Geography Reference
In-Depth Information
at the sub-national level. However, the particular way in which this Porter
(1990) argument has generally been interpreted in the international busi-
ness literature is both positive and normative (McCann and Mudambi
2005). The positive conclusion is that MNEs have a great deal to gain from
locating in clusters, and on such a basis, the additional normative conclu-
sion is that MNEs should generally locate facilities where other similar
establishments are concentrated. For example, following a combination
of the Marshall, Vernon, Porter and Alchian's arguments, 'knowledge-
intensive' MNE operations should be located in 'knowledge-intensive'
regions characterized by other similar knowledge-intensive activities and
establishments. In contrast, rather more routine and conventional factor-
intensive activities should simply locate in lower wage areas along with
similar types of activities.
There is evidence to suggest that knowledge spillovers do operate in
certain locations. For example, it is well known that R&D-intensive indus-
tries tend to be highly spatially concentrated (e.g., Castells and Hall 1994;
Saxenian 1994; Audretsch and Feldman 1996; Almeida and Kogut 1997),
and that this spatial concentration has tended to persist even in the face
of rising local labour, land and other local input costs. Recent empirical
studies have compared 'first-nature' (e.g., market size, comparative advan-
tage, and trade costs) and 'second nature' (various kinds of externalities,
e.g., knowledge spillovers, capital-market externalities, vertical produc-
tion linkages) factors driving the geographical agglomeration of MNEs
(Alfaro and Chen 2009; see also Mariotti et al. 2010). Positive externali-
ties not only play a significant role as determinants of MNE location, but
relative to domestic plants, knowledge spillovers and other externalities
seem to exert a much stronger impact on MNEs in terms of the traditional
Marshallian centripetal forces (Alfaro and Chen 2009).
However, the involvement of MNEs in clusters is not ubiquitous. There
is also evidence suggesting that the potential advantages for MNEs of
being located in clusters are very sensitive to the nature of the industry
structure in which the firm operates (Cantwell and Kosmopoulou 2002).
These rather ambivalent findings reflect an analytical weakness of many
of the clustering-agglomeration arguments as applied to MNEs. In
particular, much of the recent international business literature seems to
suggest that MNE location decisions are quite straightforward, in that
MNEs should simply locate in clusters in order to benefit from localized
spillovers. However, these decisions are anything but simple, as they also
depend on organization and knowledge internalization issues. Moreover,
a careful assessment of the empirical evidence on clustering advantages
is required, particularly when applied to MNEs, because observations of
clustered MNEs are not necessarily indirect evidence for local spillovers.
