In this paper, we concentrate on how evolutionary economics contributes to a better understanding of the spatial evolution of newly emerging industries. Inspired by evolutionary thinking, four types of explanations are discussed and tested in an empirical analysis of the spatial pattern of the software sector in the Netherlands. Traditionally, agglomeration economies provide an explanation for the spatial concentration of an industry. Firms located in a cluster of similar or related sectors benefit from cost reductions, due to lower transportation costs, a thick labour market, specialised suppliers and information spillovers. An evolutionary approach on agglomeration economies provides an alternative view. It focuses explicit attention on knowledge spillovers as a vehicle of local diffusion of organizational routines or competences from one firm to the other. Such transfers of (tacit) knowledge are facilitated by spatial proximity of firms and a common knowledge base. In addition, an evolutionary approach takes a dynamic perspective on the role of agglomeration economies. During the initial stage of development of a new industry, the surrounding environment is still directed to routines and competences related to existing industries. When the new industry concentrates in a particular area to a considerable degree, a supportive environment (specialized knowledge, labour with specific skills) may gradually come into being, and localization economies may arise. Other evolutionary mechanisms may also provide an explanation for the spatial formation of new industries. We distinguish another three of them. First of all, transfer of knowledge and successful routines between firms in an emerging industry may occur through spin-off dynamics. Secondly, (social) networks may function as effective channels of knowledge diffusion and interactive learning, because they can provide a common knowledge base and mutual understanding and trust. Thirdly, firms in new industries with organizational capabilities that can deal effectively with the lack of required resources (such as knowledge, skills and capital) may become dominant, due to selection and imitation. Based on cross-sectional data gathered among 265 software firms in the Netherlands in 2003, we have tested which factors have influenced the innovative productivity of these firms. Using regression techniques, the outcomes suggest that spin-offs and firms with organizational capabilities perform better, while networks relations do not seem to affect the performance of software firms. Geography matters as well: software firms located in a region with a labour market with more ICT-skills show a higher innovative productivity.
