Who benefits from clean air? Evidence from Beijing using housing price dataset (with Ziye Zhang)

This paper aims to investigate the customers preference for cleaner air and green space in Beijing, a city with serious air pollution issue, albeit improved, in recent years. The hedonic housing price model is applied with different fixed effects and cluster robust standard error at Jiedao level. Furthermore, we developed fancy environmental measures, i.e. AOD (aerosol optical depth) for air quality and NDVI for green space, in addition to the tradition measures of in-situ PM2.5 concentration and distance to park/open space. The results show that the PM2.5 and AOD are consistent measures of air quality. One unit increase in PM2.5 with 30-day time lag results in 0.1% decrease in housing price, and the effect is 4.6% for AOD scaled 0 to 4. If extending to 90-day time lag, the PM2.5 effect escalates to 0.3%, which suggests long-time air pollution can further reduce the housing value. We explore the temporal heterogeneity of preference to cleaner air and find that margins follow an inverted U-shape curve which indicates a trend of less preference on air quality from 2013 to 2016, and an increasing preference from 2016 to 2018. At spatial dimension, customers buying houses in Tongzhou district reveal higher preference on cleaner air, whereas in Mentougou district, the preference are at the lowest level compared with other districts.