretakers of children. In this light, the growth of women’s jail incarceration is especially troubling.

Racial disparities narrow as jail trends diverge in urban and rural areas

The recent BJS report also reveals that while racial disparities in incarceration persist, they are slowly narrowing. To be clear, Black people are still incarcerated at over 3 times the rate of white people in jails, and more than 5 times the rate of white people in prisons. But the jail incarceration rate for Black Americans is falling while the rate for white Americans is increasing. From 2008 to 2018, the jail incarceration rate for white Americans rose by 12% to 187 per 100,000, the Black incarceration rate fell by 28% to 592 per 100,000, and the Hispanic incarceration rate fell by 34% to 182 per 100,000. Meanwhile, in prisons, the Black incarceration rate declined by 28.2% from 2008 to 2018, while the white incarceration rate declined by 12.8% over the same period.

graph showing that black and American Indian or Alaskan Native Americans are disproportionately incarcerated.

The changing racial demographics of jails in particular can in part be attributed to the shifting geography of jail incarceration: Jail populations have been shrinking in urban areas, but growing in rural areas, which are disproportionately white. Many urban areas have taken steps to cut their jail populations by reducing pretrial detention, creating alternatives to incarceration, and no longer prosecuting minor drug offenses. Meanwhile, in rural counties, pretrial detention continues to grow, local authorities are increasingly leasing out jail beds to state and federal governments, and there are fewer resources to develop alternatives to incarceration. And of course, rural areas have been devastated by the opioid crisis, and have tended to respond to the crisis with incarceration.

The COVID-19 crisis has drawn unprecedented national attention to the dangers of incarceration. As jails around the country take steps to reduce their jail population, sustained decarceration seems much more possible. As we move forward, these BJS reports show us the work ahead: We must dramatically increase the pace of decarceration, minimize pretrial detention, pay special attention to women’s incarceration and rural jails, and continue to reduce racial disparities.


Our analysis finds that jails are responding to the unprecedented public health crisis by rapidly dropping their populations. In contrast, state prisons have barely budged.

by Emily Widra and Peter Wagner, May 1, 2020

This article was updated on May 14th to use a new, larger, dataset produced by the Vera Institute of Justice that contains the population reductions of 41 state prison systems and the Bureau of Prisons and also updated on August 5th with more recent jail and prison population data. The newest version should be used instead of this one.

In recent weeks, local governments across the U.S. have drastically reduced their jail populations to slow the spread of the coronavirus. Many have reduced the number of people in jail by 25% or more, recognizing that the constant churn of people and the impossibility of social distancing in jails make them inevitable hotbeds of viral transmission. But state prisons — where social distancing is just as impossible, and correctional staff still move in and out every day — have been much slower to release incarcerated people. We decided to directly compare the population cuts in local jails to those in state prisons, to highlight just how little states are doing to keep their residents (and the general public) safe:

graph comparing jail population reductions to those of prisons in the time of coronavirus. While jails continue to make quick changes in the face of the pandemic, they house only 1/3rd of the incarcerated population, while the other two-thirds are held by state and federal authorities, who are moving far too slowly. (For detailed data on 190 jails, see Table 1 below, and for the smaller changes in 15 state prison systems and the federal Bureau of Prisons, see Table 2 below.)

The strategies jails are using to reduce their populations vary by location, but they add up to big changes. In some counties, police are issuing citations in lieu of arrests, prosecutors are declining to charge people for “low-level offenses,” courts are reducing the amounts of cash bail, and jail administrators are releasing people detained pretrial or those serving short sentences for “nonviolent offenses.” (We’re tracking news stories and official announcements of the most important changes in the country on our virus response page.)

