U.S. flag

An official website of the United States government, Department of Justice.

Reentry Research at NIJ: Providing Robust Evidence for High-Stakes Decision-Making

NIJ is committed to promoting rigorous research on how best to successfully integrate individuals returning from jail or prison.
Download and print
National Institute of Justice Journal
Date Published
April 11, 2022

Reentry matters. Millions of American adults are incarcerated in local jails and state and federal prisons. Another several million are under criminal justice supervision in the community. As individuals serve their sentences and are released from custody, one thing is certain: The majority of them — approximately 95% — will return to their communities, families, and friends.[1] As they leave custody and become our neighbors, it is important that we invest in these individuals and help them succeed and contribute positively to their families and their communities. Many will leave the institutional setting with the skills necessary to become contributing members of our local neighborhoods. But, unfortunately, many will not. As an example, many individuals will return to custody. The Bureau of Justice Statistics found that 44% of individuals who left state prison were arrested [2] at least once in their first year after release.[3] Within nine years of release, 5 of 6 of those previously in state prison had been rearrested. The number of people who will reengage with the criminal justice system highlights the critical importance of reentry.

Why does reentry matter? Reentry is a critical transition for individuals returning to their communities, whether they have been away for decades or a matter of days. The difficulty, however, is that individuality can make this transition more complicated. People reentering have unique needs, and often these needs — for example, ongoing issues like mental and behavioral health — have not been addressed before release or during the reentry process. The reentry process and how long it lasts can vary from person to person.

Further complicating the matter is, at times, the inconsistent evidence for what helps individuals successfully reenter the community. The “what works in reentry” literature has consistently found that housing, employment, family unification, mental and physical health treatment, and meeting other critical needs are vital to post-release success. [4] If people reentering are able to find housing and employment and address other critical needs, they are more likely to be successful in the community and not return to custody. But what works for whom and when?

Over the past several decades, the National Institute of Justice (NIJ) has been a leader in the study of reentry. Former NIJ Director Jeremy Travis helped popularize the term “reentry” in the late 1990s when he said that interest and progress in understanding “‘prisoner reentry’ has been nothing short of remarkable.” [5] Since then, reentry has remained a priority for NIJ. In the 21st century, several federal reentry initiatives have invigorated the attention paid to the needs of persons who have been convicted of crimes as they return to society. Policymakers, practitioners, and community and criminal justice stakeholders, as well as others, have learned — and continue to learn — what works and what matters in the reentry process.

This article provides an overview of reentry, focusing on NIJ’s contributions to the field and identifying gaps in our collective empirical knowledge. It starts by discussing what is known about common barriers to successful reentry and describing federal efforts to help state and local agencies address those barriers. It then highlights NIJ’s efforts to advance sophisticated risk assessment algorithms and introduces NIJ’s evaluations of graduated sanctioning programs. This article concludes with a discussion of NIJ’s ongoing research evaluating local reentry programming.

Barriers to Reentry

Research has identified common barriers to successful reentry, including but not limited to the difficulty of obtaining gainful employment, stable housing, and education and strengthening prosocial support networks.[6] These factors represent barriers because they can inhibit treatment and the ability to overcome criminogenic needs (the triggers or situational factors that may lead someone at risk to commit a new crime).[7]

The federal government has initiated comprehensive programs to assist jurisdictions in addressing the needs of individuals returning from incarceration. The Serious and Violent Offender Reentry Initiative (SVORI) in 2003 and the Second Chance Act (SCA) in 2008 provided support to state, local, and tribal governments to address common reentry barriers. These large federal initiatives were complemented by smaller programs facilitated by individual agencies that helped address specific reentry issues.

These initiatives offered opportunities for program evaluations to generate research on how best to address criminogenic risk and needs. Findings from these evaluations also provide insight into desistance from crime and how to help individuals overcome substance use disorders, gain prosocial skills, and alter their attitudes toward crime and violence.

