Geography Reference
In-Depth Information
relations and inter-i rm learning. Third, the study of network dynamics and the role
of geography is still in a premature stage, not least because of the limited availability
of time series data on networks, and the embryonic methodological state of dynamic
network analysis. Below, we elaborate on each of these three potential contributions
more in detail.
Recent network studies have contributed to a better understanding of clusters (see, for
example, Boschma and Ter Wal, 2007; Cantner and Graf, 2006; Guiliani, 2007; Giuliani
and Bell, 2005; Morrison, 2008; Staber, 2001; Visser, 2009). Using a micro-perspective
on knowledge networks, these studies have questioned the conventional view that the
economic well-being of cluster i rms ultimately depends on extra-i rm sources of knowl-
edge, rather than intra-i rm routines. These social network studies have also challenged
the view that knowledge is 'in the air' in a cluster, from which all cluster i rms can equally
benei t. Without exception, these studies show that only a limited number of i rms in
a cluster are well connected to the local knowledge network, while many cluster i rms
are poorly connected, or not connected at all. As Elisa Giuliani demonstrates in her
study of three wine clusters in Chapter 12, this is not a trivial issue: a knowledge linkage
between two cluster i rms increases the likelihood that both i rms perform well. So, it
is not the cluster per se that matters for i rm performance, but being connected to the
local knowledge network. In addition, these network studies have shown that reliance
on external knowledge relationships does not necessarily mean these are coni ned to the
cluster. What these studies demonstrate is that the best performing i rms tend to show
a high connectivity to i rms outside the cluster. A next question then is whether these
leading i rms might still act as gatekeepers for the cluster. Network studies show that this
depends on their connectivity to other local i rms, among other things (Morrison, 2008;
Cantner and Graf, this volume Chapter 17).
Looking at the structure and nature of intra-cluster networks as a whole, in Chapter
12 Giuliani criticises the cluster literature for claiming that intra-cluster networks per se
enhance economic development in the cluster as a whole. Her study on three wine clus-
ters clearly shows that it depends on the structural properties of intra-cluster networks as
to whether these will generate benei cial ef ects throughout the cluster. More specii cally,
her analyses show that this depends on how selective knowledge networks in clusters are,
that is, how unevenly distributed these networks are among cluster i rms. Giuliani also
found that business networks (as opposed to knowledge networks) are more pervasive in
clusters, but these do not spread positive spillovers throughout the cluster.
Concerning the i rm level, network studies have recently taken up the question of what
explains the network position of i rms in clusters. In their seminal paper, Giuliani and
Bell (2005) pointed to the importance of i rm-specii c competences, like the absorptive
capacity of i rms. In Chapter 13, Stefano Denicolai, Antonella Zucchella and Gabriele
Cioccarelli draw attention to the formation of dif erent forms of reputation and trust
between local i rms, and how these af ect the formation and dissolution of network link-
ages at the cluster level. Instead of referring to i rm-specii c competences, Denicolai et al.
describe how i rms with dif erent forms of reputation may occupy dif erent network posi-
tions in clusters. They illustrate this with a number of network studies they conducted in
three Italian clusters.
Instead of focusing on clusters, network studies have also investigated the drivers of
network formation at the level of i rms. In Chapter 14, Johannes Glückler examines
