Essays on the Economics of Child Welfare and Institutional Design

Abstract

There is a longstanding interest in economics in mitigating inequality of opportunity through early childhood interventions. Foster care is one of the most far-reaching early childhood interventions with significant impacts on children and families. Despite the high prevalence of family contact with the child welfare system in the United States, economists have a relatively limited understanding of the system compared to other major early childhood interventions. This gap is particularly striking in the context of the opioid epidemic, the worst drug overdose crisis in U.S. history, which has heightened child maltreatment risk through its strong association with parental substance use. This dissertation studies the impacts of the opioid epidemic and drug policies on child welfare, examines how Child Protective Services (CPS) responded to the epidemic, and explores how the design of the child welfare system can be improved to better serve vulnerable children.Chapter 2 examines the impacts of the opioid epidemic on child maltreatment and how CPS responded to the epidemic. I use the reformulation of OxyContin, a nationwide supply shock in prescription opioids that exacerbated the opioid epidemic, as a natural experiment. Leveraging variation in states' and counties' exposure to this intervention, I find that the opioid epidemic led to an increase in maltreatment allegations, suggesting a rise in the underlying maltreatment risk. Despite heightened risk, institutional responses did not scale protective actions accordingly, resulting in more cases where children were left at home but subsequently reinvestigated for maltreatment. Foster care placement rates remained unchanged and administrative expenditures per allegation declined, pointing to a limited systemic response. This study underscores the intergenerational consequences of the opioid epidemic and highlights how public health crises can weaken protective systems, particularly when institutions lack the flexibility or capacity to adapt.Chapter 3 investigates the effects of Prescription Drug Monitoring Programs (PDMPs) on child maltreatment, distinguishing between two types that vary in policy intensity: modern and mandatory PDMPs. Leveraging the staggered rollout of PDMPs across states, I find that modern PDMPs, which are fully electronic and integrated into clinical workflows, did not yield statistically significant effects on maltreatment allegations. In contrast, mandatory PDMPs, which require prescribers to consult the database, led to an 18 percent increase in maltreatment allegations. These findings suggest that the unintended consequences of prescription opioid control policies can extend to child welfare, operating through their effects on parental behavior.Chapter 4, coauthored with Jason Baron, Richard Lombardo, Joseph Ryan, and Quitze Valenzuela-Stookey, studies how to allocate tasks among agents in the context of assigning CPS investigators to maltreatment cases, where investigators decide whether to place children in foster care. Given concerns about investigator burnout and turnover, a key constraint is ensuring that investigators are not worse off under the new system. We develop a framework that combines an identification strategy for estimating investigator performance with novel mechanism-design results that elicit investigator preferences and allocate cases to reduce prediction mistakes while respecting constraints and investigator welfare. Simulations suggest that the proposed mechanism could reduce false positives (unnecessary foster care placements) by up to 14 percent while also decreasing false negatives (missed maltreatment cases) and overall placements.