University of California Transportation Center

Although land use planning and urban design are increasingly touted as powerful tools for influencing transportation behavior, only modest empirical evidence for this relationship exists. Here, the results from a two-day activity diary are combined with innovative GIS-based measures of urban form and land use pattern to statistically test potential influences on non-commute home-based mode choice. Local measurement at multiple scales is promoted as a realistic means of quantifying an individual’s perception of the neighboring urban environment, and multinomial logit models are specified for various trip purposes. In all models tested, the inclusion of measures of urban form or land use pattern improves the model. Generally, the measurable role of physical factors is small; however, their influence is relatively large in a model predicting station access mode choice.