Density and transport mode choice in Australian, Candian and US cities


Paul Mees

Document Type

Conference Paper

Publication Date


Subject Area

land use - planning, land use - urban density, land use - urban density, ridership - mode choice, ridership - commuting, place - urban


transport planning, urban density, mode choice, journey-to-work


This paper re-examines the relationship between population density and transport mode choice, taking another look at the ideas that have come to be known as the ‘compact city’. It begins by reviewing the origins of the view that density determines mode choice, and that viable public transport cannot be provided below a density threshold variously estimated at 30 to 100 persons per hectare. The claim has been widely made, but an examination of the alleged basis reveals multiple layers of citation ultimately deriving from a single source, the Chicago Area Transportation Study 1956. The CATS analysis erroneously attributed poor suburban public transport to low densities, when the real causes were failures of planning and policy. The paper then reviews the more recent data provided by Newman and Kenworthy, who found a similar relationship to that reported in CATS. Use of the most recent census data from Australia, Canada and the United States suggests that the Newman-Kenworthy data contained errors in the estimation of urban densities. When these are corrected, the results reveal only a very weak correlation between density and public transport use, and no correlation at all with walking and cycling. The paper concludes that the ‘compact city’ notion is not substantiated by evidence.