Geography Reference
In-Depth Information
There are two reasons for this. Firstly, even without any agglomeration or
clustering effect, a random distribution of activities across space (Ellison
and Glaeser 1997) means that some activities will appear to be clustered
while other activities will appear to be dispersed. Observations of spatial
industrial concentration are not necessarily evidence of clusters (McCann
2007). Secondly, in the clustering-agglomeration type of literature the
definitions of what is 'local' and what is 'spillover' are often very vague,
in that the critical geographical dimension over which any such knowl-
edge spillover-based competitive advantage is assumed to operate is never
specified. Indeed, it has been argued that the characteristic of tacitness
commonly associated with knowledge, together with the free, uninten-
tional and disembodied nature of
pure
knowledge externalities, have
been often misinterpreted, giving rise to a mystification of the concept of
spillover which is applied indiscriminately to indicate both deliberate and
unintended exchanges, flows and dissemination channels, regardless of
the actual transmission mechanisms (Breschi and Lissoni 2001a, b, 2003,
2004; Breschi et al. 2005a, b). In the case of MNEs this is particularly
problematic, because there is much empirical evidence to suggest that the
information and knowledge transfer mechanisms of MNEs extend well
beyond the dimensions of the individual industrial cluster or metropoli-
tan areas (Audretsch and Feldman 1996; Suarez-Villa and Walrod 1997),
and may also extend well beyond a regional or even national level (Arita
and McCann 2000; Cantwell and Iammarino 2000, 2002). Therefore,
how important it is for a multinational firm to be located either within
or immediately adjacent to a particular industrial cluster or metropolitan
region is often very unclear.
In order to discuss these issues in a coherent manner, it is necessary
to reverse the logic regarding the potential advantages of clusters and to
consider another question: when should an MNE
not
locate in a particu-
lar industrial cluster or region? As we have seen in Chapter 3, the central
rationale for the firm - and in particular the MNE - is as a means of inter-
nalizing information and knowledge transactions costs within the individ-
ual firm. At the same time, as we have discussed earlier in this chapter, one
of the key rationales for industrial clustering is to internalize knowledge
and information transactions costs within the group of clustered firms,
rather than within an individual firm. As such, there may be cases where
the rationale and objectives of the MNE are in conflict with that of the
cluster. Therefore, from the perspective of an MNE, the location decision
must depend on the balance between the opportunity costs and benefits of
locating in a cluster and the opportunity costs and benefits of not locating
there, allowing for the organizational features of the firm. By adopting a
transactions-costs approach to understanding the types of inter-firm rela-
