Understanding Residential Location Choices for Climate Change and Transportation Decision Making: Phase 1 Report

This research aims to fill the gap in the knowledge between residential location decisions and preferences and the resulting travel outcomes. In this first phase, the revealed connections between residential choices and travel patterns are examined using recently collected Oregon household travel survey data. Based on distillation of these data, Oregon households are segmented into policy-sensitive markets defined by their differences in household composition, income, and age. Statistical modeling techniques were then applied to analyze the relationship between each identified market segments, their revealed travel outcomes, and three residential location decisions: housing structure (single family or multifamily), tenure (rent or own), and neighborhood type that were combined into sets of alternatives. Each residential location decision was modeled within a nested multinomial logit framework specified for the sample of households of the Portland and Mid-Willamette Valley metropolitan regions in the Oregon Household Activity Survey (OHAS) dataset. To further link the household residential location decisions to travel behavior, a set of multivariate regression models were developed and estimated to understand how the socioeconomic characterization and revealed housing, neighborhood, and tenure decisions of a household related to four travel outcomes: vehicle miles traveled, person miles traveled by mode, number of person trips by mode, and vehicle ownership. These estimates were then used to explore travel differences for households in different lifecycle stages with or without access to light rail transit. This first phase provided insight into the connection between the revealed travel outcomes of Oregon households and their neighborhood, tenure, and housing structure decisions.