Urban population density and freeways in North America: A Re-assessment

Researchers seeking to identify public policies to reduce automobile use have frequently compared characteristics such as population density and freeway provision across metropolitan regions to identify differences, their causes, and their implications for automobile reliance. One frequent comparison has been between metropolitan regions in Canada and the United States given their similar contexts but differing modal shares. However, past studies have produced inconsistent findings with respect to how Canadian and American metropolitan regions can be distinguished with respect to population densities and freeways. In response to methodological concerns with past research, this paper measures population densities and freeways in the 57 North American metropolitan regions with at least one million residents. Two spatial scales (metropolitan region and central core) are used to measure overall density, population-weighted average census tract density, and density of the top 5% of the population. These same spatial scales are also used to measure the absolute and per capita quantities of freeway lanes. The paper provides methodological explanations using Los Angeles and Vancouver as examples. While Canada's metropolitan regions are commonly portrayed as denser and less auto-oriented than their US counterparts, the six largest of Canada's metropolitan regions do not have population higher densities and fewer freeways than all of the 51 largest US metropolitan regions. Indeed, this generalization holds only when individual results are combined to produce national averages, as there are numerous US metropolitan regions that have density and freeways in the same range as Canada's metropolitan regions. The paper discusses some of the differences and recommends approaches for future comparative research into the causes of variation in transport and land use.