Reentry Research 2019–2025: A Literature Review

📥 Download the full literature review (PDF, 38 sources)

Why this page exists

Much of what gets quoted about reentry is out of date. The foundational figures most cited in policy conversations come from cohorts released a decade or more ago — pre-fentanyl, pre-§1115 Medicaid reentry waivers, pre-Ban-the-Box revisionism, pre-FCC phone-rate cap. The 2022 National Academies report The Limits of Recidivism said the field needs better measures and a refreshed evidence base. This page is a working version of that refreshed base.

It is also the canonical research home for Next Move, a decision-training platform for high-stakes life transitions. Every research claim in our other posts links here. If you want our raw position on what the evidence shows, this is the page.

How this is organized

Four domains, with a cross-cutting preface on measurement:

1. The recidivism measurement problem
2. Housing & employment
3. Mental health & substance use
4. Family & social ties
5. Systems & policy barriers
6. Cross-cutting synthesis

Sources weighted toward 2019–2025. Pre-2019 foundational work is retained only where paired with a recent replication or extension.

Preface: What the recidivism numbers actually tell us

The most-cited figures in reentry come from the Bureau of Justice Statistics’ prisoner-recidivism studies. They are useful — but they answer a narrower question than the policy conversation usually treats them as answering.

Four issues recur across recent methodological critiques:

• Rearrest ≠ reoffending. Rearrest is the broadest and most-reported BJS measure, but not all arrests result in charges, convictions, or evidence that a crime occurred.
• Technical violations get folded into “recidivism.” Many reincarceration events are returns for rule violations (missed appointments, failed drug tests, failure to pay fees) — not new crimes. BJS measures conflate the two.
• The most-cited cohort is aging. The widely quoted “83% rearrest within 9 years” figure is from the 2005 release cohort (BJS, published 2018). The more recent BJS 2012 cohort shows lower rates and a different time profile.
• Cohort studies lag reality. BJS publishes every three years; current policy discussions often rely on figures from people released 10–20 years ago — predating fentanyl, Medicaid reentry waivers, the Ban-the-Box rethinking, and the FCC phone-rate cap.

The 2022 National Academies report’s headline finding: recidivism alone is an inadequate measure of reentry success. The field should supplement it with measures across employment, housing, family stability, health, and well-being. The literature reviewed below is what that fuller picture looks like.

→ Visual companion: Five reentry research corrections worth knowing

I. Housing & Employment

Scale

The HHS/HUD 2024 Reentry and Housing Stability synthesis describes stable housing as the foundational precondition for employment, treatment continuity, and family reunification. Formerly incarcerated people are roughly 10× more likely to experience homelessness than the general U.S. population (HUD/HHS 2024; Bowman et al. 2023, Cityscape). Public-housing access is restricted by automatic federal bars on certain conviction categories and by broad Public Housing Authority discretion to deny based on any criminal history. Private-market access is constrained by background-check stigma that operates regardless of the type of offense, time elapsed, or evidence of rehabilitation.

The employment record-penalty, updated

Devah Pager’s 2003 Milwaukee audit study established the foundational finding: a felony conviction reduced employer callbacks by roughly half, and the penalty was substantially larger for Black applicants than white applicants. The 2025 update — Lageson et al. in Criminology, surveying 1,080 employers — adds three nuances: (a) the record penalty remains large for official background-check “hits,” (b) the racial gap is more complex than the 2003 framing suggested, and (c) sealing or expungement reduces but does not eliminate the penalty.

Over 30,000 occupational licensing restrictions further narrow the post-release labor market (Collateral Consequences Resource Center 2022, updated 2024).

Ban the Box: the evidence has turned

Ban-the-Box laws — which delay criminal-record questions on job applications — were initially treated as an unambiguous reform win. Recent evidence has complicated that view. Agan & Starr (2018, QJE) and Doleac & Hansen (2020, J. Labor Economics) find that in some contexts Ban the Box can amplify racial disparity in hiring through statistical discrimination: when employers cannot ask about records, some appear to use race as a proxy. Rose (2021, J. Labor Economics) reports neutral effects in Seattle. The current consensus: reform without enforcement and complementary measures backfires.

II. Mental Health & Substance Use

Scale

Behavioral-health and substance-use prevalence among incarcerated and recently-released populations is far above the general population. Continuity of care during reentry is documented as crucial.

