A Quantile Regression Analysis of Housing Price Distributions Near MRT Stations

This study uses the opening of the new Circle Line (CCL) in Singapore as a natural experiment to test the effects of urban rail transit networks on non-landed private housing values in Singapore. Using the network distance measure and a local-polynomial-regression approach, we observe discontinuity in housing prices at 600 meters radius from the closest CCL stations, and use it to divide the study area into the treatment zone and the control zone. Using the non-landed private property transaction data for the period from 2007 to 2013, we estimate the average treatment effects associated with the CCL opening at 8.96%. Using the quantile version of difference-indifferences (QDID) model, we also find significant distributional treatment effects in different price quantiles. The stronger effects are found in the 50th quantile houses at 9.26%, whereas the treatment effect are smaller in the 10th and 90th quantiles at 4.14% and 6.56%, respectively. When we adjust for spatial spillover in the model, the same distributional effects are still observed, but the magnitude of the quantile treatment effect is smaller. We next adopt the conditional quantile decomposition approach, and show that the price changes are attributed to both compositional changes and also the price elasticity changes. After the opening of CCL, transactions in the treatment area tend to be smaller, but more expensive; but the effects are weaker in the control area.