Planning and Markets: Crane : I. Index and Introduction

Endnotes

Prepared for the July 2002 Colloquium on "Sprawl in Western Europe and the United States," Chateau de la Bretesche, Missillac, France, sponsored by the Borchard Foundation, and the November 2002 meeting of the Association of Collegiate Schools of Planning in Baltimore. We are grateful to Jonathan Levine, Remy Prud'homme, Harry Richardson, and other participants for very helpful questions and comments. In addition, an anonymous referee and seminar attendees at UC Berkeley, UC Davis, and the University of Southern California provided valuable advice.

This is different from saying that commutes should or should not be reduced. Urban theory argues that traffic congestion by itself, as an externality, leads to overly decentralized monocentric cities, but that the consequences are less clear when employment also decentralizes (Anas, Arnott and Small, 1998). Thus, unpriced congestion by itself contributes to longer commutes, holding employment location fixed. However, the present study only explains commuting, not its welfare economics.
But see Chen, Ewing and Pendall (2002) for a recent effort aimed at explaining metropolitan level travel outcomes with aggregate sprawl measures.
See http://www.bea.doc.gov/bea/regional/articles/lapi6992.htm.
See http://www.census.gov/population/estimates/metro-city/99mfips.txt.
Central cities include the largest city in an MSA; any additional cities with more than 250,000 population or 100,000 workers; any additional cities of at least 25,000 in size with a large employment base (greater than 75 percent of population), 40 percent of whose employed residents work there; and secondary non-contiguous urbanized areas that meet certain criteria. See http://www.census.gov/population/www/estimates/mastand.html, Part III, Section 4.
The metropolitan areas dropped were the Bangor (ME) MSA; the Barnstable-Yarmouth (MA) MSA; the Boston-Worcester-Lawrence (MA, NH, ME, & CT) CMSA; the Burlington (VT) MSA; the Hartford (CT) MSA; the Lewiston-Auburn (ME) MSA; the New London-Norwich (CT-RI) MSA; the New York-Northern New Jersey-Long Island (NY, NJ, CT, & PA) CMSA; the Pittsfield (MA) MSA; the Portland (ME) MSA; the Providence-Fall River-Warwick (RI & MA) MSA; and the Springfield (MA) MSA.
However, this sample excludes the two largest California CMSAs, for the data availability reasons given above, which likely reduces the average commute distance for the western region.
Only one individual from each eligible housing unit is used in the analysis to avoid the need to correct the standard errors for clustering within households.

page 21