Reading for Life and Adolescent Re-Arrest: Evaluating a Unique Juvenile Diversion Program

Background:Contact with the justice system in adolescence carries lifelong consequences that may include stunted educational attainment rates, decreased job stability, and adverse substance abuse and mental health outcomes.  However, existing literature demonstrates that even among youth with a high probability of continuing criminal behavior, positive life events, or turning points, may intervene.  Juvenile diversion programs that provide youth a way to bypass adjudication and/or punishment within the criminal justice system could serve as a positive turning point that leads to a lasting trajectory change.  This study presents results of an evaluation of Reading for Life (RFL), a diversion program for nonviolent juvenile offenders in a medium-sized Midwestern county. The unique program uses philosophical virtue theory-infused literature and small mentoring groups to foster moral development in juvenile offenders.

Methods:The intervention is a randomized control trial.  Between June, 2010 and December, 2013, eligible offenders in this county were referred by their probation officers to the diversion program, where they were randomly assigned to participate in either the RFL program (the treatment group, n=194) or to 25 hours of community service (the control group, n=214).  Data on juvenile arrests are obtained from the county Juvenile Justice Center.  Recidivism impacts are measured via a regression-adjusted difference in means model and the impacts on arrest counts are measured via a negative binomial model.

Results:In the full sample, models show an 8.7 percentage point reduction in the probability of re-arrest (p= 0.066) for RFL participants, which is a 23 percent reduction from the control group mean of 0.379.  There is suggestive but imprecise evidence that the program reduces misdemeanor arrests (p=0.189), but by the two-year follow-up, misdemeanor arrests fell by a statistically significant 10.5 percentage points. Results also suggest that RFL is especially effective at reducing the chance of arrest for more serious offenses and the effect is most heavily concentrated in felony prosecutions.  The results for any prosecuted offenses and any prosecuted felonies are much larger for males than females, where analyses reveal statistically significant reductions of 68 and 85 percent, respectively.  For both white, non-Hispanics and non-white participants, results show large reductions in prosecuted felonies after one year with a 10.6 (p=0.002) and 9.7 percentage point (p=0.041) reduction for whites and non-whites, respectively. Among non-whites, for all offenses, RFL reduces recidivism rates by 18.5 percentage points, or 84 percent of the control group mean (p = .001).

Implications: These results suggest that participation in RFL greatly reduces the propensity to recidivate.  The impact is especially large for more serious offenses and for participants with observed characteristics that would predict a greater likelihood to recidivate (e.g., males, non-whites, participants from lower income families). Because RFL has a number of unique features: the focus on virtue theory, the use of literature to highlight these virtues, and the use of trained volunteer mentors, phase II of this study will feature a qualitative data collection effort that will isolate and explore the causal mechanisms that explain the program’s success.