A Leader and a Lady? A Computational Approach to Detection of Political Gender Stereotypes in Facebook User Comments

Aliya Andrich, Emese Domahidi


Voters tend to perceive female and male politicians differently, viewing women in politics through the prism of existing gender stereotypes. Although social media have become one of the key platforms for political communication, little is known about stereotypes that social media users communicate about political candidates. This study investigates how gender influences citizens’ evaluations of more than 500 U.S. politicians on social media. Drawing on a large sample of Facebook user comments (n = 13,866,507), we find that female politicians are discussed using traits describing women’s personality and appearance. We also show that users associate female politicians with leadership, competence, and empathy. However, the results are different for highly prominent politicians. Specifically, our findings support the idea of leadership roles being more strongly linked to the masculine stereotype, as we observe that Donald Trump is more strongly associated with masculinity and traits relevant for a political career than Hillary Clinton.


gender stereotypes, politicians, social media, Facebook, computational methods, word embeddings

Full Text: