Who Leads? Relative Age Effects on Social Capital

• In Event: Mentorship and Peer Effects in Education

Abstract

This paper studies the causal effect of being the oldest in a school cohort on social capital and labor market outcomes. Using a fuzzy regression discontinuity design and data on 27 million U.S. Facebook users, we find that boys who are older than their classmates make 7% more friends in high school. This social advantage matters — older boys are 40% more likely to become class president. By contrast, relatively older girls make the same number of friends as their younger peers. Men who were relatively older in school have larger and more diverse social networks, and are more likely to host events, moderate online groups and to be married. Consistent with a Roy model of occupational sorting according to comparative advantage, these social capital differences also predict career outcomes. By linking job titles on Facebook to the O*NET database, we show that relatively older males are more likely to sort into socially-intensive occupations, management and entrepreneurship. These results suggest that small age differences in school peer composition can shape long-term economic outcomes through their effects on social capital.

Author

• Matthew Jacob, Harvard University
  Presenting Author