For example, evaluations of SVORI highlighted the different reentry risks and needs for males and females. NIJ-supported researchers found that men tend to benefit from programs that promote attitudinal change, such as treatments that address criminal attitudes and anger management and promote healthy personal relationships. Women, on the other hand, tend to respond better to practical skills training, such as having a reentry case manager and going through life skills training. The evaluations found that SVORI programs helped reduce rearrests and lengthened the time to arrest; however, SVORI participation did not reduce reincarceration. SVORI programming did appear to be more beneficial to participants as time went on. This may point to the possibility that SVORI participation offered some keys to long-term success in overcoming the difficult transition from prison to the community , which is usually seen early on in missed community supervision appointments and positive drug tests.[8]

Evaluations of the SCA found that, like SVORI, it did not reduce reincarceration. The SCA did, however, successfully provide participants with greater access to programs. It increased job placement and employment outcomes, but this did not seem to translate into recidivism reduction gains.[9] These findings suggest that the process of reentry is whether complex, and that helping individuals overcome reentry barriers may not on its own be enough to lower recidivism.[10]

No Reentry Panacea

Evaluating the success of national initiatives like SVORI and the SCA is difficult. These programs were able to provide individuals with services and educational, housing, and employment opportunities that may have been previously unavailable — yet their impact on recidivism, when taken as a whole, was limited. This could be due, in part, to the myriad and complex factors that can trigger incarceration, through the revocation of parole because of a violation of the conditions of a community sentence or because of the commission of a new criminal offense. An individual’s needs may interact with the systemic disadvantages of the neighborhood and community to which they are returning. Understanding how a need will be met by a particular program or service is not straightforward, nor is knowing how the individual will respond to a program or even how open to change they may be.[11]

Reentry programs alone may not be adequate. In other words, a program needs to be delivered at the right time on the individual’s trajectory of change to promote successful reentry.[12] For example, individuals may want to change, but they might have an underlying substance use disorder and inadequate skills to thrive on their own. A job services program alone would not address a substance use disorder issue; matching appropriate services with needs at the right time would likely be more effective. The likely inadequacy of addressing just one criminogenic need may speak to the popularity of case management as a reentry tool to help ensure that individuals receive tailored services to meet all of their unique needs. In fact, case management was the most common reentry service provided under the SCA.[13] However, this does not mean that case management is widely successful, as the evaluation of SCA programming showed that it largely did not produce an effect. [14] Although having a case manager seemed to increase the time to arrest for females, it seemed to shorten the time to reincarceration for males. The research team surmised that this may have been due to increased monitoring of those working with a case manager. [15]

There are a number of programs and services available to meet the needs of an individual returning from incarceration. A case manager can tailor these services to meet specific immediate needs, though this may not prevent recidivism unless the individual is ready to change and can persevere in spite of the significant barriers that often stand in the way of successful reentry.

Assessing Risk

Services should align with a person’s criminogenic needs. In fact, a mismatch in services may inadvertently contribute to recidivism.[16] Smaller caseloads give community corrections officers a greater ability to accurately assess clients’ needs and direct them to more beneficial treatment programs based on those assessments.[17] Unfortunately, even under the best conditions, a community corrections officer’s caseload is about 50 people on probation or parole, which makes it difficult to understand each person’s triggers and barriers to success.[18] Before individuals are placed on supervision, a number of decisions must be made about their potential risk of recidivating. In the past, practitioners relied on tacit knowledge to make professional judgments about the recidivism risk of an individual leaving prison and under community supervision. Today, there are more than 400 risk assessments in use across the criminal justice system.[19] These risk assessment tools can make actuarial predictions based on risk factors that are both static (factors not subject to change, such as age and criminal history) and dynamic (factors that are amenable to potential intervention, such as negative peer associations and substance use), while still incorporating the practitioner’s professional assessment.[20]

Even with these advances, there are some critical limitations in how risk assessment tools can support reentry programming. Although more accurate than professional judgment alone, risk assessment tools may exacerbate bias and perpetuate racial and ethnic disparities. In other words, a risk assessment tool may create a negative feedback loop based on the tool’s predictive indicators. In this case, individuals from marginalized communities, including communities of color, who are more likely to have early involvement with the criminal justice system will be indicated as posing a greater risk and thus receive more intensive supervision, which may lead to increased recidivism. As stated above, matching supervision style and programming with a person’s criminogenic risk and need is critical to promoting positive reentry and reducing recidivism. In addition, using risk assessment tools that are accurate across race and gender is critical. Currently, common risk assessment tools predict risk most accurately for white men. [21] This may inhibit correctional agencies’ ability to effectively meet the needs of all individuals under supervision.

Another limitation of risk assessment tools is their passivity. Risk assessment tools make important decisions based on probabilistic averages. On their own, the tools cannot replace a meaningful relationship between an individual and a community corrections officer. These tools can help triage community corrections’ scarce resources and attempt to direct services and supervision to those who need it most, but they cannot by themselves effect change. (See the sidebar “Developing New Technology to Promote Reentry Outcomes” for more information on NIJ’s efforts to advance risk assessments.)

