Recidivism studies typically use one or more of three recidivism measures to measure reoffending:
These measures are indicators of whether and what kind of new offense might have or has occurred. Each of these ways to measure recidivism has strengths and weaknesses. Arrests are the broadest measure of crime available, but an arrest does not imply that a new offense actually occurred. Convictions indicate that a new offense did occur, but may not indicate the seriousness of the offense. Commitments to prison, the narrowest measure, do indicate that a relatively serious new offense occurred.
This report uses two recidivism measures:
Reoffense is measured as the date of the first felony offense after prison release. The data comes from inmates recommitted to the Department of Corrections for supervision or incarceration. This is a reconviction type measure because it requires that a reconviction occurred. It is called reoffense, though, because it uses the new offense date, rather than the date of conviction or of readmission to the Department. This way of measuring recidivism tells how soon after release the inmate commits a new crime.
Reimprisonment is measured as the date of return to prison after reoffense, the first offense after prison release. Rates reported on this recidivism definition—reimprisonment for a new offense—will necessarily be lower than reoffense rates, for the same follow-up period, since the reimprisoned offenders are a subset of those who reoffend. Not all reconvicted inmates are sentenced to prison. This way of measuring recidivism tells how soon after release the inmate returns to prison for a new crime.
A rearrest measure is not used in this study. The largest cohort studies of released state prison inmates show that rearrest rates vary over time while reconviction rates do not. The U.S. Department of Justice's Bureau of Justice Statistics conducted and compared studies of releases in 1983 and in 1994. The 3-year reconviction rate for state inmates remained constant at about 47%, but the 3-year rearrest rate increased from 62.5% to 67.5%. This likely resulted more from changes in policing (i.e., higher arrest probability) rather than in court processing (i.e., lower conviction probability). There is reason to doubt that rearrest is a reliable indicator of reoffending by state prison inmates.
Unlike some inmate recidivism studies, this report does not use recommitment to prison for technical violation of supervision conditions as a recidivism measure for several reasons. Technical violations result from the actions probation officers, judges, and the Parole Commission take in response to inmates' misbehavior. An in-depth study of decision-making by agents in the supervision system is beyond the scope of this report. Further, though 36.1% of released inmates have some post-release supervision, they comprise less than 4% of all offenders on supervision. This measure would be important if the results were needed to calculate all future prison costs imposed by released inmates, but the focus of this report is on repeat offending by released inmates.
Though technical violation is not used as a separate recidivism measure in this report, it does affect the rate calculation for inmates who have post-release supervision. When inmates on supervision after release commit a new crime, they are often returned to prison for a technical violation before they are convicted of the new crime. For both recidivism measures in this report, inmates who return to prison for a technical violation of supervision are counted as recidivists if they are convicted of a new offense that occurred before they returned for the technical violation. For information about how technical violation affects rates in this report, see Follow-Up Period.
For more information on these recidivism measures, see the Technical Appendix. Post-release supervision is one of several factors that reduce inmate recidivism. For information about these factors that influence recidivism rates, see Factors Affecting Rates.