Suburban subcenters and employment density in metropolitan Chicago

A detailed data set covering every quarter section in suburban Chicago is used to identify employment subcenters and to determine their influence on employment density. Proximity to the 20 identified subcenters is used to explain employment density in 1980 and 1990. We argue that density functions are subject to selection bias, and we find significant correlations between employment density and selection equation errors. Subcenter proximity has a significant influence on expected density independent of distance from the Chicago central business district, O'Hare Airport, highway interchanges, and rail lines, suggesting that subcenters offer significant advantages beyond simple access to the transportation network. (C) 1998 Academic Press.