Role of Sanctioning

The goal of community corrections is to ensure that individuals under community supervision comply with their conditions of supervision, do not commit new crimes, and are provided services that address their criminogenic needs to improve long-term outcomes. It is expected that missteps and minor violations will occur — reentry can be complex and nonlinear. The difficulty lies in deciding whether an individual may still be successful on a community sentence after a violation or whether the community sentence should be revoked. Many violations, by themselves, may not warrant the prison term that often results from revocation of a community sentence. If individuals accumulate multiple violations, however, officers may need to decide when it is time to call for revocation.

Hawaii created a probation model — Hawaii Opportunity Probation with Enforcement (HOPE) — in which every violation resulted in a short but swift sanction that did not remove the individual from probation. The theory behind HOPE was that the sanction would be severe enough to help modify behavior and introduce accountability, but it would not be as severe as revoking the probation sentence. Individuals would experience the sanction — typically a few days in jail — shortly after the violation was detected, thus reinforcing the link between the sanction and the behavior that led to the violation. This, in turn, would help promote behavioral change among those on probation and parole.[22]

Although initial findings suggested that the HOPE probation model was effective,[23] an NIJ multisite demonstration on the U.S. mainland was not able to replicate the initial successes. HOPE participants were found to have fewer arrests than those on regular probation in two of the four jurisdictions, but overall there was no difference in time-to-arrest or probation revocations.[24] Difficulties in organizational culture and communication among implementing partners, along with the model’s statutory framework, made it challenging to implement HOPE consistently across the four sites, particularly in terms of achieving a uniform definition of probation failure and a return to prison.[25] The HOPE experience illustrates the challenges of adopting a community corrections or reentry program from another jurisdiction, even when the program has some initial indication of success.

Implementing an Effective Reentry Program

Effecting change is hard, especially for those involved with the criminal justice system. The reentry field has benefited from numerous program evaluations that show which interventions are promising and which may actually make matters worse.[26]

Deciding whether to implement a particular reentry program or use a suite of best practices is a difficult choice for criminal justice agencies. First, the agency must understand the conditions that led to the success of any program that it seeks to adopt. Not only must the agency decide if it can replicate the core components of the reentry program, it must also determine if the reentry program is well aligned with its environment or target population. To further complicate matters, the role of the environment and other contextual factors that contributed to the successful implementation may not be fully understood.

Should an agency decide to adopt a reentry program, staff buy-in is critical. An NIJ-funded project examining the implementation of an evidence-based program found that it was easier to change the language community corrections officers used to describe what they were doing than to actually modify what they were doing.[27] Research shows that building staff buy-in for a new approach, although difficult, is often much easier than ensuring that staff have adopted the necessary changes to how they do their jobs.[28] Not only must staff understand what to do, they also may have to approach their jobs in an entirely different way in order to implement the program with fidelity. This is a heavy lift for any community corrections agency to undertake.

The Need for Randomized Controlled Trials

The stakes are high in reentry decision-making. It is critical for jurisdictions to implement programs that work, assess risk, and assign appropriate programming to the right individual. Individuals face numerous hurdles upon release from prison or jail. Ineffective programs are not only a poor allocation of time and resources, they may also contribute to recidivism.[29] Sound evidence is needed about what programs are effective, for whom, and under what conditions. Several years ago, NIJ dedicated a research portfolio specifically to reentry programming that relies on evaluations using randomized controlled trials (RCTs). When implemented with rigor, RCTs offer the highest degree of confidence in study outcomes. Because these studies randomly assign participants into treatment groups (those who receive the specific program under study) and control groups (those who do not receive the program being evaluated), the research team is able to isolate the effect of the reentry program (or supervision strategy) on the participants in the treatment group. All other factors that likely influence an individual’s success in reentry are cancelled out because all participants have an equal likelihood of being assigned to the treatment group or control group.[30]

In lieu of randomization, researchers often employ quasi-experimental designs, which use advanced analytical tools to mimic the treatment and control logic of an RCT by trying to compare program participants with similar individuals who did not receive the program. In some instances, program implementation can be so broad that there is no identifiable control group. Here, the research team may compare recidivism rates before and after program implementation to try to assess the program’s overall impact.