Post-release mortality, updated for the fentanyl era

Ranapurwala et al. (2022, AJPH), using 2000–2018 linked North Carolina correctional and death records, found post-release opioid overdose death rates rising sharply over the period and remaining concentrated in the first two weeks after release. Waddell et al. (2023, J. Substance Use and Addiction Treatment) confirmed elevated risk in a post-fentanyl-transition Western-state cohort and documented disproportionate risk among women, mediated by co-occurring mental-health disorders. Kouyoumdjian et al. (2023, BMJ) extended the finding to Canada: in the fentanyl-dominant era (2015–2020), 15.6% of all opioid toxicity deaths in Ontario were among people with a recent corrections history.

Mechanism

Loss of opioid tolerance during forced abstinence in custody, combined with return to pre-incarceration use patterns in a drug supply now dominated by unpredictable fentanyl analogs. The risk is mechanistically explained, not random.

MOUD evidence

Evidence for Medications for Opioid Use Disorder (methadone, buprenorphine, naltrexone) in corrections is now among the strongest in reentry research. Rhode Island’s statewide corrections-based MOUD program was associated with a roughly 60% reduction in post-release overdose mortality (Green et al. 2018, JAMA Psychiatry; subsequent replications). Yet access remains deeply uneven: a 2022 national assessment found a majority of U.S. jails and a substantial minority of state prison systems still do not offer the full FDA-approved range.

§1115 Medicaid reentry waivers

The most significant 2023–2025 policy shift in this domain: CMS approval of §1115 demonstration waivers in California, Washington, Montana, Massachusetts, and a growing list of other states permitting Medicaid coverage for specified pre-release services — behavioral-health assessment, MOUD, and short medication supplies (KFF 2024).

III. Family & Social Ties

Scale

An estimated 2.7 million U.S. children have experienced parental incarceration. The majority of incarcerated women are mothers of minor children and were primary caregivers prior to incarceration.

The visitation–recidivism link

Bales & Mears’s 2008 Florida analysis remains the most-cited single study; Mitchell et al.’s 2016 meta-analysis of 16 studies confirmed a modest but consistent recidivism-reduction effect of visitation. Duwe & McNeeley (2021, Crime & Delinquency) found that prison video visitation supplemented rather than substituted for in-person contact, with modest independent effects. Prison Policy Initiative’s 2021 synthesis reported that phone contact had a stronger association with reduced recidivism than visitation in at least one large study — a finding that matters because phone contact scales more easily.

Structural barriers

State prisons average more than 100 miles from incarcerated people’s home communities; rural facility placements push that much higher. The FCC’s 2024 Incarcerated People’s Communications Services rule capped prison and jail phone and video call rates nationally — a major recent development still being phased in.

The reintegration paradox

Bruce Western’s Homeward (2018; analyses extended through 2022) tracked 122 people released in Boston for one year post-release. The central finding: families function as both the most powerful supports and a recurring source of stress, with effects varying by family stability and the nature of pre-incarceration ties.

Disparities

ASFA timelines (Adoption and Safe Families Act, 1997) frequently exceed maternal incarceration lengths, producing parental-rights termination as a collateral consequence of sentence length rather than parenting capacity. Murray, Farrington & Sekol’s 2012 meta-analysis established the link between parental incarceration and children’s antisocial behavior, mental-health problems, and educational outcomes; subsequent work by Turney and Wildeman has extended these findings.

IV. Systems & Policy Barriers

Scale

The National Inventory of Collateral Consequences of Conviction catalogs more than 40,000 distinct state and federal restrictions triggered by a conviction. These cover employment, housing, voting, public benefits, occupational licensing, family law, and more.

Supervision violations

The Council of State Governments Justice Center’s 2024 Supervision Violations and Their Impact on Incarceration report quantifies the supervision-to-prison pipeline:

• ~200,000 prison admissions in 2023 came from probation or parole violations.
• More than 110,000 of those were technical violations — no new crime.
• ~3.7 million U.S. adults are on probation or parole at any given time.

CSG 2024 documents declines from 2018 to 2023: parole-violation readmissions decreased by ≥10% in 31 states; probation-violation admissions decreased by ≥10% in 29 states; technical-violation admissions fell in 18 states (15 of them by at least 10%). But admissions for technical violations rose by 12,000 from 2021 to 2022 — a reminder that the downward trend is uneven and partly attributable to pandemic-era caseload reductions that have since reversed.

