NCJ Number
251949
Date Published
August 2018
Length
6 pages
Annotation
This study analyzed trends in recidivism in the United States, using an administrative dataset of prison admissions and releases.
Abstract
The federal Bureau of Justice Statistics (BJS) collects data voluntarily submitted by state departments of corrections and parole, with at least 38 states providing some data going back to 2000, with additional states providing data in more recent years. This National Corrections Reporting Program (NCRP) includes records from 2000 to 2015 that link prison admission and release records for states with reliable prison inmate identifiers. This creates a panel dataset that enables analysts to track an individual's terms of prison admissions and releases over time. This term file is the dataset used in this analysis. Only 23 states provided data for the entire 2005-15 period, so this study was limited to these states. The states included in the analysis are listed in this report. Recidivism rates were analyzed by identifying whether an individual released in 2005, 2010, or 2012 had a new admission to prison after release, along with the year of this admission. Because publicly accessible data do not include the month or day of admission or release, this report bases recidivism on year of admission and year of release. Admissions in the same year as a release are considered a return to prison. The rate of return-to-prison recidivism for the 2005 cohort was 30.4 percent after 12 months, 43.3 percent after 24 months, 49.7 percent after 36 months, 52.9 percent after 48 months, and 55.1 percent after 60 months. Recidivism data are shown by selected characteristics of prisoners released from 23 states in 2005 and 2012. These characteristics are original offense (violent, property, drugs, public order, and other/unspecified/missing); gender; and age at release year. 2 tables