Geography Reference
In-Depth Information
rational and tend to be inductively based on past experience (Beinhocker, 2006). There
is some evidence that most i rms exhibit path-dependent learning (Chapman, 2005;
Glasmeier, 2007), and in many cases more radical innovations are pursued and devel-
oped by smaller start-up i rms. However, in some cases, new i rms may also stimulate
incumbent i rms to diversify and adopt new technologies (Malerba et al., 1999). Given
this, the stylised general paths of industry evolution can be interpreted as the outcomes
of a continuous tension between life-cycle processes and those that yield concrete vari-
ations and departures from standard life-cycle patterns. Furthermore, another source
of non-equilibrium tendencies may be the place-specii c interactions between dif erent
industrial paths.
In the case of a region or locality dominated by a single industry, then, the product and
technological maturity of the industry will obviously play a key role in shaping the path
of the relevant geographical economy. Yet, in all probability, such a simple and direct
inl uence is likely to be unusual, for several reasons. First, as several authors have noted,
even in the case of a single industry prone to clustering it has been found that dif erent
clusters show dif erent paths and life-cycles (Menzel and Fornahl, 2007). For example,
Saxenian (1996) showed that, because of variations in local i rm strategies and network
structures, the Boston Route 128 computing cluster declined just as the Silicon Valley
cluster expanded. Second, the translation of industrial maturity into cluster and regional
development is likely to be complicated by the heterogeneity of responses of dif erent
i rms, and by cross-scale interactions between i rms, clusters, and urban and regional
contexts. Where i rm adaptation is short-termist and primarily aimed at cost-cutting
it may ef ectively undermine the adaptability of larger scale entities (Chapman et al.,
2004; Sunley, 1992).
19
In such cases, there are local conditions and place dependencies
that shape the competitiveness of dif erent clusters so that an understanding the state of
the industrial life-cycle is necessary but not sui cient. Moreover, most high-technology
clusters are actually composed of several technological trajectories and the switching of
resources between them may be especially important in times of crisis (Bathelt, 2001).
Most life-cycle approaches, however, do not consider the possibility that the co-location
of dif erent industries in a particular region makes a real dif erence to the co-evolution of
their trajectories. It is entirely possible, of course, that the dif erent industries and activi-
ties that make up a given region's economy evolve wholly independently of one another.
It could even be that each of a region's industries evolves in a path-dependent manner
but is uninl uenced by the development paths of the other local industries: a situation we
have called 'multiple
unrelated
path dependence' (Martin and Sunley, 2006).
In reality, however, economies are typically ensembles of sectors in which productiv-
ity growth is linked by income and expenditure l ows (Metcalfe et al., 2006). In a local
or regional context there may well be interactions between industrial paths through,
for example, upstream and downstream input-output linkages, knowledge spillovers,
labour pooling, positive service and infrastructure externalities, or, conversely competi-
tion for land, i nance and skilled labour. In other words, it is more likely that there is
at least some degree of 'multiple
related
path dependence' across an urban or regional
economy (Martin and Sunley, 2006). In a dif erent context, Bassanini and Dosi (2001)
argue that systems may be composed of path-dependent entities but whether or not the
system as a whole will be path-dependent depends on the
structure of the interactions
between the constituent entities. They contend that when such interactions are strong
