Nonprofit, nonpartisan journalism. Supported by readers.


UCare generously supports MinnPost’s Second Opinion coverage; learn why.

How we use stereotypes to identify people’s political affiliations

“After a certain number of years, our faces become our biographies,” the writer Cynthia Ozick once quipped in the Paris Review.

Apparently, one of those biographical facts we wear on our faces is our political affiliation. According to a study published this month in the open-access journal PLoS ONE, people can identify with remarkable accuracy (more than by chance guessing) whether another person is a Republican or a Democrat by simply looking at that person’s headshot.

How do we do it? By relying on stereotypes, the study found. Republicans, apparently, look “powerful” in our minds, and Democrats appear “warm.”

Of course, these kinds of stereotypes can lead to perceptual errors. “Not all Democrats appear warm and not all Republicans appear powerful,” wrote the study’s authors. “However, the linearity of these effects is noteworthy: appearing warmer led to a greater chance that a target would be perceived as a Democrat and appearing more powerful led to a greater chance that a target would be perceived as a Republican.”

Experiment #1
The study, which was conducted by Nalini Ambady, Ph.D., a social psychologist at Tufts University in Medford, Mass., and Tufts doctoral candidate Nicholas Rule, involved three separate experiments.

In the first experiment, 29 undergraduates were asked to categorize the faces of 118 unnamed professional politicians (2004 and 2006 U.S. Senate candidates).The photos (cropped to be of identical size and converted to grayscale) included women candidates, but minority candidates were excluded to avoid race-based stereotypes.

After the data was analyzed, the study found that participants had categorized the photos correctly at a rate that was significantly better than chance guessing. Those results held even when the responses of 10 participants who said they recognized at least one of the candidates were excluded from the calculations.

Experiment #2
To see if the results of the first experiment could be extended to other groups of people, the researchers conducted a second experiment. This time, they asked the participants to categorize the political affiliation of photos take from the senior yearbooks of a private U.S. university — photos of students who had described themselves in the yearbook as belonging to either a Democratic or Republican campus organization. (None of the participants in this experiment recognized any of the faces shown to them.)

Again, the participants’ categorization of the political affiliations of the students in the photos was significantly greater than chance guessing.

Experiment #3
Intrigued by these findings, the researchers decided to determine what, exactly, people were using to determine if someone were a Democrat or a Republican. They set up a third experiment in which participants were asked to rate the faces of college-aged Democrats and Republicans (without knowing who was whom) on three traits: dominance, facial maturity likability and trustworthiness.

They then took the data from these responses to form two composites: “Warmth” (an averaging of the mean scores for likability and trustworthiness) and “Power” (an averaging of the mean scores for dominance and facial maturity). (Such composites are commonly used for these kinds of experiments.)

Faces perceived to be that of Republican scored higher on the “Power” scale and those perceived to be that of a Democrat scored high on the “Warmth” scale.

Judging a book by its cover
Other research has pointed out that we’re quick to make snap judgments about the people we meet based on their appearance — and often, of course, unfairly. “People are known to form impression of others from their faces instantaneously and automatically,” write Rule and Ambady. “Moreover, these perceptions can have highly consequential outcomes, such as affecting the jobs that individuals are offered, their outcomes in court, and their financial success.”

Rule and Ambady also describe some truly provocative research about how election results can be predicted by the candidates’ facial traits:

One particularly consequential judgment is political candidates’ actual electoral success based on perceivers’ naïve judgments of personality traits from the candidates’ faces. Several studies have found that judgments of competence and power from the faces of political candidates in Western cultures are significantly related to the candidates’ margin of victory. Indeed, even children’s judgments of politicians’ faces can predict their electoral success, and judgments of power and warmth from faces can predict electoral outcomes across cultures.

Something to keep in mind as we head into this year’s heated (to put it mildly) election season.

You can also learn about all our free newsletter options.

Comments (5)

  1. Submitted by Karen Sandness on 01/28/2010 - 01:52 pm.

    When I was out canvassing for Kerry in 2004, I could tell whether I’d get a warm reception the moment a resident of any given house opened the front door.

    I’m not sure how I did it, but I was never surprised when a person who looked Republican in some indefinable way said, “I’m voting for Bush.”

  2. Submitted by Ed Stych on 01/28/2010 - 01:56 pm.

    I would be curious if the photos were cropped tight enough to eliminate hair as an issue, with length of hair and color of hair creating stereotypes. Also, it would be interesting to do the study with only one gender. I think the stereotype is that women tend to be more liberal than men.

    The comment that Republicans apparently have more mature faces reminded me of one of Churchill’s more famous quotes:

    “If you’re not a liberal at 20, you have no heart. If you’re not a conservative at 40, you have no brain.”

  3. Submitted by dan buechler on 01/28/2010 - 02:16 pm.

    Great. Another idea read the 2 essays yesterday by Kline and Klobuchar quite a contrast in style. Would enjoy an analysis by a political rhetorican or a linguist. Fun.

  4. Anonymous Submitted by Anonymous on 01/29/2010 - 01:39 pm.

    Fascinating. I have categorized Republican cars,(big gas-guzzlers), Republican homes (usually very tidy with cute stuff like dwarfs in the yard)–you get the idea. But this one is news to me. But of course, Dan has nailed it in comment #3!

  5. Submitted by Beryl John-Knudson on 01/29/2010 - 04:36 pm.

    It was a few years ago and somebody got the pictures mixed up on the front page of a paper which shall remain nameless.

    One was a picture of a child molester from a wanted list. One was a photo heading a feature story on the new Lutheran minister in town. The third tied in with a political blurb on a Republican running for local office.

    All three were look-a-likes. Each had a face like a thumb and all were wearing identical, rimless glasses.

    It was a week before anyone noticed.

    Thank the gods I wasn’t planning on voting for any of them.

Leave a Reply