Can it really be true that most people in jail are being held before trial? And how much of mass incarceration is a result of the war on drugs? These questions are harder to answer than you might think, because our country’s systems of confinement are so fragmented. The various government agencies involved in the justice system collect a lot of critical data, but it is not designed to help policymakers or the public understand what’s going on. As public support for criminal justice reform continues to build, however, it’s more important than ever that we get the facts straight and understand the big picture.
This report offers some much needed clarity by piecing together this country’s disparate systems of confinement. The American criminal justice system holds almost 2.3 million people in 1,719 state prisons, 109 federal prisons, 1,772 juvenile correctional facilities, 3,163 local jails, and 80 Indian Country jails as well as in military prisons, immigration detention facilities, civil commitment centers, state psychiatric hospitals, and prisons in the U.S. territories.1 This report provides a detailed look at where and why people are locked up in the U.S., and dispels some modern myths to focus attention on the real drivers of mass incarceration.
This big-picture view allows us to focus on the most important drivers of mass incarceration and identify important, but often ignored, systems of confinement. The detailed views bring these overlooked systems to light, from immigration detention to civil commitment and youth confinement. In particular, local jails often receive short shrift in larger discussions about criminal justice, but they play a critical role as “incarceration’s front door” and have a far greater impact than the daily population suggests.
While this pie chart provides a comprehensive snapshot of our correctional system, the graphic does not capture the enormous churn in and out of our correctional facilities, nor the far larger universe of people whose lives are affected by the criminal justice system. Every year, over 600,000 people enter prison gates, but people go to jail 10.6 million times each year.2 Jail churn is particularly high because most people in jails have not been convicted.3 Some have just been arrested and will make bail within hours or days, while many others are too poor to make bail and remain behind bars until their trial. Only a small number (less than 150,000 on any given day) have been convicted, and are generally serving misdemeanors sentences under a year.
With a sense of the big picture, the next question is: why are so many people locked up? How many are incarcerated for drug offenses? Are the profit motives of private companies driving incarceration? Or is it really about public safety and keeping dangerous people off the streets? There are a plethora of modern myths about incarceration. Most have a kernel of truth, but these myths distract us from focusing on the most important drivers of incarceration.
The overcriminalization of drug use, the use of private prisons, and low-paid or unpaid prison labor are among the most contentious issues in criminal justice today because they inspire moral outrage. But they do not answer the question of why most people are incarcerated, or how we can dramatically — and safely — reduce our use of confinement. Likewise, emotional responses to sexual and violent offenses often derail important conversations about the social, economic, and moral costs of incarceration and lifelong punishment. Finally, simplistic solutions to reducing incarceration, such as moving people from jails and prisons to community supervision, ignore the fact that “alternatives” to incarceration often lead to incarceration anyway. Focusing on the policy changes that can end mass incarceration, and not just put a dent in it, requires the public to put these issues into perspective.
It’s true that police, prosecutors, and judges continue to punish people harshly for nothing more than drug possession. Drug offenses still account for the incarceration of almost half a million people,4 and nonviolent drug convictions remain a defining feature of the federal prison system. Police still make over 1 million drug possession arrests each year,5 and many of these arrests do lead to prison sentences. Drug arrests continue to give residents of over-policed communities criminal records, hurting their employment prospects and increasing the likelihood of longer sentences for any future offenses.
But at the state and local levels, far more people are locked up for violent and property offenses than for drug offenses alone. To end mass incarceration, reforms will have to go further than the “low hanging fruit” of nonviolent drug offenses. (As it happens, some of the boldest strategies for reforming the criminal justice system — such as heavy investments in social services and community-based alternatives to incarceration — benefit not only those with substance use disorders, but people at risk of incarceration for any offense.)
In fact, less than 8% of all incarcerated people are held in private prisons; the vast majority are in publicly-owned prisons and jails.6 Some states have more people in private prisons than others, of course, and the industry has lobbied to maintain high levels of incarceration, but private prisons are essentially a parasite on the massive publicly-owned system — not the root of it.
