A crisis in my community? Local-level awareness of the opioid epidemic and political consequences.

The opioid epidemic has had a profound effect on American public health, and studies suggest it has had a profound effect on American politics as well. Research suggests a relationship between the severity of the opioid crisis in a community and aggregate-level political behavior, including voting for Donald Trump in 2016, but there is very little individual-level data establishing the mechanism for this relationship. Using a question on a nationally representative survey fielded after the 2018 election, we explore Americans' perceptions of the severity of the opioid epidemic in their community, and whether these perceptions predict political behavior and attitudes. We find that relative estimates of local opioid overdoses are significantly influenced by the actual county-level overdose death rate, and that the relationship between reality and perception is strongest for the most knowledgeable and informed citizens. There is also evidence that the social construction of the opioid epidemic as a crisis primarily affecting White suburban or rural Americans affects these perceptions, as both White and non-urban respondents are significantly more attentive to objective mortality rates than others. Turning to political behavior, we find no evidence that these estimates affected validated voter turnout or U.S. House vote choice in the 2018 midterm election. Assessing public approval outcomes, we found a nonlinear effect of these estimates on support for President Trump, a pattern which is partially consistent with the existing literature. We also observed that those who believed the opioid epidemic to be worse than average in their communities reported lower approval of their state legislatures, though we found no such relationship for ratings of governors. These findings raise important questions about the micro-level mechanisms that link aggregate-level public health with political outcomes.