A nonparametric analysis of employment density in a polycentric city

Daniel P. McMillen, John F. McDonald

Research output: Contribution to journalArticlepeer-review


Nonparametric estimation procedures offer distinct advantages in modeling polycentric cities because they are flexible enough to account for functional form misspecification and incorrect subcenter sites. This paper presents locally weighted (LW) regression estimates of employment density in suburban Chicago. LW regression estimates are more accurate than OLS regression and capture the effects of missing variables. The results demonstrate that Chicago is indeed a polycentric city: although the traditional city center continues to affect employment density patterns in the suburbs, local peaks have developed around secondary employment centers.

Original languageEnglish (US)
Pages (from-to)591-612
Number of pages22
JournalJournal of Regional Science
Issue number4
StatePublished - Nov 1997
Externally publishedYes

ASJC Scopus subject areas

  • Development
  • Environmental Science (miscellaneous)


Dive into the research topics of 'A nonparametric analysis of employment density in a polycentric city'. Together they form a unique fingerprint.

Cite this