Some industries are characterised by networks of small and specialized firms with porous boundaries, while in some other industries large diversified hierarchies with concrete boundaries are a more dominant form of organization. Some other industries are combinations of these. In this paper, a simulation study is performed to explore whether, and if so how the characteristics of knowledge base influences emerging organizational structures. We focus on two dimensions of the knowledge base; breadth and depth. We define breadth of knowledge base to be the number of different knowledge types required for production, while depth refers to the extent of dominance of a single type. In the simulation study, self-interested agents form pairs to integrate their knowledge and produce together under different schemes of the knowledge base. In this way networks form. Resulting network structures are analysed to gain an insight into organizational structures that emerge. Networks are also partitioned into cohesive subgroups as proxies for the firm, and their structural characteristics are analysed. The results of the simulation study reveal that, network-based industries are likely to emerge when the products in an industry are highly specialized (employing one knowledge type intensively), but still when they have some knowledge types in common, although minor in the production process. This is the case with a broad and deep knowledge base.
