Abstract

Gender segregation in interests, activities, and lifestyles is a ubiquitous phenomenon: girls like shopping, makeup, and Victoria Secret; guys like games, sports, and Iron Maiden. In recent decades, gendered interests and activities are increasingly promoted by media, popular culture, and consumer markets. This trend animates debates about its implications for gender equality, particularly whether it reinforces gender stereotypes and brings disadvantages to women. Combining digital records of interests – Facebook Likes – and machine learning algorisms, we are able to measure how gender-stereotypical people’s interests are in a given population. We show that such gender segregation in interests is strongly associated with indices of state-level gender inequality in the US. Our research showcases the potential for using big data to advise social policy related to gender equality.