Table 1: Largest known population reductions in local jails (and a few disturbing increases)

Most jails have reduced their detained population by over 15%, and over one-third of jails have reduced their population by 25%. However, a handful of jails in Texas, Georgia, Louisiana, and others have seen troubling population increases. (Our analysis is based on a sample of 208 jails whose data is collected by the NYU Public Safety Lab, and because small changes in small jails can look more dramatic than they are, we excluded from this table the 672 jails with pre-pandemic populations under 350 people. Had we included those jails, the results would have been slightly more dramatic. For the data on all 812 jails with available data, see the appendix.)
County jail State Percentage reduction Pre-COVID-19 jail population Most recent jail population Dates data collected
Clackamas OR 66% 403 138 1/27/20 & 4/27/20
Kitsap WA 58% 401 168 3/4/20 & 4/27/20
Kenton KY 52% 722 345 1/29/20 & 4/26/20
Snohomish WA 51% 786 383 1/1/20 & 4/27/20
Scott IA 50% 464 232 2/11/20 & 4/27/20
Faulkner AR 50% 433 218 1/1/20 & 4/27/20
Washington AR 48% 714 372 1/1/20 & 4/27/20
Polk IA 47% 876 466 1/1/20 & 4/27/20
Pulaski KY 45% 371 203 1/29/20 & 4/25/20
Clark WA 45% 660 366 3/3/20 & 4/27/20
Washington OR 44% 881 497 2/28/20 & 4/27/20
York SC 43% 421 240 2/18/20 & 4/27/20
Jefferson CO 43% 1,243 712 1/28/20 & 4/27/20
Davidson NC 42% 368 213 1/7/20 & 4/27/20
Spalding GA 41% 409 240 2/26/20 & 4/27/20
Cabarrus NC 40% 360 215 2/11/20 & 4/27/20
Adams CO 40% 926 555 3/15/20 & 4/27/20
Gaston NC 40% 631 382 1/30/20 & 4/27/20
Rowan NC 39% 373 229 2/26/20 & 4/27/20
Arapahoe CO 38% 1,183 730 1/1/20 & 4/27/20
Hamilton OH 38% 1,532 946 1/30/20 & 4/27/20
Yakima WA 38% 843 524 2/27/20 & 4/27/20
Floyd GA 36% 675 429 1/28/20 & 4/27/20
Coweta GA 36% 390 249 2/26/20 & 4/27/20
Hamilton TN 36% 507 325 3/15/20 & 4/27/20
Knox TN 36% 1,415 908 1/28/20 & 4/27/20
Dougherty GA 35% 579 375 2/26/20 & 4/27/20
Minnehaha SD 35% 504 328 1/1/20 & 4/27/20
Anderson SC 35% 410 267 2/27/20 & 4/26/20
Multnomah OR 35% 1,145 747 3/9/20 & 4/27/20
San Juan NM 35% 458 299 1/1/20 & 4/27/20
Clermont OH 35% 392 256 1/1/20 & 4/27/20
Pueblo CO 35% 627 410 3/5/20 & 4/27/20
Hennepin Corrections MN 35% 486 318 4/2/20 & 4/27/20
McCracken KY 34% 567 374 2/11/20 & 4/27/20
Berkeley SC 34% 511 339 1/1/20 & 4/27/20
Salt Lake UT 34% 2,089 1,387 1/31/20 & 4/27/20
Boulder CO 34% 602 400 1/1/20 & 4/27/20
Lexington SC 34% 499 332 2/11/20 & 4/27/20
Benton AR 33% 710 473 2/11/20 & 4/27/20
Yuba CA 33% 394 263 2/3/20 & 4/27/20
Putnam TN 33% 366 245 2/3/20 & 4/27/20
Baldwin AL 33% 559 377 2/28/20 & 4/27/20
Houston AL 32% 361 245 1/23/20 & 4/27/20
Cumberland PA 32% 409 278 3/9/20 & 4/27/20
Buncombe NC 32% 525 357 1/28/20 & 4/27/20
Douglas GA 32% 614 418 2/11/20 & 4/27/20
Henderson KY 31% 439 301 2/11/20 & 4/25/20
Marion OR 31% 414 284 1/9/20 & 4/27/20
Cumberland ME 31% 354 245 1/1/20 & 4/27/20
Tippecanoe IN 31% 490 340 2/28/20 & 4/27/20
