Election Data & Demographic Factors
Perhaps the most contested question from the 2016 presidential election is: what factors motivated white working-class voters to support Donald Trump over Hillary Clinton by a margin of roughly two to one? Trump supporters often express frustration about being labeled racist. But what does the data say? Can the data confirm or deny allegations of racism?
Analysis by PRRI and The Atlantic, based on surveys conducted before and after the 2016 election, developed a model to test a variety of potential factors influencing support for Trump among white working-class voters. The model identifies five significant independent predictors of support for Trump among white working-class voters. No other factors were significant at conventional levels (Executive Summary, PRRI).
The “Economic Anxiety” Thesis Was Not Upheld
The “economic anxiety thesis” proved to be popular with pundits and journalists during the run-up to the election. Basically, the thesis posited that Americans who were suffering the most economically would cast a vote for Trump. This proved not to be the case. The PRRI model and findings reveal that things are a bit more complex.
Overall, the model demonstrates that besides partisanship (party ID) fears about immigrants and cultural displacement (i.e. “cultural anxiety”) were more powerful factors than economic concerns when it came to predicting support for Trump among white working-class voters.
That is not to say that economics was not important. Economics doubtless played a role, but these factors were not revealed in simplified terms. As it turned out, economic fatalism predicted support for Trump, though economic hardship predicted support for Clinton.
- Identification with the Republican Party. Identifying as Republican, not surprisingly, was strongly predictive of Trump support. White working-class voters who identified as Republican were 11 times more likely to support Trump than those who did not identify as Republican. No other demographic attribute was significant.
- Fears about cultural displacement. White working-class voters who say they often feel like a stranger in their own land and who believe the U.S. needs protecting against foreign influence were 3.5 times more likely to favor Trump than those who did not share these concerns.
- Support for deporting immigrants living in the country illegally. White working-class voters who favored deporting immigrants living in the country illegally were 3.3 times more likely to express a preference for Trump than those who did not.
- Economic fatalism. White working-class voters who said that college education is a gamble were almost twice as likely to express a preference for Trump as those who said it was an important investment in the future.
- Economic hardship. Notably, while only marginally significant at conventional levels (P<0.1), being in fair or poor financial shape actually predicted support for Hillary Clinton among white working-class Americans, rather than support for Donald Trump. Those who reported being in fair or poor financial shape were 1.7 times more likely to support Clinton, compared to those who were in better financial shape.
It is notable that many attitudes and attributes identified as possible explanations for Trump’s support among white working-class voters were not significant independent predictors – gender, age, region, and religious affiliation were not significant demographic factors in the model. Views about gender roles and attitudes about race were also not significant. Also noteworthy, neither measure of civic engagement—attendance at civic events or religious services—proved to be a significant independent predictor of support for Trump.
The report also provides an in-depth profile of white working-class Americans, along with analysis of this group’s worldview, outlook, and attitudes about cultural change and policy:
- Nearly two-thirds (65%) of white working-class Americans believe American culture and way of life has deteriorated since the 1950s.
- Nearly half (48%) of white working-class Americans say, “things have changed so much that I often feel like a stranger in my own country.”
- Nearly seven in ten (68%) white working-class Americans believe the American way of life needs to be protected from foreign influence. In contrast, fewer than half (44%) of white college-educated Americans express this view.
- Nearly seven in ten (68%) white working-class Americans—along with a majority (55%) of the public overall—believe the U.S. is in danger of losing its culture and identity.
- More than six in ten (62%) white working-class Americans believe the growing number of newcomers from other countries threatens American culture, while three in ten (30%) say these newcomers strengthen society.
- Nearly six in ten (59%) white working-class Americans believe immigrants living in the country illegally should be allowed to become citizens provided they meet certain requirements, while 10% say they should be allowed to become permanent legal residents. More than one in four (27%) say we should identify and deport illegal immigrants. Notably, support for a path to citizenship is only slightly lower than support among the general public (63%).
- More than half (52%) of white working-class Americans believe discrimination against whites has become as big a problem as discrimination against blacks and other minorities, while 70% of white college-educated Americans disagree.
- Fewer than four in ten white working-class Americans report they are in excellent (5%) or good shape (33%) financially, compared to six in ten who say they are in fair (35%) or poor shape (25%). White working-class Americans about as likely to say their financial situation has diminished (27%) as they are to say it has improved (29%). White college-educated Americans, in contrast, are about three times as likely to say their financial circumstances have gotten better than gotten worse (41% vs. 14%, respectively).
- A majority (54%) of the white working class view getting a college education as a risky gamble, while only 44% say it is a smart investment.
- Six in ten (60%) white working-class Americans, compared to only 32% of white college-educated Americans say because things have gotten so far off track, we need a strong leader who is willing to break the rules (Executive Summary, PRRI).
So what is “Cultural Anxiety?”
