Public Opinion in the Opioid Crisis: Essays on Salience, Policy, and Attribution

Abstract

The opioid crisis continues to be a major public health crisis in the United States. From 1999-2018, opioid overdoses claimed the lives of nearly 450,000 Americans. By 2015, drug overdose deaths became the leading cause of injury-related death in the United States, with the crisis contributing to overall declines in life expectancy. Despite bipartisan efforts by policymakers across the country, overdose mortality remains high, and many public health experts insist that bolder action is needed. In this context, my dissertation seeks to understand how public opinion has responded to the rise in drug overdose mortality. Specifically, I explore how local context, personal experience with addiction, attribution for the opioid crisis, and stigma predict perceptions of salience and attitudes about different policy solutions. First, in chapter one, I study how direct and indirect experiences with addiction influence beliefs about how serious of a problem the opioid crisis is and how much political priority it should receive. Using a variety of surveys and overdose mortality data from the Centers for Disease Control and Prevention, I find that state and county overdose mortality rates are associated with greater perceptions of problem severity and, to a lesser extent, demand for more attention to the problem. Knowing someone with addiction is a stronger predictor for perceiving the opioid crisis to be a serious problem and worth public attention. I also find that national media attention to addiction, captured by the Vanderbilt Television News Archive, moderates these relationships. Local context and personal experience with addiction are more connected to the salience of the opioid crisis when the topic is frequently covered in the national media. Overall, I show that exposure to overdose mortality can generate political salience, while national media plays a key role in facilitating public attention to the opioid crisis. Next, in chapter two, I use the state of Massachusetts as a case study to understand how different measures of the extent of the opioid crisis influence public opinion about marijuana legalization. The harms from opioid addiction may raise the salience of potential risks of legalization, such as claims that the policy could increase access to a ``gateway drug" that leads young people to use more dangerous drugs, like heroin. I show how these arguments play a prominent role in public discourse around marijuana legalization in areas affected by the opioid crisis. Then, I use municipal-level opioid overdose mortality and Emergency Medical Services data to predict support for Ballot Question 4 in Massachusetts in 2016, which legalized marijuana. I find that municipal overdose mortality is associated with increased opposition to legalization but is unrelated to support for other ballot measures or the presidential contest. I also use a 2012 ballot measure legalizing medical marijuana in the state to construct a panel fixed-effects model, where I find that within-unit increases in overdose mortality between 2012 and 2016 predict sharper declines in support for marijuana expansion ballot questions. These findings shed light on how local exposure to the harms of a public health crisis can shift public support away from related policy reforms. Finally, in chapter three, I move from contextual and experiential measures to focus on a different aspect of public opinion: attribution for the opioid crisis as a whole. Americans vary in the extent to which they blame individuals or other actors for drug addiction, and these attributions are related to views about potential policy solutions. I use a survey experiment to test whether providing news articles about doctors and pharmaceutical companies being held accountable for their role in the opioid crisis increases support for public health-oriented policies and decreases support for criminalization. I find that providing information about doctors' role in the opioid crisis increases support for stricter sentences on people who buy or sell opioids illegally, while this prime does not affect support for medication-assisted treatment or safe injection sites. This effect is concentrated among respondents with the highest level of stigma against people with opioid use disorders (OUDs). Priming blame for drug companies, however, does not change support for any policy. In the discussion, I suggest potential explanations for these findings, as well as avenues for potential future research on attribution. Taken together, my dissertation research suggests that public views about opioid addiction are complex. Communities disproportionately affected by drug overdose mortality rely on national media to connect personal experiences to political salience. Salience, however, can sometimes increase risk aversion toward policies that might affect addiction. Moreover, stigma toward people with OUDs remains an important aspect of public opinion about policy solutions, even if individuals are not fully blamed for addiction. As the salience of the opioid crisis fades, this research points to a crucial need for efforts to sustain national attention to drug addiction, while also suggesting a key role for campaigns to reduce stigma and counter perceptions of risk when advocating for new public health strategies.