Knowledge Recombination and Inventor Networks: The Asymmetric Effects of Embeddedness on Knowledge Reuse and Impact

Inventors are triply embedded. They are embedded in a network of knowledge components that they can reuse in future inventions. They are embedded in an inventor network, where internal embeddedness (the strength of relationships between focal inventors and their colleagues upon whose knowledge the team builds) and network centrality influence access to information. Finally, they are embedded in the firm, with its specific routines that favor external or internal knowledge search, what we call search orientation. Using a sample of 39,785 semiconductor patents, we study the pattern of knowledge reuse, or the recombination of technologically similar components, on invention impact. We propose that reuse of internal knowledge affects invention impact in a concave manner and posit that internal embeddedness steepens this relationship while network centrality leads to an inflection point shift. We examine whether these effects differ for subsamples of firms with inward- or outward-looking search orientation. We find that inward-looking firms' optimal pattern of internal knowledge reuse does not differ markedly from that of outward-looking firms. We find that inward-looking firms are more susceptible to internal embeddedness and that centrality in the collaborative network flattens rather than shifts the relationship between reuse and impact. These findings elevate the theoretical discourse of embeddedness from the effects of network positions on innovation outcomes to similar network positions having asymmetric effects that vary with the firm's search orientation. Our results contribute to an emergent area in innovation research on how inventor networks shape the inventive process and its outcomes.