Short-Stays in Child Welfare: Exploring State Level Factors

Background and Purpose: Child welfare research has produced relatively scant knowledge about children who experience very brief placements in out-of-home care (“short-stayers”). These removals typically last less than 30 days, and the utility and necessity of these brief removals remains to be understood. Emerging research suggests that certain child and case characteristics, such as race/ethnicity and maltreatment type, may contribute to likelihood of short-stays, but scarce empirical research has examined the possible state-level factors that contribute to these brief removals. Variation in state demographics and socio-political factors have shown associations with other child welfare outcomes; in this project we explore how these factors may contribute to risk of short-stays, or very brief engagement with child welfare systems.

Methods: We explored state-level factors associated with risk of short-stays using child welfare administrative data on 317,264 children from across the US for FY 2018. Data were primarily drawn from the Adoption and Foster Care Reporting System (AFCARS), and also incorporated state-level data from the American Community Survey (Census Bureau), and Uniform Crime Report (Federal Bureau of Investigations). State-level factors included states’ racial/ethnic composition, urbanicity, population size, regionality, political ideology, food insecurity, SNAP utilization, child poverty rate, crime rate, and police per capita. Children who exited foster care within 30 days of having entered care were compared to children who exited foster care after 30 days of involvement. A multi-level logistic regression model was tested to identify state-level factors associated with children’s risk of experiencing short-stays, controlling for child and case characteristics.

Results: Short-stayers comprised approximately 12.44% of the sample. Approximately 46% of the sample were White, non-Hispanic children, 22% Black, 21% Hispanic (any race), 8% multi-racial, 3% Indigenous, and 1% Asian. The mean age was 9-years-old (SD=5.77). Approximately 49% of the sample experienced poly-maltreatment victimization. The most prevalent removal reasons included neglect (25%), parent substance misuse (9%), and physical abuse (4%). The results of the multi-level logistic regression model showed that after controlling for child and case characteristics, several state-level factors were significantly associated with risk of being a short-stayer. Risk for short-stayer episodes increased as the percent of the state’s children on SNAP increased (OR=1.08, p<.05), and also as police per capita increased (OR=1.42, p<.05). Conversely, risk of being a short-stayer decreased as state food insecurity rate increased (OR=0.88, p<.05).

Discussion: The results of this study contribute to understanding how state-level factors influence child welfare involvement. Particularly, the relationship between police per capita and SNAP utilization and risk of short-stays suggests a connection between increased presence of surveillance of families, resulting in risk of short-stays. That SNAP utilization is associated with greater risk, but food insecurity is not, further supports this interpretation. Research supports the understanding that removals are inherently harmful to children, and the knowledge of systemic factors at the state-level being related to children experiencing these likely unnecessary removals suggests the need for child welfare policy makers to more substantively address the prevalence of short-stays, as well as the broader socio-political factors contributing to them.