Intercity Variations in the Relationship Between Urban Form and Automobile Dependence: Disaggregate Analyses of Boston, Massachusetts; Portland, Oregon; and Houston, Texas


Ming Zhang

Document Type

Journal Article

Publication Date


Subject Area

ridership - mode choice, ridership - elasticity, place - urban


Urban form, Public policy, Portland (Oregon), Mode choice, Modal choice, Land use, Houston (Texas), Elasticity (Economics), Disaggregate analysis, Choice of transportation, Boston (Massachusetts), Automobile dependence, Access


This study was motivated by the need for more empirical research on the urban form-travel connection. A two-tiered travel effect is expected from strengthening the urban form-travel connection: the enhancement of access to choices and a shift in travel mode choice from driving to nondriving. Existing studies have focused primarily on the second-tier effect but have largely omitted the first. This study attempted to fill that gap. Through joint-logit modeling of choice set formation and travel mode choice in three cities—Boston, Massachusetts; Portland, Oregon; and Houston, Texas—the study measured the degree of automobile dependence in the three cities. It also estimated elasticities of automobile dependence and of driving probabilities with respect to land use densification, transit access improvement, and control of motorization. There were large variations in the levels of automobile dependence and their elasticity estimates among the three cities. Public policies aimed at reducing automobile dependence should be formulated and evaluated based not just on the final outcome of modal split but also on the provision of travel options to travelers. As cities differ in their existing urban forms, currently available transportation services, and prevailing preferences of travel, it is important to recognize that the same set of policy strategies implemented in different cities is unlikely to generate the same level of effects in reducing automobile dependence.