Program Goals/Target Population
Community and Law Enforcement Resources Together (ComALERT) is a reentry program in
Parolees are eligible to participate in ComALERT if they are:
· Paroled to
· At least 18 years old
· In need of substance abuse treatment
Parolees are not eligible if they are a convicted sex offender or arsonist or suffer from a serious and persistent mental illness. Most program participants are on parole for either a drug crime or a crime of violence, such as robbery, assault, and homicide.
ComALERT services begin for parolees almost immediately upon release from prison. An inmate released from prison has 24 to 48 hours from release to report to the New York State Division of Parole, which is the primary source of program referrals. The parole officer may decide to refer a parolee to the division’s
After the assessment, parolees are directed to the
In addition to receiving substance abuse treatment, ComALERT participants are also referred to the “Ready, Willing, & Able” (RWA) program, which provides transitional employment opportunities, transitional housing, vocational training, and 12-step meetings, as well as courses on financial management and other life skills over a 9-month period. The program offers full-time manual labor work through the Community Improvement Project, a jobs program run by the RWA. RWA participants begin to immediately receive a weekly stipend for their work, while a portion of the money they earn is put into a mandatory savings account. Through case management and mentoring services, as well as vocational training in fields such as pest control, commercial driving, and food service, the program also provides assistance in finding a permanent job. The RWA program further provides courses to prepare for the General Educational Development test (or GED), computer/literacy classes, and individual tutoring. Participants receive meals and other services at the Doe Fund facility (RWA’s parent organization), and can also temporarily stay at a living facility (called the Stuyvesant House) if they are in need of housing. The RWA program has a zero-tolerance policy on substance abuse. Participants are administered random drug tests, and if they test positive, they are discharged from the program.
The final component of the ComALERT program is service referral. Program participants are referred to a variety of service providers, such as organizations that provide housing, vocational programs, and job readiness services. Participants can also be referred to organizations that help them enroll in government programs, such as Medicaid.
ComALERT staff works closely with the Division of Parole to monitor participants throughout the time they remain in the program. If a participant is in noncompliance with program requirements, the parole officer is informed of the violation. A law enforcement sanction, such as parole revocation, can be used at the discretion of the parole officer. For less serious violations, other sanctions, such as more frequent drug testing, can also be used. In order to graduate from the ComALERT program, participants must be drug-free for 3 months and either employed or in school.
The Kings County District Attorney’s office began the ComALERT program in 1999 as a way to bring together District Attorney staff, community-based organizations, and service providers. The collaboration also includes the Counseling Service of Eastern District New York (or CSEDNY), the New York State Division of Parole, and the Doe’s Fund, Inc., which runs the RWA program.
The ComALERT program does not require participants to have medical insurance (uninsured participants receive treatment that is paid for by a grant from the New York State Office of Alcohol and Substance Abuse Services). This allows all participants to begin receiving substance abuse treatment almost immediately after release from prison.
18 to 100
Jacobs and Western (2007) examined the recidivism rates between Community and Law Enforcement Resources Together (ComALERT) participants and a matched control group, and found statistically significant and nonsignificant treatment effects on several outcome measures.
The results of the life tables showed that ComALERT attendees were less likely to be rearrested than the matched control group members. During the first 2 years after release from prison, 29.3 percent of ComALERT graduates and 39.2 percent of ComALERT attendees were rearrested, compared to 47.6 percent of control group members, a significant difference. Attendees were 18 percent less likely to be rearrested and graduates were 39 percent less likely to be rearrested than controls.
The hazard regression analysis (or multivaried analysis) found that, controlling for race, age, gender, and criminal history, ComALERT attendees, regardless of graduate status, had a significantly lower risk of being rearrested. Attendees had a 22.9 percent lower risk of being rearrested than matched control group members. When differentiating between ComALERT graduates and discharges (i.e., noncompleters), program graduates had a 45.7 percent lower risk of being rearrested than matched control group members, a statistically significant difference after 2 years. However, ComALERT discharges did not differ significantly from controls, meaning there was no evidence that the two groups differed in the likelihood of rearrest.
