Conditions and effectiveness of land use as a mobility tool

Abstract

This dissertation examines the potential of land use as a mobility tool to affect travel, a subject of long and ongoing policy debate. Land use strategies such as densification, mixed-use development, and non-driving-oriented design have been recommended by many to reduce vehicle travel. Others argue that land use is an ineffective mobility tool; direct and effectual policies are economic measures such as pricing. This dissertation suggests that either is necessary but not sufficient. To achieve the environmental and social objectives of transportation, the two should act together as complements. The mobility role of land use is to modify transportation supply and to support expansion of travel choices, whereas pricing is to manage and redirect vehicle travel demand. This dissertation presents two case studies: Metropolitan Boston and Hong Kong. Taking a disaggregate approach, the empirical analysis builds on the economic choice theory and focuses on three aspects of travel behavior: mode choice, trip frequency and automobile dependence. Logit models of mode choice and trip frequency are estimated to examine the importance and magnitude of land use affecting travel when travel costs and socio-demographic factors are controlled for.

(cont.) The effects of densification and pricing on mode choice are extrapolated with incremental logit modeling while controlling for the impacts of these policies on individual accessibility, i.e., the utility associated with all available modes. Logit captivity models are estimated to quantify and explain automobile dependence in the process of choice set generation. The analysis shows that densification has significant influence on mode choice and automobile dependence due to the differentiated impacts of land use on modal supply. The influence of street patterns on travel is not much from the geometric difference between gridiron and cul-de-sac, but from the viability of the circulation systems for alternatives to driving. Automobile dependence in the Boston area displays certain patterns in the spatial, social and activity dimensions. The sources of automobile dependence are diverse, often lying beyond the physical environment. The Hong Kong case demonstrates that the presence of economic measures is a precondition for land use to be an effective mobility tool. Hong Kong's non-driving-dominated travel does not come by default from its unique land use pattern, but is accompanied by strong fiscal and regulatory constraints to private transportation.