Recidivism is the rate at which people released from prison are rearrested, reconvicted or reincarcerated. It is the number most reform arguments rest on, and it is frequently quoted without the context that makes it meaningful.
Key Takeaways
- 66% of people released from state prisons were rearrested within three years; 82% within ten.
- About 61% returned to prison within ten years.
- The annual arrest rate falls sharply over time — 43% in year one, 22% by year ten.
- “Rearrest”, “reconviction” and “reincarceration” are different measures and produce very different figures.
- Many returns to prison are technical violations of supervision, not new crimes.
The Headline Figures
The most substantial US dataset comes from the Bureau of Justice Statistics, which followed roughly 409,300 people released from state prisons across 24 states in 2008 for a full decade (Bureau of Justice Statistics).
| Measure | Figure |
|---|---|
| Rearrested within 3 years | 66% |
| Rearrested within 10 years | 82% |
| Returned to prison within 10 years | 61% |
| Arrested in year 1 | 43% |
| Arrested in year 10 | 22% |
Why the Definition Matters
Recidivism figures are often compared when they are not comparable. Three different things get measured:
- Rearrest — being arrested again. The broadest measure, and the highest number. An arrest is not a conviction.
- Reconviction — being convicted of a new offence. Lower.
- Reincarceration — returning to prison, whether for a new sentence or a supervision violation.
A programme reporting a “20% recidivism rate” may simply be measuring reincarceration over two years rather than rearrest over ten. Always check the measure and the window before comparing.
Technical Violations
A significant share of returns to custody follow breaches of supervision conditions rather than new offences — missing appointments, failing a drug test, travelling without permission, or losing contact with a supervising officer.
This matters for interpretation. Someone reincarcerated for missing appointments after losing housing and transport is counted the same as someone convicted of a new offence, though the policy response each implies is entirely different.
What Reduces Recidivism
The best-evidenced intervention is education. RAND Corporation research found correctional education participants were 43% less likely to reoffend and 13% more likely to be employed after release, with an estimated return of about $5 per $1 spent (RAND Corporation).
Alongside it: employment, stable housing, substance use treatment with continuity across release, and family connection. Because risk concentrates in year one, front-loading support there is where the evidence points.
One honest caveat: people who volunteer for programmes may differ systematically from those who do not, which complicates causal claims. The consistency of the direction across studies is the stronger argument, not any single figure.
Frequently Asked Questions
What is the recidivism rate in the US?
Bureau of Justice Statistics data found 66% of people released from state prisons were rearrested within three years and 82% within ten years, with about 61% returning to prison over the decade.
What does recidivism mean?
The tendency of someone previously convicted to reoffend. It is measured three different ways – rearrest, reconviction, or reincarceration – which produce substantially different figures.
What reduces recidivism most?
Correctional education has the strongest evidence base, associated with a 43% reduction in reoffending. Employment, stable housing, substance use treatment continued after release, and family connection are also consistently linked to better outcomes.
Sources & Further Reading
- Bureau of Justice Statistics — Recidivism of Prisoners Released in 24 States in 2008: A 10-Year Follow-Up
- RAND Corporation — Education and Vocational Training in Prisons Reduces Recidivism
- Bureau of Justice Statistics — Recidivism and Reentry
Last reviewed August 2026.
Part of our complete guide to prison reform.