This article addresses the nature of wine clusters in international perspective bringing together two separate literatures, the cluster/networks one and the one on the development of the wine industry, to analyze the mechanisms of innovation that have emerged in the sector. The study of innovation is often linked to economic clusters (Niosi 2005; Wolfe and Lucas 2004a; 2005), both as an analytical tool and as a policy option in the fostering of innovation and development (Feser, Renski, and Goldstein 2008; OECD 1999; 2002; Palazuelos 2005), even if not all effects of clustering are positive (Palazuelos 2005). The wine industry has received extensive attention through the use of cluster analysis (Alvarez 2007; Aylward 2004; Mytelka and Goertzen 2004; McDermott 2007; Hickton and Padmore 2005; Larreina and Aguado 2008). In general terms, clusters have been used in a broad multidisciplinary fashion and the approach has been applied to a host of issues (Boschma 2005; Malmberg and Power 2005) but this has also contributed to a reduction of its analytical sharpness and has fostered the emergence of different meanings for what should be similar concepts (Martin and Sunley 2003; see also Feser, Renski, and Goldstein 2008). However, we believe that cluster and network analysis can be used within the context of a multi-layered approach to the study of innovation policy in the wine industry. Clusters exhibit broad differences among them, by nation, by sector and by analytical approach and alongside them we often find networks. However, clusters and networks are different things with the latter embedded in the former. Both coalesce according to elective affinities, shared goals and layered interests among people. The fact that the clusters analyzed here are market clusters adds a powerful incentive to succeed, but also adds a high sensitivity on initial conditions.
Clusters coalesce geographically, while networks coalesce ideationally. This makes the latter a bit more flexible geographically: for example, global pipelines exist and networks of academics are relatively common if generally show low intensity. The 'quality' of the network embedded in the cluster may be the key to cluster success or at least a strong prerequisite of it (Eisingerich, Bell, and Tracey 2010). Therefore, we also argue that the relative degrees of success enjoyed by the various wine clusters reflect their ability to respond to shocks in the industry and the capacity of the institutional setting in which they exist to channel and foster some of the responses and activities generated by these shocks. Cluster theory offers some important fixed points from which the analysis of these agglomerations of actors can begin. Besides the notion that clusters occupy a (sometimes difficult to define) geographical area, they are internally heterogeneous in terms of the actors that compose them. Along with a broad range of firms (suppliers, service providers, banks, and so forth) we find public entities like universities and government institutions. Actors establish both traded and untraded linkages with one another, and they do so within an area of geographical proximity.i
In our analysis we are also cognizant of the fact that clusters and networks are not the sole level at which the process of innovation develops and that a multilevel approach that would include both regional systems of innovation and the national innovation policy from which many regulation, policy and economic dimensions depend (Aharonson, Baum, and Plunket 2008), is also critical in understanding the deployment of innovation. Regulation is going to affect both cluster structure (for example by creating specific organizations) and network capacity (by affecting the flow and ease of information). Regarding wine clusters, the above-noted characteristics apply with two provisos. First, the wine industry itself is highly dependant on a scarce supply of specific terroirs placed within amenable climatic zones. Second, because tradition and quality are very important, the degree of innovation that the cluster may absorbe could be different from that of other industries.
In the next section, we highlight the literature on innovation, and in section three we deal with the literature on the wine industry. In the successive section we explore the specific nature of the countries that we have chosen as subjects of our analysis: Australia, New Zealand, Canada, the United States, Argentina, Chile, France, Germany, Spain and South Africa. Finally, some conclusions are offered. Our sample covers most of the critically important old world producers, with the exclusion of Italy, and all of the important new world producers. The broad diversity in regulatory, legislative, policy and economic systems ensured by the scope of the sample is necessary to measure the effects these variables have on cluster performance.