Study
Mercado, Kaufman, and Espinoza (2019) used a quasi-experimental design to evaluate the effectiveness of the Early Intervention Diversion Program (EIDP). The study examined outcomes for education, mental health, and recidivism for youths who received the EIDP (the treatment group), compared with youths who did not participate in EIDP (the comparison group). The CrimeSolutions review of this study focused only on the program’s impact on recidivism, because data on the education and mental health outcomes were not collected for the comparison group.
To evaluate the impact of the EIDP on recidivism, arrest data were obtained from the Probation Case Management System and mined by the Probation’s Information Services Bureau. The data consisted of all first-time arrests, the dates of the arrests, and the types and levels of offense. Additionally, the researchers received scores from the Los Angeles Risk and Resilience Checkup (LARRC), which helped in generating a comparison group that was comparable with the treatment group.
To be included in the study analysis the following criteria must have been met: 1) an individual’s first arrest must have occurred between January 2017 and April 2019, 2) the case was not automatically transferred to the District Attorney, 3) the case was unsealed, and 4) at least 6 months of follow-up observational data were available. Propensity score matching was then used to identify a comparison group that closely resembled the treatment group on the following characteristics: gender, ethnicity, age at first arrest, risk level, and scores on the LARRC assessment. This resulted in a sample of 182 cases included in the analysis: 91 treatment cases and 91 comparison cases.
The treatment group and comparison group were predominately male—58 percent and 55 percent, respectively. In terms of ethnicity, nearly 72.5 percent of the treatment group was Hispanic. The remaining youths were 13.2 percent Black, 7.7 percent white, and 6.6 percent Asian. The comparison group had a similar ethnic makeup, with 71.4 percent Hispanic, 17.6 percent Black, 6.6 percent white, and 4.4 percent Asian. Average age of first arrest in the treatment and comparison group was 15.7 and 15.6, respectively. Though there were some slight differences in risk level between the two groups—with an average risk of 1.36 in the treatment group and 1.41 in the comparison group—these differences are minimal and indicate that both groups have low risk levels. The study authors did not specify whether there were any statistically significant differences between the groups on baseline characteristics.
To determine whether the treatment and comparison group had different rearrest rates, Welch t-rests (independent sample t-tests) were conducted. Though the study examined outcomes across various time periods, the CrimeSolutions review of this study focused on the difference in rearrest rates for the treatment and comparison groups within 6 months. The study did not conduct subgroup analyses.