Nevertheless, a range of private industries and even some public agencies continue to profit from mass incarceration. Many city and county jails rent space to other agencies, including state prison systems,7 the U.S. Marshals Service, and Immigration and Customs Enforcement (ICE). Private companies are frequently granted contracts to operate prison food and health services (often so bad they result in major lawsuits), and prison and jail telecom and commissary functions have spawned multi-billion dollar private industries. By privatizing services like phone calls, medical care and commissary, prisons and jails are unloading the costs of incarceration onto incarcerated people and their families, trimming their budgets at an unconscionable social cost.
Simply put, private companies using prison labor are not what stands in the way of ending mass incarceration, nor are they the source of most prison jobs. Only about 5,000 people in prison — less than 1% — are employed by private companies through the federal PIECP program, which requires them to pay at least minimum wage before deductions. (A larger portion work for state-owned “correctional industries,” which pay much less, but this still only represents about 6% of people incarcerated in state prisons.)8
But prisons do rely on the labor of incarcerated people for food service, laundry and other operations, and they pay incarcerated workers unconscionably low wages: our 2017 study found that on average, incarcerated people earn between 86 cents and $3.45 per day for the most common prison jobs. In at least five states, those jobs pay nothing at all. Moreover, work in prison is compulsory, with little regulation or oversight, and incarcerated workers have few rights and protections. Forcing people to work for low or no pay and no benefits allows prisons to shift the costs of incarceration to incarcerated people — hiding the true cost of running prisons from most Americans.
Community supervision, which includes probation, parole, and pretrial supervision, is often seen as a “lenient” punishment, or as an ideal “alternative” to incarceration. But while remaining in the community is certainly preferable to being locked up, the conditions imposed on those under supervision are often so restrictive that they set people up to fail. The long supervision terms, numerous and burdensome requirements, and constant surveillance (especially with electronic monitoring) result in frequent “failures,” often for minor infractions like breaking curfew or failing to pay unaffordable supervision fees.
In 2016, at least 168,000 people were incarcerated for such “technical violations” of probation or parole — that is, not for any new crime.9 Probation, in particular, leads to unnecessary incarceration; until it is reformed to support and reward success rather than detect mistakes, it is not a reliable “alternative.”
Finally, we come to the myth that people who commit violent or sexual crimes are incapable of rehabilitation and thus warrant many decades or even a lifetime of punishment. As lawmakers and the public increasingly agree that past policies have led to unnecessary incarceration, it's time to consider policy changes that go beyond the low-hanging fruit of “non-non-nons” — people convicted of non-violent, non-serious, non-sexual offenses. If we are serious about ending mass incarceration, we will have to change our responses to more serious and violent crime.
The data supports changing our responses to some of the crimes that scare people most: people convicted of sexual assault and homicide are actually among the least likely to reoffend after release. People convicted of homicide are the least likely to be re-arrested, and those convicted of rape or sexual assault have re-arrest rates roughly 30-50% lower than people convicted of larceny or motor vehicle theft. More broadly, people convicted of any violent offense are less likely to be re-arrested in the years after release than those convicted of property, drug, or public order offenses. Yet people convicted of violent offenses often face decades of incarceration, and those convicted of sexual offenses can be committed to indefinite confinement or stigmatized by sex offender registries long after completing their sentences.
To understand the main drivers of incarceration, the public needs to see how many people are incarcerated for different offense types. But the reported offense data oversimplifies how people interact with the criminal justice system in two important ways: it reports only one offense category per person, and it reflects the outcome of the legal process, obscuring important details of actual events.
First, when a person is in prison for multiple offenses, only the most serious offense is reported.10 So, for example, there are people in prison for violent offenses who were also convicted of drug offenses, but they are included only in the “violent” category in the data. This makes it hard to grasp the complexity of criminal events, such as the role drugs may have played in violent or property offenses. We must also consider that almost all convictions are the result of plea bargains, where defendants plead guilty to a lesser offense, possibly in a different category, or one that they did not actually commit.
