Nowadays, policy makers in charge of designing innovation policies, especially at the regional level, are more and more adopting the cluster approach either with a view to accelerate the existing clusters or for providing the basis for the emergence of new ones. In fact, not only as a consequence of their appeal as an interactive and territorially embedded vision of innovation but also owing to a lot of other reasons, clusters are usually considered as key instruments for promoting competitiveness, industrial development, innovation and growth. But, although cluster policies have a potential for generating benefits, they also entail costs and risks. Additionally, as clustering is something that has been happening spontaneously during time, the presence of potential benefits from cluster initiatives is not per se a sufficient foundation or a validation for policymakers to get involved. The key question is whether and how policymakers can add value through appropriate measures, beyond the outcomes that markets and market players produce on their own. Furthermore, if clusters are so valuable for regional productivity, and surge of particular historic, cultural and societal circumstances, why are policies needed? Which policies? Undertaken by whom, and how? In this paper we'll search answers to these questions. However, this is not an easy task as there is a lot of confusion around the cluster concept. First, as some authors have highlighted, there is a vulgarisation and a tendency to oversimplify the definition of "cluster" and, so, part of the popularity of clusters lies in its vagueness and definitional elusiveness. Second, in today's policy world, clusters are acquiring "the discreet charm of obscure objects of desire". Third, these particularities prevent an accurate policy evaluation. Furthermore, while there is an extensive literature that focuses on the cluster analysis, the connection between clusters and policy has been mainly ignored. Finally, if cluster policies differ from related and partly overlapping approaches, such as Industrial Districts, Innovation Systems and Triple Helix there may be, in practice, a conflict in the application of these different perspectives. Given the above problems, this paper aims at contributing for identifying the key features of the innovative clusters and addresses the most critical issues in the cluster policy, contributing in this way for filling the existent gap between cluster analysis and policy.
