The Myth of Female Unelectability

Perhaps nothing has been more damaging to women running for office than the idea that voters simply won’t pick female candidates. There’s just one problem: It isn’t true.

After Hillary Clinton lost the 2016 presidential election, many people, including some of her top staffers and the unsuccessful Democratic nominee herself, concluded that she had been penalized for her gender. Even two years after the election, Jennifer Palmieri, her former communications director, argued that “I think that a man would have survived” the barriers Clinton faced, such as the scandal over her emails. Clinton continues to push this idea, saying as recently as May that some voters—women voters—had held her to an impossible standard and taken a chance on Donald Trump because he’s a man.

As the 2020 Democratic presidential-primary race took shape, many in the party were apprehensive about nominating a woman. Neera Tanden, then the president of the Center for American Progress and now President Joe Biden’s Domestic Policy Council director, worried that “there’s a fear that if misogyny beat Clinton, it can beat other women.” Several female candidates, including Kamala Harris and Elizabeth Warren, competed for the nomination. When they lost, concerns about the power of sexism became even more entrenched.

But was the anxiety warranted? And is Harris, now the Democratic nominee, doomed to follow in Clinton’s footsteps? Although isolating the impact of gender is difficult—we’ll never know whether a Henry Clinton would have outperformed Hillary Clinton in 2016—existing research indicates that today’s voters do not systematically discriminate against women at the polls.

This isn’t to say that voters treat men and women the same when they run for office. Gender stereotypes abound, and women face attacks that men never would. And, of course, no woman has ever been elected president in the United States. But the research keeps getting clearer: Women can—and do—win. In large part, women win because even if voters hold sexist views, they also hold other views—on economic policy, abortion, immigration, and more. As the distance between the parties has grown on these issues, the cost of allowing sexism to turn you against your party’s nominee has also grown. Finally, a win for political polarization!

Many conversations about sexism and women in politics fail to distinguish between two questions: First, do women experience gender-based attacks when they run for office? And second, does being a woman make a candidate less likely to win an election?

The answer to the first is clearly yes. “It would be ridiculous and foolish to suggest that women don’t receive different attacks,” the Yale University political scientist Alexander Coppock told me. But, he added, “you have to hold that in your mind alongside the idea that every candidate—man, woman, nonbinary—[will be] attacked, and the precise content of that is going to vary depending on the opponent and the candidate themselves.”

People weren’t hallucinating gendered attacks on Clinton. Trump’s supporters really did wear shirts calling her a bitch. And just a brief perusal of X, TikTok, and other social-media platforms in the days following Harris’s entry into the current presidential race revealed abhorrent gender-based attacks on her too.

Historically, women did face an electoral penalty. Several studies found that, half a century ago, men tended to outpoll women in a number of Western democracies.

Yet the penalty has disappeared. When Coppock and the political scientist Susanne Schwarz reviewed more recent research on voter attitudes toward women candidates, they found that the empirical evidence of voter bias was “surprisingly thin.” In 2022, Schwarz and Coppock published the results of their meta-analysis of 67 experiments from all over the world in which researchers asked survey respondents to choose among hypothetical candidates with varied demographic profiles. Schwarz and Coppock concluded that the average effect of being a woman is not a loss; rather, it’s a gain of approximately two percentage points.

Their conclusion wasn’t unusual. When Schwarz and Coppock looked closer at studies that also failed to find voters punishing female candidates, they realized that, time and again, the original researchers had been taken aback by their own findings, even as their field was zeroing in on a consensus.

To be sure, the recent literature doesn’t show that voters treat male and female candidates identically. As the political scientists Sarah Anzia and Rachel Bernhard wrote in a 2022 paper, “Some voters infer that women candidates are more liberal than men, more compassionate and collaborative, and more competent on certain issues like education.” Anzia and Bernhard’s paper reviewed local-election results and concluded that, on average, women have an advantage over men in city-council elections, but that this advantage declines in mayoral races. Male and female mayoral candidates win at essentially the same rates; however, when the authors compared candidates with similar levels of experience, men pulled ahead, a finding replicated in other studies.

