The US has yet to elect a woman president, raising questions about the perceived electability of women candidates.Elizabeth N. Simas examines the influence of strategic sexism, where people choose not to support a woman candidate because they feel that others will be prejudiced against them. She finds that as perceived sexism increases, there is no corresponding decline in electability for women. While women may be seen as being less electable than men, these penalties do not seem to be due to people feeling that others are sexist towards them.
In US primary elections, electability, i.e. the ability to win the general election, becomes a factor in who people decide to vote for. One argument is that women are seen as less electable than men due to strategic sexism – this is the discrimination that does not stem from a person’s own sexism, but from a person’s perceptions that others are sexist. This then influences people’s vote choice, as they assume that a woman candidate may be less likely to win a general election because of other people’s sexism. While political scientists have found evidence that this occurred in the 2020 Democratic Primary, I, argue and show that these results may not hold when we look beyond just this specific case.
The role of strategic sexism in elections
First, when considering real candidates like Senator Elizabeth Warren and former Vice President Kamala Harris, it is impossible to know whether people considered them to be less electable because they were women or if these sentiments stemmed from something else. For example, with rare exception, in the 2020 election men consistently led in polls and endorsements, two crucial signals of electability. Thus, views that women were less electable may have been driven by these more objective indicators rather than any kind of gender bias.
Second, these findings may not hold in races that do not feature Donald Trump. Sexism needs to be activated, and Trump did so in a unique way. When combined with the fact that Trump did defeat a woman in the 2016 presidential election, the 2020 Democratic Primary may have provided a best-case scenario for finding effects of strategic sexism that will not be evident in other contexts.
With these and other concerns in mind, I designed a survey experiment to test whether the strategic sexism argument would hold in a more general setting. In October 2022, I recruited 1,166 US adults who identified with either the Democratic or Republican Parties. To simulate a primary election, all respondents were told that they would be shown profiles of two potential candidates for their party’s 2024 presidential nomination and then asked to indicate which of the two would have the best chance of defeating the other party’s candidate in the general election.
Following these instructions, respondents were shown five pairs of candidate profiles. To avoid the issues associated with using real candidates, profiles did not name a specific politician, but rather just contained generic information about the candidate’s sex, age, race, region of residence, political experience, ideology, and policy priority area. After viewing each set, respondents selected which of the two candidates was more electable.
Dream Big Fight Hard” (CC BY-NC-SA 2.0) by jlevinger
Determining candidates’ perceived electability
I used a statistical concept, the marginal mean, to determine the perceived electability of men and women candidates. If being a man or a woman had no impact on perceived electability, then the marginal means would be .5. This, however, is not the case. The marginal mean for being a man (.52) indicates that being male made the candidate more electable, while the marginal mean for being a woman (.48) indicates that being female made the candidate less electable. Though these differences are small, they are statistically significant and align with prior work showing that women are penalized.
But is this penalty driven by perceived sexism? To test this, I develop a measure of perceived sexism, which is the percentage of Americans that each respondent thinks would agree with the statement “women seek to gain control over men.”
If the strategic sexism hypothesis is correct, then we would expect that as perceived sexism increases, the marginal means for being male should increase (i.e., a man should become more electable) and the marginal means for being female should decrease (i.e., a woman should become less electable). But as Figure 1 shows, the marginal means remain quite constant across the range of potential values of perceived sexism. That is, those who perceive little sexism and those who perceive larger amounts of sexism are no different and penalize women in the same way. Thus, while women may be seen as less electable than men, it does not appear that these penalties are due to people feeling that others are sexist towards them.
Figure 1 – Marginal Means of Candidate Sex by Perceived Sexism
We need to look more closely at the influence of sexism and perceptions of it on women candidates
As we continue to ask whether America is ready for a woman president, my research suggests that answering this question requires going beyond the explanations solely related to perceived sexism. This is not to say that women running in the 2020 Democratic presidential primary were not subject to any form of strategic discrimination or that sexism or perceptions of it are not harming women in the contexts examined here. Rather, my central argument is that researchers need to further explore the exact role that these biases play.
And ultimately, the large number of personal and contextual factors that contribute to how individuals determine a candidate’s chances of winning complicate attempts to clearly define just what makes a candidate “electable.” But as more women enter presidential races, it should hopefully allow for more confidence in generalized estimates of any potential biases against them.