2024 Fall Seminar Series Seminar: Jordan McAllister University of Kentucky

Predicting Child Maltreatment in South Carolina and Kentucky using County-level Indicators

Abstract: Scholars investigating risk factors associated with child maltreatment generally focus on individual- and family-level determinants, leaving a need to better understand broader community-level influences. As Child Protective Services (CPS) is typically organized by county, county-level factors provide important policy implications for improving maltreatment outcomes. Using nine years of compiled child maltreatment, social determinants of health, and Family Advocacy and Support Tool (FAST) data from South Carolina and Kentucky, this study offers a unique glimpse into both county- and family-level effects. The first set of analyses use aggregated county-level factors to predict the number of maltreatment reports or FAST assessments by county, while the second tests the county-level effects on specific types of maltreatment disaggregated by family while controlling for a comprehensive set of family-specific factors. The aggregated analyses point to food insecurity and poor health as the main predictors of maltreatment in Kentucky, versus teen births and child poverty in South Carolina (when using CPS reports)or poor health and housing problems (when using FAST assessments). Then the disaggregated analyses highlight specific county-level effects in South Carolina depending on the type of maltreatment: housing problems and teen births predict family conflict; teen births, food insecurity, and excessive drinking predict family safety issues; mental distress and housing problems predict emotional abuse; mental distress and single-parent households predict physical abuse; food insecurity predicts sexual abuse; and income ratio and housing problems predict neglect. Together, these analyses help identify key county-level issues that need to be addressed in effective policies targeting child maltreatment.