Spatial econometrics principles and challenges in Jean Paelinck's research

This special issue contains a selection of many excellent papers and posters presented at the Sixth Jean Paelinck' Seminar of Spatial Econometrics, which was held at the Universidad Autonoma de Madrid (UAM) in October, 2013. The collection of papers addresses some of the major concerns and progression lines proposed by professor Paelinck along his academic life for spatial econometrics: the treatment of spatial data and the need to specify and estimate appropriate econometric models in the realm of regional and urban science. The first two papers in this issue present two methods to detect spatial clusters (Rodero-Cosano, Salinas-Perez, Garcia-Alonso and Salvador-Carulla) and concentrations (Thomas-Agnan and Bonneu) using homogeneous micro-data and individual observations, respectively, to adequately deal with the Modifiable Areal Unit Problem. The following three articles focus on the specification of spatial effectsspatial autocorrelation and spatial heterogeneityin more complex situations. Le Gallo and Chasco propose a non-parametric approach to deal with spatial effects when the model's parametric structure is unknown, as it is the case in hedonic models with large databases. Ezcurra and Rios derive a spatial Durbin panel data model from theory in order to model the spatio-temporal relationship between output volatility and regional growth. Finally, Diaz-Lanchas, Gallego, Llano and De la Mata take into account spatial complexity in trade flows introducing spatial dependence in a gravity model.