New metro and housing price and rent premiums: A natural experiment in China

Causal evidence of housing premiums of new metro lines is indispensable for financing and governing infrastructure investments. Previous studies have investigated the housing effects of urban rail transit with varying methods, while causality remains unsettled. This study used a natural experiment to estimate the causal effects of the new metro interventions on housing premiums in Shenzhen, China. We used metro planning knowledge, reasoning on pursuits in land finance and engineering efficiency to verify the as-if randomness of the treatment-control group assignment in the natural experiment to reinforce the power of causal inference. We applied hedonic difference-in-difference (DID) models to estimate the average treatment effects based on the longitudinal housing price and rent data. We found that housing rents increased significantly and consistently after the metro entered operation, but the price premium varied. In addition, the rent premiums around new metro lines showed a price gradient over the distance to stations. Our findings provide scientific evidence for designing value capture mechanisms (e.g. value-added property tax and rent revenue) to recover metro investment costs in China.