Recidivism in America 2026
Every year, approximately 600,000 people are released from state and federal prisons in the United States, walking back into communities with a question that follows them like a shadow: will they return? The answer, for too many, is yes — and the frequency with which it happens, the speed at which it happens, and the structural reasons why it happens are among the most consequential and contested data points in American criminal justice. Recidivism — typically measured as rearrest, reconviction, or reincarceration following release — is not a single, clean statistic. It is a cluster of overlapping measures that tell different parts of the same story. The Bureau of Justice Statistics (BJS), the federal government’s primary source for recidivism data, has documented that 82% of state prisoners are arrested at some point in the 10 years following their release, that 71% are rearrested within 5 years, and that 44% return before their first year out of prison is complete. At the same time, the three-year reincarceration rate — the most widely used and comparable metric across states — has fallen from 35% in 2008 to approximately 27–28% in the most recent national data, a nearly 23% reduction that represents genuine, measurable progress.
The 2026 policy environment around recidivism is shaped by this tension between persistent structural failure and documented, if uneven, progress. Nearly 75% of people released from prison today avoid reincarceration within the critical three-year window, according to the Council of State Governments Justice Center’s 2026 report — up from 65% a decade ago, translating to roughly 30,000 fewer people returning to prison each year than under the prior trajectory. Yet the economic and human costs of the remaining recidivism remain staggering: states will spend an estimated $8 billion on reincarceration costs alone for people who exited prison in 2022, a figure that excludes law enforcement, court, victim services, and social costs. The unemployment rate among formerly incarcerated people stands at 27%, housing insecurity affects them at a rate of 5,700 per 100,000 — nearly double that of the general population — and 60% of those who reoffend do so within the first year of release, when the absence of stable housing, employment, and social support is most acute. The evidence on what reduces recidivism is increasingly clear. What lags is the political will and funding to scale it.
Key Facts: Recidivism Statistics in the US 2026
RECIDIVISM FAST FACTS — US (Most Recent Data, 2025/2026)
──────────────────────────────────────────────────────────────────
📊 Rearrested within 1 year of release (state prison) ~44%
📊 Rearrested within 3 years (state prison, 2012 cohort) 68%
📊 Rearrested within 5 years (state prison, 2012 cohort) 71%
📊 Rearrested within 10 years (state prison, 2005 cohort) 82%
🏛️ 3-year reincarceration rate (national, current) ~27–28%
📉 Decline in 3-yr reincarceration since 2008 –23%
🏙️ People released annually from US prisons ~600,000
💰 Estimated reincarceration cost (2022 release cohort) $8 billion
🧑⚖️ Formerly incarcerated unemployment rate 27%
🔁 Reoffending within first year of release ~60%
──────────────────────────────────────────────────────────────────
| Key Fact | Statistic |
|---|---|
| Prisoners released annually from US state and federal prisons | ~600,000 |
| Rearrested within 1 year of state prison release (NIJ) | ~44% |
| Rearrested within 3 years of state prison release (2012 cohort) | ~68% |
| Rearrested within 5 years of state prison release (2012 cohort, BJS) | 71% |
| Rearrested within 10 years of state prison release (2005 cohort, BJS) | 82% |
| Reconvicted within 5 years (2012 cohort, BJS) | 54% |
| Returned to prison within 5 years (2012 cohort, BJS) | 46% |
| National 3-year reincarceration rate (current, most recent national est.) | ~27–28% |
| National 3-year reincarceration rate (2008 — pre-Second Chance Act) | 35% |
| Reduction in 3-year reincarceration rate since 2008 | –23% |
| People who now avoid reincarceration within 3 years | ~75% (72 per 100 released) |
| Fewer returns to prison per year (vs. 2008 trajectory) | ~30,000 fewer annually |
| Estimated reincarceration cost for 2022 state prison release cohort | $8 billion |
| Projected savings if recidivism reduced 30% (Reentry 2030) | $3.9B/yr → $44B over decade |
| Formerly incarcerated unemployment rate | 27% |
| Formerly incarcerated people without HS diploma, GED, or degree | 25% |
