ORIGINAL ARTICLE Measuring the impact of pollution on property prices in Madrid: objective versus subjective pollution indicators in spatial models Roma ´n Mı ´nguez • Jose ´-Marı ´a Montero • Gema Ferna ´ndez-Avile ´s Received: 14 March 2011 / Accepted: 12 March 2012 / Published online: 4 April 2012 Ó Springer-Verlag 2012 Abstract Much work has been done in the context of the hedonic price theory to estimate the impact of air quality on housing prices. Research has employed objective measures of air quality, but only slightly confirms the hedonic theory in the best of cases: the implicit price function relating housing prices to air pollution will, ceteris paribus, be negatively sloped. This paper compares the performance of a spatial Durbin model when using both objective and subjective measures of pollution. On the one hand, we design an Air Pollution Indicator based on measured pollution as the objective measure of pollution. On the other hand, the subjective measure of pollution employed to characterize neighborhoods is the percentage of residents who declare that the neighborhood has serious pollution problems, the percentage being referred to as residents’ perception of pollution. For comparison purposes, the empirical part of this research focuses on Madrid (Spain). The study employs a proprietary database containing information about the price and 27 characteristics of 11,796 owner-occupied single family homes. As far as the authors are aware, it is the largest database ever used to analyze the Madrid housing market. The results of the study clearly favor the use of subjective air quality measures. Keywords Housing prices  Environment  Subjective measure of pollution  Spatial hedonic models  Cluster  Kriging JEL Classification Q51  R32  C23 R. Mı ´nguez University of Castilla-La Mancha (at Cuenca), Avenida de los Alfares, 44, 16002 Cuenca, Spain e-mail: Roman.minguez@uclm.es J.-M. Montero (&)  G. Ferna ´ndez-Avile ´s University of Castilla-La Mancha (at Toledo), Cobertizo S. Pedro Ma ´rtir, SN, 45071 Toledo, Spain e-mail: Jose.mlorenzo@uclm.es G. Ferna ´ndez-Avile ´s e-mail: Gema.faviles@uclm.es 123 J Geogr Syst (2013) 15:169–191 DOI 10.1007/s10109-012-0168-x