The Effects of National Disasters on Stress and Substance Use in the United States

The relationship between national disasters, risk perceptions, stress, and substance abuse has never been comprehensively studied. The recent national disasters of the Oklahoma City bombing, the 9/11 terrorist attacks, and Hurricane Katrina provides several natural experiments to study these relationships. Behavioral Risk Factor Surveillance System data collected before and after the national disasters is used for measures of stress and substance abuse. This dissertation helps to advance the literature on the self-medication of stress with substance abuse in line with the Becker-Murphy rational addiction utility framework. The first essay of this dissertation, “Terrorism, Stress, and Smoking using Behavioral Risk Factor Surveillance System Data,” uses a regression discontinuity design to investigate the impact of the Oklahoma City bombing and 9/11 on stress and smoking. 9/11 is also used as an instrument for stress to measure the unbiased effect of stress on smoking. The second essay, “The Effects of 9/11 on High Risk Alcohol Consumption,” explores the impact of 9/11 on high risk alcohol consumption, also using 9/11 as an instrument to explore the unbiased effect of stress. In both essays, differential impacts of education and county-level characteristics such as distance to the terrorist attacks, population density, and military participation are investigated. Net costs of substance abuse attributed to 9/11 are calculated. The third essay, “The National Effects of Hurricane Katrina on Risk Perception and Substance Abuse,” investigates the impact of Hurricane Katrina on stress and substance abuse for residents at-risk of future hurricanes compared to residents that are not, using difference-in-difference analysis. Results suggest that explored disasters increased smoking and high-risk alcohol consumption, and terrorism caused increases in stress.