Geography Reference
In-Depth Information
Table 12
.
1
Expected relationship between network structure and performance
Type of networks
Network
structure
Suggested determinants
of structure
Expected impact on
economic development
Business network
Pervasive
Meso-level forces (e.g.
geographical/sectoral/
social proximity)
Progressively more even
at the local level
Knowledge network
Selective
Micro-level forces
(e.g. i rm knowledge
bases or other internal
characteristics)
Progressively more
uneven at the local level
would be spread selectively in the cluster, driving persistent heterogeneity among i rm
performance. In this case, the success of a cluster would be related to a much smaller
number of i rms and not to the entire community of i rms. In light of this, the emergence
of successful clusters could depend on the behaviour of individual i rms - quite in line
with evolutionary thinking.
3. Methodology
Context and data
This empirical study is contextualised in the wine industry. In the recent two decades,
wine consumption has dramatically changed, shifting market preferences from quantity,
non-premium wines to quality, premium wines. On the side of production, technol-
ogy and techniques of grape-growing and wine making have undergone processes of
increased codii cation of knowledge. Technical change has been strong in the industry
and the key competitive asset of wine producers is now the capacity to absorb and
manage new techniques of production.
This study has been carried out in two countries, Italy and Chile, which dif er histori-
cally but have recently undergone a similar process of wine industry growth and mod-
ernisation. In both cases, what has sparked growth is a process of technological change
aimed at improving the quality of wines. Based on this, new and successful wine clusters
have developed in both countries since the 1980s. This study considers two clusters in
Italy (Colline Pisane and Bolgheri/Val di Cornia) and one in Chile (Valle de Colchagua).
The boundaries of these wine clusters are given by their natural conditions. These types
of cluster are therefore easily identii able economic entities, whose boundaries are nowa-
days set by the Denomination of Origin regulations applied internationally by wine-
producing countries. All the three clusters are territories densely populated by i ne wine
producers and by grape growers. The degree of vertical division of labour is however
rather low, with no other relevant suppliers localised within the clusters' territory. On a
global scale, these clusters can be classii ed as 'followers', with Bolgheri/Val di Cornia
and Valle de Colchagua being more dynamic than Colline Pisane.
This study is based on micro-level data, collected at the i rm level in three clusters in
years 2002-03. The analysis has required careful data collection through face-to-face
interviews. Interviews were carried out with the skilled workers (i.e. oenologists or
