Journal on Policy & Complex Systems Volume 3, Issue 2 - Page 108

Policy and Complex Systems - Volume 3 Number 2 - Fall 2017 Modeling Complexity in Human Built Systems: New Approaches, New Findings in Foster Care Fred Wulczyn A , John Halloran B Abstract Foster care is a widely utilized intervention to provide care to children who cannot live with their parents. The substantial portion of contemporary academic work on foster care either uses individual-level analyses, or focuses on linear analysis of aggregate annual statistics. This paper argues that a complexity-based perspective will deepen our understanding of how foster care policy operates. Wulczyn (1996) argued that a class of population growth models focused on latent resource limitations held great potential to elucidate behavior within aggregate foster care dynamics. Specifically, partial adjustment models propose that population growth rates and limits are dependent on prior resource and capacity states within a system. Recent advancements in data availability, methodology, and computation have made that theory testable. Analyzing data from the Foster Care Data Archive representing 81,142 child entries into out-of-home placements from 2000 through 2014, we apply Empirical Dynamic Modeling to identify nonlinearity and detect causal relationships in coupled time series (Sugihara et al., 2012). Findings indicate nonlinear casual relationships in the coupled foster care entry and exit time series data. These findings match expectations of strength and directionality. The empirical findings and the models that they support have the potential to frame child welfare systems differently, ultimately leading to different actionable policy conclusions. Keywords: child maltreatment, child protection, human-built systems, systems theory A B Chapin Hall Center for Children at the University of Chicago Lewis University doi: 10.18278/jpcs.3.2.7 105