Economic impacts of a subway system: Exploring local contexts in a metropolitan area

Abstract

With rapid urban population growth, more subway systems have been developed in metropolitan areas around the world. It is commonly presumed that the improved mobility and accessibility provided by subway systems are capitalized as a financial premium for nearby real estate assets. This study empirically explores this presumption with a case study of Gwangju, a metropolitan area in South Korea. The study site is found to have had a negative economic impact on housing prices from a subway system in the mid-2010s; however, the sign has changed over time, although the subway has experienced no actual expansion or operational changes. This indicates that the economic impact of subway systems may not necessarily become positive or remain static. We examine a wide range of local contexts to explain the sign change. Our study adopts machine learning algorithms to precisely address the nonlinearity inherent in our dataset and quantifies the importance of housing factors in the assessment of housing prices using SHapley Additive exPlanation. We also employ a series of hedonic pricing models to determine the effects of proximity to subway lines on housing prices across years. Key changes in the later 2010s, including the increased modal share of the subway system and local politics, should have resulted in the potential benefits of the to-be-extended network being widely and fully realized. Our findings contribute to understanding the mechanisms of the economic impact of evolving transportation infrastructure on housing prices.