Punishment Beyond Prisons 2026:
Incarceration and supervision by state
by Leah Wang
April 2026
Press release
- States of mass punishment
- Fifty years of probation and parole
- Probation and parole do not “work”
- Recommendations
- Appendix: Incarceration and supervision by state, 2026
- 50-state graphs
- Methodology
Get our latest data & analysis in your inbox!
- Table of Contents
- States of mass punishment
- Fifty years of probation and parole
- Probation and parole do not “work”
- Disproportionate impacts
- Recommendations
- Right-size probation caseloads
- Reduce parole conditions
- Appendix: Incarceration and supervision by state
- 50-state graphs
- Sources and methodology
The U.S. confines nearly 2 million people in prisons, jails, and detention centers, a fact which we only know by piecing together data from siloed parts of the criminal legal system. But even this number does not capture the full picture of the U.S. system of mass punishment, which ensnares 5.5 million people in incarceration, community supervision, and other forms of control and surveillance.
The vast majority of people in the mass punishment system are on community supervision, either under probation (3 million people) or parole (536,000 people).1 Yet these massive systems do not receive nearly as much attention as incarceration despite their broad harms and high failure rates. Policymakers and the public must understand how deeply linked these systems are to mass incarceration to ensure that these “alternatives” to incarceration aren’t simply expanding it. To that end, this report is designed to help state policymakers and the public assess the scale and scope of their entire correctional systems.
Here, we present updated data for all 50 states and D.C. on federal and state prisons, local jails, Indian country jails, probation, parole, youth confinement, and involuntary commitment.2 The data suggest that states’ reliance on different systems of punishment varies widely, with supervision often designed around failure and punishment rather than support and stability. We’ve made the updated statistics available in 100+ state-specific pie charts and a data appendix.
This edition of Punishment Beyond Prisons also features two new additions: First, we include an overview of confinement and supervision population trends over the past fifty years, to show how incarceration and its purported “alternatives” have grown in lockstep with one another. Second, we draw attention to an expansive form of correctional control that is typically overlooked in discussions of punishment and/or reform: public registries listing people with sex-related convictions. These perspectives take the concept of mass punishment even further across time and space, giving advocates a stronger foundation for their work.
It’s important to note that our analysis does not include confinement systems that cannot be broken down at the state level, such as immigration confinement or military prisons. We also discourage comparisons of data in this report to prior editions, because the data for different systems vary in their consistency from year to year.
States of mass punishment
By our calculations, 1 in 61 people nationwide are under some form of “correctional control.”3 To get a sense of how massive community supervision systems are, consider: If the number of people under probation and parole alone were its own state, it would be roughly the size of Connecticut, and more populous than 21 other states and D.C. And while the massive scale of probation is more than five times the size of the parole population, there are nearly as many people on parole as there are in the United States’ 3,000-plus local jails. At the state level, seeing rates of probation and parole together with incarceration and confinement rates provides a more accurate and complete picture of punishment. Notably, some states with comparatively low incarceration rates appear much more punitive when we add other types of correctional control to the picture:
Rates of mass punishment include state prisons, federal prisons, local jails, youth confinement, Indian country jails, involuntary commitment, parole, and probation. Rates are per 100,000 state (or D.C.) residents. Scroll right to see the relative mass punishment rates for each of these systems, and see the Data Appendix for the complete data set, including counts and rates for each system in every state and the U.S. overall.
Looking closely at state variations in the use of different forms of correctional control (in the second slide above) reveals that states punish many more of their residents through supervision than through incarceration.
For example, Rhode Island, a so-called ‘progressive’ state below the national average of 566 per 100,000 residents when it comes to incarceration, is among the most punitive in the nation when you look at its full system of correctional control.
See the data for another state:
This graphic shows how many people are in different kinds of correctional control in Rhode Island. (We also made a graphic showing just the number of people in Rhode Island in different types of correctional facilities and confinement that some readers may find useful.)
Among the 50 states and D.C., Rhode Island ranks 10th in mass punishment, subjecting 1,911 per 100,000 of its residents to confinement or community supervision.
Looking at mass punishment rates by state, we find that:
- In nearly every state, more than half of people under correctional control are on community supervision;4 in 20 states, more than two-thirds are on probation or parole rather than behind bars.
- Residents of Colorado are more than twice as likely to be within the mass punishment system (1,919 per 100,000 residents) as those in neighboring Utah (901 per 100,000).
