Traditionally, in the assessment of real state values it is assumed that the observations are independent. However, data associated with the position that they occupy in the space are characterized by the spatial dependence. Following the recent international literature, this paper recognizes the importance of the space and shows how it can be incorporated in the specification of hedonic model at the same time it analyzes the real state market of the city of Recife. The data for the empirical test are derived from the IBGE Demographic Census (2000) and from the Caixa Econômica Federal database. The results indicate that considering space with base in distances to influence poles is not enough to completely explain the price variations in relation to the location, once there is a true interaction among the data, so that each building influences its neighbors' price. This means that, contrary to the traditional models hypothesis, the apartment transactions do not occur in an independent fashion. The solution for this problem is to estimate the model using the spatial econometric tools and taking into consideration all possible spatial interactions in the data.
housing price; hedonic price model; spatial econometrics
