Internal structures and external connectedness: towards a typology of French clusters

摘要

In this explorative study, we adopt a knowledge-centred perspective on clusters and investigate whether clusters? internal structure explains clusters? external (knowledge) collaborative strategies. All in all, the main analytical issue underlying this paper is the following: do internal characteristics of clusters vary (among them), and do they have any influence on the characteristics of clusters? external relationships? Our analysis focuses on the 71 French Competitiveness clusters. First, thanks to data analysis techniques, we identify 4 ideal types of clusters according to their internal characteristics. We then pinpoint variety in clusters? relational behaviours using network analysis. In a final step, we confront internal characteristics and the relational patterns of French clusters. Our data analysis combined to a network analysis on French Puf4les de compue9titivitue9 allows us to highlight several results concerning clusters? cooperation practices and to link them with internal characteristics. First, even though clusters tend to favour cooperation with a given partnering cluster, they try to diversify as much as possible their cooperation with different clusters. They thus try to strike a balance between deepened, more enduring cooperations with a one or a couple of other clusters and the search for novelty variety by initiating cooperations with new partnering clusters. Our results also suggest a significant variety in the cooperation practices among French clusters. Some of them do not play any significant role in the global network, whereas others are associated with high levels of centrality. This suggests that not all clusters implement (or are capable of implementing) similar cooperation practices. This is coherent with our first argument that clusters may not be equally endowed with the same networking capabilities. Investigating further the factors underlying clusters? cooperation practices and trying to bind them with internal characteristics, we confronted the results obtained out of our network analysis based on R&D collaborative projects. Our first results show that French Puf4les de compue9titivitue9 indeed differ according to their internal characteristics, the most discriminating factors being linked to the complexity level of clusters and their ownership structure while absorptive capacity and the nature of knowledge bases appear to be less discriminating. Considering both Puf4les' internal characteristics and cooperation practices, the internal factors determining most of their cooperation practices lie in their levels of internal complexity and their absorptive capacities. Our analysis shows that the dominant ownership structure in a Puf4le and the nature of its dominant knowledge base have less influence.