Alternative ways of obtaining Hausman's test using artificial regressions

Badi H. Baltagi; Long Liu

doi:10.1016/j.spl.2007.02.003

Alternative ways of obtaining Hausman's test using artificial regressions

Badi H. Baltagi, Long Liu

Department of Economics

Research output: Contribution to journal › Article › peer-review

6 Scopus citations

Abstract

Hausman [1978. Specification tests in econometrics. Econometrica 46, 1251-1271] showed that his specification test in panel data, which is based on the contrast between fixed effects (FE) and the random effects (RE) estimators, can also be obtained as a Wald test from an artificial OLS regression. This paper generalizes this result to the 2SLS counterpart. Instead of a contrast based on FE 2SLS and RE 2SLS as proposed more recently by Baltagi [2004. A Hausman test based on the difference between fixed effects two-stage least squares and error components two-stage least squares, problem 04.1.1. Econometric Theory 20, 223-224] artificial regressions are proposed that would generate this Hausman test using 2SLS regressions.

Original language	English (US)
Pages (from-to)	1413-1417
Number of pages	5
Journal	Statistics and Probability Letters
Volume	77
Issue number	13
DOIs	https://doi.org/10.1016/j.spl.2007.02.003
State	Published - Jul 15 2007

Keywords

2SLS
Hausman test
Panel data
Random effects
Wald test

ASJC Scopus subject areas

Statistics and Probability
Statistics, Probability and Uncertainty

Access to Document

10.1016/j.spl.2007.02.003

Cite this

@article{9d3e64279cc64b7387fc10c27fb68c01,

title = "Alternative ways of obtaining Hausman's test using artificial regressions",

abstract = "Hausman [1978. Specification tests in econometrics. Econometrica 46, 1251-1271] showed that his specification test in panel data, which is based on the contrast between fixed effects (FE) and the random effects (RE) estimators, can also be obtained as a Wald test from an artificial OLS regression. This paper generalizes this result to the 2SLS counterpart. Instead of a contrast based on FE 2SLS and RE 2SLS as proposed more recently by Baltagi [2004. A Hausman test based on the difference between fixed effects two-stage least squares and error components two-stage least squares, problem 04.1.1. Econometric Theory 20, 223-224] artificial regressions are proposed that would generate this Hausman test using 2SLS regressions.",

keywords = "2SLS, Hausman test, Panel data, Random effects, Wald test",

author = "Baltagi, {Badi H.} and Long Liu",

year = "2007",

month = jul,

day = "15",

doi = "10.1016/j.spl.2007.02.003",

language = "English (US)",

volume = "77",

pages = "1413--1417",

journal = "Statistics and Probability Letters",

issn = "0167-7152",

publisher = "Elsevier",

number = "13",

}

TY - JOUR

T1 - Alternative ways of obtaining Hausman's test using artificial regressions

AU - Baltagi, Badi H.

AU - Liu, Long

PY - 2007/7/15

Y1 - 2007/7/15

N2 - Hausman [1978. Specification tests in econometrics. Econometrica 46, 1251-1271] showed that his specification test in panel data, which is based on the contrast between fixed effects (FE) and the random effects (RE) estimators, can also be obtained as a Wald test from an artificial OLS regression. This paper generalizes this result to the 2SLS counterpart. Instead of a contrast based on FE 2SLS and RE 2SLS as proposed more recently by Baltagi [2004. A Hausman test based on the difference between fixed effects two-stage least squares and error components two-stage least squares, problem 04.1.1. Econometric Theory 20, 223-224] artificial regressions are proposed that would generate this Hausman test using 2SLS regressions.

AB - Hausman [1978. Specification tests in econometrics. Econometrica 46, 1251-1271] showed that his specification test in panel data, which is based on the contrast between fixed effects (FE) and the random effects (RE) estimators, can also be obtained as a Wald test from an artificial OLS regression. This paper generalizes this result to the 2SLS counterpart. Instead of a contrast based on FE 2SLS and RE 2SLS as proposed more recently by Baltagi [2004. A Hausman test based on the difference between fixed effects two-stage least squares and error components two-stage least squares, problem 04.1.1. Econometric Theory 20, 223-224] artificial regressions are proposed that would generate this Hausman test using 2SLS regressions.

KW - 2SLS

KW - Hausman test

KW - Panel data

KW - Random effects

KW - Wald test

UR - http://www.scopus.com/inward/record.url?scp=34447544665&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34447544665&partnerID=8YFLogxK

U2 - 10.1016/j.spl.2007.02.003

DO - 10.1016/j.spl.2007.02.003

M3 - Article

AN - SCOPUS:34447544665

SN - 0167-7152

VL - 77

SP - 1413

EP - 1417

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

IS - 13

ER -

Alternative ways of obtaining Hausman's test using artificial regressions

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this