ComALERT attendees and graduates were also less likely to be reconvicted. After 2 years, 27.8 percent of ComALERT attendees had been reconvicted, compared to 34.2 percent of control group members. Attendees were 19 percent less likely to be reconvicted than controls, although this difference did not reach statistical significance. About 19 percent of ComALERT graduates were reconvicted, compared to 34.2 percent of control group members. Compared to controls, graduates were 45 percent less likely to be reconvicted within 2 years.
The hazard regression analysis found that after 2 years, ComALERT attendees had a 22.7 percent lower risk of being reconvicted than matched control group members. ComALERT graduates had a 50.7 percent lower risk of being reconvicted on a new charge of any severity than did matched controls, a significant difference. There was no significant difference in the risk of reconviction between discharges and controls, however.
Reincarceration by Parole Violation
The reincarceration rates of ComALERT attendees for parole violations did not differ significantly from matched control group members throughout the 2-year period. However, ComALERT graduates were significantly less likely than attendees and controls to return to prison due to a parole violation. After 2 years, 15.7 percent of ComALERT graduates had been reincarcerated, compared to 24.5 percent of all ComALERT attendees and 23.8 percent of control group members.
The results of the hazard regression analysis showed there was no significant difference between ComALERT participants and matched control group members, suggesting that attending the program did not affect a parolee’s likelihood of returning to prison due to a parole violation. When the analysis differentiates between ComALERT graduates and discharge, the results showed graduates had a significantly lower risk of reincarceration, while discharges had a significantly higher risk of reincarceration by parole violation compared to control group members. Graduates had a 44 percent lower risk, while discharges had an 83.5 percent higher risk of returning to prison.
Reincarceration by New Sentence
The probability of being reincarcerated for a new sentence was low for all three groups. After 2 years, 2.8 percent of ComALERT graduates, 4.3 percent of all ComALERT attendees, and 6.5 percent of matched controls had been sent back to prison on a new sentence. None of these differences are statistically significant, however.
The hazard regression analysis found that ComALERT attendees were no less likely than control group members to return to prison on a new sentence. Although ComALERT graduates did have a 66.7 percent lower risk of reincarceration, program discharges had no significant difference in risk of a new sentence to prison compared to controls.
Since most of the reincarceration experienced by study participants occurred because of parole violations, the total reincarceration results closely resemble the reincarceration by parole violation results. ComALERT graduates were significantly less likely to be reincarcerated than ComALERT attendees and control group members. However, attendees as a whole did not have a significant difference in reincarceration rates from the matched controls. After 2 years, 18.4 percent of graduates were returned to prison for any reason, compared to 28.7 percent of all attendees and 29.9 percent of controls.
The result of the hazard regression analysis found there was no significant difference between ComALERT attendees and matched control group members on the risk of returning to prison for any reason. Distinguishing between ComALERT graduates and discharges, the analysis found that graduates had a significantly lower risk of reincarceration (49.3 percent lower) and discharges had a significantly higher risk (73.5 percent) than controls.
Employment and Earning
The analysis looked at three estimates of group differences of employment: the unadjusted estimate, the regression adjusted estimate, and the matching estimates. All three methods found that the ComALERT survey respondents had an employment rate that was about 50 percent higher than the control group. ComALERT participants also earned more money each week than those working in the control group. The analysis also took into account the effect of the Ready, Willing, & Able (RWA) program, which provides employment opportunities to ComALERT participants. RWA participants had a 90 percent employment rate, which exceeds the ComALERT average by 15 points and exceeds the employment rate for control group members by nearly 70 points.
Co-Residence and Contact With Children
All three estimates of the group differences found that the rates of marriage and cohabitation were not significantly different between ComALERT survey respondents and control group survey respondents. The rates differed by only a few percentage points. ComALERT parents were slightly more involved with their children than control group parents, but the difference was not statistically significant.
Drug and Alcohol Use
Again, all three estimates of group differences found no statistically significant difference in self-reported drug and alcohol use in the 6 months before incarceration and in the 30 days prior to the survey interview.