ATLANTA—Emotional responses to the first presidential election debate of 2016 between Donald Trump and Hillary Clinton were significantly different across genders, according to research that used a novel technique to assess the facial expressions of viewers.
The research team from Georgia State University and Arizona State University found that the impact of emotions was stronger for women viewers and that gender issues were more important for them as well.
“Women expressed internalizing emotions like sadness twice as often as men,” said Sarah Allen Gershon, an associate professor of political science at Georgia State, and corresponding author of the paper. “Men expressed externalizing emotions like anger and disgust more often. There were also differences in when men and women expressed these emotions. Our data indicated that women’s expressions of sadness and men’s expressions of anger increased during periods when Clinton spoke more.”
Gershon stressed that gender was uniquely at the fore of the event, which had the largest audience of any U.S. presidential debate in history with 84 million viewers. Clinton was the first woman to be nominated for President of the United States by a major political party. There had also been prominent discussions of gender during the campaign, including Trump’s high-profile comments about women’s appearances, and Clinton highlighting her glass-ceiling-breaking nomination.
“We found that the impact of emotional expressions was more powerful for women than men in predicting post-debate evaluations of the candidates’ debate performance, particularly in their evaluations of Trump’s performance,” Gershon said. “For example, as women showed more expressions of fear, their ratings of Trump’s performance decreased significantly. However, as men expressed more fear, their evaluations of the candidates’ performance barely budged.”
Unlike traditional political science debate assessment that relies on post-debate interviews or real-time self-reporting, Gershon and her collaborators recorded viewers with webcams as they watched the debate on desktop computers. The researchers then used facial-recognition software for a frame-by-frame analysis of emotional expression. The technology features a facial expression coding system that tracked 20 unique movements such as furrowing the brow, wrinkling the nose or frowning — markers of particular emotions like surprise, sadness, anger or disgust, Gershon said.
Though facial recognition software has been used for years in marketing research, Gershon said it is a relatively new tool for political scientists.
“Now that we have the software, it would be interesting to look at how people are looking at different types of stimuli like speeches or ads or website messaging,” she said.
“The broad question that we are concerned with for this study is, do debates matter? Are they educators for the population? The answer from our paper is that debates do matter in distinct and different ways for people watching them.”
Gershon’s co-authors were Kristina LaPlant, a political science Ph.D. candidate at Georgia State; and Kim Fridkin, professor of political science, and Jillian Courey, Ph.D. candidate, both from Arizona State University. The article, “Gender Differences in Emotional Reactions to the 2016 Presidential Debates” was published in the journal, Political Behavior.
Dr. Gershon’s research focuses primarily on the incorporation of traditionally underrepresented groups (including women, and racial and ethnic minorities) into the American political system. In seeking to explain the challenges faced by these groups, her work emphasizes the role of communication, campaigns and political attitudes.