When is it really justifiable to ignore explanatory variable endogeneity in a regression model?

Jan Kiviet

No 15-05, UvA-Econometrics Working Papers from Universiteit van Amsterdam, Dept. of Econometrics

Abstract: A conversion of standard ordinary least-squares results into inference which is robust under endogeneity of some regressors has been put forward in Ashley and Parmeter, Economics Letters, 137 (2015) 70-74. However, their conversion is based on an incorrect (though by accident conservative) asymptotic approximation and entails a neglected but avoidable randomness. By a very basic example it is illustrated why a much more sophisticated asymptotic expansion under a stricter set of assumptions is required than used by these authors. Next, particular aspects of their consequently .awed sensitivity analysis for an empirical growth model are replaced by results based on a proper limiting distribution for a feasible inconsistency corrected least-squares estimator. Finally we provide references to literature where relevant asymptotic approximations have been derived which should enable to produce similar endogeneity robust inference for more general models and hypotheses than currently available.

Date: 2015-12-05
New Economics Papers: this item is included in nep-ecm
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://ase.uva.nl/binaries/content/assets/subsites ... ics/dp-2015/1505.pdf (application/pdf)

Related works:
Journal Article: When is it really justifiable to ignore explanatory variable endogeneity in a regression model? (2016)
Working Paper: When is it really justifiable to ignore explanatory variable endogeneity in a regression model? (2016)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ame:wpaper:1505

Access Statistics for this paper

More papers in UvA-Econometrics Working Papers from Universiteit van Amsterdam, Dept. of Econometrics Dept. of Econometrics, Universiteit van Amsterdam, Valckenierstraat 65, NL - 1018 XE Amsterdam, The Netherlands. Contact information at EDIRC.
Bibliographic data for series maintained by Noud P.A. van Giersbergen ().