Abstract: We investigate whether the effect of alliance network position on innovation is causal or spurious. Do firms innovate more because of their structural position, or do they occupy that position because of their innovation strategies and capabilities? To disentangle cause and effect, we advance a theoretical concept—alliance-network externalities—that distinguishes ego-driven network changes (endogenous) from alter-driven changes (exogenous). We further develop a novel methodology to identify those two types of network change and assess the causal effect of network position on innovation. We apply our approach to re-evaluate the relationship between structural holes and firm patenting in the biotechnology industry alliance network (1995-2012) and find no evidence of a causal effect. Structural holes are positively associated with innovation only under conditions of endogenous (ego-driven) network change, and they have no discernible impact on innovation under conditions of exogenous (alter-driven) network change. We discuss the implications our findings have for the theoretical development and empirical testing of interorganizational network effects.