| Rate of homelessness/housing insecurity among formerly incarcerated | 5,700 per 100,000 |
| Who reoffend within the first year of release | ~60% of those who reoffend |
| Average cost per inmate per year (US) | ~$31,286 |
| 5-year rearrest rate for violent offenders (2012 cohort) | 69% — but only 33% for another violent offense |
| 5-year rearrest rate for property offenders (larceny/auto theft) | 79.3% — highest of any category |
| 5-year rearrest rate for drug offenders | 69.8% overall |
| 5-year rearrest rate for homicide offenders | 41.3% — lowest of any category |
| Median time to rearrest (state offenders) | ~19–21 months |
Source: Bureau of Justice Statistics — Recidivism of Prisoners Released in 34 States in 2012 (BJS, 2021); BJS — Recidivism of State Prisoners Released in 2005 (9-Year Follow-Up, 2018); Council of State Governments (CSG) Justice Center — 50 States, 1 Goal: State Recidivism Rates 2006–2022 (March 2026); Council on Criminal Justice New National Recidivism Report (2024); Reentry 2030 — Impact and Cost Savings Analysis (April 2026); Prison Policy Initiative Mass Incarceration 2025; National Institute of Justice
The key facts table above requires a note on methodology before interpretation, because no other field of criminal justice statistics is more susceptible to misleading headline comparisons. Rearrest rates are always the highest recidivism figures, because an arrest does not equal guilt and may involve technical violations of supervision or minor public-order offenses. Reincarceration rates are always the lowest, because they require a new sentence of imprisonment — the most serious possible outcome. When media headlines report “2 in 3 prisoners reoffend within 3 years,” they are typically citing rearrest data; when reform advocates cite “only 1 in 4 returns to prison,” they are citing the reincarceration figure. Both are accurate descriptions of different measured outcomes. For the purposes of policy comparison — and because it is the metric tracked by all 50 states — the three-year reincarceration rate is the most useful benchmark, and at 27–28% nationally, it tells a story of genuine but incomplete progress. The $8 billion estimated to be spent reincarcerating people from the single 2022 release cohort gives the human and policy numbers an economic weight that no abstract percentage can carry on its own.
Recidivism Rates by State in the US 2026
3-YEAR REINCARCERATION RATES BY STATE CLUSTER (Most Recent Data)
──────────────────────────────────────────────────────────────────
LOWEST RATES (≤20% — 9 states):
Oregon, Texas, California, Oklahoma,
Virginia, Utah, New York, S. Carolina,
Minnesota ████░░░░░░░░░░░░░░░ ≤20%
MID RANGE (20–30%):
Most states █████████░░░░░░░░░░ ~20–30%
HIGHEST RATES (>35%):
North Carolina,
select states ████████████████░░░ >35%
NATIONAL AVERAGE (current):
~27–28% ████████████░░░░░░░ ~27–28%
──────────────────────────────────────────────────────────────────
Source: CSG Justice Center 50 States 1 Goal, March 2026
| State / Category | 3-Year Reincarceration Rate / Trend |
|---|---|
| States with ≤20% 3-year reincarceration (most recently reported) | 9 states: Oregon, Texas, California, Oklahoma, Virginia, Utah, New York, South Carolina, Minnesota |
| States with ≥30% reduction in 3-year rate over past 10 years | 8 states: California, New York, New Jersey, Colorado, Massachusetts, Michigan, Oregon, Missouri |
| States with 3-year rates above 45% (post-Second Chance Act, most recent) | 6 states (down from 11 before the Act) |
| California — reduction in 3-year reincarceration rate | –44 percentage points (driven by Realignment, AB 109) |
| North Carolina — change in reincarceration rate (2008 to 2019 cohort) | +11 percentage points (increased — from 25% to 36%) |
| Texas — 2021 release cohort staying out 3 years | ~49,000 of 58,670 released did NOT return within 3 years |
| National 3-year reincarceration rate (2008 — baseline) | 35% |
| National 3-year reincarceration rate (2019 — most recent national cohort) | 27% |
| National 3-year reincarceration rate (current best estimate, 2026) | ~27–28% |
| Reduction in national 3-year rate (2008 to 2019) | –23% (8 percentage points) |
| States with double-digit % reduction in reincarceration rates | 9 states: CA, CO, CT, DE, IL, MD, MA, MO, SC |
| Share of states showing reduction in 3-year reincarceration rate | ~75% |
| 5-year rearrest rate (2005 cohort, 30 states) | 77% |
