SSN Basic Facts

How Voters Use Stereotypes about Potential Romantic Partners to Assess Election Candidates

Policy field

Connect with the author

University of California, Berkeley

Democracy relies on the ability of voters to assess candidates’ qualifications for office, but decades of inquiry by scholars and the media raise questions about whether voters can do this effectively. One issue is whether American voters discriminate against women running for office. My research shows that gender matters to voters, but in a different way than earlier work suggested. I find the candidates voters believe to be most qualified are those perceived to be potentially desirable mates and partners. Female candidates gain voter approval if they are physically attractive and perceived as nurturing, while male candidates gain support based on attractiveness and perceived ability to provide for and protect children. When such unconscious preferences are added up across an entire electorate, they influence election outcomes.

Why Do We Pick Partners Instead of Politicians?

Decades of research suggest that fully informed voting is costly: Rather than take the time to exhaustively research every candidate and policy, many voters, including experts, rely on mental shortcuts – called heuristics – that allow them to make assessments. Research shows that when faced with a difficult problem, people often solve an easier problem instead. Building on this insight, my research finds that when faced with the difficult task of choosing between strangers running for office, voters undertake the more familiar task of assessing whether these people seem like promising romantic partners.

People have been choosing mates for millennia, so perhaps it is not surprising that such routines unconsciously shape decisions about candidates. The criteria humans have used to evaluate romantic partners are reproductive fitness – often a matter of perceived physical attractiveness – and ability to help offspring survive to adulthood – which means female nurturing ability and male capacity to provide for or protect their families. Overall, people judge women more heavily on attractiveness, and men more heavily on provider/protector qualities. What is surprising is that these preferences still have an effect even when people are given other, more politically relevant information – such as information about candidates’ government experience and occupational background. I find that these tendencies hold true in data about legislative elections in Oregon from 2000 to 2014 and in data from large online survey experiments.

Are Good Partners Likely to be Good Politicians?

If being a good partner means that a person is more likely to be a good politician, then perhaps it is not such a bad thing for voters to use this heuristic. However, for female candidates, my research shows that these assessments have no relationship to real qualifications for public office, and for male candidates, assessments have only a slight relationship to their qualifications.

  • Compared to less qualified females, women candidates with more government experience, more education, or who came from professional backgrounds relevant to public service were not any more likely to be perceived by respondents as highly qualified. In contrast, respondents perceived men with more experience and better jobs to be slightly more qualified for office.
  • Men who held jobs that respondents say are helpful preparation for holding office, like attorney or union leader, did significantly better on Election Day than their peers; in contrast, women who held those jobs did worse. This shows a new “double bind” for female candidates: Women who come from these occupations not only do worse than men, they do worse than women who come from jobs that respondents see as less helpful preparation, such as park ranger or nurse.
  • Other information, like the candidates’ education levels or professional class, did not meaningfully predict the shares of votes respondents would give to different candidates.

These studies cannot assess whether voters are correctly predicting politicians’ actual performance in office, because some candidates lose and thus have no record. But these findings do show that voters often ignore the kinds of candidate qualifications they claim are important and instead assess the personal appeal of candidates as potential partners.

Implications and Possible Improvements

Personal stereotypes unconsciously invoked by voters can sway election outcomes. Because voter choices are important in any democratic system, considerations that influence those choices have the potential to undermine the performance of democratic governments by routinely advancing less qualified but seemingly appealing candidates. For instance, it is conceivable that voters could select a man who is less politically experienced but appears to be a good provider and protector over a woman who is more experienced but lacks maternal or physical appeal.

Policymakers could take steps to help reduce the impact of unconscious biases. Some states, for instance, include photos of candidates in their voting pamphlets; this practice should be ended because it may exacerbate reliance on mental shortcuts and unduly influence even very informed voters. Elimination of such photos could have an especially strong effect on the pool of candidates who prevail in contests for lower-level public offices, because voters rarely have extensive information about such candidates and thus may rely too much on unconscious cues. In turn, if politicians who serve at lower levels of government become more qualified, that will over time improve the quality of candidates who move up to run for state and national offices.

Finally, the findings suggest a need for creative, out-of-the-box thinking about election reforms. Many companies and organizations use blind auditions or interviews to reduce possibilities for racial, gender, and other identity biases in the hiring process. State and local elections offer perfect laboratories for similar efforts. Parliamentary elections in which voters choose party slates could reduce the sway of inappropriate stereotypes, if we presume that parties could do a better job of assessing candidate qualifications. However, we can also imagine elections where voters make choices based on candidate statements about policy positions, rather than after lengthy campaigns focused on personalities. Whatever the format of the reforms, it is time to recognize that inappropriate unconscious preferences can produce serious distortions of electoral outcomes in our democracy.


Research for this brief is drawn from the author’s ongoing projects.