One explanation for the observed effect is
that friendship networks promote knowledge transfer, allowing
managers facing similar market conditions to learn from
each other's experience. Reagans and Zuckerman (2001) also
inferred knowledge transfer from the association between
network structurea nd organizationapl erformance.I n their
analysis of corporate research and development teams, Reagans
and Zuckerman described how interactions among scientists
with non-overlappingn etworks outside of their team
improvedp roductivityC. ollaborationa mong scientists with
different external contacts bridged gaps, or "structural
holes," in the network outside the team. People on opposite
ends of a structural hole have access to distinct knowledge
and information. Bridging structural holes in the external network
enabled the scientists to access and share with each
other diverse knowledge, resulting in greater creativity and
innovation, thereby improving the team's overall productivity.
Tsai (2001) provided a third example of this approach, but
with one variation. Instead of examining how the structure of
social relations affected performance, Tsai considered how
the position of business units in the knowledge network
affected performance. Tsai found that the most innovative
and profitable business units were central. In all three cases,
knowledge transfer was assumed to be the causal mechanism
linking network structure to performance. In each
instance, however, the path from network structure to knowledge
transfer was not examined. The network effect was
inferred from the observed association between network
structurea nd some form of organizationapl erformance.F or
example, although Tsai examined how the structure of knowledge
relationships affected business unit performance, he
did not consider the impact of network structure on the ease
of transferringk nowledge.