Abstract
I study how public health policy and the social safety net shape drug use, access to treatment, and drug-related mortality in the United States. Across three chapters, I explore how Medicaid expansions and SNAP benefit generosity impact fatal drug overdoses and treatment for substance use disorder (SUD), highlighting both intended and unintended consequences.
In the first chapter, I examine the effects of the Affordable Care Act (ACA) Medicaid expansion on fatal drug overdoses and access to treatment. Using a difference-in-differences framework that accounts for staggered adoption and treatment effect heterogeneity, I find that overdose mortality increased following the expansion. At the same time, Medicaid expansion led to substantial increases in the use of medication-assisted treatment (MAT), including claims for Suboxone and Naltrexone. Treatment admissions rose most among new patients and those with extensive treatment histories, suggesting that improved access benefited both groups. The results highlight the complex relationship between expanding access to care and downstream health outcomes.
In the second chapter, I study how the ACA’s 2014 provision allowing states to expand Medicaid coverage for certain controlled prescription drugs – specifically benzodiazepines – affected overdose mortality. Benzodiazepines are rarely fatal on their own but often co-ingested with opioids in overdose deaths. I find that states that expanded both Medicaid eligibility and drug coverage experienced significantly higher overdose death rates. The effects are most pronounced among men and individuals aged 30–64, and remain robust after accounting for other drug-related policies. These findings suggest that broader drug coverage, while improving access, may also introduce new risks.
In the third chapter, I examine how income support through the social safety net affects fatal drug overdose rates. Specifically, I study the relationship between changes in Supplemental Nutrition Assistance Program (SNAP) benefit levels and overdose mortality. SNAP is the largest federal nutrition assistance program in the U.S. and a central component of economic support for low-income households. I exploit cross-state variation in how SNAP benefits are calculated and use a border-county design to compare overdose outcomes across adjacent counties facing different benefit levels. During periods of benefit increases, I find evidence of a decline in fatal overdose rates. However, this pattern does not reverse during periods of benefit decreases, where I do not observe a corresponding rise in overdose mortality. Notably, the subsequent fall in benefits is smaller than the preceding rise, lending further credence to the idea that benefit increases may help individuals reach a level of economic stability that buffers them against future reductions. This asymmetric pattern – where a period of decline follows a period of improvement – suggests a possible threshold effect in staying off drugs, whereby once a certain level of support is reached, individuals may be less vulnerable to relapse even if benefits later decrease. The overall effect is predominantly driven by males and remains robust to alternative definitions of the benefit increase period and broader definitions of local economic areas. These findings suggest that the generosity of income support programs can play a meaningful role in shaping drug-related mortality.
