Recidivism rate

The recidivism rate measures the share of people released from prison, jail, probation, or another form of supervision who commit a new offense within a defined follow-up window, commonly one, three, or five years. It is a standard outcome measure for corrections systems, diversion programs, and reentry services, computed roughly as re-offenders during window / total released during period.

The rate depends heavily on how "reoffend" is defined, a new arrest, a new charge, a new conviction, or a return to custody, each yielding a different number for the same population, and on the length of the follow-up window, since a longer window mechanically produces a higher rate. This makes recidivism figures hard to compare across jurisdictions or studies that use different definitions, a frequent source of misleading comparisons.

Recidivism rate is the primary outcome metric in program evaluation of correctional and reentry programs, and it is the outcome that a risk assessment instrument is built and validated to predict. A common pitfall is treating a lower measured rate as proof a program works without a comparison group, since recidivism also varies with the underlying crime rate and enforcement intensity, factors unrelated to the program itself.