|This centre is a member of The LSE Research Laboratory [RLAB]: CASE | CVER | CEP | SERC | STICERD||Cookies?|
Paper No' SERCDP0017: | Full paper
Save Reference as: BibTeX File | EndNote Import File
Keywords: spatial weights spatial dependence spatial models
JEL Classification: C31; O18; R11
Is hard copy/paper copy available? YES - Paper Copy Still In Print.
This Paper is published under the following series: SERC Discussion Papers
Share this page: Google Bookmarks | Facebook | Twitter
Abstract:This paper provides a survey and critique of how spatial links are taken into account in empirical analysis by applied economists/regional scientists. Spatial spillovers and spatial interrelationships between economic variables (e.g. unemployment, GDP, etc) are likely to be important, especially because of the role of local knowledge diffusion and how trade (interregional exports and imports) can potentially act to diffuse technology. Since most empirical economic studies ignore spatial autocorrelation they are thus potentially mis-specified. This has led to various approaches to taking account of spatial spillovers, including econometric models that dependent on specifying (correctly) the spatial weights matrix, W. The paper discusses the standard approaches (e.g., contiguity and distance measures) in constructing W, and the implications of using such approaches in terms of the potential mis-specification of W. We then look at more recent attempts to measure W in the literature, including: Bayesian (searching for ‘best fit’); non-parametric techniques; the use of spatial correlation to estimate W; and other iteration techniques. The paper then considers alternative approaches for including spatial spillovers in econometric models such as: constructing (weighted) spillover variables which directly enter the model; allowing non-contiguous spatial variables to enter the model; and the use of spatial VAR models. Lastly, we discuss the likely form of spatial spillovers and therefore whether the standard approach to measuring W is likely to be sufficient.
Copyright © RLAB & LSE 2003 - 2020 | LSE, Houghton Street, London WC2A 2AE | Contact: RLAB | Site updated 24 May 2020