Centrality Of Social Capital

This paper contributes to an integrated theory of knowledge sharing in multinunit firms by synthesizing the concept of network centrality from social capital theory and the notion of relatedness in skill-base from the strategic management literature. My main argument is that a high degree of network centrality among business units possessing related knowledge is needed for interunit knowledge sharing to occur and be effective. Results from a network study of 120 product development projects situated in 41 business units in a large multinunit electronics company showed that project teams obtained more software and hardware from other units and completed their projects faster to the extent that they initially had central network positions among business units possessing related knowledge. Neither network centrality nor extent of related knowledge alone explained the amount of knowledge obtained and project completion time. The results also showed that central positions attained with many direct interunit relations (as opposed to indirect relations through intermediaries) were not always beneficial. While direct relations were helpful when the teams incurred problems in transferring knowledge from another unit, they prolonged completion time when there were no transfer problems, in part because they required time to maintain and hence slowed down project work. I discuss implications for research on interunit knowledge sharing and social capital theory and also outline implications for managers of multiunit firms.