Learning from People Like Me: The Impact of Demographic Diversity on Invisible Knowledge Consumption and Performance in Online Knowledge Exchange Platforms

The proliferation of online knowledge-sharing and online learning platforms with broad accessibility to participants across the globe has instigated a shift in how people acquire knowledge and new skills. These platforms bring together a demographically diverse pool of individuals, creating opportunities for individuals to tap into diverse experiences, skills, and cultures to enhance their learning experiences and improve performance. However, a persistent question is to what extent opportunities to engage and consume content from demographically diverse peers are beneficial to individuals’ performance on the platform. To address this question, we study 2.5 million online interactions among 12,000 working professionals enrolled in an online business training program offered by a large elite U.S. university over a six-year period. Our empirical strategy exploits a natural experiment setting in which participants are randomly assigned to cohorts of peers as well as fine-grained data on individuals’ knowledge consumption patterns on the course platform. Our results show that being assigned to a cohort with more demographically similar peers causally increases performance and course completion. Then turning to the impact of individuals’ knowledge consumption patterns on the platform, we find that the positive effect of demographically similar peers on performance is mediated by individuals’ tendency to consume content from individuals who are like them in terms of age, gender, and nationality. Our study makes several important contributions to the literature on online knowledge sharing and has important implications for the effective design of knowledge exchange and learning platforms.