Race, Gender, and the Differing Effects of Out-of-Home Placement Patterns in the Child Welfare System on Juvenile Justice System Entry

Background and Purpose:

            Research consistently demonstrates a relationship between child welfare system (CWS) involvement and subsequent juvenile justice system (JJS) contact (Jonson Reid, 2002; Widom, 1989). Over 400,000 children under the age of 18 experience out-of-home placement (OOHP) in the CWS (Child Welfare Information Gateway [CWIG], 2013). There is a continuum of placement options in the CWS, from least restrictive (e.g., kinship care, foster homes) to more restrictive (e.g., group homes, residential facilities). Best practice in the CWS is to place children in the least restrictive setting (CWIG, 2012). Children often move through multiple CWS OOHPs (e.g., Jonson Reid & Barth, 2000). Most research focuses on OOHP outcomes within the CWS, but not how those patterns may be related to JJS contact. This study sought to answer: 1) What are the OOHP patterns of youth in the CWS across time?, 2) Which OOHP pattern is most predictive of JJS entry?, 3) How do these differ based on gender and race? We hypothesized most youth would follow linear placement patterns from less restrictive to more restrictive settings (Research Question; RQ1), this pattern would be most predictive of JJS entry (RQ2) and, given the dearth in research, no hypothesis was proposed for RQ3.

Methods:

            Using linked, state-level administrative data, the sample comprised children and youth with an OOHP exit from the CWS between 2008-2014 (N = 1,555). Two race categories were included, Caucasian and African American. On a continuum of six categories, the restrictiveness level for each CWS OOHP after the initial placement was averaged to indicate the following placement patterns: neutral, decreasing (more restrictive→less restrictive), or increasing (less restrictive→more restrictive). The outcome measure was a dichotomous variable of whether a youth received a JJS referral. We employed binary logistic regressions to examine main effects, test for an interaction between gender and race, and perform separate analyses for gender and race, respectively.

Results:

            The neutral placement pattern group comprised 45.5% of the sample, 29.8% were decreasers, and 24.6% increasers. Youth with increasingly restrictive OOHPs were significantly more likely to have JJS contact. Gender (p < .001) and race (p < .01) were significant in the full model, but the interaction was not. By gender, placement patterns were not significant predictors of JJS entry. By race, for Caucasians, those with increasing placement patterns were approaching significance (p = .055), but no significant differences were found for placement patterns for African Americans.

Conclusions and Implications:

            This study supports that some OOHP patterns within the CWS have an impact on JJS entry. In the CWS, research implications for practice and policy include informing placement decisions on the front-end of youths’ OOHP experiences. On the back-end of the CWS, knowing youths’ OOHP escalating history suggests flagging them for specialized/aftercare services for JJS prevention. Front-end early intervention in the JJS for CWS youth is needed that acknowledges OOHP histories. Special attention needs to be given regarding race and OOHPs.