Dissertation Title: "Essays on the Economics of Health”

This dissertation consists of three essays on the economics of health.

Chapter 1: This paper studies the effects of shutting down prescribers, dispensers, and distributors that inappropriately handle prescription opioids on local opioid supply and mortality. With competitive supply, theory suggests the effects of closing any single supplier will be offset by substitution. Closing a supplier may have an effect on overall supply, however, if the targeted supplier is more lax with prescriptions than others or if the action has general deterrence effects. To examine enforcement empirically, I exploit differential timing of initial enforcement actions across areas following a federal expansion of enforcement in 2008. I show supplier enforcement actions reduced overall opioid shipments by 20 percent in the average affected county for three years. Results further show that enforcement actions targeting distributors primarily reduced opioid shipments to pharmacies and clinics with suspicious order patterns. Enforcement interventions also reduced opioid-related mortality. In the areas with the largest and most significant interventions, enforcement reduced overdose death rates by 22 percent for a period of five years. Overall, these findings demonstrate a larger role for supply-side drug control strategies to improve health relative to theory and prior literature.

Chapter 2: Between 1990 and 2021, opioid overdose death rates in the U.S. rose nearly continuously. Such a long period of sustained increase goes against conventional theory, which suggests that the public and private policies that respond to harmful drug epidemics should reduce deaths over time. This paper examines why opioid deaths rose so greatly and for so long. We develop and estimate a model of addictive drug use dynamics which incorporates spillovers across people, which we refer to as “thick market externalities.” Thick markets may result from information spillovers across people, changes in the cost of using drugs as more people use them, and positive peer effects in consumption. When goods are addictive and there are spillovers across people, temporary shocks may lead to long-term increases in use. To test for the presence of these spillovers, we estimate our model using data on county opioid overdose death rates linked to geographic distance and peer connectedness, as measured by Facebook friendships. We find evidence of large spillovers in opioid use, particularly through friend relationships but also through geographically close areas. Our estimates imply the majority of opioid deaths in recent years are due to spillovers from historical dissemination of prescription opioids, rather than exogenous factors affecting the present.

Chapter 3: Drawing from new data which links Census and vital records for individuals born in the late 1800s and early 1900s, I study the long-term effects of economic displacement during the Great Depression on mortality. My empirical model relates the probability a person dies at different ages to their probability of being unemployed in 1930 based on the industry they started work in. I control for baseline differences in mortality risk across industries. I find that middle-aged workers with a 10 percentage point higher probability of unemployment during the Great Depression experienced a 1.5 year decrease in life expectancy. Some evidence suggests these increased rates of premature death were due to higher susceptibility to infectious diseases. Prior literature finds that aggregate mortality rates fall during economic downturns, leading people to conclude that recessions are good for health. However, these results demonstrate the longer-term negative health impacts on individuals who are displaced.