Decoupling the Effect of Land Depreciation on the Construction of Residential Property Price Indices

Unreliable property price information can lead to misallocation of land use and inefficient utilization of scarce resources. Traditional hedonic pricing models estimated by using ordinary least squares (OLS) commonly include property age as an explanatory variable, which implicitly assumes that both land and building structures depreciate over time. However, recent research argues that residential property price indices (RPPIs) should separate the depreciation of the building structure from the land component. Under this view, property age should only interact with the structural characteristics of the building, not the land. This paper estimates RPPIs by using both a standard time dummy hedonic price model and a modified version that adjusts for structural quality by interacting age with building features. Using a dataset of 451,894 observations, we find that the standard model tends to produce a downward bias in periods of mild price variation and an upward bias during periods of large price fluctuations. Finally, we compare our estimated indices with those published by the local government. The results show that turning points, price peaks, and troughs are largely consistent across both sets of indices.