Journalist Adam Serwer also wanted to know if racism the driving force behind Trump’s candidacy. He wanted to explain the contradiction of how Americans, the vast majority of whom say they oppose racism, could back an explicit and unapologetically racist candidate. He reformulates the “cultural turn” in anxiety and restates it in terms of race:
“The specific dissonance of Trumpism—advocacy for discriminatory, even cruel, policies combined with vehement denials that such policies are racially motivated—provides the emotional core of its appeal. It is the most recent manifestation of a contradiction as old as the United States, a society founded by slaveholders on the principle that all men are created equal. While other factors also led to Trump’s victory—the last-minute letter from former FBI Director James Comey, the sexism that rationalized supporting Trump despite his confession of sexual assault, Hillary Clinton’s neglect of the Midwest—had racism been toxic to the American electorate, Trump’s candidacy would not have been viable.” And the American corporate news media, by refusing to call out Trump’s virulent racism, is complicit in it.”
“What I found was that Trump embodied his supporters’ most profound beliefs—combining an insistence that discriminatory policies were necessary with vehement denials that his policies would discriminate and absolute outrage that the question would even be asked.”
What Serwer found, after talking with numerous Trump supporters after the election was that they would not change their minds about supporting Trump, because in him they have what they have always wanted: a president who embodies the rage they feel toward those they hate and fear, while reassuring them that that rage is nothing to be ashamed of (Serwer).
“These Americans,” writes Serwer, “who would never think of themselves as possessing racial animus, voted for a candidate whose ideal vision of America excludes millions of fellow citizens because of their race or religion.” “One hundred thirty-nine years since Reconstruction, and half a century since the tail end of the civil-rights movement, a majority of white voters backed a candidate who explicitly pledged to use the power of the state against people of color and religious minorities, and stood by him as that pledge has been among the few to survive the first year of his presidency” (Serwer).
Post-election discussions were abuzz with questions about “unconscious sexism,” “internalized misogyny/bias,” and “implicit bias.” But what does this all mean? Can researchers measure sexism in ways similar to the way they measured racism?
Research conducted before the election by (YouGov survey) found that Trump voters had much higher levels of sexism, on average, than Clinton voters, as measured by their level of agreement with statements such as “women seek to gain power by getting control over men.” (Bialik). The statistical effect registered as significant even after controlling for other factors like gender and political ideology. Other research reached similar conclusions.
In the case of internalized misogyny/bias, the evidence suggests that both men and women harbor bias career advancement by women. Perhaps more surprising, the research document that women hold more of this bias, on average, than men.
Caroline Heldman, a political scientist at Occidental College, said “I don’t think you can understand [Clinton’s] candidacy without understanding gender bias is baked into it.” According to her, “we don’t like women to be ambitious. It rubs men and women the wrong way” (Bialik).
The operation of implicit bias may be more subtle and difficult to measure. Nonetheless, researchers have conducted numerous studies to this end as part of an effort to understand the unconscious tendency to associate different personal attributes and behaviors with certain groups — in this case, the tendency to associate men with strength and careers and women with softness and family. Here again, this kind of bias registers stronger on average in women than in men. Among women, the bias is noted as being particularly strong among those who identify as politically conservative. Data indicate, furthermore, that this bias was especially strong among one group in 2016: women who supported Trump (Bialik).
So how do you test for implicit bias? Researchers have, for example, run online experiments to test how subjects associate words such as “boy” and “lady” with “career” or “family.” You can take a version of the test at the website of Project Implicit, a nonprofit research organization (Bialik).
Of course, most of the women who voted in the recent election did, in fact, support Clinton. Her victory margin with women was approximately 14 percentage points. Trump support, on the other hand, registered more prominently with men, by as much as 12.5 percentage points, according to exit polls — the biggest gender gap on record in a presidential election (Bialik).
Unfiltered Comments from Trump Crowds – Racism? Sexism? Maybe?
Among whites, Trump won an overwhelming share of those without a college degree; but he also won white college graduates – a group that many identified as key for a potential Clinton victory – Trump outperformed Clinton by a narrow 4-point margin – 49% to 45%.
Here, however, the widest gap was evident between those without a college degree. College graduates backed Clinton by a 9-point margin (52%-43%), while those without a college degree backed Trump 52%-44% (8-point margin).
By way of comparison, in the last Presidential election, college graduates backed Obama over Romney by 50%-48%; those without a college degree also supported Obama 51%-47% (Tyson & Maniam).
The vast majority of white voters voted for Trump. When we combine the two variables together – race and education – we see Trump’s margin among whites without a college degree was significant – the largest among any candidate in exit polls since 1980. Two-thirds (67%) of non-college whites backed Trump, compared with just 28% who supported Clinton, resulting in a 39-point advantage for Trump among this group.
