USING A SPATIAL ENDOGENOUS METHOD TO DETECT HOUSING SUBMARKETS: AN APPLICATION TO TUCSON

While tools modelling spatial autocorrelation have been unanimously adopted in the housing prices literature, there is still no consensus on the appropriate methodology to identify submarkets, i.e. on how to count for spatial heterogeneity. In this paper we propose an innovative methodology that endogenously detects submarkets while counting for spatial autocorrelation across housing prices. The advantage of an endogenous detection is to avoid arbitrariness in the sense that submarkets are defined by the variables of our model only. We apply our methodology to Tucson’s housing market for which our results provide a strong evidence of spatial heterogeneity.