The Geographic Concentration of Knowledge: Scale, Agglomeration, and Congestion in Innovation Across U.S. States

Abstract

Evidence of the importance of agglomeration economies in productivity is reported by a number of studies in regional economics. We extend the literature by looking into agglomeration and congestion in innovation and technological change using an endogenous innovation approach. It turns out that the geographic specificity of knowledge spillovers is also a central concern. Using data from U.S. states, evidence is found that knowledge spillovers are geographically concentrated but agglomeration economies far outweigh congestion effects. These results have important implications for new growth theory as well as regional economics because growth theorists have abandoned the scale implications of their models.

Urban density and the rate of invention

“…5 Rosenthal and Strange [46] consider the importance of input sharing, matching, and knowledge spillovers 2 See also Sedgley and Elmslie [48]. 3 Similar testable implications can also be derived from Jovanovic and Rob [33], if one assumes either the meeting rate or their imitation parameter (m) is increasing in the density of workers.…”

The territorial dynamics of innovation: a Europe United States comparative analysis

“…As illustrated by Ciccone and Hall (1996), average labour productivity is significantly greater where employment density is higher. In line with this approach, different studies have found that agglomeration increases innovative output even after controlling for differences in human capital, high-tech industry structure and R&D university infrastructure, both in the US (Carlino et al 2004;Sedgley and Elmslie 2004) and in some EU countries (e.g. Andersson et al 2005, for the case of Sweden).…”

Geography and High-Tech Employment Growth in US Counties‡

“…Glaeser et al (1992) and Partridge and Rickman (1999) find more evidence of Jacobian dynamic externalities than within industry externalities. 2 To be sure, agglomeration has been found to increase innovation even after controlling for other factors such as human capital and public research and development infrastructure (Sedgley and Elmslie, 2004). Nevertheless, contrary findings for high-tech plants are reported by Henderson (2003).…”