While large majorities of black and Latina women voted against Trump and for Hillary Clinton, white women didn’t. Exit polls showed that approximately 53% of white women voted for Trump; 45% of women with college degrees voted for Trump (Beckett).
As stated above, more white women (53%) voted for Trump; this includes 45% of women with college degrees.
Gender was more of a factor for men – 53% to 41%, of men supported Trump over Clinton.
Older voters voted overwhelmingly for Trump. Older voters (ages 65 and older) preferred Trump over Clinton 53%-45%. This result was similar to the previous Presidential election, where older voters backed Romney over Obama 56%-44%.
Clinton received a lower share of the vote among young voters (ages 18-29) than Obama received in 2012 or 2008. Young adults, however (ages 30-64) preferred Clinton over Trump by a wide 55%-37% margin.
While not studied as much as the preceding variables, religiosity (religious social identity) is also intertwined with politics and support for political candidates. In the United States, a religiously plural society, there are numerous religiously identified groups, many of whom hold strong views on a number of politically contentious topics.
Evangelicals, who proved to be strong supporters of Trump (exit polls showed that 81% of self-identified Evangelicals voted Trump). They stated they feel they are increasingly under attack by the secular culture – even persecuted. Evangelical leader James Dobson ( spokesperson for the interest/lobbying group Focus on the Family) was on record as one of the more prominent supporters of Trump.
According to Dr. Kristin Du Mez, “issues related to gender—from the cultural sea change on gay marriage to transgender bathroom laws to the Hyde Amendment and the contraceptive mandate—are at the center of evangelicals’ perceived victimization. The threat of terrorism, she says, also looms large.
There is an articulated fear that American power isn’t what it used to be, as nearly two-thirds of white evangelicals harbor fear that a once-powerful nation has become “too soft and feminine”(Du Mez). Here again, there is more evidence that variables are deeply intertwined (in this case religiosity and gender).
Evangelicals have a tradition of embracing literature that claims men were created in the image of a warrior God. As such, they have been receptive to sentiments like those expressed by the late Jerry Falwell in his 2004 sermon “God is Pro-War.” Researchers have documented in surveys that traditionalist evangelicals are more likely than other Americans to approve of U.S. engagement in a preemptive war, support military action against terrorism, and condone the use of torture (Du Mez).
This brand of militant masculinity, says Du Mez, “helps explain the lack of outrage on the part of many evangelicals when it comes to Trump’s character issues. Dobson himself, one of Trump’s most influential evangelical supporters, urged fellow Christians ‘to cut him some slack.’ ” More tellingly, the Rev. Robert Jeffress, pastor of First Baptist Church in Dallas and stalwart Trump supporter, explained his endorsement of the unconventional candidate in this way: “I want the meanest, toughest, son-of-a-you-know-what I can find in that role, and I think that’s where many evangelicals are.” (Du Mez)
In light of this, it would appear that Donald Trump appeared on the scene as the leader they were looking for; a man who embodies the militant masculinity that they feel certain can cure our social ills.
Race & Religiosity
Sean McElwee’s report for Demos found the following after analyzing results from the 2016 American National Election Studies (ANES) pilot survey (the survey asked respondents to rank their feelings about politicians and social groups, where a score of 0 indicated cold feelings, up to 100, which meant warmest feelings). His report offers analysis and explanation of feelings for Trump among white respondents, based on their views of government treatment and/or discrimination against whites and Christians.
According to McElwee, “among independents and Republicans, those with strong feelings that whites and Christians faced discrimination were more supportive of Trump in the Republican primaries (the ANES survey was completed in January 2016). Among white Republicans and independents who believe whites face no discrimination, 69 percent supported a Republican candidate other than Trump. Among those who believe whites face a “great deal” of discrimination, only 34 percent did. Among those who believe the government treats whites much better, 72 percent chose a candidate other than Trump, whereas among those who feel the government treats black people much better, 49 percent did.
Among Republicans, but particularly among Trump supporters, feelings of white discrimination and loss were profound. Among white Democrats, 74 percent said whites face “little or no” discrimination, compared to 51 percent of white Republicans. While 75 percent of white Democrats said Christians face “little or no” discrimination, only 30 percent of white Republicans did. As the chart below shows, there are deep partisan divides in perceptions on whether the federal government favors white people or black people (or treats both equally).”
50% of eligible voters in the United States did not vote. This one statistic alone goes a long way in explaining Clinton’s nearly 3 million popular vote victory over Trump.
Getting Under the Numbers
While some of the more obvious conclusions jump off the page, the results sometimes suggest additional questions. Why, for example, did so many white women (regardless of socio-economic status) vote for Trump? The answer as Fortune Magazine put it follows:
“The fact that 53% of white women voted for Donald Trump shows that some white women are more aligned with ideas of power and privilege than gender solidarity and racial inclusion.’
