Spatial Dynamics of Technological Evolution: Technological Relatedness as Driver for Radical Emerging Technologies

Abstract

Despite the consent of the fundamental role technological change plays for economic growth, it seems that little attention has been paid to how new technologies come into being. In particular, an understanding of the spatial and dynamic processes driving the emergence of radical technology is lacking. This paper seeks to fill out this research gap by bridging the school of evolutionary economics and the school of economic geography. Following Dosi (1988) two factors are in particular important for technological change in emerging technological paradigms: 1) accumulation of scientific and applied knowledge in firms, universities, research institutes etc., and 2) the existence of risk-taking actors who are willing and capable of implementing and exploiting radical technologies. This paper focuses on the former. Where evolutionary economics have been occupied by accumulation of knowledge over time, economic geographers have been occupied by accumulation of knowledge in space. For long, it has been discussed whether Marshallian specialization or Jacobian diversification externalities favor regional innovativeness. In the case of radical innovation, studies have found empirical support for Jacobs externalities. However, a recent stream of literature (e.g. Frenken et al., 2007, Neffke et al., 2010) has shown that knowledge does not automatically spill over in diversified regions but requires related variety in the knowledge base of the region. This paper tests empirically the overall hypothesis that the accumulation of knowledge at the regional level is an important driver for the emergence of radical technology. The paper focuses on the case of fuel cell (FC) technology, which is argued to be a radical technology, understood as a disruption and discontinuation of technological trajectories. In more details, the paper tests two opposing hypotheses: 1) A diverse regional knowledge base leads to radical innovation, and 2) Related Variety in the regional knowledge base leads to radical innovation. And two specific hypotheses for the emergence of FC technology: 3) FC technological relatedness in the regional knowledge base leads to innovation within FC technology: where FC technological relatedness is defined as the knowledge fields that form the knowledge base of FC technology. And 4) in order to generate new FC knowledge, the higher the degree of FC technological relatedness, the more important it is that the specific knowledge field is present in the regional knowledge base: where the degree of FC technological relatedness is defined as the relative importance of each knowledge field for the FC technological knowledge base. To measure knowledge production in general, and within FC and FC related knowledge fields, the paper uses patent applications as a proxy and defines knowledge