Kacie Dragan
Dissertation Title: “Place, Policy, and Health: Essays on the Intersection of Health Policy and Social Policy”
The social safety net is vital to supporting the wellbeing of many people in the United States, but it is complex. In this dissertation, I examine factors influencing outcomes of those insured by Medicaid, and I study relationships between the health system, other social service sectors, and the places in which people live. The goal of this dissertation is to contribute to the body of interdisciplinary research that can inform the design of policies to support health and wellbeing across different social service sectors.
Chapter 1: Health Shocks and Housing Instability among Urban Medicaid Enrollees
In Chapter 1, I examine the relationship between health and housing instability. Using high-frequency data on residential location and health for Medicaid enrollees in New York City from 2010 to 2019, I test whether adverse health events can cause housing instability with an event study design. I find that health shocks generate an immediate increase of 8.8 to 15.3 additional quarterly moves per 1,000 enrollees (a 15-25% relative increase), as well as increases in the risk of living in a shelter or on the street. I find that effects are smaller for those with access to subsidized housing, a usual source of outpatient care, greater social support, and higher quality inpatient care, suggesting potential areas for policy interventions to break the relationship between health problems and housing instability. These results demonstrate the ways in which people experiencing illness are uniquely vulnerable to housing instability.
Chapter 2: Major Traffic Safety Reform and Road Traffic Injuries among Low-Income New York Residents, 2009-2021 (with Sherry A. Glied)
In Chapter 2, we evaluate the impact of major transportation policy reform on the health of Medicaid enrollees. In 2014, New York City launched Vision Zero, a comprehensive traffic safety reform policy that included a citywide speed limit reduction, roadway redesign, and traffic ticketing. We conduct difference-in-differences analyses to measure changes in traffic injuries and expenditures 2009-2021, comparing NYC to surrounding counties without traffic reforms. Vision Zero led to 77.5 fewer injuries per 100,000 person-years annually (a 30% relative reduction). We also observe marked reductions in severe injuries (brain injury, hospitalizations) and savings of $90.8 million in Medicaid expenditures. Effects were largest among Black residents. Impacts were reversed during the COVID-19 period, likely due to decreases in traffic ticketing.
Chapter 3: How Do Neighborhoods Shape Pediatric Mental Health Utilization? Evidence from Within-State Movers in Medicaid (with Timothy J. Layton)
In Chapter 3, we use a “movers” design to estimate causal effects of neighborhoods on pediatric mental health utilization. Neighborhood-level variation in childhood mental health has been widely documented. Yet, it is not clear how much of this variation is driven by compositional differences in the children who live in different areas versus features of these places themselves. Using national Medicaid data, we leverage children who move within their state between 2007 and 2014 with event studies and neighborhood fixed effects models. Upon moving, enrollees’ outpatient therapy use converges 39% of the way toward the destination neighborhood’s mean, acute mental health use converges 28% of the way toward the destination neighborhood’s mean, and psychiatric prescription use converges 8% of the way toward the destination neighborhood’s mean. The neighborhood-specific effects we estimate only modestly correlate with the set of area-level characteristics we tested, like economic disadvantage, supply of providers, or school attributes. As policymakers seek to respond to rising childhood mental health utilization, attention should be paid to neighborhood contexts, even down to the ZIP code level.