The distribution of urban peoples and activities among cities of various sizes and characteristics has for a long time interested social critics, town planners, geographers, historians, journalists and some economists in western Europe and the United States. A common theme is that the distribution of urban peoples depends on the distribution of urban jobs. Thus to understand population changes, the analysis of cities usually becomes a study of trends in job location. A decentralized pattern of industrial jobs is thought conducive to a proliferation of small and medium-sized cities. Or to put it the other way, it is often suggested that the concentration of industrial jobs into large cities is a principal explanation of their size.

For more than half a century social critics and theorists have been predicting a coming era of decentralization in manufactures and, as a consequence, of urban peoples. The development of the automobile and of electricity, the use of mass production techniques in specialized single-storey buildings set on large tracts of land, improvements in telecommunications, a high income elasticity of demand for space—all of these have been suggested as forces tending to decentralize manufacturing employment. But what has in fact been the experience of the last half-century?