مقاله سال ۲۰۱۶

۱٫ Introduction

“There is certainly no unanimity on exactly what centrality is or on its conceptual foundations, and there is very little agreement on the proper procedure for its measurement.”— Freeman (1977: 217)

The burgeoning innovation literature has revealed the interest of firms in tapping into external knowledge, and has begun to understand the outside-in knowledge flows that occur as part of this, depicting that firms rely on knowledge networks to assimilate external knowledge (Dong and Yang, 2015). Knowledge networks provide ample external knowledge resources to a focal firm by allowing recombinant opportunities and innovation such as new products and services. In particular, innovation studies on new product development (NPD) suggest that firms need to be open to external knowledge resources when they search for innovation (e.g., Ancona and Caldwell, 1992; Barczak et al., 2009). On the other hand, network researchers similarly emphasize the importance for firms to assimilate external knowledge from interfirm networks to benefit their NPD activities (e.g., Krackhardt and Hanson, 1993; Tsai, 2001).

Prior innovation studies on interfirm networks have been mainly focusing on collaboration networks based on alliance partnerships (e.g., Durmusoglu, 2013; Gilsing et al., 2008, 2014; Srivastava et al., 2015; Stolwijk et al., 2013; Vanhaverbeke et al., 2012). However, knowledge networks are different from collaboration networks, as the latter are actually relationship-based rather than knowledge-based. Dong and Yang (2015) was among the first to propose that interfirm knowledge networks are constructed based on knowledge flows embedded in patent citations among firms, whereas interfirm collaboration networks are primarily developed based on alliance partnerships among firms. Within an interfirm knowledge network in an industry, it is possible that knowledge flows occur via organizational learning without formal collaborations among firms (Dong and Yang, 2015). For example, a focal firm can read and learn from another firm’s patents without entering into a strategic alliance with it.

Network centrality is thus defined as the extent to which an actor is central in a network, and is among the most important structural properties in network research (Freeman, 1977, 1979). Different centrality measures have been used in network research, including degree, closeness and betweenness centrality (Freeman, 1979), as well as the more sophisticated eigenvector centrality (Bonacich, 1972). Each of these centrality measures is conceptualized and measured from different perspectives, and thereby captures the meaning of a central position in the network in different ways. In an interfirm knowledge network, however, whether different types of centrality are all good for NPD has not been examined. Given the importance of interfirm knowledge network for NPD, we examine how different centrality measures influence NPD performance in the U.S. pharmaceutical industry. We pay particular attention to degree centrality, closeness centrality and eigenvector centrality, as betweenness centrality is not that relevant to interfirm knowledge networks.1