Kicked Out or Dropped Out? Disaggregating the Effects of Community-based Treatment Attrition on Juvenile Recidivism

Abstract

Little is known about the effects of treatment attrition on juvenile recidivism, particularly with regard to community-based treatment. Compounding this gap in the literature is a lack of studies that have disaggregated recidivism offense type or type of treatment attrition. These distinctions are believed to have both theoretical implications for understanding recidivism and practical implications for engaging youths in community-based programs. The current study analyzes a sample of 5,517 male juvenile offenders adjudicated within Philadelphia between 1996 and 2002 to determine the effects of treatment attrition, either due to dropping out or expulsion, on juvenile recidivism, distinguished in terms of property, violent, or drug offenses. Results indicate that voluntarily dropping out of treatment significantly increases the likelihood of recidivism through drug and property offending, while expulsion from treatment significantly increases the likelihood of a violent recidivism offense, suggesting that the causes of treatment attrition and reoffending may be related.

Keywords:

• juvenile recidivism
• treatment noncompletion

Notes

1. From 1996 through 2002, juvenile offenders were sent to thirty-one different community-based treatment programs as a result of their adjudication in the Philadelphia Family Court.

2. Juvenile offenders with and without treatment completion data were compared with Χ² tests on demographics, including age, race, and gender. Results of these tests indicated that no significant differences on those demographic variables exist between those who do and do not have treatment completion information. Thus, the removal of the 731 young offenders without this data is not believed to bias the remaining population analyzed, at least for the demographic variables compared.

3. The mean number of clients per year for these seven programs is 8.14.

4. For a detailed description of the weighting process, see (Garcia, Taylor, & Lawton, 2007).

5. These boundaries were created to represent distinct neighborhoods by those familiar with the city of Philadelphia to provide for more accurate spatial units compared to politically drawn boundaries, such as those provided by the US Census.

6. Black youth comprise the reference category.

7. An additional index representing social capital was also constructed to control for the effects of protective factors within the community. Consistent with prior operationalizations of this concept in research on crime (Messner, Baumer, & Rosenfeld, 2004; Sampson, Morenoff, & Earls, 1999; Sampson, Raudenbush, & Earls, 1997), this measure combined four items from the 2002 and 2004 iterations of the PHMC HHS using principal components factor analysis: the number of local groups and organizations that residents participated in, willingness to help neighbors with household-related chores, and responses (on a scale of 1–5) to two statements, “I feel that I belong and are part of my community” and “Most people in my neighborhood can be trusted.” This scale item displayed a high level of reliability in 2002 (Cronbach’s alpha = 0.785) and 2004 (0.868), with the values for each year aggregated to the 45 PHMC neighborhoods and, finally averaged to create a mean value. Social capital was not included in the subsequent models due to its close, reciprocal relationship with neighborhood disadvantage. Prior studies with these data have shown that neighborhood disadvantage and social capital exert similar, yet opposite, effects on particular juvenile justice outcomes (Grunwald et al., 2010; Lockwood, 2012); social capital has been found to decrease the likelihood of treatment noncompletion and recidivism, while increases in neighborhood disadvantage are associated with an increased likelihood that juvenile offenders will fail to complete treatment and reoffend. To confirm the relationship between these neighborhood-level processes, identical models to those reported in the “Results” section of this article were estimated with social capital in the place of neighborhood disadvantage. The results of these models show that social capital influences recidivism in a similar, but opposite manner, compared to neighborhood disadvantage. As a result, these models were not reported in this manuscript.

8. At the individual-level, 1,770 of the juvenile offenders, or approximately 32 percent of the sample, attended programs designated as aftercare treatment. 1,021, or nearly 19% of the sample attended a school-based treatment program. Relatively few of the juveniles (n = 52) attended treatment that was tailored specifically for drug and alcohol abuse or that was primarily considered counseling-based (n = 231). The remaining youths who comprise the reference category attended either day treatment (n = 921) or mentoring (n = 1584) programs.

9. The additional 26% of juveniles who failed to complete treatment did so due to several reasons unrelated to either dropout or expulsion: “juvenile not appropriate for program” (6%), “juvenile appropriate but not responding to program” (12%), “reasons unrelated to youth or program” (4%), and “other” (4%).

10. The examination of 95% confidence intervals corresponding to the odds ratios reported in the subsequent cross-classified hierarchical models also lend support for the lack of multicollinearity amongst the predictors analyzed in those models.

Additional information

Notes on contributors

Brian Lockwood

Brian Lockwood is an assistant professor in the Department of Criminal Justice at Monmouth University. His research interests include juvenile delinquency, applications of GIS, and the spatial correlates of offending. His work has appeared in The Journal of Research in Crime & Delinquency, The Professional Geographer, and Youth & Adolescence.

Philip W. Harris

Philip Harris is an associate professor of Criminal Justice at Temple University. His scholarly work focuses on juvenile justice policy, juvenile corrections, and program evaluation. His recent publications have appeared in Cityscape, the Journal of Adolescence, The Journal of Research in Crime and Delinquency, and the Journal of Juvenile Justice.