Estimating Peer Effects in Corporate Innovation : Evidence from Compensation Peer Groups

This paper uses compensation peer groups to measure peer effects in corporate innovation. This approach provides a true peer group and better leader-follower link and thus can mitigate the reflection problem suggested by Manski (1993). We find that the average innovation activity of the compensation peers is a significant and first-order predictor for corporate innovation. Further analyses show that (1) the peer effect is stronger when peer companies experience higher innovation competition and are closer to the median peer company in the peer group, (2) the results are not likely to be attributed to the knowledge spillover mechanism but rather are more consistent with the peer pressure mechanism, and (3) the SEC’s 2006 executive compensation disclosure rules can generate the peer effects. Overall, the results suggest that corporate innovation decisions are responses to innovation activities and, to a lesser extent, the characteristics of the compensation peer groups