Nonlinear public transit accessibility effects on housing prices: Heterogeneity across price segments

Document Type

Journal Article

Publication Date


Subject Area

place - asia, place - urban, place - cbd, mode - bus, mode - subway/metro, policy - equity, land use - impacts, land use - planning, planning - methods


Nonlinearity, Public transit accessibility, Residential property value, Inequity, Machine learning, Shanghai (China)


Although numerous studies have been conducted to analyze the relationship between public transit accessibility and housing prices, they often assumed a linear relationship. Using a dataset of 196,232 s-hand residential properties in Shanghai (China), this study applies the gradient boosting regression trees (GBRT) method to investigate the complicated relationships between public transit accessibility and housing prices across price segments. The results show that for low- and median-priced houses, the travel time by public transit to the central business district (CBD) contributes the most to housing prices while for high-priced houses, the systemwide metro accessibility contributes the most. Public transit accessibility has significant nonlinear and threshold effects on housing prices. Public transit accessibility has an overall positive effect on housing prices, though some negative effects occur within certain intervals. Different patterns are observed across price segments. High-priced houses receive a much larger premium from the closeness to CBD and systemwide metro accessibility but are more negatively affected by both local and systemwide bus accessibility, compared to low- and median-priced houses. The differences in transit's value-added effects across price segments may bring about inequity in accessibility across different income groups. Relevant policies on how to reduce such an inequity issue are finally discussed.


Permission to publish the abstract has been given by Elsevier, copyright remains with them.


Transport Policy Home Page: