Handbooks of Research Methods and Applications series
Edited by Charlie Karlsson, Martin Andersson and Therese Norman
Chapter 1: Spatial econometrics
Spatial econometrics is a subset of the broader econometrics discipline that deals with modeling situations involving spatially dependent sample data. Regional scientists often work with regional data samples where each observation represents a single region and the sample observations consist of a collection of information for n different regions. A key assumption made by conventional linear regression models is that each observation is independent of other observations. Spatial regression models represent a set of methods that do not make this assumption, since it seems unrealistic to assume that economic and other types of activity taking place in one region are independent of those in neighboring regions. Replacing the assumption of independence with one of spatial dependence between regions leads to a situation where spatial spillovers can arise.