Geography Reference
In-Depth Information
kind of labour, and what sets of skills (diversii ed, related, specialised) are needed for
i rms to survive over the dif erent stages of the industry life-cycle?
An evolutionary approach has the potential to bring a new perspective on the study
of regional clusters, because it meets to a considerable degree the critique advanced by
Martin and Sunley (2003) and Brenner (2004), especially the tendency to treat clusters
in too static a manner. In Chapter 8, Simona Iammarino and Phil McCann describe the
evolutionary aspects of clusters (see also Iammarino, 2005; Menzel and Fornahl, 2009;
Swann and Prevezer, 1996; Ter Wal and Boschma, 2009), and emphasise the importance
of path-dependent processes that shape cluster development, and which make clusters
look very dif erent over time (see also Iammarino and McCann, 2006). In Chapter 10,
Udo Staber provides a thorough critique on the cluster literature from an evolutionary
angle. Staber argues that the cluster literature has underestimated the origins and conse-
quences of (new) variation in clusters. Consequently, clusters have been misconceived as
coherent entities while, in practice, they consist of dif erent levels (e.g. individuals, ideas,
routines, organisations, networks) on which selection operates. According to Staber,
evolution proceeds at a faster pace at lower levels of action and, thus, the i tness of ideas
increases more rapidly than the i tness of a cluster as a whole. Thus, while at the level of
individuals, there may be much turbulence (such as exits and new network relationships),
this will not necessarily af ect the survival of the cluster as a whole. Research should
focus more on identifying the dynamics of selection mechanisms (such as competition
and imitation) at various levels to explain stability and change in clusters. This would
meet worries expressed in the literature that clusters are often treated as static, instead
of dynamic entities (Carlsson, 2003; Maggioni, 2002; Martin and Sunley, 2003). And
cluster research has to recognise that clusters can develop dysfunctional features that
often give rise to diseconomies and negative economic ef ects, a point also made by
Martin and Sunley (2003). Staber proposes an alternative evolutionary view on regional
clusters. His basic line of argument is that, while evolution may unfold very dif erently
across clusters, the general Darwinian processes (variation, selection, retention) operate
in all units and at all levels of actions.
Another critique on the cluster literature is that it almost ignores the existence and
ef ects of collaborative relationships among constituent i rms. In Chapter 11, Phil Cooke
and Carla de Laurentis begin by explaining how an evolutionary view fundamentally
dif ers from a neoclassical approach. They then present the results of an empirical
research project on collaborators and non-collaborators in cluster and non-cluster set-
tings in the UK ICT industry. The data show that collaborators in clusters perform
better than non-collaborators in clusters, suggesting a premium ef ect of local network-
ing. There is also some evidence that collaborators in clusters perform better than col-
laborators in non-cluster settings. However, non-collaborators in clusters perform worse
than non-collaborators in non-clusters, suggesting a diseconomies of scale ef ect in clus-
ters for non-collaborators. The study by Cooke and de Laurentis shows how the analysis
of networks of collaboration is important for understanding the dynamic performance
of clusters, and links to the topic that is the focus of the next part of the topic.
4. Network evolution and geography
Part 3 of this topic concentrates explicitly on the importance of networks in an evolu-
tionary approach to economic geography.
There is increasing awareness that networks
