Regional economists, planners, and geographers for more than half a century have drawn a useful distinction in characterizing the properties of spatial agglomerations, or growth centers (or, to use the currently fashionable term, ''clusters''). They write on the one hand of the presence of same-sector businesses and employees (''localization''), and on the other of a diverse complex of economic and social institutions (''urbanization''). While both processes-sameness and diversity-are relevant to making sense of how economic activity sorts itself out across space, empirically oriented economists and students of organizational behavior are just now providing scientific support for the hypothesis that urbanization is more important than localization in explaining spatial patterns of innovation and economic development. In this paper, we report on the results of research conducted at the level of individual companies and plants, rather than on the aggregate economies of cities and legions. Across a national size-stratified random sample of almost one thousand manufacturing establishments, we find that the likelihood that managers will adopt new technology-in this case, programmable automation-is significantly associated with the degree of ''urbanity'' of the counties in which their factories are situated (metropolitan versus nonmetropolitan, suburban rather than downtown, urban rather than rural). But, after controlling for establishment size, scale of those production operations for which this type of automation is relevant, product mix, labor relations, and the industry's dependence on sales to the U.S. Department of Defense, we found that innovation was not systematically related to the density of clusters of similar businesses. Subsequent research by Kelley, which resolves certain technical problems encountered in the present work, reports significant effects of both urbanization and localization, with the former continuing to be relatively more important than the latter in influencing innovation. The difficulty in measuring such firm- and plant-specific effects probably explains why the empirical (as distinct from the theoretical) research literature has thus far offered such a weak basis for exploring the nexus between urbanization and localization processes, on the one hand, and technology and business organization, on the other.
