Opioids and Elections

• In Event: New Developments in Economic Voting

Abstract

It is well-documented that voters punish incumbent politicians for poor economic conditions, crime, war deaths, and education outcomes. We extend this literature and test whether voters punish politicians for opioid usage and overdoses in their local communities. The opioid epidemic in the U.S. is real – and getting worse. According to the CDC, about 30 people died of an opioid drug overdose per 100,000 in 2019. In 1999, that number was less than 10 per 100,000. While studies in economics have sought to explain opioid usage, and how different public policies may slow or weaken the crisis, no study to date considers the political consequences of the opioid epidemic.

To do so, we link county-level elections and turnout data going back to 2000 to county-level data on the number of pills shipped to each county from 2006-2014, as well as CDC fatality and overdose data dating back to the 1990s. To test the causal effects of opioid usage and deaths on elections, we employ a series of generalized difference-in-difference models. Throughout, we also pay particular attention to heterogenous effects across different communities (e.g., rural vs. urban), and how the presence or absence of state policies designed to address the crisis might explain relationships between opioids and elections. Ultimately, our paper breaks new ground in a growing literature on the opioid crisis, and more broadly, offers new evidence on when, why, and how voters punish incumbent politicians for the world around them.

Authors

• Ryan Baxter-King

• Justin de Benedictis-Kessner, Harvard University

• Brian T. Hamel, Louisiana State University

• Michael Hankinson, George Washington University

• John Holbein, University of Virginia