Donald Trump just pulled off one of the most stunning upsets in American political history, capturing the presidency last Tuesday night. It was a shock to political pundits, and to pollsters – a group I belong to – who had predicted a tight, but safe Clinton victory. But Trump’s win was reportedly a surprise even to those inside the campaigns, including his own.
How did it happen? One of the best places to start is by looking at the results of the exit polls. Exit polls are coordinated by the National Election Pool (NEP), a group of media organizations including ABC News, the Associated Press, CBS News, CNN, Fox News, and NBC News. I serve as ABC’s representative on the NEP’s Survey Committee. While most polls are conducted by phone or online, exit poll interviewers literally catch people as they leave polling places and ask them to fill out a short paper and pencil survey asking them how and why they voted the way they did.
My initial take on these results is that Trump’s success had little to do with traditional attributes normally used to judge candidates, nor did it have much to do with issues, except for how his positions related to people’s identities. This is true both on foreign policy and domestic issues. Indeed, this election was all about identity politics, with Trump able to connect with non-college whites, especially white men without a college degree.
His election was a rejection of political and media elites, by his strongest base of support – whites without a college degree, who flipped states northern states like Wisconsin, Pennsylvania, Ohio, Iowa, and, possibly, Michigan into the Republican column. They also made other states thought to be safe for Hillary Clinton – Minnesota, New Hampshire and Virginia – agonizingly close.
Personal Qualities – What the Campaign Seemed to be About
First, let’s deal with candidate attributes. On all counts, Hillary Clinton was rated at least no worse than Donald Trump, and often much better. Clinton’s overall favorability rating was only 44 percent, but Trump’s was even lower at 38 percent. Just 36 percent of voters saw her as honest and trustworthy, a big problem for her throughout the campaign, yet only 33 percent said Trump was honest. Both were a wash.
Fifty-two percent said Clinton was qualified to be president versus just 38 percent who said so of Trump. Fifty-five percent thought she had the right personality and temperament, whereas many fewer, 35 percent, felt that way about Trump. On these attributes that one might consider basic hurdles for the presidency, Clinton was well ahead of Trump, and yet she lost.
Let that sink in: Americans just elected a president whom majorities say is unlikable, dishonest, unqualified and has the wrong temperament.
On the two big scandals of the campaign season, Clinton was again at no disadvantage. Forty-five percent of voters were bothered a lot by her use of a private email server while secretary of state, but half were also highly bothered about Trump’s treatment of women. Again, at worst, this is a wash.
Lastly, thinking about a possible Clinton or Trump presidency, 29 percent said they were scared of a Clinton administration, but so did 36 percent of a Trump administration. Based on candidate qualities, there’s no way Clinton should have lost the election.
On the Issues
The exit poll results on the issues are not so clear-cut, but they also don’t point to a Trump victory. Perhaps the most glaring of these is voters’ views on immigration.
Seventy percent of voters supported a path to citizenship for undocumented immigrants and 54 percent opposed Trump’s signature policy on the issue, a wall along the entire border with Mexico. On the most important issue in the election – the economy/jobs, terrorism, foreign policy or immigration – Trump won those who picked immigration as most important by a wide 2-1 margin over Clinton. But just 13 percent picked immigration as their top issue, the same as the number of foreign policy voters (13 percent), slightly behind terrorism voters (18 percent) and far behind economy voters (52 percent).
Trump also prevailed among terrorism voters, 57-39 percent Trump-Clinton, but Clinton did just as well among foreign policy voters, 60-34 percent Clinton-Trump, and won by 10 points among economy voters 52-42 percent Clinton-Trump.
It’s no surprise that Trump won voters most concerned about terrorism. In a related question, he also won voters who thought the US war with ISIS was going badly by 45 points. They outnumbered those who thought it was going well, 52 to 42 percent.
