President Trump is known for his antagonistic and controversial comments aimed at political opponents, and the marginalized social groups who support them. From referring to Mexicans as “rapists” and “child smugglers,” to blocking immigrants from “shithole countries” and enacting a Muslim ban, Trump’s insensitive comments on race and ethnicity are now par for the course. Yet, his critics often state that this behavior is not “normal” politics.
But is President Trump actually appealing to motives that differ from “normal” partisan battles? In our preliminary and ongoing research, we find evidence that the appeal of Donald Trump is different from the appeal of the Republican (and, obviously, Democratic) Party.
In particular, support for Trump is characterized more by out-group hatred than by in-group affection. That is, Trump approval is best explained by a disdain for the Democratic groups he so often attacks. Republican Party approval, on the other hand, is best explained by affection towards the groups who actually vote for Republican candidates.
Intolerance: the sordid consequence of social sorting
Certain social groups tend to be Republican (i.e., whites, Christians, rural voters, men, etc.), while others tend to be Democratic (i.e., non-whites, non-Christians, metropolitan voters, women, etc.). Partisan identities are now matched up with ideological, religious, racial, and other cultural and geographical identities. In Lilliana Mason’s recent book, this is referred to as “social sorting,” and it generates an increasingly emotional and visceral dislike of partisan opponents.
The social psychologist Marilynn Brewer, along with a number of her colleagues, repeatedly found that white individuals with a very well-sorted set of national, religious, political, occupational, and recreational identities are significantly less accepting of cultural diversity and affirmative action, and more intolerant of social out-groups including African Americans, Mexican immigrants, Muslims, and gays and lesbians, even when accounting for age, education, and ideology.
This increasing alignment between our partisan and other identities pushes us toward intolerance of out-groups. This is particularly true among the Republican Party, which is made up mostly of voters with fewer, and better-aligned religious and racial identities. We have previously shown that while Democrats come from a variety of racial, ethnic, and religious backgrounds, nearly all Republicans fit the white Christian profile.
Further, as voters’ racial and religious identities align with their party, their partisan identity strengthens. For Republicans, this means that the closer they feel to fellow whites and Christians, the closer they feel toward the Republican Party: an ideal recipe for intolerance of “the other.”
In-group love and out-group hate
However, Brewer also argues that favoring in-group members does not always coincide with hostility toward out-group members. Sometimes a desire for our own group to win motivates us, and other times, we are motivated by a wish for the hated “other” to lose.
Does this in-group love and out-group hatred dynamic exist in current American politics? Are some partisans motivated by affection for the groups in their own party, while others motivated by animosity toward the groups in the opposing party? And, furthermore, do feelings toward Trump himself — often characterized by distinctly hostile rhetoric towards Democratic groups — differ from feelings toward the two parties?
Some well-timed data
Conveniently, the Democracy Fund’s Voter Study Group survey includes a random sample of several thousand US citizens who completed the initial online survey in 2011 and who took it again in 2017. Importantly, respondents in 2011 were not yet familiar with Trump as a political candidate, which means their feelings towards various social groups at that time were not influenced by Trump’s rhetoric or policies.
We can also use these 2011 respondents as a baseline against which we can compare their 2017 survey responses. In other words, those who would become Trump supporters today were (largely) not yet Trump supporters in 2011. This enables us to, in a sense, travel back in time to figure out which pre-Trump feelings towards social groups contribute to later President Trump approval — and identify the types of people who eventually turned to Trump.
In order to do this, we used this Democracy Fund data to predict 2017 Trump approval (as well as Republican and Democratic Party approval) based on 2011 feelings toward a number of party-linked groups, controlling for party and ideological identification in 2011, along with 2011 measures of socio-demographic variables (i.e., political interest, race, religion, educational attainment, gender, age, and income).
We measured feelings toward each group using a “feeling thermometer,” a scale in which individuals can rate how favorable or unfavorable they feel towards certain social groups including: African Americans, Hispanics, Muslims, Christians, whites, and LGBT people. Values on these feeling thermometers range from “least hostile” to “most hostile” feelings. In the graphs below, higher scores on the feeling thermometer mean that people feel more hostile towards that group in 2011.
We predict 2017 approval ratings for Trump, the Republican Party, or the Democratic Party using 2011 feelings toward the specified group, controlling for all of the other variables. Higher approval values represent more approval, and the scores range from 0 to 1.
Why support for Trump is not “normal” party support
We expect that warmer feelings in 2011 toward African-Americans, Hispanics, Muslims, and LGBT people – all groups aligned with the Democratic Party – will predict higher approval of the Democratic Party in 2017, as a form of “in-group love.” Put simply, when you like the people who make up the party, you like the party. As shown in the graphs below, less hostile feelings toward these groups predicts higher support for the Democratic Party.
More importantly, can dislike of these Democratic groups lead to approval for Trump and the GOP?
We find mixed evidence of a link between out-group hatred and Republican Party support. Hostile feelings toward African Americans and Hispanics do not drive future Republican Party support (effects are not statistically significant), but hostile feelings toward Muslims and LGBT people do slightly drive future Republican Party support.
But hostility toward these groups very consistently predicts Trump support. For African-Americans, Hispanics, Muslims, and LGBT people, support for Trump is strongly associated with dislike of these Democratic groups. In every case, the people who felt hostile towards Democratic groups in 2011 are most likely to be Trump supporters today. The same cannot be said of Republican partisans.
