Geography Reference
In-Depth Information
Consequently, one ends up in a paradoxical situation. In section 2, we have gone at
length to explain that a high degree of proximity is considered a prerequisite to make
agents connected. However, when assessing the economic ef ects of networks, we argue
that proximity between agents in networks does not necessarily increase their innovative
performance, and may even harm it (Boschma, 2005; Broekel and Meder, 2008). We refer
to this as the proximity paradox. When incorporating a proximity framework in network
analysis, one should therefore make a distinction between the drivers of network forma-
tion on the one hand (in which the forms of proximity positively af ect the establishment
of networks), and the ef ects of network on innovative performance on the other hand (in
which it is uncertain what the ef ects of proximity on network performance are).
We claim it depends on the level of proximity between agents whether their connection
will lead to a higher level of innovative performance or not. The success of a network
relation may be related to optimal levels of geographical proximity (Camagni, 1991),
social proximity (Fleming et al., 2007; Uzzi, 1996), institutional proximity, organiza-
tional proximity (Grabher, 1993; Grabher and Stark, 1997) and cognitive proximity
(Nooteboom, 2000). When thinking about an optimal level of geographical proximity,
this does not mean determining an optimal geographical distance between two agents.
Instead, one should think of a balance of local and non-local linkages. Similarly, the
optimal social distance consists of a balance between embedded relationships within
cliques and strategic 'structural hole' relationships among cliques. For institutional
proximity, an optimal level consists of operating simultaneously in dif erent institu-
tional regimes, such as multinationals operating in dif erent countries or high-tech
labs cooperating with industry, government and academia. Concerning the optimal
level of organizational proximity, loosely coupled networks that consist of weak ties
between autonomous agents combine the advantages of organizational l exibility
and coordination. The optimal level of cognitive proximity follows from the need to
keep some cognitive distance (to stimulate new ideas through recombination) but also
secure some cognitive proximity (to enable communication and ef ective knowledge
transfer).
Besides looking for the optimal level of proximity on all dimensions, one can think of
other solutions to the proximity paradox. The negative impact of excessive proximity in
one dimension on innovative performance may be counteracted by lower levels of prox-
imity in other dimensions. For instance, regions may confront the problem of regional
lock-in by having a (related) variety of dif erent technologies in the region (Frenken
et al., 2007), or by having loosely coupled networks, as rel ected in regional networks
consisting of agents with weak ties (Grabher and Stark, 1997). In sum, optimal levels of
proximity may enhance network performance, but the location of an optimum along one
proximity dimension depends most likely on the location along other proximity dimen-
sions at the same time.
Though the concept of optimal level of proximity balancing pros and cons has become
well established (Boschma, 2005), to test these propositions empirically is not straight-
forward. There are two ways to go about this. First, classifying relationships into rela-
tions with high and low proximity, one can assess whether a mix of the two types of
relationship leads organizations to perform better than organizations relying primarily
on relations with low proximity or on relations with high proximity. This methodologi-
cal strategy was followed by Uzzi (1996) to test the hypothesis of an optimal social prox-