| 5-year rearrest rate (2012 cohort, 34 states) | 71% — 6-point improvement |
| Reentry Alabama program — reported recidivism reduction | 86% reduction: from 30% to 4% among participants |
| Estimated annual reincarceration costs (states, current) | $8 billion for 2022 release cohort alone |
Source: CSG Justice Center — 50 States, 1 Goal: Recidivism Rate Trends (March 2026); CSG Justice Center — 50 States, 1 Goal: Second Chance Act Era (2024); Council on Criminal Justice New National Recidivism Report (2024); BJS Recidivism of State Prisoners; World Population Review Recidivism Rates by State 2026
The state-level data is where the national recidivism narrative becomes most complex and most instructive. California’s 44-percentage-point reduction in its measured 3-year reincarceration rate is the most dramatic shift in any state’s data — but it was driven primarily by the 2011 Public Safety Realignment Act (AB 109), which transferred responsibility for certain lower-level felony offenders from state prison to county jail. This definitional change means California’s prison reincarceration rate dropped without necessarily reflecting an equivalent change in individuals’ actual reoffending behavior. North Carolina’s increase from 25% to 36% over the same period reflects the opposite: policy changes that expanded who could be returned to state prison, making the state look worse statistically even without a commensurate worsening of behavior. These examples illustrate why the Council of State Governments cautions against ranking states against each other — comparisons are meaningful over time within a state, not across states using different definitions.
The nine states achieving 3-year reincarceration rates at or below 20% represent the frontier of what is possible with sustained policy commitment. Texas, despite its reputation as a tough-on-crime state, has achieved significant reductions through a combination of drug courts, diversion programs, community supervision enhancements, and reentry investment, keeping nearly 50,000 of the 58,670 people released in 2021 out of prison within three years. The eight states achieving 30% or greater reductions over the past decade — a geographically and politically diverse group spanning California, New York, Colorado, Oregon, Missouri, Michigan, Massachusetts, and New Jersey — demonstrate that meaningful recidivism reduction is not the exclusive property of any one political philosophy. What they share is sustained investment in evidence-based programming, supervision reform, and reentry infrastructure rather than reliance on incarceration alone as a public safety strategy.
Recidivism by Offense Type in the US 2026
5-YEAR REARREST RATES BY MOST SERIOUS OFFENSE (2012 Release Cohort, BJS)
──────────────────────────────────────────────────────────────────
Larceny / Motor Vehicle Theft ████████████████████████████████ 79.3%
Weapons Offenses ███████████████████████████████░ 73.3%
Drug Offenses (overall) ████████████████████████████░░░░ 69.8%
Drug Possession ████████████████████████████████ 73.0%
Drug Trafficking ██████████████████████████████░░ 67.2%
Violent Offenses (overall) ████████████████████████████░░░░ 69.0%
Robbery ███████████████████████████░░░░░ 68.0% (est.)
Burglary ████████████████████████████░░░░ ~74%
Murder / Manslaughter ████████████████░░░░░░░░░░░░░░░░ 41.3%
Sex Offenses (rape/sex assault) ███████████████░░░░░░░░░░░░░░░░ ~49%
──────────────────────────────────────────────────────────────────
Source: BJS Recidivism of Prisoners Released in 34 States in 2012 (2021)
| Offense Type | 5-Year Rearrest Rate (2012 Cohort) | Note |
|---|---|---|
| Larceny / Motor Vehicle Theft | 79.3% | Highest of any category |
| Burglary | ~74% | Driven by addiction/financial need |
| Weapons offenses | 73.3% | Higher than many expect |
| Drug possession | 73.0% | Higher than trafficking |
| Drug offenses (all) | 69.8% | Reflects addiction cycle |
| Drug trafficking | 67.2% | Lower than possession |
| Violent offenses (all) | 69% | High rearrest, but low re-violent-offense rate |
| Violent offenders rearrested for another violent crime (5 years) | Only 33% | Most rearrested for non-violent offenses |
| Robbery | ~68% | Among violent subcategories |
| Sex offenses (rape / sexual assault) | ~49% | One of the lower categories |
| Murder / nonnegligent manslaughter | 41.3% | Lowest of any offense category |
| Homicide offenders rearrested for any violent crime (5 years) | Lower than property offenders rearrested for same | Age factor dominant |
