ABSTRACT
Eighty-five United States Standard Metropolitan Statistical Areas (SMSAS) were classified by the B C TRY cluster analysis system using 68 1960 census variables. Fifteen city types were derived that had greater discriminatory power than 10 other city classifications.
There were two “super-types” of cities: Those within a relatively small geographical area and those that covered a large part of the United States. This finding suggests that United States cities can be arranged along a bipolar continuum with purely local geographical factors at one end and economic-functional factors at the other. The 15 city types define empirical sub-units of this continuum.
Results are only tentative and must be verified by re-analysis using 1960 census data. Yet speculation on the significance of certain findings leads to important questions.
The chief discriminating set of variables define older households. Age of households as measure of environment quality has significant bearing on design policy. Building without regard for maintenance costs may be a principal cause of urban problems, contributing to decreasing revenues, racial impaction, and housing abandonment. The design solution is to create the maintenance-free, low cost house. But will economic and social considerations permit this, or will designers have the courage to create designs that change national policy?
