Geography Reference
In-Depth Information
The embeddedness of i rms in local networks has often been considered to be a way
through which transaction costs are reduced, because networks breed trustworthy
relations among i rms (Granovetter, 1985). Because of geographic proximity, i rms in
clusters or industrial districts are often described as being embedded in local business
networks, dei ned as 'an integrated and co-ordinated set of ongoing economic and non-
economic relations embedded within, among and outside business i rms' (Keeble and
Wilkinson, 1999, p. 299). Business networks are important also because they enhance the
dif usion of knowledge and the generation of local spillovers (Audretsch and Feldman,
1996; Caniëls and Verspagen, 2001; Simmie, 2003). A reason why knowledge spillovers
are believed to be highly localised is that a relevant part of the knowledge that is trans-
ferred between i rms has a tacit component, which is informally transmitted by face-to-
face interactions.
As mentioned, economic geographers and cluster scholars have tended to give rele-
vance to the
density
of networks (Amin and Thrift, 1992; Camagni, 1991; Garofoli, 1991;
Keeble et al., 1999; Nadvi, 1999; Schmitz, 1995; Yeung, 2000), but still limited analysis
has been carried out on how dif erent types of network - in term of their content and
especially of their structure - af ect the emergence of successful and vibrant clusters. This
is largely because of the fact that disentangling dif erent networks' ef ects on perform-
ance is a dii cult exercise, which may also require a shift in the method of analysis (Boggs
and Rantisi, 2003; Markusen, 2003).
Recent works have advanced in this direction, applying
social network analysis
(Wasserman and Faust, 1994) to look at whether the structural positions of i rms within
dif erent types of intra-cluster network af ect their performance (see Bell, 2005; Boschma
and Ter Wal, 2007; Giuliani, 2007b). In these works, the interest of the authors was
to understand whether more central i rms in dif erent types of intra-cluster network
achieved higher performances. In this chapter, instead, the interest is to explore whether
the existence of a linkage between any dyad of cluster i rms increases the probability of
them both being 'good performers', that is, of achieving similarly high performances.
This is elaborated in the following section.
Business networks, knowledge networks and economic development
In this chapter I explore the impact of two dif erent intra-cluster networks (i.e. the busi-
ness and the knowledge networks) on the likelihood that cluster i rms perform similarly
well. Thus, the interest here is not to explore whether a i rm performs better than another
by way of its better structural positioning, but to look at whether the presence of a given
linkage between any two i rms in the cluster af ects the likelihood that these two i rms
are both good performers. This is not done with the objective of identifying the factors
that lead the i rms to be good performers but just to compare the relative power of two
dif erent networks in af ecting this process.
6
As explained later in this section, this is done
with the aim of investigating (1) which of the two networks inl uences most the perform-
ance of i rms and, above all, (2) whether they fuel processes of even (or uneven) economic
development within the cluster.
In order to carry out this analysis, I draw on a previous work (Giuliani, 2007a) carried
out on three wine clusters (Colline Pisane and Bolgheri/Val di Cornia in Italy and Valle
de Colchagua in Chile), in which I have compared the structural characteristics of two
types of network: the business and the knowledge networks. The former is dei ned as the