| Property offenders rearrested for violent crime (5 years) | 51.2% | Cross-category reoffending is common |
| Violent offenders rearrested for property crime (5 years) | 28.9% | Criminal behavior is not specialized |
| Public order offenses — most common category of rearrest | 54–58% of those rearrested | Includes DUI, disorderly conduct, weapons |
| Drug rearrests (2005 cohort vs. 2012 cohort) | Declined by 6 percentage points | Meaningful improvement |
| Property offense rearrests (2005 vs. 2012 cohort) | Declined by 3 percentage points | Some improvement |
| Violent rearrest rate (2005 vs. 2012 cohort) | Virtually unchanged | Persistent challenge |
Source: Bureau of Justice Statistics — Recidivism of Prisoners Released in 34 States in 2012: A 5-Year Follow-Up (July 2021); BJS Recidivism of State Prisoners Released in 2005 (9-Year Follow-Up, 2018); Council on Criminal Justice New National Recidivism Report (September 2024); LegalClarity Criminal Recidivism Rates, Causes and Consequences (May 2026); Prison Policy Initiative Mass Incarceration 2025
The offense-type breakdown of recidivism rates produces results that are genuinely counterintuitive and have important policy implications. The conventional assumption — that violent offenders are the highest recidivism risk — is directly contradicted by the data. Murder and nonnegligent manslaughter offenders have the lowest 5-year rearrest rate at 41.3%, lower than any other category. Sex offenders’ rearrest rate of roughly 49% is also lower than the overall average, despite being the group most associated in public discourse with high recidivism risk. The primary explanation is age: people serving time for serious violent crimes often serve longer sentences and are released significantly older, at an age when criminal activity has naturally and substantially declined for virtually everyone.
At the top end, property offenders — particularly those convicted of larceny and motor vehicle theft — have the highest 5-year rearrest rates at 79.3%, followed closely by drug possession offenders at 73.0%. The mechanism here is also straightforward: property crime and drug possession are disproportionately driven by addiction and poverty, which are conditions that prison does not treat and that persist — or worsen — upon release. Someone who was stealing to support a heroin habit before incarceration will, absent treatment for the underlying addiction, face the same pressure to steal after release. The critical additional finding is that criminal behavior is not highly specialized: a person convicted of robbery is almost as likely to be rearrested for a property crime (28.9%) as another violent crime (32.4%), and many property offenders are rearrested for violent offenses (51.2%). The labels “violent offender” and “non-violent offender,” while legally meaningful, are poor predictors of future offense categories.
Recidivism by Age & Demographics in the US 2026
REARREST RATE BY AGE AT RELEASE — Federal Offenders (US Sentencing Commission)
──────────────────────────────────────────────────────────────────
Under age 21 ████████████████████████████████████████ 67.6%
21–24 ████████████████████████████████████████ 66.4%
25–29 ████████████████████████████████████████ 62.2%
30–34 ████████████████████████████████░░░░░░░░ 55.3%
35–39 ████████████████████████████░░░░░░░░░░░░ 48.8%
40–44 ████████████████████████░░░░░░░░░░░░░░░░ 42.4%
50–59 ██████████████░░░░░░░░░░░░░░░░░░░░░░░░░░ 24.7%
60+ ████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 16.0%
──────────────────────────────────────────────────────────────────
Source: US Sentencing Commission Recidivism Study (Federal Offenders, 2016)
| Demographic / Age Factor | Recidivism Data |
|---|---|
| Rearrest rate — released under age 21 (federal, USSC) | 67.6% |
| Rearrest rate — released age 21–24 (federal) | 66.4% |
| Rearrest rate — released age 25–29 | 62.2% |
| Rearrest rate — released age 30–34 | 55.3% |
| Rearrest rate — released age 35–39 | 48.8% |
| Rearrest rate — released age 40–44 | 42.4% |
| Rearrest rate — released age 50–59 | 24.7% |
| Rearrest rate — released age 60+ | 16.0% |
| People released age 24 or younger vs. 40+: reincarceration risk at Year 5 | 64% more likely (56.8% vs. 36.3%) — Council on CJ |
| Median time to rearrest — offenders under 30 | 17 months |
| Median time to rearrest — offenders 60+ | 28 months |
| Reconviction rate — offenders under age 21 | 48.5% (highest of any group) |
