查看文章

This paper quantifies how arbitrary new transportation technologies–so-called “magic carpets”–that reduce the burden of travel may affect accessibility and, in turn, residential location choices, using the Puget Sound Region (PSR) as a case study. The work focuses on quantifying how changes in the value of travel time could affect the probability of households choosing to live in different types of neighborhoods. The general approach is to estimate a multinomial logit residential location choice model that incorporates the value of travel time via two measures of accessibility. This paper describes the results of this process for the PSR, which has involved calculating and mapping accessibility under current and future conditions with a reduced value of travel time, then calculating a multinomial logit residential choice model based on these results. The authors found that replacing personal cars with magic carpets would tend to shift population to the urban fringes, while adding magic carpets to the mode mix as an on-demand service would tend to increase the attractiveness of living in more urbanized areas. The authors also learned, despite adhering to peer reviewed methods and following what mainstream literature on residential location choice modeling suggested, considerable effort was required to obtain results that were not prima facie absurd in terms of finding “right” model specifications.