Testing an Integrative Theory of Beliefs and Emotions Predicting Trump Support

• In Event: The Causes and Consequences of the Rise of Populist Sentiments

Abstract

What accounts for the intense devotion of Donald Trump’s supporters? Pettigrew (2017) has cautioned against single factor explanations. Using data from the 2016 Cooperative Congressional Election Study, we tested a theory of attachment to strongly-held belief systems that integrates multiple influential variables, such as partisanship (e.g., Jacobson 2017), racism (e.g., Schaffner et al. 2018), and authoritarianism (e.g., Ludeke et al. 2018) within a narrative structure that includes 5 belief components and related emotions:

The evaluative component (describing alternative outcomes differing in desirability) was measured by 4 items assessing outcomes of the 2016 election as described by Trump: opposing abortion, gun control, gay rights, and (what he asserted was) a deteriorating economy (α=.63);

The behavioral component (encompassing actions said to determine which outcomes will occur) by 3 items: deporting illegal immigrants, repealing the ACA, and opposing EPA regulation (α=.63);

The identificational component (specifying supporters and opponents who take the actions and experience the outcomes) by a 3-item Identification-With-Trump index, comprising Republican partisanship, conservative ideology, and the perception Trump cares about people like you (α=.78); and a 7-item Identification-Against-Blacks index, comprising items on racism and racial resentment (α=.79);

The normative component (beliefs relevant to morality) by the 4-item ANES authoritarianism scale (α=.64);

The explanatory component (arguments and evidence showing the belief system is true) by an item asking how qualified Trump was to be president.

Emotions were measured by ANES-format questions on how often respondents felt enthusiasm, hope, pride, fear, anger, and contempt when thinking about Trump[Clinton] because of the kind of person he[she] is or something he[she] has done.

According to the theory, strongly-held belief systems tend to incorporate the belief components in a dramatic story: our opponents (because of bad motives or inferior abilities) are taking unjustified actions that will lead to ruinous outcomes, unless we act; but if we do what we should, the threats can be averted and desirable outcomes attained. Belief strength may be intensified by emotions, such as fear of the threats; anger and contempt toward enemies; hope for victory; and affection and pride toward leaders who would help attain it.

Analyses. We ran simultaneous regressions to assess whether each of these belief system components and emotions contributed to predicting evaluations of Trump and vote choice. For these analyses we dichotomized each variable as absent (scores up to its midpoint) vs. present (above the midpoint).

Predicting Trump evaluations. An OLS regression predicting Trump thermometer ratings from the 5 belief components found that each accounted for significant variance. Largest influence came from the identificational component (in particular, identifying with Trump, ß=.37), the explanatory component (believing Trump was qualified, ß=.29), and the behavioral component (endorsing actions Trump would take, ß=.21; all ps<.001). The adjusted R2 for this model was .66.

Adding emotions to this model found significant variance contributed by hope (ß=.11, p=.001) and enthusiasm (ß=.08, p=.021) elicited by Trump, and contempt elicited by Clinton (ß=.06, p=.036). Betas for the belief components were somewhat reduced by the inclusion of emotions, though all but one remained significant (ps<.001). With emotions included, significance of the normative component (authoritarianism) was marginal. This model accounted for 70% of the variance in thermometer support for Trump.

Predicting vote choice. A logistic regression predicting the two-party vote from belief components and emotions found similar results (Nagelkerke R2=.88). Voting for Trump (against Clinton) was predicted by endorsing Trump-advocated actions, believing he was qualified, identifying with him, and endorsing outcomes he favored. Voting for Trump was also predicted by contempt elicited by Clinton; voting for Clinton, by fear elicited by Trump.

Voting for Clinton was also predicted by enthusiasm elicited when thinking about Trump. This could reflect ambiguity in ANES-format affect questions, which do not specify the object of elicited emotions (Crowell et al. 2019). For example, thinking about the kind of person Trump is could have inspired enthusiasm for Clinton.

These results provide significant support for the integrative theory of strongly-held beliefs, and for the importance (in the 2016 race) of hope, enthusiasm, contempt, and fear. Support for Trump was predicted not by just one factor, but by beliefs and emotions that may be elements of a story in which Trump, caring about his supporters, would take strong actions opposing immigration, Obamacare, government regulation, abortion, gun control, and gay rights, and reversing economic decline—all of which, as a successful businessman, he could accomplish.

Authors

• Ira Roseman, Rutgers University, CCAS

• Kyle Mattes, Florida International University

• David P. Redlawsk, University of Delaware