| Reconviction rate — offenders age 21–24 | 48.4% |
| Reincarceration rate — offenders age 21–24 | 38.6% (highest reincarceration of any age group) |
| Males vs. females — rearrest rate | Males consistently higher; women more likely rearrested for property/drug than violent |
| White offenders — rearrest rate (youngest age group, USSC) | 59.1% |
| Black offenders — rearrest rate (youngest age group, USSC) | 72.7% (highest of any age-race combination) |
| Formerly incarcerated — unemployment rate (all demographics) | 27% |
| People with 2+ arrests/year who have incomes below $10,000 | 49% |
| Black unemployment rate (2025) | 1.9x higher than White unemployment rate |
| Latino unemployment rate (2025) | 1.5x higher than White unemployment rate |
Source: US Sentencing Commission — Recidivism Among Federal Offenders: A Comprehensive Overview (2016, reference cohort); Council on Criminal Justice New National Recidivism Report (2024); BJS Recidivism of Prisoners Released in 34 States in 2012 (2021); CSG Justice Center Data Snapshots 2025; Prison Policy Initiative Reentry and Recidivism Research 2025
Age is the single most powerful predictor of recidivism — more powerful than offense type, criminal history, program participation, or nearly any other factor researchers have studied. The rearrest rate falls from 67.6% for those released before age 21 to just 16.0% for those released at 60 or older, an extraordinary 51.6-percentage-point span that represents a decline in criminal activity that occurs naturally with age — what criminologists call “desistance” — largely independent of intervention. This finding has profound policy implications: if age drives desistance, then long sentences served by people who have already aged out of criminal risk impose enormous costs on taxpayers and the incarcerated individual while contributing little to public safety. People convicted of homicide who are released in their 50s or 60s after decades of incarceration have extremely low rearrest rates — yet they have often served the longest sentences in the system.
The racial disparity in recidivism data is real, but it must be interpreted alongside the structural context that drives it. Black offenders’ rearrest rates are significantly higher than white offenders’ at every age group, with the youngest Black offenders reaching 72.7% compared to 59.1% for young white offenders. But these disparities reflect — and are substantially driven by — downstream consequences of deeper structural inequities: the Black unemployment rate being 1.9 times the white rate, housing discrimination that limits access to stable post-release housing, greater exposure to neighborhoods with concentrated poverty and law enforcement presence, and the compounding effects of criminal records on employment and housing applications that make reintegration significantly harder for communities of color. 49% of people with two or more arrests per year have incomes below $10,000 — poverty and recidivism are not merely correlated, they are mechanically linked through the deprivation of the material conditions that make successful reentry possible.
Recidivism Reduction Programs & Outcomes in the US 2026
EVIDENCE-BASED PROGRAM IMPACTS ON RECIDIVISM (Meta-Analysis Research)
──────────────────────────────────────────────────────────────────
Correctional education programs
(RAND meta-analysis):
→ 43% lower odds of recidivating ████████████████████░░░░ −13 pct pts risk
→ 13% higher odds of post-release employ ██████░░░░░░░░░░░░░░░░░░ +13%
Prison education & workforce (Mackinac):
→ Recidivism likelihood reduced by ████████░░░░░░░░░░░░░░░░ −14.8%
→ Employment likelihood increased ████░░░░░░░░░░░░░░░░░░░░ +6.9%
Secondary-degree programs (Brookings):
→ Recidivism reduction ████████████████░░░░░░░░ −30%
Cognitive-Behavioral Therapy (CBT):
→ Recidivism reduction range ████████████░░░░░░░░░░░░ 10–30%
──────────────────────────────────────────────────────────────────
Source: RAND; Mackinac Center; Brookings; NIJ evidence reviews
| Program / Intervention | Recidivism Impact / Outcome |
|---|---|
| Correctional education programs (RAND meta-analysis) | 43% lower odds of recidivating; 13% higher post-release employment odds |
| Correctional education — reduction in recidivism risk (absolute) | −13 percentage points (not just odds ratio) |
| Prison workforce and education programs (Mackinac Center largest meta-analysis) | −14.8% reduction in likelihood of recidivism; +6.9% employment; +$131/quarter wages |
