This paper shows the importance of using spatial econometrics in the studies related to the analysis of housing market's behavior. In empirical analysis carried out on the object of estimating a Housing Demand Function for the city of Recife, Brazil, based in information from the Demographic Census (2000) and data from real state financed by Caixa Econômica Federal, strong indications of spatial dependency were verified in all explored economical variables. One verifies that the manner of considering the spatial issue as a function of distances of housing to influence poles or space division in regions, like currently occurring in literature, is not capable of explaining entirely the behavior of the housing demand. The results indicate that the demand equations traditionally estimated, without considering the effects of spatial dependency may generate biased results, where the price-elasticity by the traditional model represents less than 50% of the estimates of the spatial model.
housing demand; spatial econometrics