Secondly, many of these categories group together people convicted of a wide range of offenses. For violent offenses especially, these labels can distort perceptions of individual “violent offenders” and exaggerate the scale of dangerous violent crime. For example, “murder” is an extremely serious offense, but that category groups together the small number of serial killers with people who committed acts that are unlikely, for reasons of circumstance or advanced age, to ever happen again. It also includes offenses that the average person may not consider to be murder at all. In particular, the felony murder rule says that if someone dies during the commission of a felony, everyone involved can be as guilty of murder as the person who pulled the trigger. Acting as lookout during a break-in where someone was accidentally killed is indeed a serious offense, but many may be surprised that this can be considered murder in the U.S.11
Looking more closely at incarceration by offense type also exposes some disturbing facts about the 63,000 youth in confinement in the United States: Too many are there for a “most serious offense” that is not even a crime. For example, there are over 8,100 youth behind bars for technical violations of their probation, rather than for a new offense. An additional 2,200 youth are locked up for “status” offenses, which are “behaviors that are not law violations for adults, such as running away, truancy, and incorrigibility.”12 Nearly 1 in 10 youth held for a criminal or delinquent offense is locked in an adult jail or prison, and most of the others are held in juvenile facilities that look and operate a lot like prisons and jails.
Turning to the people who are locked up criminally and civilly for immigration-related reasons, we find that 13,000 people are in federal prisons for criminal convictions of immigration offenses, and 10,600 more are held pretrial by U.S. Marshals. The vast majority of people incarcerated for criminal immigration offenses are accused of illegal entry or illegal re-entry — in other words, for no more serious offense than crossing the border without permission.13
Another 49,000 people are civilly detained by U.S. Immigration and Customs Enforcement (ICE) not for any crime, but simply for their undocumented immigrant status. ICE detainees are physically confined in federally-run or privately-run immigration detention facilities, or in local jails under contract with ICE. An additional 11,800 unaccompanied children are held in the custody of the Office of Refugee Resettlement (ORR), awaiting placement with parents, family members, or friends. While these children are not held for any criminal or delinquent offense, most are held in shelters or even juvenile placement facilities under detention-like conditions.14
Adding to the universe of people who are confined because of justice system involvement, 22,000 people are involuntarily detained or committed to state psychiatric hospitals and civil commitment centers. Many of these people are not even convicted, and some are held indefinitely. 9,000 are being evaluated pre-trial or treated for incompetency to stand trial; 6,000 have been found not guilty by reason of insanity or guilty but mentally ill; another 6,000 are people convicted of sexual crimes who are involuntarily committed or detained after their prison sentences are complete. While these facilities aren't typically run by departments of correction, they are in reality much like prisons.
While this report provides the most inclusive view of the various systems of confinement in the U.S. justice system available, these snapshots can’t capture all of the important systemic issues. Once we have wrapped our minds around the “whole pie” of mass incarceration, for example, we should zoom out and note that confinement is just one piece of the larger system of correctional control. There are another 840,000 people on parole and a staggering 3.6 million people on probation. Given the onerous conditions of probation and the steep consequences for technical violations, policymakers should be wary of "alternatives to incarceration" that can easily lead to incarceration for people who pose no threat to public safety.
Beyond identifying the parts of the criminal justice system that impact the most people, we should also focus on who is most impacted and who is left behind by policy change. Poverty, for example, plays a central role in mass incarceration. People in prison and jail are disproportionately poor compared to the overall U.S. population.15 The criminal justice system punishes poverty, beginning with the high price of money bail: The median felony bail bond amount ($10,000) is the equivalent of 8 months' income for the typical detained defendant. As a result, people with low incomes are more likely to face the harms of pretrial detention. Poverty is not only a predictor of incarceration; it is also frequently the outcome, as a criminal record and time spent in prison destroys wealth, creates debt, and decimates job opportunities.16
It's no surprise that people of color — who face much greater rates of poverty — are dramatically overrepresented in the nation's prisons and jails. These racial disparities are particularly stark for Black Americans, who make up 40% of the incarcerated population despite representing only 13% of U.S residents. The same is true for women, whose incarceration rates have for decades risen faster than men's, and who are often behind bars because of financial obstacles such as an inability to pay bail. As policymakers continue to push for reforms that reduce incarceration, they should avoid changes that will widen disparities, as has happened with juvenile confinement and with women in state prisons.