| Secondary-degree (GED/HS diploma equivalent) prison programs | Reduce recidivism by ~30% (Brookings review) |
| Postsecondary education in prison | Even stronger recidivism reduction than secondary education |
| Cognitive-behavioral therapy (CBT) programs | 10–30% recidivism reduction (risk-needs-responsivity model) |
| Drug courts (evidence-based) | Reduce recidivism vs. traditional prosecution — among strongest evidence bases |
| Reentry Alabama program | Reported 86% recidivism reduction: from 30% to 4% among participants |
| Housing assistance for formerly incarcerated (MOVE pilot, Maryland) | Lower rearrest rates vs. those without housing; movers to new communities had best outcomes |
| Substance use disorder treatment in custody | Reduces recidivism; unmet SUD needs a top driver of reoffending |
| Vocational training programs | Positive employment and recidivism outcomes |
| Risk-needs-responsivity (RNR) model | Leading evidence-based framework for tailoring interventions to individual risk/need level |
| Second Chance Act (2008) — national reincarceration rate change | 3-year rate fell from 35% → 27% since passage (−23%) |
| Reentry 2030 goal | Additional 30% reduction in recidivism; keep 500,000+ out of prison over decade; save $43B+ |
| States setting bold recidivism reduction goals (2025–2026) | Missouri, Alabama, North Carolina, Nebraska — targeting 2030 goals |
| Second Chance Act Reentry Education & Employment grants (FY 2026) | Active — up to $900K/project; 36-month programs starting June 1, 2026 |
Source: RAND Corporation — Evaluating the Effectiveness of Correctional Education (meta-analysis); Mackinac Center for Public Policy — Prison Education Programs Meta-Analysis (2023, largest to date, 148 results from 78 studies); Brookings Institution — A Better Path Forward: Training and Employment for Correctional Populations (July 2024); CSG Justice Center — 50 States, 1 Goal (March 2026); Reentry 2030 Impact and Cost Savings (April 2026); National Institute of Justice evidence reviews; Second Chance Act FY 2026 grants via fundsforNGOs (April 2026)
The research on what actually reduces recidivism has reached a level of consistency and depth that is rarely seen in social science. Correctional education is the most studied and most robustly evidenced intervention in the field. RAND Corporation’s landmark meta-analysis — the most comprehensive review of the evidence at the time of publication — found that inmates who participate in correctional education programs have 43% lower odds of returning to prison than those who do not, translating to a 13-percentage-point reduction in absolute recidivism risk and a 13% improvement in post-release employment odds. The Mackinac Center’s subsequent and even larger meta-analysis, drawing on 148 results from 78 high-quality research papers, found a 14.8% reduction in recidivism likelihood and meaningfully better employment and wage outcomes. Secondary-degree programs reduce recidivism by approximately 30%, with postsecondary education showing even stronger effects — findings that have led Congress to restore Pell Grant eligibility for incarcerated students under the Second Chance Pell program.
Cognitive-behavioral therapy, drug courts, housing assistance, and substance use disorder treatment each have their own evidence bases pointing in the same direction: structured interventions that address the underlying drivers of criminal behavior — addiction, cognitive distortions, lack of employment skills, housing instability — produce better outcomes than incarceration without such programming. The Reentry 2030 initiative, launched by the CSG Justice Center and its partners, has set an ambitious national goal of a 30% additional reduction in recidivism on top of the gains since 2008, projecting that achieving this would keep more than 500,000 additional people out of prison over a decade and avert close to $44 billion in corrections costs — a figure that does not include reduced costs for law enforcement, prosecution, victim services, or the economic value of the formerly incarcerated people who would instead be working, paying taxes, and supporting families. The Second Chance Act’s active FY 2026 grant round, with awards of up to $900,000 per project for education and employment reentry programs beginning June 2026, represents the federal government’s ongoing investment in the evidence base — even as political headwinds have threatened broader correctional funding.
