Recidivism, Race, and Reentry Services among Local Jail Inmates

Background: During the last decade, the number of men held in local jails has greatly increased to reach 661,329 in 2006, indicating that 457 in every 100,000 male residents in the U.S. are under the custody of local jails (Sabol et al., 2007). Almost 60% of jail inmates are from racial minority groups, 69% of whom are blacks, unchanged from 1996 (Sabol et al., 2007; James, 2004). Across age groups black men were between 5.7 and 8.5 times more likely to be incarcerated in prison or jail than white men (Sabol et al., 2007). Furthermore, recidivism rate among prisoners is known racially disparate. Blacks are more likely to be rearrested, reconvicted, and reincarcerated than whites (Bonczar, 2003; Kansal, 2005; La Vigne et al., 2004; Langan & Levin, 2002; Sabol et al., 2007). However, recidivism studies among jail inmates, either involving race as a factor or not, are extremely rare (Yamatani, 2008), let alone studies of recidivism in context of reentry services evaluation. This study examines overall recidivism rates among jail inmates released in 2003 from Allegheny County Jail (ACJ in the following) in Pennsylvania. Racial comparison was made (1) overall as well as (2) in a sub-sample of ACJ reentry services beneficiaries. Overall, racial disparity in recidivism was expected, but the disparity was expected to decrease among services beneficiaries.

Methods: The sample, generated from ACJ's historical inmate datasets, consisted of ACJ inmates released in 2003 and followed up for 3 years since release (N=12,553). The sub-sample of ACJ reentry services beneficiaries was generated from face-to-face interviews with former inmates followed up for a year after release in 2005 (N=140). Data were analyzed using survival analysis: Kaplan-Meier Estimation and Cox Proportional Hazard Model.

Results: Results showed that overall recidivism rate within 3 years was 52.2%: blacks 60.4% and whites 44.9%. Survival analysis of Kaplan-Meier Estimation presented that blacks were likely to return to ACJ in a shorter time than whites (p<.001). The result that blacks return to ACJ in a shorter time than whites persisted even after controlling for ‘age at release' and ‘length of stay at ACJ' by Cox Proportional Hazard Model (p<.001). No racial difference in recidivism was found, however, among ACJ reentry services beneficiaries, indicating the racial equity impact of the services.

Implications: Recidivism rate among jail inmates carries significant implications in public safety not only because prisoners tend to pass through local jails before and/or after serving time in state or federal prisons but also because offenders sentenced to a year or less are most likely to be incarcerated in and be released from local jails. More recidivism studies among local jail inmates with risk and/or protective factors incorporated are necessary. The fact that the ACJ reentry services eliminated the racial disparity in recidivism suggests that jail reentry services to assist jail inmates' adjustment after release can serve an important role in decreasing or eradicating racial disparity in recidivism as well as promoting public safety.