When the two candidates were compared head-to-head on issues, there were no clear Trump advantages. On who’d best handle the economy, it was a close 46-49 percent Clinton-Trump. And on two foreign policy-related issues, it was a slim 49-46 percent for Clinton on who would be a better commander in chief a wider 52-43 percent for Clinton on better handling foreign policy in general.
Trade was not a clear winner for Trump either. Forty-two percent of voters said trade with other countries takes away more American jobs and these voters supported Trump by about a 2-1 margin. But more voters felt trade either had no effect (11 percent) or created American jobs (38 percent), and they voted for Clinton by a similar margin.
On other issues, voters split fairly evenly on whether Obamacare had gone too far (47%) or whether it did not go far enough or was about right (48% combined). And voters were fairly divided on whether the justice system is unfair to blacks (48%) or fair to all (43%). Neither was a decisive advantage for either candidate.
While Trump did well among immigration and terrorism voters, these issues alone are not enough to explain his victory. That’s because it was more about who he was – an outsider – than the attributes that were measured, and more about how the issues he talked about made non-college whites feel about themselves than the actual policies themselves.
It’s How People Identify, Stupid!
Overall, the continued demographic shift in the United States was evident in the exit polls. Seventy percent of voters were white, a record low for whites in exit polls going back to 1974, and 30 percent non-white. That should have provided an advantage to Clinton, except for two reasons – the vote margins within the groups changed drastically and the concentration of non-college whites in states that Trump managed to flip such as Ohio, Iowa, Wisconsin and Pennsylvania.
Overall, Trump won whites by 21 points, similar to Mitt Romney’s 20-point margin versus Barack Obama in 2012. Clinton’s 53 percentage-point victory among non-whites was smaller than Obama’s in 2008 (64 points) and 2012 (61 points), neutralizing their slightly increased turnout.
And, most importantly, Trump over-performed expectations in his key group, non-college whites, especially non-college white men, while Clinton underperformed in a demographic she needed to counter Trump, college-educated whites, especially college-educated white women.
Trump beat Clinton 67-28 percent among non-college whites, a 39-point margin that was the largest ever in exit polls back to 1980. Clinton managed only a four-point loss to Trump among college-educated whites.
Breaking it down further, Trump won non-college white men 72-23 percent, a record 49-point gap, and won non-college white women 62-34 percent, an impressive 28-point margin. He also won college-educated white men by 15 points, 54-39 percent, typical for a Republican. Clinton won college-educated white women, a vital group where she was expected to make big gains, by just 6 points, 51-45 percent. Combine that with her somewhat smaller margin among nonwhites and it wasn’t enough to overcome Trump’s huge margins among non-college whites. That mattered tremendously across the Midwest and in states like New Hampshire, Virginia and Pennsylvania.
Where do a lot of these non-college whites live? Many are in rural areas and the exit poll results bear this out. In cities, where 34 percent of voters live, Clinton beat Trump by 24 points, 59-35 percent. In the suburbs, where 49 percent of voters live, it was close, but they went narrowly for Trump by 5 points, 50-45 percent. And rural voters, though where just 17 percent of voters live, broke overwhelmingly for Trump, 62-34 percent, a 38-point margin. These rural voters include a lot of the people who are disillusioned with political and economic elites and also feel uncomfortable with the cultural changes taking place nationally, changes that are taking place to a large extent in big cities.
This point brings us back to candidate qualities. When asked which candidate quality mattered most to their vote, a plurality said it was someone who could bring “needed change” (39 percent). Fewer voters picked experience (21 percent), judgment (20 percent) or empathy (15 percent). Clinton won the vote among all these three latter categories. But among change voters, Trump won 83-14 percent.
In the end, it was voters’ associations of change with Donald Trump that ruled the day. This really doesn’t have that much to do with any personal qualities of Trump, but shows that what mattered most was that he was viewed as an outsider running against perhaps the quintessential political elite in Hillary Clinton.He successfully channeled the frustrations of non-college whites all the way to the presidency.