Ostermann (2009) examined the ways in which two community programs in New Jersey affected the recidivism rates of parolees reentering the community. The two programs were Day Reporting Centers (DRCs), also known as Community Resource Centers (CRCs), and the Halfway Back (HWB) program. Three measurements for recidivism were used: rearrest, reconviction, and reincarceration (except for parole violations). Rearrest was measured if study participants were arrested for a new crime after their release date. Reconviction was counted if participants were found to be guilty of one of their charges. Reincarceration was measured if participants served a custodial term in either prison or jail after their release. Time spent in the community was calculated by finding the difference in days between the study participant’s date of release from prison and the date of data collection (May 15, 2007). This allowed for a 3-year follow-up period. Data was gathered from multiple sources including the New Jersey State Parole Board, the New Jersey Department of Corrections, and the New Jersey State Police. Recidivism was determined by analyzing the study participant’s criminal case history and Interstate Identification Index.
The study sample included all individuals released from New Jersey Department of Corrections in 2004. A total sample of 714 participants made up four groups:
- Offenders who maxed out their prison sentence and did not receive any community supervision following release from prison (n=200)
- Offenders who were paroled but did not participate in any community programs (n=198)
- Offenders who were paroled to a DRC on release (n=135)
- Offenders who were paroled to a HWB program on release (n=181)
This CrimeSolutions.gov review focused on the comparison between offenders who were paroled to a CRC on release and the offenders who maxed out their prison sentences and did not receive any community supervision following release from prison. Control variables used were age, gender, race, number of previous arrests, and type of crime for which they were incarcerated and then released in 2004. The sample's average age was 35 years old and was 93 percent male, 67 percent African American, 17 percent white, and 16 percent Hispanic. Fifty-three percent of study participants were incarcerated for drug related offenses, 26 percent for property offenses, and 21 percent for violent offenses. No significant differences were found among groups.
The study used Chi-square tests and analysis of variance (ANOVA) analyses for variance to establish differences between the groups in relation to control variables. Sequential logistic regression was used to examine the impact of program membership on recidivism outcomes, including control variables. Further analyses were run by program types. Kaplan–Meier survival analyses were conducted to measure the differences in time to rearrest, with Cox-proportional hazards tests used to predict recidivism with control variables.