Innovation across cities

This paper examines the distribution of patenting activity across cities in the OECD, using a sample of 218 cities from 2000 to 2008. We obtain three main results. First, patenting activity is more concentrated than population and GDP. Second, patenting activity is less persistent than population and GDP, especially in the middle of the distribution. Third, in a parametric model, patenting does not exhibit mean-reversion, and is positively associated with GDP and population density. Our results suggest that policymakers can influence the amount of innovative activity through the use of appropriate policies.