Assessed Juvenile Dependency Court Decision Making and Children's Child Welfare Outcomes

Background: Juvenile dependency court judges play a central role in determining the disposition and timing of children's permanency outcomes. Dependency court judges are charged with reviewing placement decisions, establishing permanency plan goals, determining whether reasonable efforts are made to help parents meet those goals, and deciding whether parental rights should be terminated. Although studies have examined the case-level predictors of juvenile court outcomes, the potential influence of judicial-level factors has yet to be rigorously examined. Nevertheless, a better understanding of the relationships of these factors to children's outcomes could inform efforts to improve the effectiveness of dependency courts and address disparities in the timing and disposition of permanency outcomes. The current study links administrative and survey data to examine the relationships among a measure of judicial quality and the disposition and timing of children's juvenile court outcomes.

Methods: The study sample includes approximately 14,100 children who were in protective custody between May 2012 and January 2014 under the supervision of 37 distinct juvenile court circuits in Georgia and Washington State. Data sources include state administrative databases, 2010 Census data, and a survey of 208 attorneys representing children within these jurisdictions. This survey contained a series of questions about the quality of juvenile court judicial decision-making within attorneys' respective jurisdictions. Attorneys were asked whether judges in their jurisdictions make sound legal decisions, have clear understandings of child welfare law, and understand the service needs of children. Based on multilevel confirmatory factor analysis (MLCFA), a set of plausible values of jurisdiction-level "judicial quality" were estimated. Using multilevel discrete-time hazard models, the respective relationships among case-level characteristics, judicial quality, and the rates of various juvenile court milestones (e.g., disposition, TPR, exit to permanency) were examined.

Results: The results of the MLCFA indicate that the reliability of the jurisdiction-level latent variable for judicial quality is high (.952). Based on the intraclass correlation coefficients (ICC) from the unconditional multilevel hazard models of court milestones, the percentage of the overall variability in outcomes attributable to the jurisdiction level is generally modest (<10%). The one exception is the transition between TPR and adoption, which has an jurisdiction-level ICC ranging from 0.15 to 0.30. After controlling for child demographics, assessed judicial quality is found to be negatively associated with exit to reunification, but is not found to be related to any other milestone. However, several other jurisdiction-level characteristics, including the child poverty and per capita foster care entry rates, are found to be significantly associated with several milestones.

Implications: Although there appears to be significant variation across jurisdictions in assessed judicial quality, the findings of the hazard models suggest that most of the variation in court outcomes is attributable to the child level. These findings underscore two important issues. First, a primary objective of dependency court judges is to protect due process rights. Thus, higher quality judicial decision making may actually run contrary to the realization of timely court outcomes. Second, there may be inherent limitations on the ability of judicial decision making to affect child outcomes.