Does knowledge base complexity affect spatial patterns of innovation? An empirical analysis in the upstream petroleum industry

Using network analysis, we investigate if an industry's complex and integrated knowledge base leads to a higher spatial concentration (or dispersal) of innovative activities. This is important because the extant literature provides competing claims about how knowledge base complexity impacts on the spatial distribution of industrial innovation. To help empirically resolve this issue, we draw on longitudinal data (1970–2010) on the upstream petroleum industry and build indexes of entropy and complexity to render knowledge base dynamics, assess the spatial concentration of innovation, and study industry structural transformations. We first find a correlation – once a crucial distinction between variety and systemic complexity is drawn - between increasing knowledge base complexity and higher concentrations of innovation at national level. In addition, we find that this increase was accompanied by a rising share of non-country of origin inventions owned by multinational companies, and that globally integrated service multinationals were best placed to manage this complexity and the integration of complementary knowledge fields. These findings, therefore, nuance the competing claims about the relationship between knowledge base complexity and spatial patterns of innovation. Furthermore, they reveal that although leading innovators may operate globally, their core innovative activities still remain located in a few key countries.