How to save clusters from dying

The aim of this paper is to contribute to the discussion on clusters competitiveness and paths dependence. Italian clusters offer a recognized example of territorial production model. However, recent studies in the field showed that it is passing trough a crisis. Several enterprises are delocalizing their manufacture abroad loosing the shared knowledge that characterizes the cluster culture.The paper is part of the area of studies that analyses cluster dynamics (Krackhardt, 1994; Ahuja, 2000; Hansen, 2002; Zaho & Aram, 1995; Nooteboom, 1992, 2004; Gulati and Gargiulo, 1999) and explores the inter-organisational relationship in a cluster situated in the Campania region and concentrated, in particular, in the urban area of Naples. This is the first part of a study that aims to analyse the evolution of a traditional industry cluster operating in a moment of maturity/crisis, analysing the impact of possible actions on the part of a leading organization or guide (Lorenzoni, 1992) and an institutional actor (Antonelli, 2005).The focus of the empirical analysis are the companies in the railway industry, which have many elements of homogeneity: geographical and socio-cultural nearness, size, geographical trade market and end clients. The cluster has two central actors: an institution (the Campania Region, in the combined work of three assessorates) and a company (Alfa) linked to other companies in the cluster as their main final client.In the first stage we analysed the strategic motivations that stimulate these actors to intervene and support processes of increasing the level of inter-organisational collaboration between the companies in the cluster. The first stage of the analysis that we present below is based upon structural indicators peculiar of the network analysis methodology. These first results over a short period allow us to highlight several limits of an aseptic analysis of the main structural indicators of a network. However, the qualitative analysis is a useful complement to understanding the strategic behaviour of actors that operate in a very cohesive cluster, conditioning the evolution of the life cycles of the cluster and inter-organisational relationships between its actors. The conclusion of the first part of the paper enlightens those limits. We then try to go over them by using computer simulations and we speculate on the role that a meta-organisation has to save the cluster from dying in the long run. In the last part of this work we pose the argument upon which to construct the simulation model we will test in a further research.