The Complementary Dynamic Effects of Clusters and Networks

Abstract

Over the past decades, researchers and policymakers around the world have been paying attention to the concept of clusters of related firms, industries and institutions, with a view to the presumably positive effects of clustering for learning, innovation and the productivity of firms. More recently, a network approach to learning and innovation emerged, which emphasizes strategic, preferential, repeated and at the same time temporary knowledge exchange (i.e. dynamic cooperation) between firms and other organizations. This may, however, go at the expense of the attention for the important, different and complementary learning effects of the mainly spatial process of concentration and clustering of related firms, industries and institutions. This paper argues that clusters and networks are two separate concepts that both merit attention, especially—albeit not exclusively—with a view to learning, knowledge development and innovation. A first argument is that spatial clustering has quite different effects for the development of knowledge, learning and innovation in and by firms, as compared with network settings. A second point is that in some cases, clustering yields a governance advantage over networks. Taking into account the risks of cognitive, technological, organizational and institutional lock-in associated with both processes, this paper concludes that both clustering and networking have advantages and disadvantages for the firms involved. The two concepts are potential dynamic complements, as clustering and networking have different but complementary effects for learning, although they are also static substitutes, as firms may opt to switch between the two processes, for example, leaving a local or regional cluster to engage in a network endeavor at a higher spatial scale.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 43 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.