Association for Public Policy Analysis & Management

Substance use can have damaging effects on families, including reducing parents’ ability to safely care for their children. Increases in child abuse and neglect can result in children being removed from home and placed in foster care. After years of decline, foster care caseloads has been rising over the past five years. Between 2011 and 2015, the number of children in care rose from 397,000 to 428,000, an 8 percent increase. This increase has caused concern among child welfare practitioners and policy makers, and parental substance use is often cited as the cause. In particular, media reports and child welfare administrators describe the growing opioid epidemic as a major contributor to increases in caseloads.

Despite the anecdotes, there is no nationwide empirical evidence to support this assumption. This study provides the most rigorous evidence yet of a link between substance use and child welfare. We identify a positive, causal relationship between county-level indicators of substance use prevalence and indicators of child welfare caseloads, including reports of child abuse and neglect, substantiated reports, and foster care placements. Using ten years of data and instrumental variables methods, we find that a 5 percent increase in death rates due to drug overdose – the median annual increase for counties – is associated with a 1 percent increase in reports and substantiated reports, and a 2 percent increase in foster care placements. Similarly, a 4 percent increase in prescription opioid sales is associated with a 2 percent increase in reports, a 3 percent increase in substantiated reports, and a 4 percent increase in foster care placements. Our models take account of various potentially mediating factors, including population and urbanicity, racial/ethnic composition, income levels, measures of local practice, and state fixed effects. Robustness checks and falsification tests support the strength of the findings.

While the positive relationship supports widespread beliefs on the role of parental substance use in child abuse and neglect, how community prevalence of substance use may cause increases in child welfare caseloads is not well understood. We suggest three mechanisms: 1) substance use prevalence is an indicator of parental substance use, which can lead to child abuse and neglect; 2) reporters of abuse and neglect, caseworkers, and courts respond differently to cases in areas where substance use is prevalent; 3) local policies shift to reflect changing substance use prevalence. Our analysis cannot definitely point to the mechanism, but the differential effect of substance use prevalence on reporting, substantiation, and foster care placement suggest that differences in practice and policies play an important role.

Finally, we discuss preliminary findings from an ongoing qualitative study exploring these proposed mechanisms and how the use of different substances affects caseloads. The study also examines strategies communities use to mitigate the effects of parental substance use on caseloads, in particular through the incorporation of substance use treatment. This qualitative study is based on interviews and focus groups at ten sites in different parts of the country, conducted over the spring and summer of 2017.