- Georgia has long led the nation in its use of probation (a staggering rate of 3,272 per 100,000 residents), but also has a top-ten incarceration rate (848 people in prisons and jails per 100,000), making it the most punitive state by a long shot.
- Louisiana uses all forms of punishment at consistently high rates, with similarly high rates of prison and jail incarceration, parole, and probation.5
Widening the net of social control: Fifty years of probation and parole
Community supervision populations nationwide have generally been shrinking for the past 15 years, driven by changes in probation. But over a longer timeframe, the use of both parole and probation have actually ballooned to the point of being commonplace. In other words, the era of mass supervision began in parallel with the more well-known era of mass incarceration. Unsurprisingly, the same tough-on-crime policies that led to a drastic increase in imprisonment also brought millions of people into the “net” of “formal social control” — that of supervision, and particularly probation.
We compiled incarceration and supervision data from as early as the 1970s (to be published in a forthcoming publication) and found that these populations climbed precipitously throughout the 1980s and ’90s, with all populations finally peaking around 2007. At that point, they started inching downward until the COVID-19 pandemic accelerated this trend. Over the last ten years of data collection (2013 to 2023), the probation population has dropped by nearly 22%6 — the equivalent of over 880,000 people — with most of the drop happening in 2020 and 2021. This change, however, had more to do with pandemic-related changes to the criminal legal system’s capacity than any significant change in policy. After this shock to the system subsided, probation trends reversed course, growing again between 2021 and 2023. At most recent count, a few states have greater supervision levels now than ever before:
- Arkansas’s probation population is the largest it’s ever been, with over 50,000 people in 2023.
- Montana and Nevada’s probation populations peaked in 2022 (and dipped only marginally in 2023).
- Utah’s probation population reached a record high in 2023.
If these populations continue to climb, they will return to peak levels in a few years’ time, again reflecting an overuse of probation rather than any shift away from punitive measures. Meanwhile, jail, prison, and parole populations remain at elevated levels compared to decades prior, suggesting the era of mass punishment is far from over.
It’s important to note that changes in probation and parole populations over time are not inherently good or bad; they can result from a wide range of reforms throughout the criminal legal system. For example, growth — particularly in parole rates — may signify that states are moving people out of prisons. Declining supervision rates would be an intended outcome of policies lowering barriers to successful supervision completion, like reducing the number of conditions or the use of carceral sanctions for violations. In our two-part report on parole systems, we explain that parole hearings and releases have been slowing down in dozens of states, lowering parole rates as people remain locked up instead. Understanding changes in individual state policy or practice (which is beyond the scope of this report, though we provide a few impactful examples later on) will provide context to these trends.
Public registries: An overlooked system of control and punishment
Public registries and notification systems listing people convicted of certain sex-related offenses (as well as some “violent crime” registries) proliferated in the 1990s and early 2000s, and now include hundreds of thousands of people. There is scant evidence that these registries do anything to prevent harm, but a lot of evidence that the strict regulations imposed on people who must register lead to homelessness and joblessness, which in turn can drive violations of supervision and recidivism. Although registries are administrative in nature — that is, they are separate from any court-ordered sentence — they share the same logic as other forms of punishment.
Public registries and notification systems listing people convicted of certain sex-related offenses (as well as some “violent crime” registries) proliferated in the 1990s and early 2000s, and now include hundreds of thousands of people. There is scant evidence that these registries do anything to prevent harm, but a lot of evidence that the strict regulations imposed on people who must register lead to homelessness and joblessness, which in turn can drive violations of supervision and recidivism. Although registries are administrative in nature — that is, they are separate from any court-ordered sentence7 — they share the same logic as other forms of punishment.
People who must register have to comply with dozens of conditions or risk arrest and incarceration, making them just as much part of the mass punishment system as anyone else discussed in this report — even after they have completed their sentences. (However, because registry data are difficult to obtain and analyze for all 50 states, they are not a part of our core state-by-state dataset.)8 By our estimates, of the nearly 800,000 registered people nationwide,9 about 16% (just over 125,000) are simultaneously on community supervision.10 Excluding this “dual status” population, then, over 668,000 registrants who are not on any other form of community supervision must endure this separate form of administrative punishment.
