Agglomeration economies, knowledge spillovers, technological diversity and spatial clustering of innovations.

This paper explores the spatial patterns of innovative activities from an empirical perspective and with reference to the Italian case. Using patent and other economic data at the NUTS 3 level (provinces), it borrows methodology and techniques from spatial statistics in order to analyse the way innovative and economic activities are arranged in space. Results show that innovative activities are considerably more spatially concentrated than production, but that there are also large differences across sectors in the spatial patterns of innovation. In mechanical engineering, industrial equipment and instruments sectors innovative activities tend to cluster around local systems of contiguous provinces, while in most chemical and electronic sectors innovative activities tend to concentrate in few metropolitan provinces surrounded by other non-innovative provinces. Regression analysis is also carried out to evaluate the impact of agglomeration economies, knowledge spillovers and technological diversity on the innovative performance of provinces.