Study
Hassoun Ayoub and Pooler (2015) conducted a randomized controlled trial to examine the impact of the Harlem Parole Reentry Court on recidivism. The study took place in Upper Manhattan, in New York City, between 2010 and 2012. Parolees were considered eligible if they lived in Upper Manhattan; however, parolees diagnosed with Axis I mental health disorders, persons who have committed sex offenses, and parolees convicted of arson were excluded from participating because the reentry court could not properly address their needs.
A total of 504 parolees were randomly assigned to either the treatment group, who participated in the Harlem Parole Reentry Court (n = 213), or to the control group, who participated in traditional parole (n = 291). The total study sample was predominately male (97.2 percent) and with an average age of 30 years. Most were Black (69 percent) or Hispanic (30 percent), and the remaining 1 percent was white. The majority of participants were born in the United States (93 percent). The average participant had four prior custodial sentences, including an average of two in prison. Most had been arrested on felony charges (91 percent), drug charges (84 percent), drug felonies (73 percent), or violent felonies (67 percent) at least once. The treatment and control groups differed on 4 of 51 baseline measures. Parolees in the treatment group were more likely to have been 1) arrested, 2) convicted of a violent felony offense, 3) born in the United States, and 4) released in a different year. The study authors conducted preliminary logistic regression models to examine the impact of these variables and determined they were unlikely to influence the resulting impact findings; therefore, they did not make adjustments to account for these differences.
The outcomes of interest were the number or rearrests, number of reconvictions, number of parole revocations (failure to comply with the conditions of parole), number of months employed, the receipt of a high school diploma or GED, and rate of substance use. Official records were collected from the New York State Division of Criminal Justice Services (DCJS). Substance abuse was measured using the Texas Christian University (TCU) Drug Screen, which produced two scores, a drug score and a measure of severity. All measures were collected at a 12- and 18-month follow-up. Logistic regression models were conducted to determine statistically significant differences in outcomes. The study did not conduct subgroup analyses.