Equipped with the full picture of how many people are locked up in the United States, where, and why, our nation has a better foundation for the long overdue conversation about criminal justice reform. For example, the data makes it clear that ending the war on drugs will not alone end mass incarceration, though the federal government and some states have taken an important step by reducing the number of people incarcerated for drug offenses. Looking at the “whole pie” also opens up other conversations about where we should focus our energies:
Now that we can see the big picture of how many people are locked up in the United States in the various types of facilities, we can see that something needs to change. Looking at the big picture requires us to ask if it really makes sense to lock up 2.3 million people on any given day, giving this nation the dubious distinction of having the highest incarceration rate in the world. Both policymakers and the public have the responsibility to carefully consider each individual slice in turn to ask whether legitimate social goals are served by putting each group behind bars, and whether any benefit really outweighs the social and fiscal costs.
Even narrow policy changes, like reforms to money bail, can meaningfully reduce our society’s use of incarceration. At the same time, we should be wary of proposed reforms that seem promising but will have only minimal effect, because they simply transfer people from one slice of the correctional “pie” to another. Keeping the big picture in mind is critical if we hope to develop strategies that actually shrink the “whole pie.”
People new to criminal justice issues might reasonably expect that a big picture analysis like this would be produced not by reform advocates, but by the criminal justice system itself. The unfortunate reality is that there isn’t one centralized criminal justice system to do such an analysis. Instead, even thinking just about adult corrections, we have a federal system, 50 state systems, 3,000+ county systems and 25,000+ municipal systems, and so on. Each of these systems collects data for its own purposes that may or may not be compatible with data from other systems, and that might duplicate or omit people counted by other systems.
This isn’t to discount the work of the Bureau of Justice Statistics, which, despite limited resources, undertakes the Herculean task of organizing and standardizing the data on correctional facilities. And it’s not to say that the FBI doesn’t work hard to aggregate and standardize police arrest and crime report data. But the basic reason why so many simple-sounding questions about criminal justice are in fact unanswerable is the “system” wasn’t set up to answer them.
Similarly, there are systems involved in the confinement of justice-involved people that might not consider themselves part of the criminal justice system, but should be included in a holistic view of incarceration. Juvenile justice, civil detention and commitment, immigration detention, and commitment to psychiatric hospitals for criminal justice involvement are examples of this broader universe of confinement. The “whole pie” incorporates data from these systems to provide the most comprehensive view of incarceration possible.
To produce this report, we took the most recent data available for each part of these systems, and where necessary adjusted the data to ensure that each person was only counted once, only once, and in the right place.
Before explaining the data sources, we want to explain two methodology changes that make this report not directly comparable with past reports. Unlike past years, in this report:
This briefing uses the most recent data available on the number of people in various types of facilities and the most significant charge or conviction. This year, several planned government reports were not published on their anticipated schedule, delayed in part by the government shutdown of December 2018 and January 2019. We sought out alternative data sources where possible, but some data simply has yet to be updated. Furthermore, because not all types of data are collected each year, we sometimes had to calculate estimates; for example, we applied the percentage distribution of offense types from the previous year to the current year’s total count data. For this reason, we chose to round most labels in the graphics to the nearest thousand, except where rounding to the nearest ten, nearest one hundred, or (in two cases in the jails detail slide) the nearest 500 was more informative in that context. This rounding process may also result in some parts not adding up precisely to the total.