The unique and extreme social exclusion people listed on registries experience — since these are typically public, searchable databases — actually makes our communities less safe: The scarlet letter of registration makes landlords, employers, and even nursing facilities wary of providing essential services, and takes a toll on registrants’ mental health. Among those required to register, some are children as young as eight, some have convictions that are not of a sexual nature at all, and many have statistically very low probabilities of committing future crimes. Moreover, while people on registries struggle to navigate life under registry restrictions, almost all sexual violence (over 90%) is perpetrated by people who are not registered, very often by people already known to the victims.
The conditions imposed on registrants are sometimes even stricter than community supervision conditions. For example, while people on probation and parole are often prohibited from “associating” with others who have felony convictions or meet other criteria, registered people may be prohibited from having contact with anyone under the age of 18 — even their own children. Furthermore, residency restrictions can banish people from entire neighborhoods, pushing registrants further away from resources and support, and often into homelessness.
Other conditions that are more onerous under registries than under probation or parole include:
| Supervision condition examples | Registry condition examples |
|---|---|
| Nightly curfew | No overnight guests |
| No “associating” with certain people or groups | No contact with anyone under the age of 18, including one’s own children; no dating anyone with children under 18 |
| No entering bars or restaurants serving alcohol | No entering or loitering near areas “frequented” by minors, like schools, parks, arcades, swimming pools, parades, carnivals, or “adults-only” establishments where sexually oriented material is available |
| Get permission to travel out of state | Get permission to travel out of state, and register in the other state, following residence restriction laws |
We hope that bringing in these statistics and comparisons helps advocates better understand and explain the heightened scrutiny that registered people in their states face — not because of their true risk to public safety, but due to draconian, counterproductive laws.
Probation and parole do not “work” — in fact, they often lead to incarceration
Time and again, we’ve explained why community supervision tends to be poorly designed as a way to balance accountability with connecting people to the services and treatment they need to achieve stability. There are two primary ways that parole and probation set people up to fail:
- Both probation and parole impose an excessive number of conditions on people that can be nearly impossible to comply with all the time;
- When someone on supervision is accused of violating a condition — particularly through behaviors that are not crimes — they are often incarcerated as a response, which only further destabilizes their lives.
And “failure” is indeed common: Only about 43% of people successfully completed a term of probation, and 56% completed parole, in 2023.
Conditions. People under supervision and their loved ones know all too well that the conditions imposed on them are numerous, make daily activities difficult, and are sometimes irrelevant. In a prior report where we analyzed the “standard conditions” of probation, we found that conditions often make reentry harder and conflict with one another. For example, typical conditions like curfews, travel restrictions, and prohibiting certain establishments can severely limit employment or educational opportunities — even though maintaining full-time employment or study is almost always mandatory. Supervision fees are also a poorly-conceived condition: Although people on probation are statistically some of the poorest, monthly fees that can be several hundred dollars force them to choose between necessities like rent, supporting dependents, and staying in compliance with their supervision.11
Living under the thumb of community supervision means constant scrutiny from parole and probation officers, from regular phone calls and check-ins to surprise visits at home or work. These interactions often lead to the detection of behaviors that are considered violations, like breaking a curfew. Instead of connecting their clients with resources to help to address underlying issues, supervision officers and other system actors often use their significant discretion to issue harsh and inequitable sanctions, including incarceration. And while the average probation term averages about two years,12 parole supervision can last far longer, if not indefinitely.13
Violations. Of the 1.9 million “exits” from supervision in 2023, over 125,000 people were put back behind bars for breaking a rule, not as a result of a new criminal sentence. In many cases, people are incarcerated on alleged violations, and must wait behind bars for the whole revocation process to play out. At a more granular level, some states’ incarceration data reveals a very alarming reliance on incarceration as a violation sanction. For example:
- Nearly 1 in 3 people in Idaho’s local jails were held on alleged parole violations in mid-2025.14
- In Utah, people with probation violations are being sentenced to jail at a higher rate than ever before, and the increase is not fully explained by a shift toward higher-risk individuals in the probation population.15
- According to annual reports from Hawaii’s parole board, over 99% of returns to prison for parole violations during fiscal year 2024 were for technical violations.16
- A multi-site study of probation jurisdictions found that more than 1 in 10 jail admissions involved a probation violation. Importantly, people detained for a probation violation had a length of stay that was longer than those without a violation in every jurisdiction. In some cases, the length of stay was nearly 50 to 60 times longer.17
To find those most impacted by supervision, look to society’s most marginalized groups
It’s important to understand how community supervision extracts the most wealth, stability, and autonomy from people on the lowest rungs of the socioeconomic ladder. For example, supervision disproportionately impacts Black Americans: over 30% of the community supervision population is Black, despite Black people making up only 14% of the overall population.18 Research shows that probation can drive rising jail populations through the racially-disparate imposition of probation conditions and rearrests of Black people on probation. People on probation and/or parole are also in demonstrably poorer health than the general public, reporting higher rates of mental illness, chronic conditions, and substance use disorders.
