Methods: Two of the papers examine questions related to the decision to place a child in residential care. The first of these uses propensity score matching to compare the behavioral profiles and outcomes of children placed in group homes to those placed in family-based care. The second involves an ethnographic case study of one child welfare/mental health county system and examines the decision processes leading to placement. Another paper in the symposium uses latent class analysis to identify and characterize the modal placement histories of adolescents entering residential care for the first time. This paper also compares the characteristics and residential care outcomes of adolescents experiencing different placement histories. The final paper presents an agent-based model of residential care utilization. Using empirically-based estimates of the rules governing the decisions of residential care provides and other "agents," this paper describes how simulation that can be used to explore how changes in the operating context affect the overall level of residential care utilization and the supply of beds.
Results: Results from the papers examining the decision to place children in residential care suggest that decisions may often be driven by factors other than children's clinical needs. In brief, a comparison of the behavioral profiles of children placed in group homes and family-based care reveal a striking equivalence between these two groups of children. The ethnography found that decisions about residential care placement are based on many considerations, some of which have little to do with the clinical needs of the child. Findings from the study of adolescents' placement histories reveal the existence of between 4 and 6 modal placement pathways preceding placement in residential care. Placement pathways are also found to be related to adolescents' characteristics and discharge outcomes. Finally, the simulation results illustrate how bed supply and provider preferences interact with utilization to shape the complex dynamic that drives whether beds are available when needed.
Conclusions: Residential care utilization may be only partially explained by the clinical needs of children. To the extent to which non-clinical influences drive residential care decisions, the system as a whole will be less effective and efficient in its efforts to serve children, Thus, efforts to reform the child welfare system would be well-served by identifying and minimizing these influences.