Plant‐level data from the Longitudinal Research Database of the US Bureau of the Census are employed to estimate the impact of agglomeration economies on industry productivity across US metropolitan areas. This analysis seeks to remedy three shortcomings of previous empirical studies of agglomeration economies: reliance on aggregate spatial or sectoral data; lack of attention to spatial dependence in data; and representation of agglomeration economies with vague proxies such as city‐size. We show how a number of establishment‐, industry‐, and city‐specific factors influence labor productivity across US cities, and we pay particular attention to separating the influence of different kinds of agglomeration economies on firm efficiency. Here we follow Marshall's Principles of Economics in examining the spatial concentration of input–output linkages, the character of local labor pools and embodied technological spillovers.

You do not currently have access to this article.