For women, compliance with the rigid rules and schedules of supervision is nearly impossible. Anything involving travel and program participation can be particularly difficult due to family caregiving obligations, such as finding or paying for child- or elder-care. And the vast majority of women on probation have limited access to abortion care due to their state’s laws and movement restrictions. Finally, unhoused or housing-insecure people may find themselves in violation of a standard condition requiring a residential address, but they are also likely to have trouble finding stable housing if association restrictions extend to people with whom they would consider living. (Unhoused people are also more likely to interact with police, and contact with law enforcement must typically be reported to supervision officers.)
Only sweeping change can make community supervision a true alternative to incarceration
Our analysis of state punishment systems, along with decades of research and advocacy, finds that probation and parole are double-edged swords: They could be important tools for moving and keeping people out of harmful correctional facilities, but they are currently designed to anticipate failure rather than success, resulting in a “revolving door” to prisons and jails at significant cost to families and taxpayers.
Probation: Reduce caseloads and prioritize treatment and services
It is time for every jurisdiction to reduce and rethink its use of probation. When probation is reserved for people who are at a high risk of committing a crime and who require the most support, the savings derived from a smaller probation population can allow for reinvestment in voluntary, community-based services that address social or health concerns (such as drug use or homelessness) that are too often criminalized.
In addition to substantially reducing the use of probation, jurisdictions must also dial back the excessive number of standard and “special” conditions imposed on people. (To find out the standard probation conditions in your jurisdiction, see this comprehensive table we compiled in 2024.) At the same time, probation terms must be capped, and violations must be dealt with fairly.
States have been chipping away at the harms and overuse of probation. For example:
- In 2025, Virginia passed bipartisan measures that make it easier to reduce probation terms by helping people secure housing, employment, healthcare, and other key components of stability, rather than imposing irrelevant and burdensome conditions on them.
- In 2023, Pennsylvania’s Act 44 became law, overhauling one of the most punitive probation systems in the country. The new law limits incarceration for non-criminal violations, uses an individualized approach to setting conditions, and allows for termination of supervision for those who no longer need it.19
- In 2022, a diverse coalition of Florida lawmakers transformed probation through Senate Bill 752, which created remote reporting options and implemented credit systems to reduce probation terms through education or employment.
Parole: Pare down conditions and expand opportunities for success in the community
Astute readers will point out that we recently put out a report, Parole in Perspective, which highlights the woeful decline in the use of parole in most states, and argues that more people should be released on parole. Indeed, the “front end” of parole should be a functional release mechanism for people who can demonstrate readiness to succeed in the community.20 States should focus on making the “back end” of parole functional as well — reducing the overall burden of conditions, capping the length of parole terms, and mitigating collateral consequences, which will stem the tide of supervision-related admissions to prisons.
For advocates who are not sure where to start, the Prison Policy Initiative and MacArthur Justice Center’s National Parole Transformation Project teamed up to publish a set of Parole Principles that offer policy guidance for any state, no matter the circumstances of their current parole system. Several dozen organizations have signed onto the Principles, which state that parole supervision should:
- Have few (if not zero) conditions, and allow for the removal or modification of conditions;
- Allow multiple options for checking in with parole officers to limit disruptions to work and other obligations, and minimize the chance of violating conditions; and
- Not be indefinite: The length of supervision should be proportional to the original sentence, but no longer than five years.
And those facing an alleged violation of parole should be guaranteed that:
- They will not be put behind bars for non-criminal conduct, particularly as they await a hearing; and
- If supervision is ultimately revoked, there must be a cap on the length of time to be served, and it should be proportionate to the seriousness of the violation.
Fortunately, states are realizing that parole should not feed into prison populations. For example, New Jersey’s Substance Abuse Recovery and Supervision Accountability Act, introduced in late 2025, would ensure that non-criminal violations of parole, such as a positive drug test, do not automatically lead to reincarceration. And New Mexico passed a bill in 2025 eliminating one of the most insidious conditions of parole: monthly fees, totaling up to $1,800 per year.
