Now that the two major parties have all but selected their nominees for president, it’s a good time to ask who is supporting Donald Trump — and how they differ from who is supporting Hillary Clinton. Which issues and beliefs does each group hold especially dear — and therefore, what will the United States likely be debating in the coming months?
We find that this election will be driven by many of the same things as our previous two elections: party identification, ideology and racial resentment.
Here’s how we came to that conclusion. We used data from an Amazon Mechanical Turk survey of 1,543 U.S. adults collected in April 2016 and the American National Election Studies 2016 Pilot Study. While the MTurk sample is not perfectly representative of the U.S. adult population, the ANES sample is representative, and both surveys reveal nearly identical relationships.
We took the difference between “feeling thermometer” survey questions that asked respondents to rate how “hot” (positive) or “cold” (negative) their feelings were toward the candidate on a 0-100 scale. In the graphs below, someone who registered the highest level of support toward Trump and the lowest toward Clinton would be positioned at the “Trump” marker on the vertical axis. Just the opposite is true in the case of perfect positive support for Clinton and perfect lack of support for Trump.
We measured populism using a combination of three questions used by Eric Oliver and Wendy Rahn in a previous Monkey Cage post. Respondents were asked how much they agreed or disagreed with the following statements:
- It doesn’t really matter who you vote for because the rich control both political parties.
- Politics usually boils down to a struggle between the people and the powerful.
- The system is stacked against people like me.
We measured authoritarianism by combining a set of commonly used questions asking respondents to reveal which characteristics are most desirable in a child:
- Respect for elders or independence
- Good manners or curiosity
- Obedience or self-reliance
- Well-behaved or considerate
We measured racial resentment with four specific questions asking respondents about the extent to which they agreed or disagreed with the following statements:
- Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.
- Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.
- Over the past few years, blacks have gotten less than they deserve.
- It’s really a matter of some people not trying hard enough; if blacks would only try harder, they could be just as well off as whites.
Finally, partisan and ideological identification are two scales that range from strong/extremely Democrat/liberal to strong/extremely Republican/conservative, with independent/moderate directly in between the two extremes.
The first set of graphs offers no big surprise: Being strongly identified with one party or ideology is tightly connected with support for one candidate or the other. Stronger Republicans and more extreme conservatives are more likely to support Trump and less likely to support Clinton. Unlike previous work that looked at how Trump supporters greatly differed from supporters of his fellow Republicans, authoritarianism doesn’t seem to predict much of the difference in support for Trump and Clinton. Trump supporters were shown to be more authoritarian than at least Marco Rubio and John Kasich supporters. But strong authoritarians only slightly favor Trump over Clinton.
That’s true for populism as well: Trump supporters are only slightly more populist than Clinton supporters. All these relationships and non-relationships hold in a multivariate setting controlling for the effects of other variables.
On the other hand, levels of racial resentment are just as strongly correlated with supporting one candidate or the other as identifying with one party or ideology over another. High levels of racial resentment are correlated with supporting Trump; low levels of racial resentment are associated with supporting Clinton.
His supporters apparently feel the same way. Indeed, racist Google searches are one of the strongest predictors that someone will support Trump.
The power of racism was apparent in the past two elections as well. In the graphs below you can see the relationships between racial resentment, authoritarianism, and partisan and ideological self-identifications and differential support for both Romney/Obama in 2012, and McCain/Obama in 2008. These data come from the 2012 and 2008 American National Election Studies, respectively.
These graphs show us another way that the Trump/Clinton matchup should be like the contest between Barack Obama and either John McCain or Mitt Romney. In those two previous elections, as well, partisans or ideological adherents supported one or the other, quite strongly. Then, too, authoritarianism didn’t seem to matter much in the elections. And most important, racial resentment correlates just as strongly with support for either Romney/Obama and McCain/Obama as it does with support for either Trump/Clinton.
That may seem surprising. In 2008 and 2012, the U.S. election included Obama, the first presidential candidate of color from one of the two major parties. That’s not true now — but the racial resentment associated with Obama’s historic presidency remains with us.
With Trump in the race, the Democratic candidate will be inheriting a highly racialized political environment, in which the power of racism is nearly the same as that of ideology and partisanship even though both major candidates in the general election will be white.
Trump might not be the great uniter he says he is, or that the GOP now needs him to be. But his followers’ attitudes toward sociopolitical opposition and toward their perceived enemies — attitudes driven by resentment of people from other races — appears consistent with supporters of “establishment” Republican Party candidates for the past two elections, if not before.