The Boston Reentry Initiative (BRI) is an interagency public safety initiative that helps adult offenders who pose the greatest risks of committing violent crime when released from jail transition back to their neighborhoods. The goal of BRI is to reduce recidivism among recently released high-risk violent offenders by providing mentoring services, case management, social service assistance, and vocational development to program participants.
Each month, the Boston Police Department (BPD) selects 15 to 20 high-risk inmates committed to Suffolk County House of Correction (the local jail) to participate in the BRI. The initiative targets male inmates between the ages of 17 and 30 who reside in Boston and are considered by law enforcement to be at high risk for continuing their involvement in violent crime following release from jail.
Both subjective and objective criteria are used during the selection process. The objective criteria include consideration of an offender’s age, sex, current conviction offense, and past criminal arrest history. The BPD’s Boston Regional Intelligence Center examines the list of offenders entering the Suffolk County House of Correction and makes subjective recommendations about who should be enrolled in the program based on several factors, including gang membership, criminal history, likelihood to recidivate, and expectation that the inmate will return to a community with high rates of violent crime.
Within 45 days of entering the House of Correction, program participants attend a BRI panel session. The session includes representatives from criminal justice agencies, and social service providers who present inmates with information about the program from the unique perspective of his or her organization. Representatives from these providers educate BRI participants about the institutional programs and community resources available both in prison and postrelease that can help their reintegration back into the community. Representatives from prosecution, probation, and parole departments also inform participants that they will be held accountable for staying away from further criminal activity upon release or they will face serious consequences if they are caught committing crime upon their return to their neighborhoods.
The initiative is voluntary, so inmates must choose to participate. If inmates choose to participate, following the panel session, program participants are assigned a case manager who begins working and meeting with inmates immediately. A “transition accountability plan” helps to set out a coordinated regimen of treatment and supervision beginning at the House of Correction and continuing after release. The plan includes an array of wraparound services customized to address participants’ individual needs. Services address immediate issues, such as identification/drivers licenses, health insurance, shelter, transportation, or an interim job, as well as long-term issues, such as substance abuse treatment, mental health treatment, education, or permanent housing.
On the day of release, the facility arranges for either a family member or a case manager to meet the individual at the door. After release, participants are encouraged to continue to work with their case manager and social service providers while reentering the community. For those inmates who leave jail on conditional supervision, such as probation, the supervising agency usually incorporates participation in BRI as part of their stipulation of release. Participants can receive supervision and support services through a local day reporting center. The case managers provide support to offenders for up to 18 months post release.
The BRI was implemented by the BPD in partnership with the Suffolk County Sheriff’s Department (SCSD). The BPD and SCSD developed partnerships with other law enforcement and criminal justice agencies such as the State Department of Probation, the State Department of Correction, the State Parole Board, the Suffolk County District Attorney’s office, and the U.S. Attorney’s office. These partnerships helped the BPD and SCSD to identify the most serious offenders, provide coordinated postrelease supervision whenever possible, and vigorously prosecute BRI–identified inmates who recommit offenses.
The initiative also involves the collaborative efforts of social service providers. The State Department of Revenue, Youth Options Unlimited Boston, Dorchester Bay Economic Development Center, and Whittier Street Health Center are the main social service partners in the BRI. Many of the services provided to BRI participants are made available by social service providers, including local one-stop career centers, health commissions, community colleges, and halfway houses.
Arrests for Any Crimes
The first analysis by Braga, Piehl, and Hureau (2009) found consistently and significantly lower failure rates for the participants of the Boston Reentry Initiative (BRI) relative to the control group, though the differences between the two groups narrowed somewhat over time.
· At 1 year postrelease, 36.1 percent of BRI participants had been arrested for a new crime, compared to 51.1 percent of the control group.
· After 2 years, 67.6 percent of BRI participants had been arrested for a new crime, while 78 percent of the comparison group had been arrested.
· After 3 years, 77.8 percent of BRI participants had been arrested compared to 87.7 percent of the control group.
Arrests for Violent Crimes
Arrests for violent crimes presented a similar pattern over the study time period.
· After 2 years, 20.4 percent of BRI participants had been arrested for a violent crime compared to 34.6 percent of the control group.
· After 3 years, 27.8 percent of BRI participants had been arrested while 39.2 percent of the control group had been arrested for a violent crime. Even after 3 years, BRI participants were still 30 percent less likely to have been rearrested for a violent crime.
The second analysis that looked at the effects of treatment on time to failure also found statistically significant differences between the treatment and control groups. The BRI treatment was associated with a statistically significant 32.1 percent reduction in the subsequent overall arrest hazard rate. The BRI treatment was also associated with a significant 37.1 percent reduction in subsequent violent arrest hazard rate.
The study by Braga, Piehl, and Hureau (2009) used a quasi-experimental design to compare recidivism rates among Boston Reentry Initiative (BRI) participants to an equivalent control group that did not receive reentry services. Propensity score matching techniques were used to attempt to create equivalent treatment and control groups: relevant pretreatment characteristics of each subject were summarized into a single-index variable (the propensity score); subjects in the treatment and control group were then matched, based on values of the single-index variable. The matching technique resulted in a treatment group of 108 offenders (who participated in BRI in 2002) to a comparison group of 309 offenders (who were committed to the House of Correction in 2002, prior to the existence of BRI).
Statistical tests revealed no significant differences between the groups. The average age of the treatment group was 21.3 years, and participants were 97.2 percent non-white. The control group was 98.1 percent non-white, with an average age of 22.1 years.
Recidivism was measured as arrests for any crime and arrests for violent crime following release from jail. Data was collected from the Massachusetts Criminal Offender Record Information (CORI) records on prior offending and subsequent recidivism. CORI criminal history arrest data was examined up to 3 years (1,095 days) immediately following release from jail on the current conviction offenses. Arrest events, the types of arrest crimes, and the time in days between release and a subsequent arrest event were recorded for all subjects.
To assess the effects of BRI on time to recidivism for program participants released to the community, the authors used simple Kaplan–Meier group comparisons of survival times and Cox proportional hazard models.