Conclusion
Wherever possible, jurisdictions should be reducing the use of incarceration and confinement, which tears apart communities and families. But this report shows where each state falls within the national landscape of mass punishment, including the overly used and poorly designed systems of probation and parole. Our state-specific breakdowns suggest where state advocates and policymakers might start when developing proposals for meaningful justice reform, especially when probation and parole feed people right back into incarceration.
Footnotes
It’s possible to be on both probation and parole at the same time. At yearend 2023, there were 11,790 people reported to be under both types of supervision; since they are counted in both populations separately, these figures do not add up exactly to the total community supervision population. ↩
Youth confinement and involuntary commitment are technically outside of the adult criminal legal system, but because they involve detention and control, we consider them part of the big picture of mass punishment. For some of these systems, data disaggregated by state are not publicly available; for others, the available data only cover some, not all, states (such as subcategories of the “involuntary commitment” population). Data for probation, parole, and state prisons are primarily from year-end 2023, while other data are from other years; see the methodology section for more on how we compiled state-level populations for each system. ↩
This figure is different than what you’ll find reported in the Bureau of Justice Statistics’ (BJS) report Correctional Populations in the United States for two reasons. First, BJS counts only counts people in adult correctional systems (prisons, jails, probation, and parole) and in the U.S. adult population. Second, we adjusted some states’ counts to correct for under- or over-reporting (see the methodology section for more information). In this report, as in our Mass Incarceration: The Whole Pie report, we take a more comprehensive view and include other types of confinement, like juvenile justice system “residential placement” facilities and prison-like “civil commitment” facilities. ↩
In Alaska and West Virginia, 42% and 43% of the mass punishment populations, respectively, are under community supervision. ↩
Louisiana is one state that holds a large number of people serving state prison sentences in local jails; other examples include Kentucky and Mississippi. ↩
This figure is different than what the Bureau of Justice Statistics has published in its report, Probation and Parole in the United States, 2023 due to data corrections and adjustments; see our methodology section for state-specific information. ↩
“Sex offense registration” laws are civil laws — not criminal sentencing laws — and apply only after someone has been convicted of a qualifying crime. However, failure to register or comply with registry-related restrictions is a criminal offense subject to criminal sentencing. ↩
We plan on publishing the state-specific data we collected and analyzed showing the number of people on sex offense registries, under community supervision, or both, in a forthcoming publication. ↩
This figure represents our estimate of people listed on sex offense registries in 42 states and D.C., based on public records requests and publicly available data. Where data were available on both registry populations and at least one form of community supervision, we found that roughly 125,600 people are on some form of state community supervision and listed on the state’s sex offense registry. In some states, more than half of people listed on the sex offense registry were also on community supervision. ↩
This exclusion is to avoid double-counting people in our full accounting of the scope of mass punishment. Because community supervision and “sex offense” registration both originate with criminal convictions, there is inevitable overlap between these two populations. For example, someone may be released from a state prison on parole, but they are also required to be listed on their state’s sex offender registry; each system is overseen by a different agency, so the individual will have to comply with two distinct sets of conditions. See the methodology section to learn how we adjusted each population so that each person is counted only once, even when they are under multiple forms of correctional control. ↩
Maintaining a residential address is a standard probation condition in most jurisdictions; supporting dependents is a standard probation condition in 30 of the 76 jurisdictions we analyzed. ↩
This nationwide average, last measured by Pew Charitable Trusts using data through 2018, obscures just how widely probation terms vary by jurisdiction. For example, an analysis of probation in Minnesota jurisdictions found that over a two-year period, probation in the Fourth Judicial District averaged 38.5 months (just over three years), while in the Seventh Judicial District it averaged 86.4 months — more than seven years. In D.C., probation terms can be set for up to five years. The Pew study found that about half of all states allow probation sentences of over 5 years (see Table 2 and Fig. 7). ↩
Generally, people released on parole can expect to be on supervision for a term equal to the remainder of their criminal sentence. But for people serving life with parole, the remainder of their sentence is indefinite unless a statute or policy allows termination at some point. Some states have made headlines for having notoriously difficult pathways to parole termination. ↩