Chatham NC 30% 1,743 1,213 2/2/20 & 4/27/20
St Joseph IN 30% 613 427 1/29/20 & 4/26/20
Carroll GA 30% 442 308 1/24/20 & 4/27/20
Tulare CA 30% 1,548 1,080 2/11/20 & 4/27/20
Shawnee KS 30% 530 370 1/28/20 & 4/27/20
Lafayette LA 30% 936 658 1/1/20 & 4/27/20
Bergen NJ 30% 573 403 1/31/20 & 4/27/20
Racine WI 30% 753 530 2/28/20 & 4/27/20
Worcester MA 29% 753 533 2/11/20 & 4/27/20
Galveston TX 29% 1,002 710 1/28/20 & 4/27/20
Knox KY 29% 384 273 3/15/20 & 4/27/20
Blount TN 29% 537 383 2/26/20 & 4/27/20
Whitfield GA 28% 474 340 3/4/20 & 4/27/20
Daviess KY 28% 704 505 1/29/20 & 4/25/20
Franklin OH 28% 1,923 1,383 1/1/20 & 4/27/20
Lafourche LA 28% 458 330 1/1/20 & 4/27/20
Bell TX 27% 857 624 1/1/20 & 4/27/20
Washington NC 27% 455 332 3/9/20 & 4/27/20
Ellis TX 26% 410 302 1/25/20 & 4/27/20
Saginaw MI 26% 368 272 3/17/20 & 4/27/20
Campbell KY 26% 604 447 2/11/20 & 4/26/20
Midland TX 25% 474 355 3/13/20 & 4/27/20
Lancaster NE 25% 606 454 2/11/20 & 4/27/20
Bonneville ID 25% 376 284 1/1/20 & 4/27/20
Weber UT 24% 1,030 779 3/16/20 & 4/10/20
New Hanover NC 24% 454 344 1/28/20 & 4/27/20
Tom Green TX 24% 438 333 1/1/20 & 4/27/20
Will IL 24% 739 562 1/27/20 & 4/27/20
Milwaukee WI 24% 1,890 1,442 1/1/20 & 4/27/20
Christian KY 24% 759 580 1/30/20 & 4/27/20
Norfolk VA 23% 961 738 1/31/20 & 4/27/20
Ware GA 23% 406 312 1/25/20 & 4/27/20
Houston GA 23% 683 526 4/12/20 & 4/27/20
Pamunkey VA 23% 361 279 2/11/20 & 4/25/20
Monroe FL 23% 507 393 1/7/20 & 4/27/20
Spartanburg SC 22% 742 576 2/11/20 & 4/27/20
El Dorado CA 22% 389 302 1/21/20 & 4/27/20
Warren KY 22% 684 532 2/29/20 & 4/26/20
Guilford NC 22% 1,060 826 2/11/20 & 4/27/20
Las Vegas NV 22% 371 290 3/30/20 & 4/27/20
Shasta CA 22% 466 365 2/11/20 & 4/27/20
Tangipahoa LA 22% 587 461 2/19/20 & 4/27/20
Walton FL 21% 471 370 1/1/20 & 4/27/20
Yellowstone MT 21% 454 358 3/18/20 & 4/27/20
Hopkins KY 20% 397 316 1/29/20 & 4/27/20
Dauphin PA 20% 1,121 899 1/1/20 & 4/27/20
Madera CA 20% 631 507 3/4/20 & 4/26/20
Travis TX 20% 2,119 1,704 3/18/20 & 4/27/20
Bernalillo NM 19% 1,573 1,274 1/1/20 & 4/27/20
Ouachita LA 19% 1,173 954 2/15/20 & 4/27/20
Kenosha WI 19% 533 434 2/16/20 & 4/27/20
Virginia Beach VA 18% 1,486 1,213 1/31/20 & 4/27/20
Terrebonne LA 18% 647 531 1/28/20 & 4/27/20
Forsyth GA 18% 394 324 2/26/20 & 4/27/20
Lancaster PA 18% 781 644 2/11/20 & 4/27/20
Santa Rosa FL 17% 681 563 2/4/20 & 4/2/20
Laurel KY 17% 672 556 3/15/20 & 4/27/20
Canyon ID 17% 420 348 1/1/20 & 4/27/20
Escambia FL 17% 1,450 1,204 2/28/20 & 4/27/20
Blue Ridge Lynchburg VA 17% 492 410 2/11/20 & 4/27/20
Boone KY 17% 427 356 1/1/20 & 4/27/20
Wayne MI 16% 2,069 1,733 1/1/20 & 4/27/20
Iberia LA 16% 409 343 1/28/20 & 4/27/20
Prince Georges MD 16% 848 713 1/1/20 & 4/27/20
Webster LA 16% 668 563 2/19/20 & 4/27/20
Alachua FL 16% 690 583 1/1/20 & 4/27/20
Rapides LA 15% 875 740 1/31/20 & 4/26/20
Oakland MI 15% 917 776 4/9/20 & 4/27/20
Avoyelles LA 15% 424 359 2/11/20 & 4/27/20
Clark Henderson NV 15% 394 334 3/15/20 & 4/5/20
