Organization Science Vol. 23, No. 4, July–August 2012, pp. 1024–1039 ISSN 1047-7039 (print)  ISSN 1526-5455 (online) http://dx.doi.org/10.1287/orsc.1110.0688 © 2012 INFORMS Bridging the Knowledge Gap: The Influence of Strong Ties, Network Cohesion, and Network Range on the Transfer of Knowledge Between Organizational Units Marco Tortoriello IESE Business School, 28023 Madrid, Spain, mtortoriello@iese.edu Ray Reagans MIT Sloan School of Management, Cambridge, Massachusetts 02142, rreagans@mit.edu Bill McEvily Rotman School of Management, University of Toronto, Toronto, Ontario M5S 3E6, Canada, bill.mcevily@rotman.utoronto.ca P rior research has emphasized the importance of boundary spanners in facilitating the transfer of knowledge between organizational units. The successful transfer of knowledge between organizational units is critical for a number of organizational processes and performance outcomes. The empirical evidence on the success of boundary spanners is mixed, however. Research findings indicate boundary spanners can either facilitate or inhibit the flow of knowledge between organizational units. We develop and test a theoretical argument emphasizing the importance of the broader network context in which boundary spanning occurs. In particular, we consider how tie strength, network cohesion, and network range affect the level of knowledge acquired in cross-unit knowledge transfer relationships. An analysis of knowledge transfer relationships among several hundred scientists indicates that each network feature had a positive effect on the level of knowledge acquired in cross-unit knowledge transfer relationships. Our findings illustrate how network features contribute to the flow of knowledge between organizational units and, therefore, how network context contributes to heterogeneity in boundary-spanning outcomes. Key words : social networks; organizational boundaries; knowledge management History : Published online in Articles in Advance August 24, 2011. Introduction The ability to transfer knowledge between organiza- tional units is critical for a host of organizational pro- cesses and performance outcomes, including creativity, innovation, and new product development (Argote 1999, Carlile 2004, Bechky 2003, Hargadon 2003). Organi- zational units often represent distinct stocks of knowl- edge and expertise. Consequently, acquiring knowledge from outside the focal unit provides unit members with an opportunity to see a problem or task from an alternative perspective, which can stimulate creativ- ity and innovation (Leonard-Burton 1995, Hargadon and Sutton 1997, Perry-Smith 2006, Hansen 1999). Although cross-unit transfers can be beneficial, insuffi- cient knowledge overlap between the knowledge source and recipient makes cross-unit transfers more difficult than within-unit transfers. Cross-unit transfers can be even more problematic in a knowledge- or technology- based organization, where unit members often develop a “local” language, which makes it easier for them to learn from each other but also makes it more diffi- cult for unit members to share knowledge with col- leagues outside their unit (Dougherty 1992). Indeed, in knowledge- and technology-based organizations, it is generally understood that cross-unit transfers are simul- taneously a source and a barrier to innovation and new product development (Carlile 2002, p. 442; Seidel and O’Mahony 2010). Prior research has highlighted the importance of indi- viduals who maintain relationships with colleagues in different organizational units (i.e., “boundary spanners”) in facilitating cross-unit transfers (Allen and Cohen 1969). Yet the empirical evidence on the success of boundary-spanning outcomes is mixed. On the one hand, units with boundary spanners are more produc- tive (Tushman and Katz 1980, Ancona and Caldwell 1992, Hansen 1999), and individuals that maintain a boundary-spanning position do better in terms of their technological and knowledge proficiency (Tushman and Scanlan 1981). On the other hand, boundary spanners can restrict the flow of knowledge throughout organiza- tions (Gould and Fernandez 1989, Cross et al. 2002), sometimes because they desire to maintain their power and influence (Crozier 1964) or because they are unwill- ing to invest the time and effort that is required for suc- cessful knowledge transfer. The mixed empirical evidence suggests that what- ever benefits cross-unit transfers can create often go 1024