When Joe Biden decided to end his campaign for president and Kamala Harris quickly seized the nomination, Democrats solved the concern about Biden’s age. But these events raised a different question – could a woman finally take the White House?
After all, many voters express sexist sentiments in surveys. For example, according to the 2022 Cooperative Election Study, one in four American adults agreed that “women seek to gain power by getting control over men.” And 35 percent agreed that “women are too easily offended.” Social scientists use statements like this to measure “hostile sexism” in voting and other decisions.
With many Americans holding sexist attitudes, it is reasonable to assume that they would be less willing to vote for a woman candidate for president than a man. But this is not really how sexism relates to vote choice in contemporary American politics. Sexist voters, the surveys show, tend to vote for Republicans while non-sexist voters support Democrats. This pattern holds regardless of candidate gender. For this reason, Harris won’t win over the vast majority of sexist voters – but it’s also unlikely that any Democratic nominee would.
Sexist voters opposed Clinton in 2016 – but not necessarily because she was a woman
To understand this point, it’s useful to go back to 2016. Political scientists have published a wealth of studies showing that sexist attitudes were a strong predictor of whether voters supported Hillary Clinton or Donald Trump in 2016. It might be natural to assume that this finding arose from the fact that Clinton was seeking to become the first female president. But a deeper dive into the data suggests a more nuanced story.
The graph below uses data from a study one of us co-authored showing how sexism predicted presidential vote choice in 2016. Here, we focus on respondents’ favorability ratings of Barack Obama and Hillary Clinton.
Notably, people who agreed with sexist statements in the survey also rated Obama and Clinton unfavorably at nearly identical rates. To the extent that Obama was more popular than Clinton, it was actually among voters who rejected sexist statements. This is not what we would expect if sexist voters were penalizing Clinton, in particular.
Biden improved on Clinton’s rating among non-sexist voters
Fast forward to 2020, when Joe Biden’s nomination to the Democratic ticket seemed, at least partly, to be a response from Democrats to the idea that Clinton lost in 2016 because of her gender.
For instance, political scientist Regina Bateson demonstrated that some Democratic primary voters engaged in what she termed “strategic discrimination” – voting to nominate a white man because they worried a woman couldn’t win the general election. Jon Green and one of us also found evidence for this dynamic.
But it turns out that Biden’s improvement over Clinton’s favorability didn’t have much to do with sexism. According to data from the 2020 Cooperative Election Study, Biden won about 4 percent of the vote among voters who agreed with sexist statements and did not support Clinton in 2016. But among voters who disagreed with sexist statements and didn’t vote for Clinton in 2016, Biden won 17 percent. In other words, Biden’s gains among Clinton opponents were bigger among non-sexist voters in 2020 than among sexist ones.
Harris fares no worse than Biden among sexist voters
But what about Harris? Polling data from the last four years suggests that sexism plays virtually the same role in explaining her support as it does for Biden’s.
In October 2020, toward the end of the presidential campaign, Biden and Harris’ favorability ratings were virtually identical for respondents at various levels of sexism. For instance, among respondents who expressed the strongest agreement with sexist statements, both Biden and Harris had favorability ratings of approximately 12 on a 100-point scale.
And after nearly four years with Harris as vice president, that pattern has persisted. The graph below presents data from the George Washington University Politics Poll conducted in June, just before the first debate and before Biden dropped out.
The favorability ratings for Biden and for Harris were much the same among voters who expressed varying levels of support for a series of sexist statements. Once again, both Democrats get their highest ratings from the least sexist respondents and the lowest from the most sexist. But at every level of sexism among survey respondents, the scores for Biden and Harris are statistically indistinguishable.
Sexism is baked into partisanship
Does all of this suggest that gender attitudes and sexism are irrelevant in American politics? Hardly.
In fact, sexism plays a substantial role in explaining whether people vote for Republicans or Democrats – but only a limited role in explaining whether they vote for men or women.
For example, in an analysis of U.S. House elections in the 2018 midterms, one of us found that voters expressing more sexist views were more likely to vote for Republicans than Democrats. But the magnitude of that effect was the same regardless of whether the candidates were men or women.
The reason is that attitudes about gender have become baked into people’s partisan attachments. Most people with more conservative attitudes about gender are Republicans, and most people with more liberal attitudes are Democrats.
These attitudes already marked a partisan dividing line before 2016. Trump’s candidacy and presidency appeared to further strengthen that relationship. Studies have shown that sexism was a stronger predictor of vote choice in House elections in 2018 than it had been in 2016. And the correlation between people’s partisanship and their views towards women has increased substantially since the 2016 election.
The fact that Democrats have nominated a woman for president for the second time in three election cycles is one illustration of the way the parties have adopted differing views about gender. But it is precisely because of those existing cleavages that sexism is not likely to be the threat to Harris’ candidacy that some observers have speculated it to be.
