LinkedIn: A New Frontier for Gender Bias in Hiring Practices

Abstract

In the realm of employment, the importance of recommendations is well-established. Yet, the impact of gender-based differences in recommendation content remains underexplored. Further, prior research has largely neglected the use of LinkedIn for professional recommendations. To address these gaps in literature, I conducted an original study with 1,000 recent university graduates to investigate gender differences in the quality and quantity of recommendations given and received by men and women. I used natural language processing methods and textual analysis to extract and classify recommendation content. My findings reveal mixed results with regards to gender homophily in recommendations. While women were similarly likely to receive recommendations from men and women, men were more likely to receive recommendations from men than women. In terms of content, the key difference I find is that men are described as more agentic and meaningfully managerial than women, with fairly minimal differences in other categories. For instance, both men and women received similar amounts of praise and positive valence in their recommendations. Together, these findings may reinforce gender stereotypes and have implications for organizations, hiring managers, and LinkedIn users evaluating job candidates.