Partisan Bias in Bipartisan Places?

Abstract

Abstract An extensive literature examines how partisanship divides public opinion on hot-button political issues, but little is known about its potential to polarize attitudes on bipartisan issues. Recent work shows that while Americans hold strong preferences for bipartisanship, their attitudes toward bipartisan issues quickly become polarized when associated with partisan identities. While prior research has examined the effect of these associations in lab settings, tests outside the lab are far more rare. In this research note, we aim to provide such a test by leveraging a bipartisan issue that became associated with partisan identities suddenly in 2018: the presidential alert. While the presidential alert—a product of bipartisan efforts to improve the government's capacity to send emergency communications in the wake of Hurricane Katrina—received little notice when signed into law, it gained widespread media attention during its inaugural test in 2018. We rapidly recruited a sample of US adults immediately before the alert was sent, such that participants in our study received the alert on their phones while completing a survey. We exploited the timing of the alert to randomize whether respondents answered questions about the alert moments before or after receiving it. Across two experiments we find little evidence that associating the alert with the Trump administration had any polarizing effect on attitudes, even when explicitly associated with a partisan cue, suggesting that at least some bipartisan attitudes are not as easily polarized as prior work implies.