Barriers to Successful Reentry & Structural Drivers of Recidivism in the US 2026
REENTRY BARRIERS — FORMERLY INCARCERATED PEOPLE, US 2025/2026
──────────────────────────────────────────────────────────────────
Unemployment rate (formerly incarcerated) ████████████░░░░░░░ 27%
Without HS diploma / GED / degree █████████░░░░░░░░░░ 25%
Homelessness / housing insecurity rate ██████████████████░ 5,700/100k
People below $10k income (2+ arrests/yr) ████████████████████ 49%
Probation conditions avg per day ████████████████████ 12+ requirements
% who reoffend in Year 1 after release █████████████████████ ~60%
60% reoffend in first year — access to
housing, jobs, support is most acute here
──────────────────────────────────────────────────────────────────
Source: Prison Policy Initiative; CSG Justice Center; NIJ
| Reentry Barrier / Structural Driver | Data |
|---|---|
| Unemployment rate — formerly incarcerated people | 27% |
| Formerly incarcerated without HS diploma, GED, or any degree | 25% |
| Rate of homelessness or housing insecurity after release | 5,700 per 100,000 (far above general population) |
| People with 2+ arrests/year with income below $10,000 | 49% |
| % of those who reoffend who do so within year 1 of release | ~60% |
| Average number of conditions people on probation must comply with daily | 12+ |
| Probation/parole technical violation — contributor to reincarceration | Significant — return to prison does not always mean new crime |
| Background check barriers to employment | Mandatory employer access in most states; limits job access for 68M+ Americans with records |
| Americans with a criminal record | More than 68 million (a quarter of the US population) |
| People on probation for violent, property, and drug offenses (2023) | Roughly equal numbers in each category |
| Substance use disorder among formerly incarcerated | Higher than general population; SUD is major driver of reoffending |
| Mental health conditions — prevalence among incarcerated (national) | Significantly elevated vs. general population; prison rarely provides treatment |
| Formerly incarcerated people in Illinois prisons — without housing plan at release | 28% of men; 42% of women (Loyola Chicago study, 2025) |
| Clean Slate laws — impact on formerly incarcerated (Clean Slate Initiative, 2025) | 42% of those with records sealed reported improvement in employment, finances, or public assistance |
| Racial disparities in post-release outcomes | Black unemployment 1.9x, Latino unemployment 1.5x White rate (2025) |
| Annual cost of incarceration — correctional system total | ~$182 billion (correctional + policing + courts) |
| Estimated cost of harm to victims of aggravated assaults alone (2023) | ~$43 billion |
Source: Prison Policy Initiative — Reentry and Recidivism Research 2025; Prison Policy Initiative Mass Incarceration: The Whole Pie 2025; CSG Justice Center Data Snapshots 2025/2026; Loyola Chicago Center for Criminal Justice 2025; Clean Slate Initiative 2025; Brennan Center for Justice; American Action Forum — Economic Costs of the US Criminal Justice System
The structural barriers to successful reentry are not incidental to the recidivism problem — they are its primary engine. When 27% of formerly incarcerated people are unemployed at rates far above the general labor market, when 25% lack a high school diploma or GED, when housing instability affects the formerly incarcerated at a rate of 5,700 per 100,000, and when 49% of the most persistent reoffenders have annual incomes below $10,000, the recidivism rate is not mysterious. It is the predictable output of releasing people from prison into the same conditions — or worse conditions — that preceded their incarceration, without the material support needed to stabilize and rebuild. The first year post-release is the most dangerous window: approximately 60% of those who reoffend do so within 12 months of release, precisely when employment, housing, and social support are most difficult to secure and the pull of familiar environments and associates is strongest.
The 68 million Americans who carry a criminal record — roughly a quarter of the entire US population — face a labor market in which background checks are ubiquitous and disqualifying, a housing market in which criminal history is a standard screening criterion, and a social support system in which benefits eligibility is frequently restricted based on conviction type. Clean Slate laws — which automate the sealing of qualifying records after a crime-free period — have shown meaningful early results: in Pennsylvania, Utah, and Michigan, the Clean Slate Initiative found that 42% of people whose records were automatically sealed reported subsequent improvements in employment, personal finances, or access to public assistance. This kind of structural intervention — removing the criminal record as a permanent employment and housing barrier for those who have demonstrated successful reintegration — operates at the policy level rather than the individual intervention level, and it addresses the reentry obstacle at its source rather than downstream.
