Does context matter? Exploring the effects of productive structures on the relationship between innovation and workforce skills’ complementarity

Innovation is often seen as a crucial factor for regional economic growth. Many strands of literature have investigated the role of agglomeration externalities or workers’ skills in the innovation capacity of the territory without ever considering their interaction. Using fixed-effects Poisson regression models on official data for 2014–2019 in the Italian regions, this paper aims twofold. First, by controlling for region-specific variables, the paper explores how workforce skills’ complementarity (overlapped skills, connected skills, unlinked skills) and productive structure (i.e. MAR specialised or Jacobs diversified structures) foster innovation. Second, the paper investigates how innovation processes depend on the alternative configurations deriving from the interactions between the productive structure and workers’ skills. While all types of skills participate in the innovation generation process in MAR specialised contexts, only connected skills positively affect innovation performance in Jacobs contexts. Guidelines are provided to help policymakers and managers who increasingly require regional place-based approaches to stimulate innovation.