According to the Idaho Department of Correction’s most recent (FY2025) population report, 395 of 1,232 people (32%) held in local jails on a single day were categorized as “Parole Violators,” defined as “usually a temporary status until a hearing can be conducted to determine if their parole will be reinstated or revoked.” ↩
The Utah Commission on Criminal and Juvenile Justice’s annual report on Justice Reinvestment Initiative shows that something called the “risk-adjusted population intensity” — a calculation of a population weighted to account for calculated risk levels — is also at a record-high level, but this only helps to explain “some of the increase” in jail sentences for probation violations. ↩
Of 189 parole revocations held between July 1, 2023 and June 30, 2024, 188 were for technical violations of parole, and only one was for a new felony conviction. In FY2023 and FY2022, 100% of revocations of parole were for a technical violation, not a new felony conviction. ↩
Length of stay data were available for eight out of the nine jurisdictions in this study. ↩
In 2024, an estimated 14% of adults in the U.S. identified as Black or African American alone, according to the U.S. Census Bureau’s American Community Survey 2024 1-Year Estimates, Table S2603. ↩
The ACLU of Pennsylvania opposed the bill that became Act 44, arguing that reforms to the termination process would not be enough to counteract overall probation terms, which could still be lengthy. ↩
The two-part report focuses on 35 states with broad-based discretionary parole. In it, we explain how the “front end” of parole — where parole boards meet to consider releasing certain eligible people from prisons — is dysfunctional and overly focused on static factors that lead to denying release. Ultimately, our analysis finds that parole board members should be more diverse, and parole boards should hold more hearings and increase grant rates. ↩
According to resources from National Association of Counties and Human Rights Watch, misdemeanor probation involves private probation agencies in Arkansas, Florida, Georgia, Kentucky, Missouri, Montana, North Dakota, Tennessee, Utah, and Wyoming. There is also evidence of private probation in Colorado, Oklahoma, Alabama, and Mississippi. ↩
About the author
Leah Wang is a Senior Research Analyst at the Prison Policy Initiative and is the author of multiple reports including Parole in Perspective: A deep dive into discretionary parole systems, Punishment Beyond Prisons 2023: Incarceration and supervision by state and Chronic Punishment: The unmet health needs of people in state prisons. She has also authored numerous briefings on topics including the imposition of disciplinary fines in prisons and the negative impacts of “association” restrictions for people on probation and parole. She also oversees our work on the intersections of incarceration and environmental justice.
About the Prison Policy Initiative
The non-profit, non-partisan Prison Policy Initiative was founded in 2001 to expose the broader harms of mass criminalization and spark advocacy campaigns to create a more just society. Through big-picture reports like Mass Incarceration: The Whole Pie, as well as in-depth reports on issues such as probation and parole, the organization helps the public more fully engage in criminal justice reform. The organization also launched, and continues to lead, the national fight to keep the prison system from exerting undue influence on the political process (a.k.a. prison gerrymandering).
Acknowledgments
All Prison Policy Initiative reports are collaborative endeavors, and this report is no different. The author would like to thank Jacob Kang-Brown and Emily Widra for research assistance, as well as authors and contributors to past editions of this report. We credit sociology Professor Michelle Phelps for developing the concept of “mass probation,” which informs much of our thinking about how these systems interrelate. Finally, we would like to thank our donors and funders who make this work possible.
Get our latest data & analysis in your inbox!
Additional graphs
The graphs made for this briefing are included in our profiles for each state:
and are available individually from this list:
Methodology and about the data
For all the data we use in this report and the 100+ pie charts, see our data appendix and read below for more information about how this data was collected and prepared.
- State prisons: Bureau of Justice Statistics’ (BJS) Prisoners in 2023 — Statistical Tables, Table 2 (jurisdictional population, by state). We used counts as published, and calculated rates by using U.S. Census Bureau population data instead of using BJS’ published rates, which are based on the sentenced population only.
- Federal prisons: The number of people in Bureau of Prisons (BOP) facilities comes from the BOP website on January 22, 2026. From this population, we removed the people under US Marshal Service (USMS) jurisdiction held in BOP facilities, to avoid double counting. The number of people under USMS jurisdiction (in FY 2025) comes from their 2026 “Facts and Figures” report. We estimated state-level counts by using percentages for each state provided in a 2025 FOIA request regarding state of origin, multiplied by the combined BOP and USMS populations. We calculated rates using U.S. Census Bureau population data.
