Learning and Structural Properties in Small Firms’ Networks: A Computational Agent-Based Model

This paper explores how structural properties of small firms’ networks emerge from the ways firms exchange knowledge. In particular we are interested in analysing if and under which conditions the need for knowledge exchange within a set of co-located small firms is able to generate a more or less stable structure of links among fi rms. We focus on a specific kind of small firms’ networks called Industrial Districts (IDs). One of the peculiar characteristics of IDs is flexible specialization: small firms specialize in given phases of the production processes and join up production chains in a flexible and dynamic way depending on market opportunities. Consequently, knowledge exchange is mainly related to the matching of complementary knowhow and competencies. To explore the relationship between the exchange of complementary knowledge assets and network structure we developed a computational model of an ID. The results obtained through computer simulations of the model show that the exchange of complementary knowledge assets is able to generate stable networks and that, even with different conditions, such networks evolve toward a hub and spoke configuration with a few firms becoming key actors in the network.