Mechanisms of system-driven failure

• Conditions of supervision. Human Rights Watch (2020) and ACLU document that typical conditions — maintain employment, remain drug- and alcohol-free, keep all appointments, avoid all police contact, pay fees — set a floor of compliance many people cannot meet under reentry-window scarcity.
• Legal financial obligations. The Brennan Center’s Steep Costs research line documents that post-release LFOs (restitution, court fees, supervision fees, drug-testing fees) compound and trigger violations independent of any new criminal conduct.

Disparities

CSG 2024: Black adults are approximately 25% more likely than white adults to be readmitted for parole violations and 50% more likely to be imprisoned for the violation. The architecture of collateral consequences remains radically inconsistent across states, producing dramatically different reentry environments for otherwise-similar individuals.

Cross-cutting synthesis

Reading across the four domains, four patterns recur across the 2019–2025 literature.

1. The first 30 days concentrate risk across all four domains. Overdose mortality peaks in the first two weeks. Housing instability is most acute in the first month. Supervision violations cluster early. Reentry risk is heavily front-loaded.

2. Challenges compound; interventions often don’t. Single-domain programs (housing-only, employment-only, SUD-treatment-only) show modest effects in evaluation. Multi-domain interventions with pre-release planning and warm handoffs perform substantially better.

3. Disparities are structural, not residual. Race- and gender-based disparities appear in every domain: employment callbacks, housing access, revocation rates, overdose mortality mediated by mental-health burden. The 2019–2025 literature is consistent that these are produced by system architecture, not by underlying differences in behavior.

4. We need better measurement, and we know it. The 2022 National Academies report is a watershed. Its core argument — that recidivism alone is inadequate and should be supplemented across employment, housing, health, and well-being — is increasingly being adopted in evaluation frameworks.

What this means for practice

The cluster of evidence above points consistently in the same direction:

• Pre-release planning outperforms post-release scrambling.
• Warm handoffs to community providers outperform binders of resources.
• MOUD initiated in custody and continued post-release is among the strongest interventions in the entire reentry literature.
• Housing secured before the gate outperforms housing pursued after.
• Medicaid pre-enrolled through §1115 waivers is the most significant 2023–2025 policy lever.
• Supervision reform that distinguishes rule violations from new crimes is moving — unevenly.

What’s missing from this picture, and what most reentry programming has not addressed: the cognitive-bandwidth bottleneck at the moment of decision. That gap is the subject of our manifesto.

Frequently asked questions

What time period does this review cover?

Sources are weighted toward 2019–2025 research. Pre-2019 foundational work is retained only where paired with a recent replication or extension. Each section flags where older sources still anchor the evidence.

Where can I download the full version?

Reentry literature review PDF (38 sources).

Is this a peer-reviewed publication?

No. This is a synthesis blog post drawing on peer-reviewed work, federal statistical agencies, the National Academies, and established research organizations. Every empirical claim is sourced.

Can I cite this page?

Yes. Suggested citation: Milman, O. (2026). Reentry research, 2019–2025: a literature review for practitioners and policymakers. Next Move Technologies. https://www.nextmove.one/blog/reentry-literature-review-2026/

Will you update it?

Yes — at least annually. The next scheduled update is May 2027, or sooner if a major BJS cohort or National Academies follow-up changes the picture.

Reference list

Full numbered reference list with hyperlinks lives in the downloadable PDF. Headline sources cited above:

• National Academies of Sciences, Engineering, and Medicine (2022). The Limits of Recidivism. https://doi.org/10.17226/26459
• Council of State Governments Justice Center (2024). Supervision Violations and Their Impact on Incarceration.
• HUD / HHS (2024). Reentry and Housing Stability: Final Report. ASPE.
• Bowman, S. W., Ely, K., et al. (2023). Addressing Barriers to Housing in Reentry Programs. Cityscape 25(2).
• Lageson, S. E., et al. (2025). The mark or trace of a criminal record. Criminology.
• Ranapurwala, S. I., et al. (2022). Opioid Overdose Deaths Among Formerly Incarcerated Persons. AJPH 112(2): 300–303.
• Green, T. C., et al. (2018). Postincarceration Fatal Overdoses After Implementing Medications for Addiction Treatment. JAMA Psychiatry 75(4): 405–407.
• Agan, A., & Starr, S. (2018). Ban the Box, Criminal Records, and Racial Discrimination. QJE 133(1): 191–235.
• Doleac, J. L., & Hansen, B. (2020). The Unintended Consequences of “Ban the Box.” J. Labor Economics.
• Western, B. (2018). Homeward: Life in the Year After Prison. Russell Sage.
• Mullainathan, S., & Shafir, E. (2013). Scarcity. Henry Holt.