Spatial dependence and spatial heterogeneity in the analysis of regional economic performance and house price developments

When analyzing economic aspects at regional levels, the spatial dimension of the data plays a crucial role. This is the case as spatial data and models are often characterized by two spatial effects, namely spatial dependence and spatial heterogeneity. The field of spatial econometrics is a separate field in econometrics that explicitly incorporates these spatial effects into econometric models. Following Anselin (1988), spatial dependence implies that observations in one region depend on observations in neighboring regions. This is in contrast to the conventional assumption in cross-sectional samples, where it is assumed that observations made in one regions are independent of observations in other regions (LeSage and Pace, 2009). Spatial heterogeneity implies that model coefficients vary with location, which is again in contrast to the conventional assumption of homogeneous parameters (Anselin, 1988). This dissertation consists of three chapters, with each chapter analyzing a different spatial aspect of regional data. The first chapter deals with spatial dependence, with a special focus on the construction and the varying effects of different spatial weight matrices. The second chapter is about spatial heterogeneity. The third chapter combines spatial dependence, spatial heterogeneity, and differences in spatial dependence across time and space into one model. The three chapters analyze different economic aspects but they are linked by the spatial setting.