Redefining the Value of Accessibility: Toward a Better Understanding of How Accessibility Shapes Household Residential Location and Travel Choices

Abstract

Accessible locations in a metropolitan region afford individuals who occupy them greater convenience to interact with activities distributed across the region. This convenience may translate into a range of economic benefits: reduced time-plus-money spending on travel to reach desirable destinations (termed here travel-cost savings), welfare gains resulting from enhanced social and economic interactions, consumer satisfaction due to a greater choice of activities to engage with, and so on. Yet many urban researchers have either implicitly or explicitly equated the benefits afforded by accessible locations to travel-cost savings (TCS), excluding other forms of benefits from their purview. An exclusive focus on TCS underestimates the value of accessibility and in many policy contexts constitutes a conceptual barrier that impedes the promotion of accessibility-based planning practice and policymaking. For instance, observations of excess commuting are frequently used as evidence refuting the merits of job-housing balance strategies.

This three-paper dissertation challenges this TCS-based view of accessibility benefits. In the first paper, I trace the origin of TCS-based view of accessibility to classic urban economic theories and review its application in residential location studies. In order to test the hypothesis that individuals value accessibility beyond the benefit of travel-cost savings, I develop residential location choice models for two U.S. regions (Puget Sound and Southeast Michigan) to examine if transit accessibility remains a significant predictor of residential location choice after controlling for all possible travel-cost savings associated with it. The results do not support a TCS-based view of accessibility benefits. Considering that only a small fraction of Americans regularly use transit, I conclude that it is probably the option value of transit access that attracts people to transit-accessible neighborhoods.

Building on the idea that individuals value accessibility beyond the benefit of TCS, the second paper critiques the common practice of using VMT reduction as the main empirical measure to represent the transportation benefits of accessibility-enhancing compact-development strategies. I argue that VMT-reduction measures blur the impact that compact development has on the utility that people receive from their environment because compactness can shape personal VMT in opposite directions: a desire for TCS would make people reduce their VMT consumption, but people can end up traveling more if they make more trips and/or travel to more remote destinations in order to gain greater destination utility. I test these ideas by fitting trip-frequency models in the Puget Sound region and in the Southeast Michigan region. Empirical analysis supports my hypothesis by suggesting that compact development has countervailing effects on driving. I thus conclude that VMT-reduction measures underrepresent the transportation benefits of compact development.

To facilitate accessibility-based planning policy implementation, the third paper empirically evaluates the relative importance of walkability, transit accessibility, and auto accessibility in residential location choice across three U.S. regions (Puget Sound, Southeast Michigan, and Atlanta). I find that, in general, transit accessibility is a more important determinant of resident location choice than walkability and auto accessibility. The results further suggest that the “preferred behavior” of households can be different from their actual choice because of housing supply constraints. This implies that if the conditions of housing supply change, estimates of accessibility preferences may change accordingly. This finding challenges the standard practice of land-use and transportation modeling which forecasts future land-use patterns based on presumed stability of historical or present estimates of accessibility preferences.