| This dissertation employs a discrete-choice model of residential choice for the Metropolitan area of Portland, Oregon, using microdata from the 1994–1995 “Travel Activity and Household Behavior Survey”. Individual and neighborhood characteristics, together with accessibility measures, are used to capture microbehavior regarding residential preferences. Explicit school quality and safety measures are used to reflect neighborhood quality, while life cycle is used to determine changes in demanded levels of locational quality. In addition, this research develops an explicit approach to choice set generation based on travel-time tolerances, and demonstrates how the lift-curve plot can be useful in nonbinary choice models.; The findings of this study indicate that locational quality significantly affects residential choice. In addition, these findings show that modeling microbehavior, using explicit neighborhood quality factors together with the life cycle concept, can improve the explanatory power of conventional residential choice models. |
