Threat or Reassurance? Framing Midterms Results among Latinos and Whites

• In Event: Race and Identity in Political Behavior

Abstract

Results of the 2018 midterm election have been described as a Democratic wave, with the Democratic Party winning back control of the U.S. House of Representatives after gaining 40 seats. The Republican Party, however, maintained firm control of the U.S. Senate, gaining two seats. Depending on one’s perspective and their information environment, these results could represent a threat or reassurance about the future direction of the country.

In this paper, we investigate the effects of framing the 2018 midterm results as a win for the Democrats or for the Republicans, and test how these contrasting frames affect emotional and psychological distress responses on a national sample of Latino and white registered voters surveyed about 3 weeks before and again 3 weeks after the 2018 midterm election. In an online survey experiment, subjects were randomized into conditions that informed them of the Democratic majority in the House, the Republican majority in the Senate, or provided no information in a control condition.

Subjects were then asked about their sense of threat or reassurance, emotional responses to the election outcome, and mental health responses. Initial results show that Latinos informed of the Democratic majority in the House display lower levels of psychological distress and greater belief that life as a Latino in the U.S. will become less difficult, compared to those in a control condition that received no information. Conversely, Latinos informed of the Republican majority in the Senate display higher levels of psychological distress, and are more likely to have a pessimistic outlook about life as a Latino in the U.S. Whites are also found to believe life as a Latino will improve under House Democratic control, and worsen under Senate Republican control. These results provide causal evidence of electoral effects on the psychological and life outcomes of a salient identity group frequently targeted by both Democratic and Republican campaigns.

In addition, the survey asked questions on racial/ethnic identity strength, political news media consumption frequency and source, and exposure to the campaign environment that we will test as moderators of our treatment effects. Related, because subjects were asked about their candidate preferences before the election and about their vote choice afterwards, we can test whether subject candidate preferences also moderates our treatment effects. We have begun collecting these auxiliary data and look forward to sharing our results.

Authors

• Ali A. Valenzuela, Princeton University

• Kassra AR Oskooii, University of Delaware

• Loren Collingwood, UC Riverside