Trump’s Coalition: Broad, Loyal and Conflicted

There’s a popular portrait of a “Trump voter.” He’s a white man without a college degree, and so loyal that he would stick by President Donald Trump no matter what.

There’s a reason the stereotype exists: Trump’s strength among white working-class voters, particularly men, put him over the top in the decisive battleground states in 2016. And his approval ratings have been very steady, despite a year of controversial tweets and policy decisions. But it’s not the whole story.

Yes, white voters without a college degree shifted decisively from Barack Obama to Trump in 2016. But these voters actually made up only a slightly larger share of Trump’s coalition than they did of the previous three Republican nominees’ coalitions.

Trump has a large and resilient base of supporters, but a not-insignificant share had reservations when they cast their ballots for him and continue to have reservations. A small but meaningful number of his voters, particularly women, appear to have soured on him since the election.

Understanding the breadth of Trump’s coalition is important to understanding the Republican Party’s position heading into the 2018 midterms. Trump’s most enthusiastic supporters were vital to his victory in the primary, and Obama-Trump voters in old industrial towns were decisive in the general election.

But the midterms could be decided by voters at the edge of Trump’s coalition and of the public’s imagination: stereotype-defying female, college-educated or nonwhite Trump supporters, who are somewhat likelier to harbor reservations about the president. They may have been reluctant to back him, but they were still essential to his 2016 victory and are essential to the Republican Party’s chances today.

This more nuanced picture emerges from a survey of validated voters on Pew’s American Trends Panel, a representative sample of American adults who agreed to take Pew surveys every month. The panel allows a rare direct measurement of how voters have shifted over time.

Pew asked panelists how they voted in November 2016, and the responses were matched to voter records that indicate whether a panelist actually cast a ballot. It’s a big advantage over typical polls, which struggle to distinguish shifts in public opinion from the effect of a new set of respondents in each poll. It offers perhaps the clearest picture yet of who supported Trump and how his voters feel about him today.

If you want to understand why Trump won the presidency, there’s one big reason: white voters without a college degree. They put Trump over the top in disproportionately white working-class battleground states where Obama fared relatively well, like Pennsylvania, Wisconsin and Michigan.

But Trump’s supporters aren’t monolithic. Nor is his coalition necessarily dominated by the groups that broke most strongly for him.

Thirty-three percent of Trump’s supporters were white men without a college degree. A majority of Trump’s supporters defy the stereotype: They were either women, nonwhite or college graduates (or some combination of those).

Overall, 47 percent of Trump’s voters were women. And though he struggled among affluent college-educated whites for a Republican, he still won 44 percent of voters making more than $150,000 per year, according to the Pew data, and nearly 40 percent of college-educated white voters.

Perhaps surprisingly, Trump’s voters were about as likely as the supporters of other recent Republican nominees to hold a college degree.

How did the number of white working-class Republican voters stay so constant? Republicans have been winning a progressively larger share of white voters without a degree, but the group is shrinking overall. The result is that the two trends have basically canceled each other out.

The shift among college-educated white voters was particularly sharp, and the Pew data is one of the strongest pieces of evidence indicating that Hillary Clinton did far better among this group than initially believed. In the Pew data, she carried college-educated white voters by 17 percentage points, a huge shift from 2012, when Mitt Romney won that group.

It’s a very different story from the exit polls, which showed Trump winning college-educated white voters. There’s little doubt the exit polls were wrong. Almost all other survey data, along with the precinct-level election results, suggest that Clinton won college-educated white voters and probably by a big margin.

There has been little change in Trump’s approval rating in the last 18 months, and so it’s often assumed that nothing can erode his base of support. The Pew data suggests it’s not so simple.

Yes, nearly half of Trump’s voters have exceptionally warm views toward him: 45 percent rated their feeling toward him as a 90 or higher out of 100, a figure that is virtually unchanged since his election. But a meaningful number of his voters had reservations about him in November 2016, and even more Trump voters in 2018 hold a neutral or negative view of him.

Overall, 18 percent of Trump’s voters gave him a rating of 50 or less, on a scale of 0 (coldest) to 100 (warmest), up from 13 percent in November 2016.

It is worth noting that the November 2016 Pew survey was taken after Trump won the presidency, at the height of his postelection honeymoon. But even when you consider the slightly lower ratings voters gave him in the months before the election, the big picture is the same: A modest number of Trump’s voters didn’t like him that much then, and don’t like him much now.

Women, and especially college-educated women, are the likeliest Trump voters to have serious reservations about him in 2018: A striking 14 percent of the college-educated women who voted for him hold a very cold impression of him, up from just 1 percent in November 2016.

The resilience of Trump’s winning coalition gives Republicans a chance to hold Congress, even in a midterm election year when the president’s overall approval rating is well beneath 50 percent. It helps explain why he has retained almost unified support from Republican members of Congress despite his unorthodox stances on trade and Russia.

But while it’s clear that most of Trump’s supporters stand by him, there are big questions about whether they’ll be as helpful to Republicans in 2018 as they were to him in 2016.

Trump’s coalition was enough to win a presidential election in large part because white working-class voters were overrepresented in the presidential battleground states. That’s not so true in the most vulnerable Republican-held House districts.

At the same time, white voters without a college degree typically turn out in smaller numbers in midterm elections. And turnout among college-educated voters has been unusually high in special and general elections held since Trump won the presidency.

If the patterns hold, the combination of a better-educated battleground and lower turnout among less educated voters could mean that House control is decided in districts where college-educated voters make up around 47 percent of voters, rather than the 34 percent share of such voters in Pennsylvania, Wisconsin and Michigan in 2016, according to Upshot estimates based on census data.

Trump’s most fervent supports might stick with him through November and beyond. But in the midterms, Republicans will have the burden of fighting in many districts where the base might not be enough.