Minority women have voted by double-digit margins
for Democratic presidential candidates for decades.
Conversely, since 2004, non-college-educated white women have voted Republican by double-digit margins, and by a seven-point margin in 2000. In 2016, their loyalty to the GOP looked like it could falter. Close to the end of the race, they swung toward Hillary Clinton. But in the last weeks of the campaign, they moved decisively for Donald Trump.
For this group, when it comes to female candidates, relatability is crucial. But when trying to relate to these voters, female candidates face a particular challenge: If non-college-educated white women can’t see themselves — their life choices and values — reflected in a woman who’s running for president, they’re probably not going to vote for her. They “tend to be even more judgmental along some of these dimensions,” says Celinda Lake, a prominent Democratic pollster. “There’s a judginess there in a class as well as a gender way.”
Christine Matthews, a Republican pollster, has seen this in her work. When polling against Georgia Democratic senatorial candidate Michelle Nunn in 2014, she found that non-college-educated white women wanted to know whether Nunn was married and had kids. “ ‘She looks like a college professor,’ they’d say, which was damning praise,” Matthews said, recalling her focus groups. “What they were essentially saying was, ‘I cannot relate to her.’ ” Nunn, now chief executive of a national nonprofit — whose husband stayed home to raise their kids — lost that race.
The same dynamic plagued
Clinton in 2016, Matthews said. “For college-educated white women, relatability wasn’t as big a deal because essentially they themselves had college degrees. They were, or had been, in the [white-collar] workforce,” Matthews says. “But for non-college-educated women, they were intimidated.” Clinton, notably, won college-educated white women — a swing of 13 percentage points from Barack Obama’s loss of that demographic to Mitt Romney in 2012. Despite this, her loss of non-college-educated white women was so severe that, overall, she lost white women by 11 percentage points, a far cry from her husband, who in 1996
was the last Democrat to win the white female vote.
In the 2016 election, most demographic groups were predictably committed: White men and older voters favored
Trump. Minorities and college-educated voters went for Clinton. The only demographic that moved back and forth dramatically during the campaign was non-college-educated white women. After the uproar over Trump’s disparagement of Gold Star parents Khizr and Ghazala Khan during the Democratic National Convention, Trump dropped from a statistical tie
the day of Khan’s convention speech to around an eight-point
deficit a little more than a week later, according to RealClearPolitics’ average of national polls. The most pronounced change was among women: A CNN/ORC poll taken right before Khan’s speech showed Clinton only four points ahead among women; the same
poll taken immediately after the speech showed Clinton stretching her lead to 23 points.
Enter then-campaign manager and current White House adviser Kellyanne Conway, Trump’s ubiquitous cable TV surrogate, who cut her pollster teeth studying non-college-educated, white female voters. She helped Trump make up that lost ground by highlighting his history of hiring women in the male-dominated construction industry, including daughter Ivanka, and pushing him to
for policy proposals targeted at female voters in the campaign’s final stretch.
By the first general-election debate, Trump was just 2.3 points behind Clinton in the RealClearPolitics average. Then came the “Access Hollywood” recording on Oct. 7, and by their second debate on Oct. 9, Clinton was ahead by 4.6 points. An Atlantic/PRRI poll found that among non-college-educated white women, Clinton and Trump were even at 40-40.
If Clinton had held steady among this group, she probably would have won.
Non-college-educated white women picked George W. Bush by 18 points in 2004. They chose John McCain by 17 points and Mitt Romney by 20. The idea that they’d swing 20 points from 2012 to 2016 was wild. And then FBI Director James Comey reopened the investigation into Clinton’s email in the final weeks of the campaign. Non-college-educated white women ended up voting for Trump by a historic margin of 27 percentage points.
At the heart of it was the perception that Clinton considered herself above it all. “It is true that many female voters tend to ask of candidates, ‘Do I like you?’ and more importantly,” Conway wrote in an email to me, “ ‘Are you like me?’ . . . The latter is the ‘connective tissue’ of standing in one’s shoes, sharing one’s struggles. Bill Clinton was seen this way; Hillary Clinton was not.”
In the 2018 midterm elections, non-college-educated white women swung again. They still voted Republican, but by a margin of only 14 percentage points, helping to hand Democrats control of the House and elect a record number of women to public office.
Can this be repeated in 2020 by a woman running for president? Voting down-ballot is very different from voting at the top of the ticket. Women face unique challenges seeking executive office that men don’t, according to
research by the Barbara Lee Family Foundation, which specializes in helping elect women to executive office. Women’s strengths of collaboration, compromise and civility help them win down-ballot offices where group decision-making is key. But women face a competency test for executive decision-making — essentially a yes-or-no question about whether they can be commander in chief — that men don’t face for the highest offices, which is one reason women make up only 21 percent
of the nation’s mayors, 18 percent of governors and
of Fortune 1000 chief executives. For women, it’s a thin line between proving oneself capable and being seen as a bitch — essentially unlikable.
This is the conundrum for the 2020 female candidates. On the competency side, those with law enforcement experience, such as Harris and Sen. Amy Klobuchar (D-Minn.), have an inherent advantage. But Klobuchar has battled to balance her too-tough-boss image with “Minnesota nice,” while Harris is running on something of a unity platform. Warren has demonstrated her skills by providing the most detailed and comprehensive set of policy proposals among the candidates, but she “will face relatability issues in the same way Hillary Clinton did. She is, after all, a liberal New England Ivy League professor, and that’s probably why she got a golden retriever puppy” last year, said Matthews, the GOP pollster.
Clearly, men also face a relatability test — it’s why John Kerry was mocked as elitist in 2003 for putting Swiss cheese on his Philly cheesesteak. Relatability, or the lack thereof, also hurt President George H.W. Bush’s reelection chances when he seemingly couldn’t come up with the typical price of milk
during a debate. But for non-college white women, that relatability test is far more personal. It’s more about seeing themselves, or not seeing themselves, in the woman in front of them.
Former senator Barbara Mikulski (D-Md.), the longtime dean of Senate women, used to joke that we’ll know women have made progress when mediocre women are elected to Congress. All of the women running for president are exceptional, from former Harvard professors to former state attorneys general to television stars. That they’ve had to be twice as good to succeed could, paradoxically, hurt them with an important voting bloc.
Ultimately, it won’t be about winning non-college-educated white women as much as mitigating their loss. A female candidate’s best chance is to seem both capable and relatable, a difficult task until enough of these voters
relate to female ambition and leadership. In other words, until enough women aspire to power, or at least feel that it’s not beyond their reach, America may not be able to elect a powerful woman to the presidency.