Lilliana Mason | Vox

Trump support is not normal partisanship

2019-06-21T19:23:41-04:00

A rally during Donald Trump’s campaign for president in the 2016 election. | Scott Olson/Getty Images

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.

Base politics

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.

Donald Trump attracts poor voters with multiple Republican social identities

2019-03-06T00:00:49-05:00

Donald Trump’s early electoral success has exceeded the expectations of virtually everyone. His success among a number of constituencies with varying interests invites the question: How does someone known to clearly switch his positions within an hour-long debate, who has virtually no political experience, and who brags about the size of his biological features gain this much traction with voters?

A number of explanations have been proposed. Individuals who score high on authoritarianism, a personality trait that indicates a psychological preference for order and a fear of outsiders, helps explain Trump support. So, too, does white identity. Hostility toward out-groups, coupled with the perception that government is increasingly discriminatory toward whites, offers considerable insight into individual voting intentions.

These explanations all hold merit but, on their own, may be incomplete. Instead, our research predicts that as individuals’ social and political identities converge (or become “sorted”), support for someone like Donald Trump is predictable — particularly when you consider that most Republicans are not orthodox ideologues. In fact, given what we know about the structure of belief systems, namely that the average voter holds a considerable number of policy preferences that are paradoxical, this social sorting offers an explanation for Trump’s support that draws on the power of group attachments.

In a primary race where the main narrative seems to be that the Republican Party is being broken into pieces, recent data from the first wave of the American National Election Studies 2016 Pilot Study suggests that this division is not the source of Trump’s support. In fact, it appears Trump is popular despite the splintering of the Republican Party.

We find that the three major components of the modern Republican Party (traditional Republicans, conservative Christians, and white identifiers) are most likely to predict Trump support when these separate identities are working together. So, in a sense, Trump’s support isn’t an anomaly of Republican splintering — it’s the natural outcome of these identities that have been coalescing over time. Thus, for many Republicans who strongly identify with one group (i.e., Republican) but not another (i.e., white), this primary season may feel alienating and frustrating, but for those who strongly identify with all three of these groups, Trump is their guy.

When we look at the “affect” or feelings toward the three front-running Republican candidates, we find that “feeling thermometer” ratings, or descriptions of how warm or cool someone feels toward a candidate, are related to the alignment of a person’s Republican, racial (white), and religious (Christian) identities. When a person does not strongly identify with all three social groups, she feels quite negative toward Trump. However, when all three identities are well-sorted, Trump benefits more than either Ted Cruz or Marco Rubio.

Moreover, not only do such individuals feel more positive about Trump, but they are also far more likely to vote for him.

Social psychology research has found that when most of the members of one social group are also members of another social group, those who identify with both groups are more likely to view outsiders with intolerance and to see them as very different from themselves. Once people feel so isolated from outsiders, for example, they are more sensitive to threats from these potential enemies, and they are more easily angered by threats from anyone who is not in the group.

In other words, the more our identities are aligned, the more foreign any outsider will seem, regardless of logical reasoning (or, perhaps, policy preferences). These sentiments are familiar from Trump’s supporters, and the effect of combining these identities may fit perfectly with a Trump candidacy, as he speaks directly to the anger and intolerance that are results of this type of social sorting.

These angry and intolerant responses, however, should be the most concentrated among those who are in a perpetual state of feeling like their status is being threatened. This is because social identities provide people with a sense of esteem that they may not be capable of getting from their own individual lives. When their group wins, they feel like winners, even if they are not winners on their own.

In order to examine this feature of the groundswell of Trump support, we split the sample into “poor” (the people facing the most status threat in their lives in general, coded as the bottom fifth of self-reported income distribution within the ANES sample) and “not poor.”

It turns out that support for Trump is most powerful among those identifying as Republican, religious, white, and poor. This “coalition” built by the Republican Party is particularly strong among the poorest voters, ostensibly because they feel the most status threat and therefore are the ones who most desperately need a sense of group victory. (In effect, this cleavage may be emblematic of the awkward relationship between Dixiecrats and the larger GOP.) They therefore respond most strongly to their religious, partisan, and racial identities by voting for Trump (who makes them feel most empowered and is the most aggressive attacker of outsiders).

According to these results, Trump is capable of capturing the imaginations of well-sorted poor voters far more effectively than either Cruz or Rubio can. Not only do poor voters feel the most warmly toward Trump, as shown in their “thermometer” ratings, but they are also far more likely to vote for him.

These results call into question the idea that Trump is only garnering the votes of poor white people. In fact, he is winning the votes of poor, white, religious Republicans better than either Cruz or Rubio. This could also help explain why Cruz has been performing well in caucus states, where the cost of participation is high and poorer voters have less of a voice.

As we see it, Trump is likely to remain a formidable candidate, as long as he provides a strong identity for those who need to buttress their sense of victory in their own lives by gathering their group identities around themselves. As the primary season wears on, this social sorting is likely to further push the well-sorted in the direction of Trump as questions of viability threaten the other candidates.

Lilliana Mason is an assistant professor of government and politics at the University of Maryland College Park.

Nicholas Davis is a doctoral candidate in political science at Louisiana State University.