Geography Reference
In-Depth Information
In the model of pure agglomeration, the bulk of knowledge is explicit and codii ed,
available to any actor and organisation, and generated outside i rms' boundaries, being
largely created in public institutions. Variety and promiscuity are distinctive features of
cities: the combination of dif erent streams of knowledge occurs across a broad range
of sectors (Jacobs externalities) and individual linkages or relations are unpredictable,
because of the low degree of cumulativeness. However, even though in many cases the
critical distance over which urban agglomeration externalities operate may be that of
the broad city-metropolitan area (Gordon and McCann, 2005a, 2005b), as is assumed
by many theoretical models of agglomeration, there is also much evidence to suggest
that for many i rm-types and industries, the critical distances over which agglomeration
externalities operate may be very much larger than that of the city-region (Arita and
McCann, 2000; Caniëls, 2000; Cantwell and Iammarino, 2003; Simmie, 1998; Suarez-
Villa and Walrod, 1997). These considerations come from observations of regional
innovation systems, the location and performance of MNE R&D facilities, the behav-
iour of local and regional labour markets, and the behaviour of transportation systems,
particularly air-transport systems.
The combination of diverse kinds of knowledge into an interdependent economic
and technological base needs crucially a plurality of sources and networking among
them. The features of economic systems - and particularly their communication
opportunities - play a major role in assessing the conditions of the production of new
technology (Antonelli, 2000; Patrucco, 2001; see also Antonelli, Chapter 7 in this topic).
In this respect, urbanised and metropolitan regions have proved to of er highly positive
institutional contexts explaining the features of the collective dynamic of technological
progress, as a result of the mix of variety and complementarity of economic activities,
science and technology infrastructures and communication and network mechanisms.
3
However, these i ndings are further complicated by the fact that while there is much
evidence to suggest that the link between innovation and cities can be strong in certain
sectors (Acs, 2002), the evidence on this particular issue is also not always conclusive.
Cities do not always appear to be centres of innovation, and nor does innovation neces-
sarily appear to be centred on cities (Simmie and Sennet, 1999).
As described in Table 8.2, the industrial complex model is instead associated primarily
with cumulative learning from sources inside the industry and the i rm, such as in-house
R&D, and on the basis of knowledge that is specii c to industrial applications. Such cases
generally exhibit low entry possibilities and high industry concentration, which is likely
to display a complementary strong concentration at the spatial level. Large incumbents,
often MNEs, account for most of the sector's innovative activity, and these i rms can
proi t from their innovations in part because they have the potential to exclude rival
i rms from using the new products and processes they have generated. In these situations,
knowledge based on non-transferable experience is an important input in generating
innovative activity, and the incumbent i rms tend to have the innovative advantage over
new i rms because innovation is relatively routinised and processed within the exist-
ing hierarchical bureaucracies. As such, leading i rms, particularly large MNEs, play
a crucial role, and power asymmetry is central to the value chain and the governance
of innovation in the cluster or region (Cooke, 2001). Regions with a strong specialised
industrial structure and composite and advanced knowledge-production basis may
provide the most suitable environment for technology creation and experimentation, as
