Networks in the board of directors
Working Paper, 2022. Using over 70,000 first-time director appointments from 1990 to 2020, I reassess the role of professional connections in the board selection process. While previous literature suggests that direct ties to incumbent directors increase the likelihood of appointment, I show that this effect is overstated when firms draw from segmented candidate pools, such as industry-specific networks. I develop a two-stage model in which the probability of being considered depends on observable characteristics, and the board selects from among the considered candidates. Empirically, I find that candidates with broader professional networks are significantly more likely to be shortlisted and appointed, consistent with the hypothesis that networks serve as screening tools. In contrast, direct personal ties to the board have no positive effect once consideration is modeled, and are associated with relatively lower odds of appointment. Supporting evidence from board committee appointments suggests that personal connections and favoritism are unlikely to explain the observed patterns. The results imply that prior estimates conflate a candidate’s relevance with their probability of appointment. In practice, boards appear to prefer candidates with broad, well-documented experience over those with personal ties to insiders.