Date of Award

Summer 7-1-2022

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Public Administration


Heflin, Colleen


Child Welfare, Foster Care, Public Policy

Subject Categories

Public Affairs, Public Policy and Public Administration | Social and Behavioral Sciences


This dissertation is comprised of three essays which examine policy relevant issues in the child welfare system. In the United States the goal of the child welfare system is to protect the well-being of children by ensuring their safety and promoting permanency. Children typically become involved with the child welfare system due to an allegation of neglect or child maltreatment. If the allegation is substantiated, the government can take custody of the children, remove them from their homes, and place them in foster care. All three of these papers examine foster care outcomes. Removing a child from their family is one of the most intrusive government interventions in the lives of citizens. As such, research examining governmental policies in the child welfare sphere is of the utmost importance. The first two papers in this dissertation evaluate the impact of a policy change on foster care outcomes. The first paper examines how a change in federal fiscal policy, which provides reimbursement to states for payments made to guardians, influences how foster children exit care. The second looks to see if the enactment of a supply-side drug intervention, the reformulation of OxyContin, affects the number of children in foster care. The third paper shows the extent of a problem identified by policymakers, children transferring schools when they enter care, and provides insights into which children are most likely to experience educational instability.

In Chapter 1, I evaluate the effect of becoming eligible for guardianship assistance, monthly payments paid to caregivers who agree to permanently look after the child in their care, on foster care exits. The primary goal of guardianship assistance is to allow for more children to exit care, specifically to guardianship. I leverage the staggered implementation of the Federal Guardianship Assistance Program (GAP) using a difference-in-difference approach and a novel estimator. I find that after becoming eligible for guardianship assistance, children are five times as likely to exit to guardianship (though they were very unlikely to exit to guardianship prior to GAP.) However, I find no evidence that becoming eligible for guardianship assistance changes the overall probability a child exits from care, suggesting that guardianship assistance primarily changes the composition of exit types.

Chapter 2 examines the unintended consequences of the OxyContin reformulation on child welfare utilization by exploiting cross-state variation in rates of prereformulation OxyContin misuse. A growing literature suggests that after the reformulation, a subset of people with opioid use disorder transitioned from using prescription opioids to using heroin. If these individuals were caregivers of children, there would be a higher need for child welfare services, such as foster care. Although foster care case and entry rates increased after the reformulation, I find that it was not due to an increase in Child Protective Services (CPS) inquiries, such as reports of maltreatment. These findings suggest that though children were investigated by CPS at similar rates after the reformulation, more children were removed from their homes due to higher true or perceived risk. The unintended consequences of the reformulation were not limited to the drug users themselves but also affected their children and the state agencies that serve families and children.

Chapter 3 explores educational instability among children entering foster care. Federal legislation, such as the 2008 Fostering Connections to Success and Increasing Adoptions Act (FCA), highlights policymakers' interest in promoting educational stability for foster youth by aiming to keep children in the same school when they enter foster care. Using administrative data from a large and diverse state, I examine the incidence of instability among children entering foster care as well as which characteristics are associated with a higher risk of experiencing it. I find that 3 in 10 children entering foster care experience a nonpromotional school transfer and about 1 in 4 children change school districts. My analysis provides evidence that the child's race/ethnicity, age at entry, disability status, reason for removal, initial case goal, initial placement type, Title I school status, and their urbanicity are all associated with educational instability.


Open Access