The Dynamics of Postwar Industrial Location

Abstract

T HE location of industrial employment is a critical factor in virtually all models of urban spatial structure. However, most theoretical and empirical models of the urban economy say little about the determinants of industrial location. In the extreme, all employment is assumed to be centrally located (e.g., Muth, 1969, or Mills and Mackinnon, 1973), but, even when this assumption is relaxed as in transportation and land use studies (see Brown et al., 1972) or the recent vintage of urban simulation models (e.g., Ingram et al., 1972), employment location remains exogenously determined.' Theoretical analyses of industrial location have provided few specific insights into the underlying behavioral relationships, and existing empirical work has failed to identify much more than aggregate trends in employment dispersal. This paper develops a conceptual framework for analyzing the dynamic changes in industrial location and presents tests of several major hypotheses about the determinants of employment location. The empirical analysis models the spatial structure of employment for three selected industries in the Boston metropolitan area for the period 1947 through 1968.