Spatial fixed effects and spatial dependence in a single cross-section
We investigate the common conjecture in applied econometric work that the inclusion of spatial fixed effects in a regression specification for a single cross‐sectional data set removes spatial dependence. We demonstrate analytically and by means of a series of simulation experiments how evidence of...
Saved in:
Published in | Papers in regional science Vol. 92; no. 1; pp. 3 - 17 |
---|---|
Main Authors | , |
Format | Journal Article |
Language | English |
Published |
Blackwell Publishing Ltd
01.03.2013
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | We investigate the common conjecture in applied econometric work that the inclusion of spatial fixed effects in a regression specification for a single cross‐sectional data set removes spatial dependence. We demonstrate analytically and by means of a series of simulation experiments how evidence of the removal of spatial autocorrelation by spatial fixed effects may be spurious when the true data generating processes (DGP) takes the form of a spatial lag or spatial error dependence. In addition, we also show that spatial fixed effects correctly remove spatial correlation only in the special case where the dependence is group‐wise, with all observations in the same group as neighbours of each other.
Resumen
Se investiga la conjetura utilizada frecuentemente en econometría aplicada de que la inclusión de efectos fijos espaciales en una especificación de regresión para un único conjunto de datos transversales elimina la dependencia espacial. Demostramos analíticamente, y por medio de una serie de experimentos de simulación, cómo las pruebas de la eliminación de la autocorrelación espacial por efectos fijos espaciales pueden ser engañosas cuando los verdaderos procesos de generación de datos (DGP, siglas en inglés) consisten en un retardo espacial o una dependencia del error espacial. Además, mostramos también que los efectos fijos espaciales eliminan correctamente la correlación espacial sólo en el caso especial de que la dependencia sea en sentido grupal, en la que todas las observaciones del mismo grupo sean vecinas entre sí. |
---|---|
Bibliography: | We are grateful to acknowledge the many useful suggestions by Julia Koschinsky, David Folch, Pedro Amaral and the referees of Papers in Regional Science. David Folch developed the original architecture of the GeoDa Center spatial econometrics workbench upon which the current simulations were built. The research behind this paper was supported in part by US National Cancer Institute Grant 1R01CA126858-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the US National Cancer Institute, or the US National Institutes of Health. ArticleID:PIRS480 ark:/67375/WNG-ZDZPZ9DK-3 istex:EFCA1602309FB7D561EF7EA2824DE77A3B16F04E ObjectType-Article-2 SourceType-Scholarly Journals-1 ObjectType-Feature-1 content type line 23 |
ISSN: | 1056-8190 1435-5957 |
DOI: | 10.1111/j.1435-5957.2012.00480.x |