Juvenile courts play a critical role in the identification of youth with mental health needs and in effecting the delivery of services to them. The long-range goal of this research is to understand better the nature of the relationship between the mental health and juvenile justice systems as it affects the delivery of mental health care to youthful offenders. Specifically, the current applica-tion focuses on the propensity of juvenile courts to refer youth to mental health services or to commit youth to state care, and the factors that influence those decisions. Based on resource-dependence theory, it is proposed that court dispositions are a function of the combined influ-ences of three organizational properties: input (i.e. individual and aggregate characteristics of the youth), organizational structure, and the external environment in which courts operate. To shed light on service-related impli-cations of courts' decisions, service histories of youth placed in state care are also examined. Of particular interest is the degree to which services rendered to youthful offenders differ from those of other custodial youth with similar mental health profiles, and the factors that explain variability in services. Finally, the service profiles of youth who return to state care are examined to shed light on the relationship between prior service use and whether the reasons why youth first enter then return to state care represent a """"""""step up"""""""" or """""""" step down"""""""" in their offense history. Secondary, statewide data on youthful offenders referred to the juvenile courts in Tennessee and on those committed to state care are supplemented by a survey of court officials and county-level data on community resources. To model the determinants of court decisions for referred youth, hierarchical linear modeling is used to improve the estimation of individual-level effects and measure the interaction between higher-order and individual-level factors (i.e., structure and environment). To assess service utilization of youth committed to state care, cluster analysis is used to derive common types of service profiles and common types of youth based, for example, on race and reason for custody. Service history and youth types are cross-classified to identify significant associations between the two measures. A repeated measures analysis with categoric dependent and independent variables is used to address the significance of prior service history for youth's outcome or change in status upon returning to state care.
