Geography Reference
In-Depth Information
ties, which allow solving a high degree of technological complexity, are a prerogative
of large i rms, and in particular of MNEs. On the other hand, external economies are
usually benei cial particularly to spur incremental innovative processes of small and
medium sized enterprises (SMEs). The nature and extent of externalities cannot be
separated from the dominant i rm dimension in the cluster, and from the patterns of col-
laboration between small local i rms, large indigenous i rms and MNEs. Another issue
to be considered is that MNEs can be at the same time 'internal' actors, contributing to
the creation and dif usion of new technical knowledge within the cluster, and 'external'
players, channelling knowledge created elsewhere (within the i rm) into the local system,
thereby assuming the role of technological gatekeepers, but even inl uencing the cluster
knowledge base without being necessarily located
in situ
(as in the case of global buyers
or distributors).
Even though the international business literature has largely acknowledged the geo-
graphical concentrated dispersion of production and technology networks as a chief
mechanism to achieve sustainable competitive advantages in both MNEs and host
regions (e.g. Almeida, 1997; Cantwell, 1989; Cantwell and Iammarino, 1998, 2003;
Cantwell and Janne, 1999; Frost, 2001; Kogut and Chang, 1991; Kuemmerle, 1999;
Pearce, 1999; Zander, 1999), the operational mechanisms of such technological linkages
tend to be mainly depicted as a merely micro-to-micro (intra- and inter-i rm) interaction,
where the space itself is just an exogenous variable hardly explained in its internal struc-
ture and dynamics. As emphasised by Rugman and Verbeke (2002), the introduction of
operational space into the study of MNE locational advantages is becoming essential.
In other words, while the country level of analysis has been extended to (macro)regional
trade and investment blocks (e.g. Ohmae, 1985; Rugman, 2005), as far as the (meso)sub-
national level is concerned (i.e. innovative clusters, regional innovation systems, local
production systems, etc.) the reference literature is still rather scanty.
In the next section, we interpret the 'global versus local' interaction from the angle of
the logic of spatial agglomeration by using a transaction costs-based perspective. This
latter is especially useful to show that many stylised cluster notions found in the current
literature are predicated on the basis of assumptions that are often incompatible with
MNE behaviours.
3. Clustering, MNEs and knowledge spillovers: a transaction costs classii cation of
spatial models
The most popular explanations of the benei ts of spatial clustering have focused either
on the role played by specialised external agglomeration economies (Marshall, 1920) -
namely local information and knowledge spillovers, a local supply of non-traded inputs,
and a skilled local labour pool - or on the role played by industrial variety in agglomera-
tion, as in the models of the proponents of the new economic geography (e.g. Fujita et
al., 1999; Krugman, 1991) and in traditional approaches to urbanisation and diversii -
cation economies (e.g. Glaeser et al., 1992; Jacobs, 1960). This literature suggests that
inter-industry spillovers may be generally more important than intra-industry spillovers
in explaining local growth (Martin and Ottaviano, 1999), although intra-industry ef ects
may dominate in certain manufacturing activities (Henderson et al., 1995).
Nevertheless, the relationship between geography and economic growth appears to
be rather more subtle, complex and varied than these agglomeration models suggest,
