Approximate Bias Correction in Econometrics

James MacKinnon () and Anthony A. Smith

No 919, Working Papers from Queen's University, Department of Economics

Abstract: This paper discusses ways to reduce the bias of consistent estimators that are biased in finite samples. It is necessary that the bias function, which relates parameter values to bias, should be estimable by computer simulation or by some other method. If so, bias can be reduced or, in some cases that may not be unrealistic, even eliminated. In general, several evaluations of the bias function will be required to do this. Unfortunately, reducing bias may increase the variance, or even the mean squared error, of an estimator. Whether or not it does so depends on the shape of the bias functions. The techniques of the paper are illustrated by applying them to two problems: estimating the autoregressive parameter in an AR(1) model with a constant term, and estimation of a logit model.

Keywords: bias function; mean squared error; simulation; finite samples (search for similar items in EconPapers)
JEL-codes: C13 (search for similar items in EconPapers)
Date: Written
View list of references

Published in Journal of Econometrics, 85, 1998

Downloads: (external link)
http://www.econ.queensu.ca/working_papers/papers/qed_wp_919.pdf First version 1995 (application/pdf)

Related works:
Working Paper: Approximate Bias Correction in Econometrics
Working Paper: Approximate Bias Correction in Econometrics (1996)
Journal Article: Approximate bias correction in econometrics (1998)
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: http://EconPapers.repec.org/RePEc:qed:wpaper:919

Access Statistics for this paper

More papers in Working Papers from Queen's University, Department of Economics
Contact information at EDIRC.
Series data maintained by Mark Babcock ().