Invention in the United States City System

Abstract

This article draws on several empirical regularities underlying central place theory (CPT) to enhance understanding of the uneven distribution of invention in the U.S. city system, especially the immense array of specializations that comprise national technological advance. CPT depicts city systems as collections of places in which functions expand in number as city size increases. Small cities have few functions and large cities many. A long-term hierarchical system is successively inclusive if large cities have all of the functions of smaller cities and some additional ones. The functions investigated here are 399 patent classes distributed across 366 U.S. metropolitan areas in the period from 2000 to 2011. Evidence is strong that patent classes with large numbers of awards are widely spread across the city system. This leads to the average sizes of places active in generating patents in the robust classes to be significantly smaller compared with the average sizes of areas that generate patents in unusual classes. Small cities are tied to national technological advance through the generation of patents in the most active and ubiquitous inventive specialties. Inventors in large cities are more likely to invent in unusual domains. Bigger areas are significantly more diversified compared with smaller ones. The system is not, however, strictly successively inclusive. Whereas 88.3 percent of all patent class–area pairs are generated in at least 50 percent of equally sized and bigger areas, only 20.5 percent of pairs are 100 percent strictly hierarchical.