Franklin LA 15% 833 707 1/1/20 & 4/26/20
Aiken SC 15% 631 536 2/26/20 & 4/27/20
Riverside VA 15% 1,368 1,165 1/25/20 & 4/27/20
Stanislaus CA 15% 1,305 1,112 2/5/20 & 4/27/20
Wake NC 15% 1,288 1,099 2/11/20 & 4/27/20
Brown WI 14% 721 617 1/31/20 & 4/27/20
Yavapai AZ 14% 473 405 1/1/20 & 4/27/20
Fulton KY 14% 497 426 1/29/20 & 4/26/20
Monroe NY 14% 758 651 2/28/20 & 4/15/20
Middle River VA 14% 884 761 1/31/20 & 4/27/20
Claiborne LA 14% 581 502 1/1/20 & 4/27/20
Sarasota FL 13% 883 769 1/30/20 & 4/27/20
Shelby MO 13% 512 446 3/15/20 & 4/27/20
Shelby TN 13% 1,819 1,588 1/1/20 & 4/27/20
Bartow GA 12% 589 516 1/1/20 & 4/27/20
Jackson MO 12% 737 646 1/1/20 & 4/27/20
Richmond GA 12% 1,003 884 2/28/20 & 4/27/20
St Charles LA 12% 469 414 1/28/20 & 4/27/20
Morgan AL 12% 600 531 2/26/20 & 4/27/20
Morgan TN 12% 600 531 2/26/20 & 4/27/20
Washington UT 11% 371 329 4/7/20 & 4/27/20
Pike KY 11% 400 355 1/29/20 & 4/26/20
Randall TX 11% 389 347 2/22/20 & 4/27/20
Western Virginia VA 10% 880 792 1/25/20 & 4/27/20
Mohave AZ 10% 351 316 4/8/20 & 4/27/20
Kings CA 10% 488 441 4/8/20 & 4/27/20
Kemper MS 9% 381 345 1/1/20 & 4/27/20
Martin FL 9% 429 389 4/8/20 & 4/27/20
Virginia Peninsula VA 9% 378 344 3/4/20 & 4/26/20
Caldwell LA 9% 612 559 2/19/20 & 4/27/20
Morehouse LA 9% 484 443 1/29/20 & 4/27/20
Meherrin River VA 8% 421 388 2/11/20 & 4/26/20
Broward FL 8% 1,685 1,557 1/1/20 & 4/27/20
St Lucie FL 7% 1,291 1,196 1/30/20 & 4/27/20
Hardin KY 7% 644 597 4/10/20 & 4/24/20
Denver CO 7% 1,216 1,130 4/10/20 & 4/26/20
Cascade MT 7% 419 391 3/21/20 & 4/27/20
Comanche OK 6% 358 336 2/11/20 & 4/27/20
Lubbock TX 6% 1,243 1,170 1/28/20 & 4/27/20
Dane WI 6% 580 546 4/2/20 & 4/27/20
Baltimore City MD 5% 1,459 1,387 4/1/20 & 4/27/20
St Johns FL 5% 412 393 1/28/20 & 4/27/20
Northwest OH 4% 526 504 4/17/20 & 4/27/20
Pierce WA 4% 686 662 4/12/20 & 4/27/20
Summit OH 3% 401 388 4/20/20 & 4/27/20
Henrico VA 3% 1,133 1,098 4/12/20 & 4/27/20
Roanoke City VA 3% 374 363 4/9/20 & 4/20/20
Yazoo MS 3% 553 538 1/29/20 & 4/24/20
Clay FL 3% 397 387 1/30/20 & 4/27/20
Orleans LA 2% 817 797 4/8/20 & 4/26/20
Osceola FL 2% 690 675 4/16/20 & 4/27/20
Bibb GA 2% 795 779 4/26/20 & 4/27/20
Collin TX 2% 942 925 4/9/20 & 4/27/20
Western Tidewater VA 2% 736 724 4/15/20 & 4/26/20
Fort Bend TX 1% 722 718 4/9/20 & 4/27/20
Mobile AL 1% 1,086 1,080 4/17/20 & 4/27/20
Tarrant TX 0% 3,484 3,474 4/6/20 & 4/27/20
Westchester NY 0% 364 363 4/17/20 & 4/27/20
Newberry SC 0% 514 513 4/8/20 & 4/27/20
Kern CA 0% 733 732 4/20/20 & 4/27/20
Ector TX 0% 592 592 2/21/20 & 4/27/20
Kaufman TX increased by 1% 380 382 4/20/20 & 4/27/20
Mecklenburg NC increased by 1% 1,397 1,409 4/8/20 & 4/27/20
Highlands FL increased by 1% 368 372 4/20/20 & 4/27/20
Walton GA increased by 1% 367 371 4/9/20 & 4/27/20
Pennington SD increased by 5% 445 469 4/9/20 & 4/27/20
Montgomery TX increased by 5% 584 616 4/8/20 & 4/27/20
Yuma AZ increased by 7% 356 382 1/1/20 & 4/27/20
Hennepin Jail MN increased by 10% 451 494 4/3/20 & 4/27/20

