Toward an improved causal test of network effects: Does alliance network position enhance firm innovation?

Research Summary: Assessing whether network position causes firm outcomes is difficult because networks result from firms' strategic choices. To address this issue, we develop a research design that separates self-driven network change from other-driven change. Because other-driven change suppresses the focal firm's agency, outcomes resulting from it can be interpreted as causal—under verifiable assumptions. We apply this method to reassess whether alliance network position affects firm innovation. Using data from the biotechnology industry, we find that structural holes (but not closure) increase firm innovation when self-driven network change is involved but not when other-driven change is involved. This raises two possibilities: network effects spuriously capture unobservable firm characteristics or network theories must be updated to account for agency. Managerial Summary: We examine a crucial question for managers making decisions about alliance networks: Does a firm's position in its alliance network drive innovation, or is it a reflection of the firm's inherent qualities? Our research explores this question by examining network changes outside a firm's control. We find that bridging disconnected partners (structural holes) enhances innovation, but only when firms actively create these positions and not when these positions result from others' actions. Our findings suggest that network position alone does not drive innovation. Acknowledging that network position, per se, is not advantageous provides nuanced insight to better guide managerial decisions.