SHIFTING SIZE DISTRIBUTION OF METRO AREAS

Abstract

This paper reviews traditional theoretic arguments for metro size distribution as an outcome of technical characteristics of production and the nature of demand for commodities and compares theoretical expectations with actual population growth experience in metro areas. Rigorous comparisons of actual with expected growth, however, are very difficult due to problems in making operational definitions of “place”, “intermediate size” and “metropolitan”. Despite these practical difficulties the actual record of growth by size class is reviewed for the U.S. since 1900 and for the countries of the European Community since 1950. The theoretical distribution is computed from the frequency distribution deduced by Losch which is calibrated to discrete size intervals according to Zipf and compared with actual distributions for U.S. counties, U.S. Standard Metropolitan Areas and EEC Functional Urban Regions. On the basis of these multiple comparisons it is concluded that somewhere perhaps as early as the 1940s and certainly by 1970 there was a strong shift in favor of intermediate size areas, but that by the early 1980s that this shift was weakening with indications of a lag in the adjustment in Europe compared with the U.S.