- Local jails: To estimate state-level counts of people in local jails, we started with figures from Vera Institute of Justice’s publication People in Jail and Prison in 2024. We then used the total number of people held in jails for other authorities from Bureau of Justice Statistics’ Jail Inmates in 2023 — Statistical Tables, Table 12, and produced state-level estimates by using the same proportions as displayed in the last available dataset of this type, Census of Jails 2005-2019, Table 12. We subtracted the estimated number of people held for other authorities by state from the Vera data. We calculated rates using U.S. Census Bureau population data.
- Indian country jails: Bureau of Justice Statistics (BJS)’ Jails in Indian Country, 2024; state-level counts of jails in Indian Country were available through a restricted-used data access agreement with the Inter-university Consortium for Political and Social Research, which hosts the data behind the BJS publication Jails in Indian Country. We calculated rates using U.S. Census Bureau population data.
- Youth confinement: State-level counts are published in the Easy Access to the Census of Juveniles in Residential Placement (EZACJRP), 2023. Includes all types of youth confinement (committed, detained, diversion, and other/unknown) for youth up to age 20. State counts may not sum to total for two reasons: To preserve the privacy of the juvenile residents, state level cell counts have been rounded to the nearest multiple of three. Additionally, state of offense was not reported for 1,380 youth, so we could not allocate those youth to specific states. We calculated rates using U.S. Census Bureau population data.
- Involuntary commitment: This category includes two populations: people in state psychiatric hospitals on a “forensic” status and people civilly committed due to conviction for a sex offense.
- Forensic psychiatric patients: People involuntarily committed by the court to state psychiatric hospitals for evaluation and after being found Not Guilty by Reason of Insanity. We used Treatment Advocacy Center’s report Prevention Over Punishment: Finding the Right Balance of Civil and Forensic State Psychiatric Hospital Beds, Appendix C, Tables 5 and 6.
- Civil commitment: People convicted of sexual offenses who are detained or committed in “civil commitment centers” (typically prison-like facilities) under civil commitment laws after their sentences are complete. Counts for New Hampshire, North Dakota, Pennsylvania, and South Carolina are from the 2024 Sex Offender Civil Commitment Programs Network (SOCCPN) annual survey, shared by former SOCCPN President Shan Jumper; the count for Wisconsin is from the 2023 SOCCPN survey, and the count for Nebraska is from the 2018 survey. The other 14 states provided civil detention and commitment data in response to public records requests: ten states — California, Florida, Illinois, Iowa, Kansas, Missouri, New Jersey, Texas, Virginia, and Washington — provided counts as of February or March 2026, two states (Arizona and Minnesota) provided counts as of 2025, and two states (Massachusetts and New York) provided counts as of 2024. The civil commitment count for the federal Bureau of Prisons (BOP) comes from a 2020 Williams Institute report because it is more recent than the SOCCPN survey data from 2019. Note: In Massachusetts, civil commitment via Section 35 refers to people committed for an alcohol or substance use disorder; those held under Section 35 are not included in this count.
- Parole: We started with counts from Bureau of Justice Statistics’ Probation and Parole in the United States, 2023, Appendix Table 11. To address double counting, we added up all people who were on parole and also in jail, state prison, or federal prison, and subtracted them from the total population. Then, we took the ratio of this smaller population to the total parole population, which was 0.953, and applied it to each state’s reported parole population, in order to estimate state-level probation populations while not counting anyone twice. These counts exclude federal supervised release. We calculated rates using US Census Bureau population data.
- Probation: We started with counts from Bureau of Justice Statistics’ Probation and Parole in the United States, 2023, Appendix Table 6. To address double counting, we added up all people who were on probation and also in jail, state prison, or federal prison, or parole, and subtracted them from the total population. Then, we took the ratio of this smaller population to the total probation population, which was 0.976, and applied it to each state’s reported probation population, in order to estimate state-level probation populations while not counting anyone twice. We’ve chosen to include people on both probation and parole in the parole population. These counts exclude federal supervised release.
- Exceptions to probation and parole figures: Because probation and parole are administered by several (or sometimes hundreds) of local agencies within a state, the data published by the Bureau of Justice Statistics (BJS) are sometimes rougher estimates based on partial reporting and imputation as opposed to a comprehensive, standardized count.
By looking at notable but unexplained year-to-year changes from 2022 to 2023 in the BJS data, we identified several states where we suspected that probation or parole populations were imprecise due to errors in reporting and/or imputation. Staff at BJS also told us that they do not include data on privately-operated probation agencies unless a state agency is already collecting and reporting it, which alerted us to data gaps in states where private agencies commonly supervise probation.21 Finally, we learned that some states (e.g., Virginia) have a mix of both state-operated (handling felony cases) and locally-operated probation agencies (handling misdemeanor cases), but the local probation data aren’t always included in BJS’ data.
