Marshallian Externalities And The Emergence And Spatial Stability Of Technological Enclaves *

Technological dualism often is found to be associated with the geographical clustering of firms that use the same techniques. To shed further light on these localization phenomena, we analyze the long-run dynamic behavior of a system in which firms' choices among alternative production methods (each of which requires a technique-specific input) are influenced by both firm-specific random shocks and Marshallian ‘industrial neighborhood’ effects. The latter are local factor market externalities that tend to lower the relative marginal costs d those inputs that are used most extensively in the immediate locale. The model developed here focuses on labor market externalities affecting the supply conditions for workers with technology-specific skills, and their effect on the choices made by producers at various sites whose choice of technique is subject to periodic revisions. A special structure familiar in the applied theory of Markov random fields, the stochastic Ising model. provides a reduced-form representation of this dynamic spatial system. The general properties of models of this type and their application in economics are considered. Discrete time numerical simulations of the behavior of an ensemble of firms (located at the nodes of a finite lattice formed on a two-dimensional (orus) shows that positive neighborhood externalities effects do not necessarily result in the uniquitous diffusion of one of the two available technologies. Instead. this system exhibits a spatially localized form of ‘technological dualism," in which at least two technological enclaves emerge and undergo path-dependent evolution. The temporal durations of these spatial patterns in technology adoption are affected by parameters of the Ising model that can be given a straightforward economic interpretation