Housing studies typically use parcel level distance variables or some variant of the spatial weights matrix approach to incorporate spatial effects into hedonic regression models. In this paper, using very detailed data on household attitudes and parcel attributes, hedonic regression residuals are used in a structural equations framework to check for additional spatial effects in the hedonic coefficients beyond those captured in the hedonic regression itself. In this way, a ‘‘nearest neighbors’’ approach utilizing parcel level distance variables is compared directly to OLS estimation using spatial variables, showing the relative efficiency of the estimates in the former approach.
Author: Clifford A. Lipscomb
Originally published in the Journal of Housing Research, Vol. 15, No. 2 (2006)