Meanwhile, state Departments of Correction have been announcing plans to reduce their prison populations — by halting new admissions from county jails, increasing commutations, and releasing people who are medically fragile, elderly, or nearing the end of their sentences — but our analysis finds that the resulting population changes have been small.

Table 2: Most state prison systems show only very modest population reductions (showing 15 states — and the Federal Bureau of Prisons — where the data was readily available)

Table 2. Data collected and analyzed by the Prison Policy Initiative. Fifteen state prison systems and the Federal BOP have readily available and frequently updated populations counts, including pre-pandemic and mid-to-late April counts. For other state-level prison reforms, see our COVID-19 response tracker here: https://www.prisonpolicy.org/virus/virusresponse.html.
*Of note, Vermont is one of six states with a combined jail and prison system. Because of this, we cannot be certain how much of Vermont’s incarcerated population reduction is due to the release of pretrial detainees (who would be in jail in other states) or people sentenced to state prison, which suggests that even the state with most drastic prison population reduction is still too far behind the typical jail.
Prison system Percentage reduction Pre-COVID-19 prison population Most recent prison population Dates data collected
Vermont* 16.0% 1,649 1,385 3/13/20 & 4/27/20
Maine 7.9% 2,161 1,991 March 2020 & 4/27/20
Utah 7.9% 6,626 6,101 Feb 2020 & 4/27/20
Iowa 3.2% 8,495 8,222 March 2020 & 4/27/20
Kansas 2.5% 10,051 9,797 2/27/20 & 4/27/20
Kentucky 4.3% 12,240 11,708 2/28/20 & 4/27/20
South Carolina 1.9% 18,074 17,735 2/1/20 & 4/27/20
Mississippi 1.7% 20,879 20,519 2/3/20 & 4/1/20
Wisconsin 4.1% 23,471 22,506 2/28/20 & 4/24/20
Oklahoma 3.8% 24,994 24,042 2/24/20 & 4/27/20
North Carolina 3.5% 35,010 33,714 3/31/20 & 4/27/20
Arizona 2.0% 42,282 41,440 2/29/20 & 4/27/20
Pennsylvania 2.8% 44,756 43,500 2/29/20 & 4/27/20
Georgia 3.6% 53,523 51,618 2/28/20 & 4/24/20
California 4.0% 123,105 118,161 2/26/20 & 4/22/20
Federal Bureau of Prisons 2.1% 164,440 160,979 3/5/20 & 4/23/20