Given how few people are willing to admit overtly racial motivations for their vote, the numbers nonetheless betray a strong racial correlation for Trump support. This is one of the reasons that underlie accusations that Trump voters are racists.
Doubtless, a fair amount of economic resentment fuels racial anxiety. Some of this anxiety, as the video demonstrates, bleeds into open racism. Despite this, to write off the anger of white working class people as nothing more than racism is misleading, for it fails to account for the intersectional complexity of the problem.
In the absence of racial antipathy, one might surmise at the very least significant numbers of white men and women did not feel vulnerable on the basis of their race (they were not a declared target of Trump’s rhetoric). Maybe they found it easy to dismiss Trump’s racist rhetoric due to an implicit understanding that they were not likely to suffer any consequences. For white women, the story differs somewhat.
Some scholars and analysts have explained this particular vote as a function of “internalized misogyny” – an internalized form of self-hatred, where they have become so accustomed to being seen as second-class citizens and having their bodies objectified (dismissing “locker room” talk as boys will be boys, that they were willing to overlook Trumps misogynistic rhetoric and behavior in an effort to maintain social solidarity with the men in their life, who are a source of comfort and support for them.
In light of all of this, the numbers constitute evidence that suggests both overt racism and at the very least implicit bias proved to be a significant predictor of Trump support. When a candidate advocates a Muslim registry, thinks Hispanics are rapists and wants to reinstate unconstitutional practices like “Stop and Frisk” (which research proves disproportionately targets people of color), it’s hard to escape allegations of racism, even if one personally feels the label does not apply to them.
When racism isn’t a “deal-breaker,” tolerance of racism and/or support for policies that produce racist outcomes makes one complicit in the racist project.
(Photo by Alex Wong/Getty Images)
Toward an Intersectional Analysis
Sociologists draw from “intersectional theory” – a theory that looks at the intertwined nature of social identities based on race, class, and gender – in their efforts to explain things like political party affiliation and candidate support among other things. But here you might be wondering – Why is it important to take into account interpenetrating social identities?
To be sure, it was not only the U.S. electoral politics of 2016 that saw an increase in appeals t populism (exemplified by the election of Trump as US President), England reflected similar sentiments as seen in the highly contentious ‘Brexit’ vote – the UK’s referendum vote to leave the European Union.
Explanations thus far have cited the myriad failures of globalization, which exacerbated white working class anxiety and hostility toward immigrants and those perceived as outside “Others.”
While gender, as was pointed out here, figured prominent in analyses, the bulk of critique remained focused on ‘alienated working class’ nationalism and patriotism. The resulting “whitelash” in these cases caused gendered complexities to recede into the background. But as we saw with the recent Women’s March, there remains much to be said about women’s voices and perspectives.
It is only by calling upon intersectional theory that we can attempt to move forward and arrive and a more nuanced analysis of what happening; an intersectional analysis that takes into account very complex understandings of social identity and representation and how these are further interpreted by political economic factors.
“American Horror Story,” by Jared Yates Sexton – published in The New Republic as “American Horror Story: A Donald Trump Rally is a Homophobic, Misogynistic, Racist Nightmare.” He is an Assistant Professor of Creative Writing at Georgia Southern University and the author of three collections of stories and a crime novel.
The Nationalist’s Delusion, by Adam Serwer, The Atlantic Magazine, November 2017.
“Beyond Economics: Fears of Cultural Displacement Pushed the White Working Class to Trump” | PRRI/The Atlantic Report by Daniel Cox, Rachel Lienesch, and Robert P. Jones, May 2017.
“A Morning with ‘Adorable Deplorables’: Why Trump Supporters are Optimistic,” by Louis Beckett
“Behind Trump’s Victory: Divisions by race, gender, education,” by Alec Tyson and Shiva Maniam
“How Unconscious Sexism Could Help Explain Trump’s Win,” by Carl Bialik
“Donald Trump and Militant Evangelical Masculinity,” by Kristen Du Mez
Trump’s supporters believe a false narrative of white victimhood — and the data proves it; Trump voters believe that whites and Christians face discrimination — but they call the left sensitive snowflakes” by Sean McElwee
What does the data tell us here about voter preferences?
Was education an important predictor of how people voted in this election?
What about socio-economic status? Did having/not having money (a proxy for economic anxiety) make a difference in voting patterns? Put differently, was it shown that people who were suffering the most economically were the ones that voted for Trump?
Of the three variables – race, class, gender – which one predicted voting outcomes the best?
What do you think is the source of the rage on display in the videos? Do you think the people yelling at the rallies are potentially racist? Does their rage (and the things they are saying) indicate that they feel hate, disrespect, or feel threatened by women that operate outside traditional gender roles? Or are they simply angry about economic issues (economic anxiety thesis)? What does the research/data say?
What are we to make of the fact that 50% of voters in the United States did not vote?