How Voters’ Education Levels Factored Into Trump’s Win

November 10, 2016

On the heels of Donald J. Trump’s surprise victory in the presidential election, experts are, like the rest of us, still working to make sense of exactly what happened.

A natural place to turn for insight is exit polls — even in an election cycle that has cast serious doubt on the polling industry. "Exit polls are pretty good," said John T. Scott, professor and chair of political science at the University of California at Davis ­— certainly better than those conducted ahead of an election. "They’re not perfect." They’re also the first data that’s available.

“The surprise ... is how many college graduates voted for Trump.”
Throughout the election cycle, observers have been interested in the important role that education appeared to play in shaping voter preferences. What do the exit polls suggest about how that materialized? Let’s take a look:

How did educational attainment play into voting patterns?

More or less as expected. Over all, exit polls are "telling us what was predicted," Mr. Scott said. "Namely, people with college degrees — and especially higher degrees — are far more likely to vote for Clinton than Trump, and vice versa."

That expectation marked a departure from recent history. Mitt Romney had a slight edge with college graduates in 2012, though President Obama won among those with more education than that.

But as anticipated, in 2016, voters with a college degree — and especially those with more education — broke for Hillary Clinton, while Mr. Trump was favored by those with lower levels of education.

"The surprise, however," Mr. Scott said, "is how many college graduates voted for Trump." That margin was narrow: 49 percent of college graduates voted for Mrs. Clinton, 45 percent for Mr. Trump.

All along, Mr. Trump’s support was believed to be concentrated among white voters with lower levels of education, and the exit polls bear this out: 67 percent of such voters went to Mr. Trump, 28 percent to Mrs. Clinton. But Mr. Trump also won the white college-educated vote, with 49 percent to Mrs. Clinton’s 45 percent.

 Throughout the campaign, polls suggested that Mrs. Clinton had a solid lead among white, female, college-educated voters, said William H. Frey, a demographer at the Brookings Institution. Her lead with that group was so sizable, he said, that some expected her to carry the white college vote outright.

But those polls, Mr. Frey said, "may have spun up and down depending on what Trump was saying at the time." Mrs. Clinton’s support among white, college-educated women was especially high, he said, following the release of the Access Hollywood tape in which Mr. Trump bragged about grabbing women by their genitals and made other aggressive comments about women. "It just goes to show you," Mr. Frey said, "the polls are one thing, and the election itself is another."

What accounts for Mr. Trump’s support among white, college-educated voters?

Mr. Trump clearly resonated beyond the white voters with less education, whom he was expected to win. Why? If Mr. Scott had to hazard a guess, he’d focus on the narrative of Mr. Trump’s campaign: that the loss of jobs to overseas left many people vulnerable and seeking change. That narrative, he says, "extended from non-college-educated people into college-educated people."

White voters with lower levels of education, in other words, were not the only ones who may have felt left behind, economically or otherwise.

It’s not really fair to say that political scientists had bought into the idea that Mr. Trump had support only among white voters with less education, said Hans Noel, an associate professor of government at Georgetown University. Republicans tend to do well among more-affluent voters, Mr. Noel said, and income and education are correlated. Republicans have made their gains among the white working class, he said, but white educated voters have been their base.

What ultimately drives voters’ behavior, Mr. Noel said, is identity. "At the end of the day, he said, "it’s how people are thinking of themselves that matters."

Voters’ identities can be linked to their demographic characteristics, but no voter can be reduced to a one-dimensional description.

"It is really somewhat simplistic to say white working-class voters are for Trump," Mr. Noel said. That’s a squishy category at best, he said. Take a plumber making $150,000. He might identify with the white working class even though his income suggests otherwise.

This could be a problem of categorization, Mr. Noel said. Historically, Roman Catholics, Protestants, and Jews once had distinct voting patterns. Today, those differences are present but are less important than how religious a voter of any religion is. Perhaps, Mr. Noel said, something similar is now happening with class.

"It might be, Mr. Noel said, "‘white working class’ is a misleading category to call these people."

What made education level such a focus in this election?

Education has become a more-important predictor of voting behavior. We are "seeing the declining role of income and the increasing role of education in determining how people identify with political parties," said Joshua Dyck, an associate professor of political science at the University of Massachusetts at Lowell.

“The thing that swamped all of this was race.”
Traditionally, the parties broke down along the lines of "haves" and "have-nots," said Mr. Dyck, who also co-directs the university’s Center for Public Opinion. But "education has over the last 20 years become the more important mover in determining which party you belong to," he said.

Being a more important factor than income, however, does not make education the most important driver of voting behavior. Other demographic details matter, Mr. Dyck said, and in 2016, "I think the thing that swamped all of this was race."

This helps explain that while "white noncollege men are really, really, driving Trump’s support," Mr. Dyck said, "he also won white men with college degrees."

Both Mrs. Clinton and Mr. Trump had high unfavorability levels, and putting together what voters made of Mr. Trump and who they voted for, Mr. Dyck said, 11 percent thought he was unfavorable, but still voted for him, suggesting what these voters wanted most was a change. That group, Mr. Dyck suspects, is heavily white and male.

"I know education plays part of this story," Mr. Dyck said. "I think a lot of the time, we told this as a story on education and left out the race and gender part." It’s now apparent, he said, that race and gender were the more important factors.

Beckie Supiano writes about college affordability, the job market for new graduates, and professional schools, among other things. Follow her on Twitter @becksup, or drop her a line at

Joshua Hatch contributed to this report.