Some states’ prison population cuts are even less significant than they initially appear, because the states achieved those cuts partially by refusing to admit people from county jails. (At least Colorado, Illinois, California, and Oklahoma are doing this.) While refusing to admit people from jails does reduce prison density, it means that the people who would normally be admitted are still being held in different correctional facilities.

Other states are indeed transferring people in prison to outside the system, either to parole or to home confinement, but these releases have not amounted to significant population reductions. For example, the Iowa Department of Corrections has released over 800 people nearing the end of their sentences since March 1st, but the overall net change in Iowa’s incarcerated population has only been about 3%. Kentucky Governor Andy Beshear commuted the sentences of almost 200 people convicted of felonies in early April, and the state also planned to release 743 people within 6 months of completing their sentences. But since February, the Kentucky prison population has only decreased by a net 4.35%.

Of the states we analyzed, those with smaller pre-pandemic prison populations appeared to have reduced their populations the most drastically. The prison population has dropped by 16% in Vermont and almost 8% in Maine and Utah. But the median percentage of people released from jails hovers around 20%, still surpassing Vermont’s state prison reduction of 16%.

States clearly need to do more to reduce the density of state prisons. For the most part, states are not even taking the simplest and least controversial steps, like refusing admissions for technical violations of probation and parole rules, and to release those that are already in confinement for those same technical violations. (In 2016, 60,000 people were returned to state prison for behaviors that, for someone not on probation or parole, would not be a crime.) Similarly, other obvious places to start are releasing people nearing the end of their sentence, those who are in minimum security facilities and on work-release, and those who are medically fragile or older.

If the leadership and success of local jails in reducing their populations isn’t enough of an example for state level officials, they may find some inspiration in the comparative success of other countries:

Table 3: Countries reducing their incarcerated populations in the face of the pandemic (showing 13 countries where current population data was readily available)

Table 3. The United States incarcerates more people than any other country, and all U.S. states incarcerate at higher rates than most countries. Countries around the world are recognizing that public safety includes protecting society from the unnecessary spread of COVID-19, and are reducing their prison populations in order to meet that goal. (Release counts collected by Prison Policy Initiative from news stories covering international prison and jail releases. Percentage of reductions calculated by the Prison Policy Initiative based on pre-pandemic populations — including pretrial and remand detainees — from the World Prison Brief.)
Country Percentage reduction Pre-COVID-19 prison population Number released Dates data collected
Afghanistan 33% 30,748 10,000 2018 & 3/26/20
Turkey 31% 286,000 90,000 2019 & 4/14/20
Iran 29% 240,000 70,000 2018 & 3/17/20
Myanmar 26% 92,000 24,000 2018 & 4/17/20
South Sudan 20% 7,000 1,400 2019 & 4/20/20
The Gambia 17% 691 115 2019 & 4/26/20
Indonesia 14% 270,387 38,000 3/31/20 & 4/20/20
France 14% 72,000 10,000 3/2020 & 4/15/20
Ireland 13% 3,893 503 2018 & 4/22/20
Italy 11% 61,230 6,500 2/29/20 & 4/26/20
Kenya 9% 51,130 4,500 2018 & 4/17/20
Colombia 8% 122,085 10,000 2/29/20 & 3/31/20
Britain 5% 83,189 4,000 3/27/20 & 4/4/20

Prisons and jails are notoriously dangerous places during a viral outbreak, and public health professionals, corrections officials, and criminal justice reform advocates agree that decarceration will help protect both incarcerated people and the larger communities in which they live. It’s past time for U.S. prison systems to meaningfully address the crisis at hand and reduce the number of people behind bars.



Stay Informed


Get the latest updates:



Share on 𝕏 Donate