Our data sources were:
To avoid counting anyone twice, we performed the following adjustments:
All Prison Policy Initiative reports are collaborative endeavors, but this report builds on the successful collaborations of the 2014, 2015, 2016 and 2017, and 2018 versions. For this year’s report, the authors are particularly indebted to Heidi Altman of the National Immigrant Justice Center for feedback and research pointers on immigration detention, Todd Minton for helping us understand the Bureau of Justice Statistics’ data on jails, Mack Finkel, Alexi Jones, and Maddy Troilo for research support, Wanda Bertram for her invaluable edits, and Shan Jumper for sharing updated civil detention and commitment data. This year, we are again grateful to Jordan Miner for making the report interactive, and Elydah Joyce for her help with the design. Any errors or omissions, and final responsibility for all of the many value judgements required to produce a data visualization like this, however, are the sole responsibility of the authors.
We thank the John D. and Catherine T. MacArthur Foundation Safety and Justice Challenge for their support of our research into the use and misuse of jails in this country. We also thank the Public Welfare Foundation and each of our individual donors who give us the resources and the flexibility to quickly turn our insights into new movement resources.
Wendy Sawyer is a Senior Policy Analyst at the Prison Policy Initiative. She is the author of the 2018 reports Youth Confinement: The Whole Pie and The Gender Divide: Tracking women’s state prison growth, as well as the 2016 report Punishing Poverty: The high cost of probation fees in Massachusetts.
Peter Wagner is an attorney and the Executive Director of the Prison Policy Initiative. He co-founded the Prison Policy Initiative in 2001 in order to spark a national discussion about the negative side effects of mass incarceration. He is a co-author of a landmark report on the dysfunction in the prison and jail phone market, Please Deposit All of Your Money. Some of his most recent work includes Following the Money of Mass Incarceration and putting each state’s overuse of incarceration into the international context in States of Incarceration: The Global Context 2018.
He is @PWPolicy on Twitter.
The non-profit, non-partisan Prison Policy Initiative was founded in 2001 to expose the broader harm of mass criminalization and spark advocacy campaigns to create a more just society. Alongside reports like this that help the public more fully engage in criminal justice reform, the organization leads the nation’s fight to keep the prison system from exerting undue influence on the political process (a.k.a. prison gerrymandering) and plays a leading role in protecting the families of incarcerated people from the predatory prison and jail telephone industry and the video visitation industry.
The number of state facilities is from Census of State and Federal Correctional Facilities, 2005, the number of federal facilities is from the list of prison locations on the Bureau of Prisons website (as of March 14, 2019), the number of youth facilities is from the Juvenile Residential Facility Census Databook (2016), the number of jails from Census of Jails: Population Changes, 1999-2013, and the number of Indian Country jails from Jails in Indian Country, 2016. We aren’t currently aware of a good source of data on the number of the facilities of the other types. ↩
10.6 million jail admissions includes multiple admissions of some individuals; it does not mean 10.6 million unique individuals cycling through jails in a year. According to a presentation, The Importance of Successful Reentry to Jail Population Growth [Powerpoint] given at The Jail Reentry Roundtable, Bureau of Justice Statistics statistician Allen Beck estimates that of the 12-12.6 million jail admissions in 2004-2005, 9 million were unique individuals. More recently, we analyzed the 2014 National Survey of Drug Use and Health, which includes questions about whether respondents have been booked into jail; from this source, we estimate that approximately 6 million unique individuals were arrested and booked into jails in 2014. ↩
The local jail population in the main pie chart (611,506) reflects only the population under local jurisdiction; it excludes the people being held in jails for other state and federal agencies. The population under local jurisdiction is smaller than the population (731,300) physically located in jails on an average day in 2016, often called the custody population. (For this distinction, see the second image in the first slideshow above.) The “not convicted” population is driving jail growth. ↩
The data doesn’t show how many people are convicted of drug law violations and are held in territorial prisons or Indian Country jails. ↩
