Divide to Connect: Corporate Spin-Outs as Linking Contexts of Intraorganizational Clusters

Understanding how and why intra- and interorganizational collaborations evolve over time are relevant questions for both organization and management scholars (Ahuja, Soda, and Zaheer, 2012; Phelps, Heidl, and Wadhwa, 2012). Empirical research indicates that collaboration networks tend to evolve into cohesive cliques or clusters (i.e., groups of nodes, densely connected to one another but sparsely connected to others; Moody and White, 2003; Upham, Rosenkopf, and Ungar, 2010). For large, corporate, research and development (R&D) departments, this dynamics represents both benefits and threats. Clustered intraorganizational networks can enhance knowledge transfers and diffusion, as well as facilitate learning in specific knowledge domains (Fleming, Mingo, and Chen, 2007; Reagans and McEvily, 2003). Clustering, however, may embed researchers and reduce the diversity and variety of knowledge available in a cluster (Gargiulo and Benassi, 2000; Uzzi and Spiro, 2005). Whereas the emergence of distinct clusters in corporate R&D departments enables independent explorations of different knowledge domains, through cluster specialisation, embeddedness may reduce the organization’s ability to recombine knowledge domains and develop innovative solutions (Sytch, Tatarynowicz, and Gulati, 2012). Explicating the formation of bridging ties across such clusters thus has critical importance in terms of organizational performance (McEvily, Jaffee, and Tortoriello, 2012). Specifically, an important question pertains to how corporations can loosen individual constraints (embeddedness) in intraorganizational clusters.

Building on recent contributions regarding the interplay between incentive and cognitive structures (e.g., Kaplan and Henderson, 2005) and theory from organizational learning studies (e.g., Siggelkow and Levinthal, 2003), we propose that the reorganisation of R&D units into corporate spin-outs—corporate-backed ventures created by the voluntary incorporation of a unit or division of the parent organization (e.g., Chesbrough, 2003)—allows bridging ties to form across loosely connected inventor clusters within an intraorganizational co-invention network.

Using longitudinal data about the patenting activity of a sample of inventors employed in 11 U.S. information and communication technology (ICT) corporations and their spin-outs, we provide evidence that inventors who join a corporate spin-out are more likely to form bridging ties across clusters, whereas a control group of matched inventors who stay with the parent firm do so to a lower extent. Furthermore, these results are robust to several econometric specifications that account for the non-random assignment of inventors to spin-outs, corporate strategy, and diverse brokerage opportunities induced by structural changes in work environments.