Stated Preferences in Modelling Accessibility

Abstract

The stated preference (SP) experimental design problem in measuring and representing the role of accessibility in modelling urban residential location is considered. The preferences of households between certain accessibility attributes and the monthly rent of the house, ceteris paribus with respect to the rest of the location attributes, is analysed in order to isolate the effects of interest. A review of the conceptual definition of accessibility, its relationship with the transportation system, its functionality and the different measures used up until now, allowed generation of an SP data collection instrument. This was employed to subject a sample of households to the location decision based on several transport system variables. The experiment made several novel contributions. First, the whole household was considered as the choice-making entity, with decisions shared by all its members. Second, the alternatives considered different accessibility levels without modifying the household's current location. Using the data gathered, models were estimated with consistent parameters in terms of their signs and significance tests, and some functional forms associated with the perception of the access concept were defined. Furthermore, a valuation of time index was generated and this allowed a new discussion to be opened regarding its relation to the subjective value of time traditionally derived from mode-choice models.