An artificial intelligence model in a recent study conducted in Denmark found conservative female politicians more attractive and happier in photos than their liberal counterparts.
Published in the journal Scientific Reports in March, the research revealed that AI is 61% accurate in predicting a person’s political affiliation after analyzing one headshot.
Approximately 3,200 publicly submitted photos of political candidates who ran in the 2017 Danish municipal election were used in the study. The researchers inputted the photos into Microsoft Azure’s Face API tool to evaluate the person’s emotional state, with 80% of the faces analyzed read as happy and 19% read as neutral.
“For females (though not males), high attractiveness scores were found among those the model identified as likely to be conservative,” the findings read. “These results are credible, given that previous research using human raters has also highlighted a link between attractiveness and conservatism.”
The study also found that left-leaning male politicians displayed more neutral faces that were less happy than conservative politicians.
“Attractiveness was not the only correlate of model-predicted ideology,” the scientists said. “We also found that expressing happiness is associated with conservatism for both genders.
“Previous work has found smiling in photographs to be a valid indicator of extraversion,” they continued. “And while extraversion is not broadly associated with ideology, some studies have found that right-wing politicians are more extraverted.”
Because attractiveness helps boost the chances of electoral success, the scientists noted that “all candidates are incentivized to provide an attractive photograph.”
“Politicians on the left and right may have different incentives for smiling — for example, smiling faces have been found to look more attractive, which is comparatively important for conservative politicians,” the paper read.
Further research is needed to determine “the extent to which happy faces are indicative of conservatism outside of samples of politicians,” the researchers concluded.