Skip to main content
ABC News
How Unconscious Sexism Could Help Explain Trump’s Win

A woman has never come closer to the presidency than Hillary Clinton did in winning the popular vote in November. Yet as women march in Washington on Saturday, many of them to protest the presidency of Donald Trump, an important obstacle to the first woman president remains: the hidden, internalized bias many people hold against career advancement by women. And perhaps surprisingly, there is evidence that women hold more of this bias, on average, than men do.

There has been lots of discussion of the role that overt sexism played in both Trump’s campaign and at the ballot box. A YouGov survey conducted two weeks before the election, for example, 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.” An analysis of the survey found that sexism played a big role in explaining people’s votes, after controlling for other factors, including gender and political ideology. Other research has reached similar conclusions.

Two recent studies of voters, however, suggest that another, subtler form of bias may also have been a factor in the election. These studies looked at what’s known as “implicit bias,” the unconscious tendency to associate certain qualities with certain groups — in this case, the tendency to associate men with careers and women with family. Researchers have found that this kind of bias is stronger on average in women than in men, and, among women, it is particularly strong among political conservatives. And at least according to one study, this unconscious bias was especially strong among one group in 2016: women who supported Trump.

There is no way to know for sure how implicit gender bias affected voters’ behavior at the ballot box. Most women voters did support Clinton: She won among them by 14 percentage points, while Trump won among men by 12.5 percentage points, according to exit polls — the biggest gender gap on record in a presidential election. The impact of other kinds of implicit bias is disputed by researchers, and gender bias’s effect hasn’t been studied extensively.

But while quantifying the effect is difficult, experts said there is little doubt that gender bias — both explicit and implicit — played a role in the election.

“I don’t think you can understand [Clinton’s] candidacy without understanding gender bias is baked into it,” Caroline Heldman, a political scientist at Occidental College who has written about internalized sexism, said in a telephone interview. “We don’t like women to be ambitious. It rubs men and women the wrong way.”

Measuring implicit bias is tricky — because the bias is unconscious, you can’t just ask people how they feel about women in a professional context, or about sexist attitudes. So researchers instead run online experiments that test how quickly subjects associate typically male or female 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.)

Project Implicit ran this experiment with more than 700,000 online test takers from 2006 to 2015. The results show that conservative women had higher implicit gender bias than women with other political ideologies and than men of any political ideology, according to data provided by Colin T. Smith, the project’s director of education, at my request. At every level of self-reported political ideology, women had a higher level of implicit gender bias than men.1

Project Implicit’s data set doesn’t yet cover this general-election campaign,2 but another study found consistent results. In October, HCD Research, a New Jersey research firm, enlisted about 500 likely voters, including about 100 in each of four groups — Clinton and Trump supporters of both genders — in a test of implicit gender bias modeled after Project Implicit’s test. The result was striking. The group with the highest level of implicit bias against linking women with careers was women who planned to vote for Trump. More than 80 percent of them showed a bias toward linking men with careers more quickly than women, compared with 74 percent of women supporting Clinton and a little more than 50 percent of men supporting either candidate. The average bias of women planning to vote for Trump was the highest of any group, to a statistically significant degree.


You can try HCD’s test for yourself, though it won’t tell you your individual result. The study, which also included nearly 100 people who intended to vote for candidates other than Trump or Clinton, also found that all groups of voters have implicit bias against black people and that the level of bias is higher among Trump voters.

Gender bias scores vary by age, race and year the test was taken as well as by gender and political ideology. Teenagers had the least bias, while people 55 and older had the most. Black women show a stronger implicit bias than white women do, and their bias is fairly flat across the political spectrum, while the more conservative black men are, the less bias they exhibit. Among white, black and Hispanic men, self-described moderates show less bias than liberals.3 And bias scores were lower for more recent test takers, suggesting that this kind of societal bias is diminishing over time.4 The findings for each demographic factor were consistent even when controlling for the others.5


It is hard to know how implicit bias affected the 2016 election specifically. People don’t necessarily act — or vote — on their implicit biases. Although implicit bias against black people has been studied extensively and has been shown to predict prejudiced actions, researchers dispute the strength of the association. Implicit bias against women has been researched less — a gap researchers say they intend to address.

But there is evidence that it has an impact on politics more generally. As girls grow up and spend more time exposed to social pressures, they make up a smaller share of aspiring politicians. At age 7, about the same share of girls and boys say they want to be president. But far more 15-year-old boys than girls want to be president, according to a study by Roberta Sigel cited by Heldman in the film “Miss Representation.”6

“Internalized sexism isn’t a heavily trafficked area of study, but if we’re going to successfully run a woman for president, it needs to be,” said Jess McIntosh, who was director of communications outreach on the Hillary Clinton campaign. In November, McIntosh pointed to “internalized misogyny” as one factor behind Clinton’s loss. (Spokespeople for Trump and for his transition team did not respond to requests for comment.)

When women do run for office, they usually win about as often as men — and an unprecedented number of women have shown interest in running for office since Trump’s win, according to groups that recruit women candidates. But Heldman cautions against extrapolating to the presidency from the experience of other women candidates. “The presidency is different from every other office in ways that are gendered,” Heldman said, citing sexism in media coverage of women’s bids for the presidency, such as questioning whether a woman commander-in-chief could handle a major international crisis.

HCD would like to do follow-up studies on the topic, including tracking bias levels over time and examining whether exposure to certain segments of the media affects bias. Michelle Niedziela, scientific director of HCD Research, said she initially found the finding surprising but has since rethought her position.

She recounted a conversation she had after the election with a woman she knows who had voted for Trump. “She was saying things I found to be really shocking, that I had never heard before,” Niedziela said. The woman, a professional, told Niedziela that she didn’t think men and women should get equal pay. “This study reflects these beliefs; they are out there,” Niedziela said. “Maybe we shouldn’t be shocked.”


  1. The participants in Project Implicit’s study aren’t representative of the general population — they are visitors to the project’s website, and 19 percent of them live outside the U.S. More of them are women than men, and test takers’ self-reported education level is, on average, higher than the general population’s. Nonetheless, Smith says there is no particular reason the sample shouldn’t be representative within a specific subgroup, such as conservative black women. And the study has such a big sample size that it has enough people in many subgroups to make comparisons.

  2. Project organizers hope to share 2016 test results with researchers soon.

  3. The data shown for white people and black people is only for those who say they are not Hispanic.

  4. There are far more people who describe themselves as politically neutral among implicit-bias test takers than those who call themselves strongly conservative or strongly liberal, so the sample sizes for the data in the above chart vary widely, from a couple of hundred to tens of thousands (there were more than 1,000 test takers for most subgroups).

  5. In a regression.

  6. Sigel didn’t publish the details of the study before her death in 2008, Heldman said.

Carl Bialik was FiveThirtyEight’s lead writer for news.