Unfortunately, recent NIJ-funded research has shown that quasi-experimental designs in criminal justice research tend to overestimate the treatment effect.[31] In other words, quasi-experimental evaluations may show that a program has a significant impact on reentry outcomes when, in fact, the impact may be minimal. By the same token, this type of evaluation may show that a program has a moderate or minimal impact when it has no effect at all.

There are times when RCTs are not feasible. In these cases, many program evaluations are based on robust quasi-experimental designs. However, because the barriers to successful reentry are so great and the resources of community corrections agencies are so scarce, NIJ has made a commitment to support reentry evaluations that use RCT designs to provide practitioners and the individuals who rely on them with the soundest possible evidence to guide reentry programming.

In 2018, NIJ issued its first solicitation that prioritized RCT evaluations of promising reentry initiatives. Corrections researchers and practitioners responded overwhelmingly to the call, recognizing the critical need for sound research in this area. In the first year, NIJ funded five projects totaling more than $5 million. Given the success of the first year, NIJ quickly institutionalized this solicitation as a standing program. After three years of funding, the program has now grown to 13 ongoing research projects totaling approximately $17 million in research funding. To put this in context, over the same three-year period, NIJ funded 25 corrections-related research projects from four different portfolios (including reentry), representing an investment of approximately $27.7 million. Of this, the reentry portfolio represents more than half of the projects and 60% of the research investment.

Themes of Ongoing Reentry Research

NIJ’s 13 ongoing reentry research projects can be grouped together under a few general themes. These projects were submitted independently of one another, and NIJ purposely offered significant flexibility to the field to guide the selection of topics. Thus, the themes of these projects can be considered a snapshot of the areas of critical research need in reentry today. The general themes include evaluations of:

  • Young adult reentry programs.
  • Treatment for individuals on parole with past traumatic brain injuries.
  • Risk-need-responsivity strategies.
  • Emerging technology.
  • Innovative treatment modalities.

Young Adult Reentry Programs

Three ongoing evaluations are examining interventions for moderate- to high-risk young adults.[32] One program seeks to address past trauma and its likely consequences, particularly impulsivity and aggression, among individuals released from prison. The remaining two programs focus on interventions for young adults in jail, mainly cognitive behavioral therapy with case management. One program also includes a subset of participants who will live together in a dedicated housing pod. With these projects, NIJ hopes to understand how treatment and dedicated housing interact to impact reentry outcomes.

Traumatic Brain Injury

The study of traumatic brain injury is an emerging area for NIJ. There is an increasing awareness that many individuals who are incarcerated suffer from past traumatic brain injuries, some unknowingly.[33] These injuries often make it more difficult to perform functions necessary for successful reentry, such as securing employment or building technological literacy. NIJ has two ongoing projects that focus on screening individuals for traumatic brain injury and providing tailored service— both case management and therapy — to meet their needs.[34]

Risk-Need-Responsivity Strategies

NIJ is supporting two projects that evaluate risk-need-responsivity strategies.[35] Both programs under evaluation seek to improve outcomes on the major post-release reentry barriers mentioned above. The first evaluation is taking place in a prison setting and examining a stand-alone 16-week curriculum administered to medium- and high-risk individuals before release. The second evaluation includes multiple treatment methods of a similar program given to individuals both pre- and post-release from jail, pre-release only, and post-release only, as well as no programming. Taken together, findings from these evaluations will provide critical evidence on the impact of continuity of programming pre- and post-release for individuals returning to the community.

Emerging Technology

NIJ is also supporting two projects that examine the use of technology to enhance reentry outcomes.[36] One is a virtual reality job interview program that seeks to enhance individuals’ post-prison employment chances. The second will evaluate the impact of a web-based reentry planning and management tool. Medium- to high-risk individuals will receive a yearlong subscription to the web-based tool prior to their release from prison; their subscription will continue through early reentry. The study will assess how the web-based interactive tool can enhance standard reentry programming. At the conclusion of these projects, NIJ hopes to provide the field with evidence on the possibilities and limitations of advanced technological tools to aid conventional reentry programming. (See the sidebar “Developing New Technology To Promote Reentry Outcomes.”)

Innovative Treatment Modalities

The remaining projects cover a variety of reentry topics. [37] The evaluations include studying how expungement affects employment outcomes, how restorative justice can enhance therapeutic community treatment for individuals with substance use disorders, and how parole officer home visits can be used as a rehabilitative tool and not just a surveillance strategy. These evaluations cut across the myriad functions community corrections agencies and their partners provide for individuals reentering their communities. These projects — coupled with those discussed above and corrections-related projects from other NIJ research portfolios — will provide robust evidence on how best to promote successful outcomes among individuals returning from jail or prison.


