Urban Productivity and Factor Growth in the Late Nineteenth Century

This paper uncovers a series of empirical facts regarding the sources of U.S. urban growth in the 1880s. We use a large theoretical literature to provide motivations for a number of potential sources of growth, particularly those based on geographical proximity externalities. These sources are characterised and linked to empirical proxies. Then we estimate the covariation of these empirical proxies with the growth rate in output, capital and labor respectively. We find that traditional (neoclassical), several geographic externality, and socio-political factors all covary significantly with aggregate growth, though in very specific ways. For example, the size of a city (a measure of the degree of urbanization) is uncorrelated with output growth, positively correlated with labor growth, and negatively correlated with capital growth. No one extant theory of growth accounts simultaneously for all the phenomena that we observe.