Populist anger, Donald Trump, and the 2016 election

ABSTRACT

The election of Donald Trump as president in 2016 arguably represents the most unanticipated election outcome in contemporary American politics. Using three nationally representative surveys of the American electorate, this article analyzes the predictors of candidate support during the 2016 presidential campaign. At issue in the analysis is the question of whether Trump’s popularity was tied primarily to ideology and principled support for Trump’s issue positions or whether instead it was inspired by emotions and populist anger at the federal government. The results of the analysis inform our understanding of candidate support in two important ways. First, the results show that individuals’ policy preferences on issues, particularly immigration, played an influential role in shaping their attitudes toward Republican and Democratic candidates during the presidential campaign. Second, the results show that while support for Trump was strongly associated with anger toward the federal government, such sentiment failed to increase support for any other candidates in either party. Taken together, these results suggest that Trump was able to leverage populist anger for political advantage in ways that other candidates were not and that such anti-elitism played an important and underappreciated role in explaining the outcome of the presidential election.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 All three surveys were sponsored by the Pew Center for the People & the Press and were conducted by Abt SRBI. The population under study is non-institutionalized adults who are 18 or older and reside in the United States. A probability sample was drawn using both landline and cellular random digit dialing. For the October 2015 survey, the cooperation rates (AAPOR COOP) were 17.0% and 19.4% for the landline and cellular samples respectively. For the March 2016 survey, the cooperation rates were 17.1% and 21.6%. For the October 2016 survey, the cooperation rates were 17.3% and 18.0%. Within each survey, some questions were asked to random half samples.

2 Full question wording and frequencies are reported in the appendix.

3 This instrument more closely reflects the concept of political populism rather than the concept of economic populism (see e.g. Oliver and Rahn 2016).

4 An anti-immigration index was created for the March 2016 survey using the two available instruments.

5 An economic anxiety index was created for the March 2016 survey using the two available instruments.

6 Given the cross-sectional nature of the design, it is, of course, impossible to make strong claims about populist anger causing candidate support. Fortunately, however, the longitudinal evidence presented in Table 2 provides some assurance that neither candidate support nor the 2016 campaign is causing populist anger in the electorate.

7 In the Democratic model, the predicted decline in support for Clinton due to anger is matched by a 17 percentage point increase in the likelihood of respondents preferring “none” of the Democratic candidates.