Life history-oriented residential location choice model: A stress-based two-tier panel modeling approach

Abstract

This paper presents a life history-oriented modeling framework to investigate residential location decisions as a two-tier process of location search and location choice. In the first tier, a stress-based location search model is developed by assuming that households search for a new location due to continual stress at different life-domains. The search model adopts a fuzzy logic-based modeling method that mimics the inter-dependencies between push and pull factors. In the second tier, a location choice model is developed that accommodates how location decisions interact with life-cycle events at different life-domains. The model utilizes a latent segmentation-based logit modeling technique to address the panel effect of the households’ housing career. The model results suggest that households in general show preference for larger lots, and locations closer to work place, transit stop, and health service. Location choice is found to be significantly influenced by the life-cycle events as well as the lead and lagged effects. For example, the birth of a child magnifies the need of larger lots. The life-history effects, however vary across two segments. Suburbanite households in segment two prefer larger lots following a job change; whereas, urbanite households in segment one show a negative relationship. The adjustment period for a job change is longer than that of addition of a new job. A longer adjustment time is also found in the case of the first time vehicle purchase than acquisition of a vehicle. Presence of children influences suburbanite households to reside closer to work place. Urbanite households with children prefer to live closer to school.