In this article, we discuss NIJ’s investment in the field of reentry writ large, from evaluating federal initiatives, advancing risk assessments, and examining innovative programming such as HOPE, to evaluating local reentry work. All of this work is aimed at helping individuals succeed during the reentry process. Reentry matters. How the criminal justice system addresses the risk and needs of individuals returning to our communities matters. We know there is a critical need to accurately align the community supervision strategy and reentry programming with the individual’s criminogenic needs. We also know it is difficult to understand the factors that contribute to positive behavioral change on an individual basis, along with which carrots and sticks (including sanctioning) are available to correctional agencies. We know that assessing risk has the potential to exacerbate criminal justice biases and racial and ethnic disparities and thus potentially inhibit successful reentry. And importantly, we know it is difficult to find a program that works and to know how to implement it successfully.

Supporting the field of criminal justice by promoting rigorous research is critical to understanding what works for whom and under what circumstances. For more than three decades, NIJ has invested in reentry research in order to help the field answer these important questions — and it will continue to do so with a renewed focus on rigorous research designs. Understanding and building evidence about what works and what does not work in reentry programming and community supervision will enable the criminal justice system and its stakeholders to help individuals succeed when they return to the community.

About This Article

This article was published as part of NIJ Journal issue number 284.

While NIJ continues to pursue its research agenda of rigorously evaluating innovative reentry programming, the Institute is also actively engaged in technology development to improve reentry outcomes. NIJ facilitated the creation and application of enhanced artificial intelligence (AI) and machine learning to refine our understanding of the risk of recidivism and to explore the possibilities of timely intervention with individuals who are most at risk for recidivating. NIJ’s efforts began in 2019 with a solicitation seeking AI solutions to assist community corrections and continued in 2021 with a Challenge to develop enhanced risk assessment models. Both are discussed below.

Artificial Intelligence Technology Development

NIJ’s 2019 solicitation Artificial Intelligence Research and Development To Support Community Supervision resulted in two funded research projects aimed at developing AI tools to improve reentry outcomes for individuals on community supervision. These projects mark an important step in fielding this technology to help community corrections officers positively intervene with individuals in their caseloads.

The first award, made to Purdue University, will facilitate the development of a new AI-based support and monitoring system (AI-SMS). The AI-SMS will consist of a deployed smartphone application distributed to and worn by a sample of individuals under parole supervision in Tippecanoe County, Indiana. This application is intended to allow community supervision officers to interact with individuals under parole supervision, augment their program, and send alerts to officers about individuals who are at risk of recidivating.

In the second project, RTI International — in partnership with Applied Research Services and the Georgia Department of Community Supervision — will implement the Integrated Dynamic Risk Assessment for Community Supervision, which will enhance risk assessments with dynamic, real-time risk factors and guide community supervision officers’ interventions. For more information on these projects, see https://nij.ojp.gov/topics/articles/tapping- artificial-intelligence .

Enhanced Machine Learning Predictive Tools

In fiscal year 2021, NIJ released the NIJ Recidivism Forecasting Challenge, which sought to promote the development of enhanced and accurate risk assessments in community corrections while, at the same time, mitigating potential racial bias and considering unique gender differences. Challenge participants used a de-identified dataset from the Georgia Department of Community Supervision to predict recidivism among individuals on parole over a 12-, 24-, and 36-month period. Participants were also encouraged to use other datasets or information deemed useful to predict recidivism.

The Challenge received more than 150 submissions forecasting recidivism for each year over a three-year period. In total, 28 teams won prizes. The winning teams submitted the most accurate models predicting male and female recidivism separately in each of the three years. Awards were also given to participants who created models that were equally accurate for both Black and white individuals under parole supervision. Challenge winners submitted papers to NIJ discussing their models and what information they deemed most useful or predictive of recidivism. Challenge winners also participated in a symposium on how to advance risk assessments so that community corrections can have a greater understanding of risk across race and gender. Learn more about the results of the Challenge from:

For more information on the Challenge, go to https://nij.ojp.gov/funding/recidivism-forecasting-challenge .

For More Information

Learn more about NIJ’s reentry research at https://nij.ojp.gov/topics/corrections/reentry .

Date Published: April 11, 2022