The way to see this is to look at the slope of the lines. While the “GOP Approval” and “Dem Approval” lines tend to go downward or stay relatively flat, the “Trump Approval” line consistently goes upward. This indicates that, for African-Americans, Hispanics, Muslims, and LGBT people, support for Trump is strongly associated with great dislike of these Democratic groups. You can also compare levels of approval at the “least hostile” and “most hostile” points of the graphs.
For Democrats, in-group love drives party support. For Republicans, out-group hatred plays a mixed role. But for Trump supporters, out-group hatred wins the day.
Feelings toward Republican-linked groups
The Republican-linked groups with feeling thermometers in 2011 were whites and Christians. For these groups, we expected to see in-group love affecting Republican Party and Trump support. We didn’t know whether dislike of these groups would drive Democratic Party support.
We found that Republican support is powerfully fueled by in-group love. Support for the GOP is strongly related to feelings of warmth toward whites and Christians. On the other side, Democratic approval is not at all related to feelings towards whites or Christians. If anything, hostility to these Republican-aligned groups decreases support for the Democratic Party.
We should note that there is a key difference between Republican and Democratic party-linked groups. Very few African-Americans, Hispanics, Muslims, and LGBT people (the Democratic-linked groups) identify as Republican. However, many whites and Christians (the Republican-linked groups) identify as Democrats. For instance, in the 2018 midterm election, 44 percent of white voters reported voting for a Democratic candidate. Ninety percent of black voters voted for a Democratic candidate, and nearly 80 percent of African-Americans identify as Christian.
This makes it more difficult for Democrats to consider Republican-linked groups to be completely outside of their group. For Republicans, most Democratic party-linked groups truly are social out-groups, in the sense that they tend not to be affiliated with the Republican Party.
Even with these differences, Democratic and Republican Party support is strongly motivated by in-group love. For President Trump, this isn’t quite the case. Trump support is rooted somewhere in between in-group love and out-group hate. His support does increase as feelings toward whites and Christians grow warmer, just not as strongly as the effect for the Republican Party. But more importantly, Trump support is uniquely dependent upon out-group hatred.
What drives Trump support?
A consistent hatred towards Democratic-aligned groups is what differentiates Trump approval from overall approval of the Republican Party. As social sorting promotes both in-group love and out-group hatred, we see that both are used to Trump’s advantage. In contrast, it is mainly in-group member positivity that relates to support for the Republican and Democratic parties.
We emphasize that the feelings towards racial and religious groups shown in our graphs predate Trump’s candidacy. Trump, in other words, did not create out-group animosity. He simply tapped into these existing reservoirs of negative out-group feelings and activated them to his benefit. As Trump’s rhetoric about blacks, Hispanics, and Muslims repeatedly exploited anger towards these groups – much to the chagrin of his own party’s leaders – he solidified support among voters who already held disdain and resentment towards a diversifying American landscape.
This work is not the first to find that Trump support is uniquely driven by out-group hatred. In the new book Identity Crisis, political scientists find that even in the 2016 Republican presidential primary, out-group hatred predicted support for Trump, but not for Cruz, Rubio, or Kasich. This means that support for Trump is distinct from the support that Republicans have traditionally received. Trump was uniquely attracting voters who had already been feeling increasingly uncomfortable with racial, religious, and social diversity. We tend to think of partisans as being generally intolerant of outsiders, but our findings suggest that Trump supporters are unique in terms of their out-group hatred.
This all raises the possibility that the Trump era is ushering in a new “normal” to American politics, in which support is derived not just from appealing to in-group affection, but from the outright belittlement and denigration of the opposition party and the social groups that comprise its members.
When support for a party’s candidate comes from active hatred for certain social groups (as we have found in the case of President Trump), this fundamentally changes how political candidates construct their campaign strategy and messaging. In this new “normal” of American politics, a candidate’s coalition building isn’t about finding groups of people who support their cause, but finding people who actively hate and want to harm the other side. The chance for national cross-partisan coalitions recedes as this type of politics proliferates.
Our findings also imply that Trump could encourage more compromise from Democrats on common ground issues such as infrastructure spending if he stopped committing wanton affronts to Democratic groups (e.g., by delaying the Harriet Tubman $20 bill indefinitely). This is because support for the Democratic Party comes from support for marginalized groups. Should Trump change his behavior towards African-American, Latinos, and Muslims, even symbolically, it could potentially go a long way in building bridges with Democrats.
However, there is less cause for optimism when it comes to Trump-loyal Republicans who comprise 90 percent of the party. In a new “normal” of out-group hatred-driven partisanship, there is little incentive to sit at the negotiating table with Democratic leaders like Sen. Schumer or Speaker Pelosi. This is precisely because a key ingredient driving support for Trump is antipathy toward marginalized groups and, presumably, anyone who represents the interests of those groups.
Certainly, politicians are sometimes willing to go against the wishes of their own political base (see recent research on this here and here). But Trump doesn’t just eschew bipartisan compromise. He actively focuses on belittling his opponents to score points with his most loyal supporters.
Until Trump is willing to tamp down his adversarial rhetoric and behavior against Democrats and their voters, the prospects for political compromise on any matters of substance appear intractably grim. And from all the available evidence, Trump’s political incentives point him further down the road of identity-based conflict. The voters who support Trump today are the same people who hated the idea of racial, religious, and cultural diversity well before Trump appeared on a debate stage.
Lilliana Mason (@LilyMasonPhD) is an assistant professor of government and politics at the University of Maryland, College Park. Julie Wronski (@julie_wronski) is an assistant professor of political science at the University of Mississippi. John V. Kane (@uptonorwell) is an assistant professor at New York University’s Center for Global Affairs.
Update: This post has been revised since its original publication.