When voters can almost instantaneously determine a female politician’s gender from her facial features, that politician is more likely to win her election — above and beyond any issues of competence.

And that outcome is especially true in conservative areas of the country.

Those are the intriguing (and discouraging) key findings of a study published Thursday in the journal Social Psychological and Personality Science.

“Though perceptions of competence and attractiveness from politicians’ faces have been previously linked to political success, the current research demonstrates that gendered cues uniquely predict female politicians’ electoral success beyond these factors,” write the study’s authors.

This finding suggests “a discrepancy between traits used to evaluate male and female politicians,” they add.

A variety of cues

For the study, researchers used a software technology called Mouse Tracker, which was developed by the study’s senior author, Jon Freeman, an assistant professor and director of the Social Cognitive & Neural Sciences Lab at Dartmouth University. Mouse Tracker can be used to measure real-time computer mouse movements as people make snap decisions — in this case, the decision of whether a photo of a face flashed on a computer screen was male or female.

Previous research has shown that people use specific biological and social cues to identify gender from facial features. Larger eyes, for example, tend to suggest femininity, as do longer hair and makeup (at least in Western cultures). Prominent upper brows and wider jaws, on the other hand, tend to convey masculinity.

And, yes, most faces do contain some ratio of overlap between these gender categories. That’s why people can encounter a split second of uncertainty about another person’s gender upon seeing their face.

Following the mouse

Freeman and his colleagues decided to assess if and how that initial gender uncertainty could predict real-world electoral outcomes. First, they assembled almost 200 images of men and women who had run in a gubernatorial or U.S. Senate race between 1998 and 2010. Each image was rated for perceived attractiveness and competence by an online panel recruited through Amazon’s Mechanical Turk website. The researchers excluded from the images any well-known politicians, such as former Gov. Sarah Palin (R-Alaska), and included only a few non-white candidates to avoid introducing racial bias in the participants’ responses.

The researchers then asked 32 Dartmouth undergraduates (including 21 women) to use a computer mouse to categorize each of the images as either female or male — and to do it as quickly as they could. The students were also asked if they would vote for the person in the image (on a scale of 1 to 6), although they were not provided any information about the politicians’ political views or background.

The results showed that the longer it took for the students to select the correct gender of a female politician’s face, the less likely they were to say they would vote for that candidate. The same was not true for the male politicians.

Predictive of election results

Freeman and his colleagues then recruited 260 people (119 women) through the Mechanical Turk site to complete the same categorization of the images online. The researchers compared the speed of these participants’ mouse-tracking responses to the politicians’ latest election results — specifically, their margins of victory or defeat.

They found that the split-second speed at which a female politician’s gender was selected could predict her electoral outcome. The longer it took for the respondents to process her gender identity, the less likely she was to win her election.

“These effects were independent of other social dimensions previously implicated in political decision-making, such as competence, attractiveness, and familiarity,” Freeman and his co-authors write.

We’re not talking here about a long hesitation, either.

“A female politician’s success was related to how feminine or masculine her face was perceived less than on half-second after its initial exposure, suggesting that the way a face’s gender is rapidly processed may translate into real-world political outcomes,” said Freeman in a statement released with the study.

The influence of political leanings

The researchers also found that the effects seen in these experiments were more pronounced among participants in conservative states. (They determined a state’s conservative/liberal leanings by looking at the results of the last five presidential elections.)

“It is important to note that participants across these different regions were equally sensitive to gendered facial cues,” write Freeman and his colleagues. “Rather, what varied was the relationship between this partial activation and electoral success, as more masculine female politicians received a lower percentage of votes in more conservative constituencies.”

“This finding could be explained either by conservatives’ appreciation of traditional gender norms, their aversion to uncertainty, or both,” they add.

You can read the study in full online. And you can view a video demonstrating the Mouse Tracker technology in action on the website of Dartmouth’s Social Cognitive & Neural Sciences Lab. The video shows the differences in results when study participants were shown the faces of Nikki Haley, the current Republican governor of South Carolina, and Judy Bear Topinka, the comptroller of Illinois who ran on the Republican ticket for governor of that state in 2006. Haley won her gubernatorial election by a four-point margin; Topinka lost hers by 11 points.