The table below describes sources and methods to find more complete 2023 (or the closest available date) probation or parole data, replacing the BJS estimates for these states. In a forthcoming publication about historical supervision populations, we will present other prior years of data we suspect are inaccurate and offer new estimates using the below sources and methods. (Note that the over-time charts in this report reflect all states and years of adjusted data, while the table below only reflect states for which we adjusted 2023 data.)
State Population type Sources of more accurate state-level estimates for 2023 Delaware Probation Count from Delaware Open Data, using the number of people by type of institution and including people with community supervision statuses ranging from probation to home confinement, restitution, and work release. We use this source because the BJS reports show an unexplained 85% increase in the probation population from 2022 to 2023 that is not reflected in the Delaware Open Data on community supervision. Illinois Probation Count from 2023 Annual Report of the Illinois Courts statistical summary, which reports active adult caseloads for felony, misdemeanor, DUI, traffic, and administrative probation. Maryland Probation The Maryland probation number is calculated using data provided by the Maryland Department of Public Safety & Correctional Services in response to a public information request. We applied the percentage of overall cases supervised that were probation to the total number of supervision cases at the end of fiscal years 2023 and 2024, and took the average of these. To this figure, we added the average of the Drunk Driver Monitoring Program’s caseload counts for end of fiscal years 2023 and 2024. Massachusetts Probation Count from the Massachusetts Probation Services Research Department’s Caseload Review dashboard (excluding juvenile cases). Minnesota Probation Count from Department of Corrections’ 2024 Probation Survey (excluding juvenile cases). Oklahoma Probation Count combines the BJS data on state felony probation and some misdemeanor probation agencies with our estimate of District Attorney probation based on reports by the Oklahoma District Attorneys Council provided to the Oklahoma legislature on DA Supervision Fee revenue. (Fees are $40 per month, and the most recent available data indicated $7,980,989 in revenue, suggesting 16,627 people paying in an average month.) Oregon Parole Count from Oregon Department of Corrections’ Community Profile Report for October 1, 2024 (using the “postprison” population). Pennsylvania Probation Estimate from data published by two state agencies: The Pennsylvania Commission on Crime and Delinquency’s annual report on Pennsylvania County Adult Probation and Parole (which includes 117,000 people on probation, probation with restrictive conditions, accelerated rehabilitative disposition, pretrial bail supervision, as well as roughly 12,000 absconders from probation), and the Department of Correction’s Monthly Parole Statistics report (which includes about 5,000 people on special county probation). Tennessee Probation State probation data for felony charges is from the Tennessee Department of Correction’s Felon Population Update for Probation and Community Correction. Misdemeanor probation is run by local governments, who can run their own probation agency or contract with a private firm; BJS started collecting publicly-run local probation data, but they do not collect private probation data. Leveraging sample changes in BJS data collection from 2022 to 2023 (when they began collecting data from local public agencies), we were able to estimate local public probation in the state for yearend 2023. To estimate local private probation, we used a 2024 state audit of Private Probation Services, utilizing data about fees paid to the Tennessee Department of Commerce and Insurance’s Private Probation Services Council. Virginia Probation Estimate accounts for both state and local supervision. BJS began collecting local probation data in recent years, but we knew it was only partial, so we attempted to collect more. State probation counts are from Department of Corrections (DOC), and local probation counts were obtained via FOIA from the Department of Criminal Justice Services. - U.S. population (to calculate rates): U.S. Census Bureau’s American Community Survey 2023 1-Year Estimates, Table DP05, ACS Demographic and Housing Estimates. We elected to use 2023 data, even though later data are available, because the majority of correctional control populations analyzed in this report — including probation, the largest of those populations — are from 2023.
Recommended Reading:
50 state incarceration profiles
All of our recent reports about prison/jail growth, racial disparities, and more, re-organized by state.
The rapid & unregulated growth of e-messaging in prisons
It can strengthen connections, but instead has been used to sap money from incarcerated people.
Incarceration and supervision by states
Probation and parole are the leading forms of mass punishment in America.
Prison Gerrymandering Project
We are leading the movement to protect our democracy from the Census Bureau's prison miscount.