In 2017, there were 1,632,921 drug arrests in the U.S., the vast majority of which were for drug possession or use rather than for sale or manufacturing. See Crime in the US 2017 Table 29 and the Arrests for Drug Abuse Violations table. ↩
At yearend 2016, six states held at least 20% of those incarcerated under the state prison system’s jurisdiction in local jail facilities: Kentucky (48%), Louisiana (58%), Mississippi (26%), Utah (26%), Tennessee (24%), and Virginia (21%). For more on how renting jail space to other agencies skews priorities and fuels jail expansion, see the second part of our report Era of Mass Expansion. ↩
According to the most recent National Correctional Industries Association survey that is publicly available, an average of 6% of all people incarcerated in state prisons work in state-owned prison industries. However, the portion of incarcerated people working in these jobs ranges from 1% (in Connecticut) to 18% (in Minnesota). For a description of other kinds of prison work assignments, see our 2017 analysis. ↩
According to the Bureau of Justice Statistics report Probation and Parole in the United States, 2016, Appendix Table 3, 98,698 adults exited probation to incarceration under their current sentence; Appendix Table 7 shows 69,855 adults were returned to incarceration from parole with a revocation. The number of people incarcerated for technical violations may be much higher, however, since nearly 78,000 people exiting probation and parole to incarceration did so for “other/unknown” reasons, and some states did not report data. ↩
The federal government defines the hierarchy of offenses with felonies higher than misdemeanors. And “[w]ithin these levels, … the hierarchy from most to least serious is as follows: homicide, rape/other sexual assault, robbery, aggravated assault, burglary, larceny/motor vehicle theft, fraud, drug trafficking, drug possession, weapons offense, driving under the influence, other public-order, and other.” See page 13 of Recidivism of Prisoners Released in 1994. ↩
The felony murder rule has also been applied when the person who died was a participant in the crime. For example, in some jurisdictions, if one of the bank robbers is killed by the police during a chase, the surviving bank robbers can be convicted of felony murder of their colleague. For example see People v. Hudson, 222 Ill. 2d 392 (Ill. 2006) and People v. Klebanowski, 221 Ill. 2d 538 (Ill. 2006). According to a recent New York Times article, the U.S. is currently the only country still using the felony murder rule; other British common law countries abolished it years ago. A small but growing number of states have abolished it at the state level. ↩
Most children in ORR custody are held in shelters. A small number are in secure juvenile facilities or in short-term or long-term foster care. With the exception of those in foster homes, these children are not free to come and go, and they do not participate in community life (e.g. they do not attend community schools). Their behaviors and interactions are monitored and recorded; any information gathered about them in ORR custody can be used against them later in immigration proceedings. And while the majority of these children came to the U.S. without a parent or legal guardian, those who were separated from parents at the border are, like ICE detainees, confined only because the U.S. has criminalized unauthorized immigration, even by persons lawfully seeking asylum.
The number of children in ORR custody shot up dramatically under the Trump administration’s “zero tolerance” and family separation policies, as ORR was tasked with taking separated children into its custody. Additionally, the recent involvement of the Department of Homeland Security in the agency’s background checking process has scared off many potential family sponsors, slowing down the placement process for these children. ↩
Our report on the pre-incarceration incomes of those imprisoned in state prisons, Prisons of Poverty: Uncovering the pre-incarceration incomes of the imprisoned, found that, in 2014 dollars, incarcerated people had a median annual income that is 41% less than non-incarcerated people of similar ages. Our analysis of similar jail data in Detaining the Poor: How money bail perpetuates an endless cycle of poverty and jail time found that people in jail have even lower incomes, with a median annual income that is 54% less than non-incarcerated people of similar ages. ↩
This is the most recent data available until the Bureau of Justice Statistics begins administering the next Survey of Inmates in Local Jails. ↩
Notably, the number of people admitted to immigration detention in a year is much higher than the population detained on a particular day. The immigration detention system took in 396,448 people during the course of fiscal year 2018. ↩