Federal Recidivism Statistics in the US 2026
FEDERAL OFFENDER RECIDIVISM (US Sentencing Commission — 8-Year Follow-Up)
──────────────────────────────────────────────────────────────────
Rearrested within 8 years ████████████████████████░░░░ 49.3%
Reconvicted within 8 years ████████████████░░░░░░░░░░░░ 31.7%
Reincarcerated within 8 years ████████████░░░░░░░░░░░░░░░░ 24.6%
Median time to rearrest 21 months
REARREST RATE BY CRIMINAL HISTORY CATEGORY (USSC):
Category I (lowest) ████████████░░░░░░░░░░ 33.8%
Category II ██████████████████████░ 54.3%
Category III ████████████████████████ 63.3%
Category IV ████████████████████████████░ 74.7%
Category V █████████████████████████████░ 77.8%
Category VI (highest) ██████████████████████████████ 80.1%
──────────────────────────────────────────────────────────────────
Source: US Sentencing Commission Recidivism Study
| Federal Recidivism Metric | Data |
|---|---|
| Federal offenders rearrested within 8 years (2005 cohort) | 49.3% |
| Federal offenders reconvicted within 8 years | 31.7% |
| Federal offenders reincarcerated within 8 years | 24.6% |
| Rearrest rate — released from federal incarceration (2005) | 52.5% |
| Rearrest rate — placed directly on federal probation (2005) | 35.1% |
| Median time to rearrest (federal offenders) | 21 months |
| Largest proportion first rearrested within Year 1 | 18.2% of all federal offenders rearrested in Year 1 |
| Most common rearrest charge (federal offenders) | Assault (23.3% of those rearrested) |
| Other frequent rearrest charges | Drug trafficking, larceny, public order offenses |
| Rearrest rate — federal offenders under age 21 at release | 67.6% |
| Rearrest rate — federal offenders over age 60 at release | 16.0% |
| Criminal history Category I rearrest rate (lowest history) | 33.8% |
| Criminal history Category VI rearrest rate (highest history) | 80.1% |
| Effect of sentence length on recidivism (federal data) | Very little variation — 50.8% (6 months–2 years) to 55.5% (5–9 years) |
| Federal recidivism — 2005 vs. 2010 release cohort comparison | Virtually identical — 49.3% both cohorts (unchanged despite Booker decision and evidence-based practices increase) |
| Federal offenders rearrested — 2010 median time to arrest | 19 months |
Source: US Sentencing Commission — Recidivism Among Federal Offenders: A Comprehensive Overview (2016); US Sentencing Commission — Federal Offenders Released in 2010: Recidivism Study; Crime in America / CrimeInAmerica.net Offender Recidivism and Reentry 2026
The federal recidivism data from the US Sentencing Commission offers several findings that challenge widely held assumptions about what works in reducing reoffending. Most striking is the near-identical rearrest rate between federal offenders released in 2005 (49.3%) and those released in 2010 (49.3%) — despite the intervening Supreme Court decision in United States v. Booker that gave judges more sentencing flexibility, and despite the documented increase in evidence-based supervision practices during that period. This null result does not mean evidence-based practices are ineffective — it may reflect the time lag required for new practices to penetrate a large bureaucratic system, or limitations in the specific practices adopted — but it does underscore that reform at the federal level has been slower to translate into measurable recidivism improvements than in some state systems.
The criminal history data is among the most actionable in the federal dataset. The rearrest rate rises almost linearly from 33.8% for those with the least criminal history (Category I) to 80.1% for those with the most extensive records (Category VI) — a nearly 47-percentage-point spread that confirms criminal history as the strongest predictor of future reoffending in federal data, even stronger than age. The sentence length finding is equally important for policy: rearrest rates vary very little across sentence lengths — from 50.8% for sentences of 6 months to 2 years to 55.5% for sentences of 5–9 years — suggesting that longer sentences do not reduce subsequent criminal behavior and may be primarily serving a short-term incapacitation function at significant financial cost. This evidence base underlies the growing bipartisan consensus that targeted sentencing reform, particularly for lower-level and first-time offenders, can reduce prison populations without meaningfully increasing recidivism rates.
Disclaimer: The data research report we present here is based on information found from various sources. We are not liable for any financial loss, errors, or damages of any kind that may result from the use of the information herein. We acknowledge that though we try to report accurately, we cannot verify the absolute facts of everything that has been represented.