Anzia and Bernhard also found that the more Republicans within a constituency, the greater the disadvantage to women. Schwarz and Coppock identified a similar effect. But are Republicans discriminating against women because they are women, or because they correctly intuit that, on average, women are more liberal than men? Interestingly, this dynamic may be strongest in downballot races—which are more likely to be nonpartisan and may receive little rigorous news coverage. The effects of stereotyping, Anzia and Bernhard wrote, “are largest in elections when voters tend to know less about local candidates.”

Harris, by skipping the traditional primary process in 2024, avoided one lingering way voters may punish female candidates at the ballot box. In a paper about “pragmatic bias” in the 2020 Democratic presidential-primary season, the Stanford sociologist Christianne Corbett and three colleagues surveyed likely Democratic-primary voters and found that respondents indeed expected that then-candidates Harris and Warren would have more difficulty beating Trump than Biden or Bernie Sanders would. The researchers found evidence that some voters who labeled Harris or Warren as their favorite candidate nevertheless said that they intended to vote for Biden or Sanders, at least in part because they thought a female candidate was unelectable.

But Corbett told me she can’t see how this would affect Harris in the general election. “The two candidates are so different,” she said. Party identification, she predicted, will overcome pragmatic bias. Although primary voters might reasonably conclude that Warren and Sanders would support most of the same policies if elected, few voters who support Harris’s positions would see Trump as an acceptable substitute.

The political-science literature has some limits. Many of the most applicable studies that best control for the effect of gender in elections ask voters about hypothetical candidates. Until now, the sample size for real-life American major-party female presidential nominees was … one. The boundary between fair and unfair scrutiny of specific female candidates is difficult to define. For instance, one recent Time article saw evidence of sexism in criticism of Amy Klobuchar’s treatment of her aides and of Warren’s claims of Native American ancestry. But doubts about a candidate’s judgment are a legitimate reason not to like her.

And worries about Harris’s electability hinge on not just her gender but also her identity as a Black and South Asian woman. Here, too, research should reassure her supporters. A 2022 meta-analysis of 43 candidate experiments from the preceding decade could “not find any evidence for voter discrimination against racial/ethnic minority candidates.” Rather, underrepresentation of groups in elected office likely comes from “supply-side effects”—disadvantages crop up against racial and ethnic minorities before they ever run for office.

Yes, some voters may be sexist, racist, or both. But that doesn’t mean they won’t vote for Harris. You can be sexist and racist but still prefer her party and her platform to her opponent’s.

Although women make up slightly more than half of the population, the 118th Congress, convened in January 2023, was 72 percent male. That’s still a problem, but it isn’t voters’ fault. The research on women’s electability focuses on a very specific question: Do voters penalize women for their gender when they run for office? But even if they don’t, other parts of the electoral process clearly do. Women themselves don’t run for office at the rates of men. When they do run, they may face barriers behind the scenes from donors who discriminate against women—or even just plain incumbency advantages that lock in longtime elected officials, who are more likely to be male. The political scientist Jennifer Lawless has described a large pipeline issue for women—men are more likely to say they want to run for office and be encouraged to run for office, and are more likely to perceive themselves as qualified for the job.

Harris has already cleared all of these hurdles, and she appears to be doing her best to inoculate herself against attacks that she is too liberal, one way that voters may penalize female candidates. Already, she is seeking to moderate her image by promising to sign border legislation and breaking with past views about fracking.

But although the Democratic nominee and her team see a path to victory, they still seem to believe that being a woman is an electoral liability. Harris’s running-mate shortlist included only white men. Two talented midwestern electoral overperformers—Klobuchar and Gretchen Whitmer—didn’t make it.

The irony is that the most consequential gender-based critiques that women candidates face aren’t coming from sexists and bigots. They’re coming from many people who desperately want to see a female candidate elected.

In other words, stop telling voters a woman can’t win. They might start to believe you.