Geography Reference
In-Depth Information
dif erent forms of proximity as driving forces of network formation in space. This would
not only shed light on the question of whether the dif erent proximities are substitutes or
complementarities, but also on the question of in which stage of the network formation
some dimensions play a more prominent role. For example, if geographical proximity
af ects network formation, is this inl uence persistent over time? This concerns both the
study of: (1) the creation of relations by new i rms and by incumbent i rms linking up
with other nodes; and (2) the break-up of existing relations because of the exit of i rms or
because incumbent i rms dissolve their relations with other nodes. Doing so, the study of
network formation is not only about who connects with whom and why (being depend-
ent on proximity), but it should also be grounded in i rm dynamics (which concerns the
formation and dissolution of nodes).
Second, a dynamic network approach should make explicit how the evolution of
a network structure may be seen as a path dependence process, and how that may be
tested. When the current structure of a network is af ecting its future structure, network
evolution becomes an endogenous process: the creation of a new tie is not only inl u-
enced by the structure of the network but it also causes changes in the network (Glückler,
2007; Kilduf and Tsai, 2003). Path dependence in network evolution is shown in the
persistence of existing ties and the path-dependent formation of new ties. In the latter
case, new ties replicate or reinforce the existing structure of the network (Gulati, 1999).
The dif erent proximities (besides preferential attachment) may induce path dependence
in network evolution, and may cause retention in the network. These 'retention mecha-
nisms' (Glückler, 2007) may take place at the local level because geographical proximity
plays a role, both directly and indirectly (through its ef ect on the other proximities). If
geographical proximity matters a lot in this respect, another crucial research question is
under what conditions local network retention leads to regional lock-in, and how that
may be broken apart.
Third, a dynamic network approach should also account for that fact that the evolu-
tion of network structures may, in turn, af ect the degree of the dif erent forms of prox-
imity. The study on the dynamics of proximities during network formation is an (yet)
unexplored but promising i eld of research: it would account for the ef ect of networks on
the attributes of nodes in the network and thus their degree of proximity in their dif erent
dimensions over time, and how that might feedback on the structure of the network (Ter
Wal and Boschma, 2009b). Moreover, a change in one proximity dimension can also
have consequences for the other dimensions of proximity (Menzel, 2008).
In sum, we proposed an evolutionary perspective on the spatial evolution of network
formation that is i rmly grounded in a proximity framework. There are still many prob-
lems to be tackled before a dynamic proximity framework can be fruitfully applied to the
spatial formation of networks. Having said that, we i rmly believe it opens up a whole
new research agenda that will contribute to a better understanding of the spatial evolu-
tion of innovation networks. In that respect, it may be a considered a crucial part of the
further development of an evolutionary approach in economic geography (Boschma and
Frenken, 2006).
References
Agrawal, A., I. Cockburn and J. McHale (2006), 'Gone but not forgotten: knowledge l ows, labor mobility,
and enduring social relationships', Journal of Economic Geography , 6 , 571-91.
Search WWH ::




Custom Search