Page - (000278) - in Knowledge and Networks

275 Cluster The cluster concept is associated with two elementary components in the definition: a geographical concentration of firms and a functional interrelation between them. Porter (1998) defines geographic clusters as regional concentrations of interlinked companies that perform similar activities in a common field. These two defining elements need to be assessed separately. Local concentration gives firms traditional localization advantages deriving from the joint use of infrastructure, labor markets and specialized services. The greater the number of a location’s firms that require specialist employees, the cheaper and more probable it is that a corresponding seg- ment of the labor market will form. These traditional localization advantages result, in particular, from external economies of scale. Local externalities evoke the theory of the club good (Buchanan, 1965), a reminder of why geographic concentrations are often referred to as regional club goods (Capello, 1999). The localization advan- tages work irrespective of any interorganizational action and require only that sev- eral firms with the same activities be colocated (Malmberg & Maskell, 2002). The second part of the definition distinguishes between the narrow and the wide senses of the term cluster. The former predicates not only a geographic concentra- tion but also functional links between the firms in a cluster. Concepts for industrial districts (Belussi & Pilotti, 2002; Sforzi, 1989) or the creative milieu (Maillat, 1998), for example, note the importance of cooperation relationships that benefit from their proximity and that are often based on trust (Bathelt, 1998). One can use transaction-cost theory (Scott, 1988) and the embeddedness approach (Uzzi, 1996) to argue that geographic proximity reduces communication costs and that face-to- face communication promotes the development of binding, trusting, and reciprocal relationships (Sabel, 1994). In this regard learning processes are due, in particular, to cooperation between local companies along the value chain. Despite the plausi- bility of the argument, the tendency for a firm to cooperate is often just as strongly geared to partners outside its cluster as to those within it. Empirical studies such as the software cluster in Darmstadt, southern Germany, show that lead firms and tech- nology SMEs in the region attach substantially greater importance to strategic alli- ances with partners outside the region than to local opportunities for cooperation (Angelov, 2006). It is clear that functional links in a cluster are not as strong or important as supposed in traditional concepts. Firms in a cluster are consequently a concentration of related activities based on a social division of labor between different stages of the value chain and in competi- tion within the same stage. Malmberg and Maskell’s (2002) knowledge-based the- ory of clusters thus incorporates the concept of rival learning. The two researchers explicitly explore the relative advantage of having a multiform cluster rather than a single integrated firm in one place. In the case of full internalization, a single firm could exploit internal economies of scale through the reduced unit costs of large production capacity, minimize external transaction costs through an authority-based governance mode, and smoothly organize the transfer of knowledge under a regime of hierarchical control. By contrast, multiple and colocated firms engaging in 13 Connectivity in Contiguity