A more powerful subvector Anderson Rubin test in linear instrumental variables regression

Patrik Guggenberger, Frank Kleibergen, Sophocles Mavroeidis

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

We study subvector inference in the linear instrumental variables model assuming homoskedasticity but allowing for weak instruments. The subvector Anderson and Rubin (1949) test that uses chi square critical values with degrees of freedom reduced by the number of parameters not under test, proposed by Guggenberger, Kleibergen, Mavroeidis, and Chen (2012), controls size but is generally conservative. We propose a conditional subvector Anderson and Rubin test that uses data-dependent critical values that adapt to the strength of identification of the parameters not under test. This test has correct size and strictly higher power than the subvector Anderson and Rubin test by Guggenberger et al. (2012). We provide tables with conditional critical values so that the new test is quick and easy to use. Application of our method to a model of risk preferences in development economics shows that it can strengthen empirical conclusions in practice.

Original languageEnglish (US)
Pages (from-to)487-526
Number of pages40
JournalQuantitative Economics
Volume10
Issue number2
DOIs
StatePublished - May 1 2019

Fingerprint

Instrumental variables regression
Critical value
Development economics
Weak instruments
Instrumental variables
Risk preferences
Inference

All Science Journal Classification (ASJC) codes

  • Economics and Econometrics

Cite this

Guggenberger, Patrik ; Kleibergen, Frank ; Mavroeidis, Sophocles. / A more powerful subvector Anderson Rubin test in linear instrumental variables regression. In: Quantitative Economics. 2019 ; Vol. 10, No. 2. pp. 487-526.
@article{8166a62477db4c33b6899d86f9779f53,
title = "A more powerful subvector Anderson Rubin test in linear instrumental variables regression",
abstract = "We study subvector inference in the linear instrumental variables model assuming homoskedasticity but allowing for weak instruments. The subvector Anderson and Rubin (1949) test that uses chi square critical values with degrees of freedom reduced by the number of parameters not under test, proposed by Guggenberger, Kleibergen, Mavroeidis, and Chen (2012), controls size but is generally conservative. We propose a conditional subvector Anderson and Rubin test that uses data-dependent critical values that adapt to the strength of identification of the parameters not under test. This test has correct size and strictly higher power than the subvector Anderson and Rubin test by Guggenberger et al. (2012). We provide tables with conditional critical values so that the new test is quick and easy to use. Application of our method to a model of risk preferences in development economics shows that it can strengthen empirical conclusions in practice.",
author = "Patrik Guggenberger and Frank Kleibergen and Sophocles Mavroeidis",
year = "2019",
month = "5",
day = "1",
doi = "10.3982/QE1116",
language = "English (US)",
volume = "10",
pages = "487--526",
journal = "Quantitative Economics",
issn = "1759-7323",
publisher = "The Economic Society",
number = "2",

}

A more powerful subvector Anderson Rubin test in linear instrumental variables regression. / Guggenberger, Patrik; Kleibergen, Frank; Mavroeidis, Sophocles.

In: Quantitative Economics, Vol. 10, No. 2, 01.05.2019, p. 487-526.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A more powerful subvector Anderson Rubin test in linear instrumental variables regression

AU - Guggenberger, Patrik

AU - Kleibergen, Frank

AU - Mavroeidis, Sophocles

PY - 2019/5/1

Y1 - 2019/5/1

N2 - We study subvector inference in the linear instrumental variables model assuming homoskedasticity but allowing for weak instruments. The subvector Anderson and Rubin (1949) test that uses chi square critical values with degrees of freedom reduced by the number of parameters not under test, proposed by Guggenberger, Kleibergen, Mavroeidis, and Chen (2012), controls size but is generally conservative. We propose a conditional subvector Anderson and Rubin test that uses data-dependent critical values that adapt to the strength of identification of the parameters not under test. This test has correct size and strictly higher power than the subvector Anderson and Rubin test by Guggenberger et al. (2012). We provide tables with conditional critical values so that the new test is quick and easy to use. Application of our method to a model of risk preferences in development economics shows that it can strengthen empirical conclusions in practice.

AB - We study subvector inference in the linear instrumental variables model assuming homoskedasticity but allowing for weak instruments. The subvector Anderson and Rubin (1949) test that uses chi square critical values with degrees of freedom reduced by the number of parameters not under test, proposed by Guggenberger, Kleibergen, Mavroeidis, and Chen (2012), controls size but is generally conservative. We propose a conditional subvector Anderson and Rubin test that uses data-dependent critical values that adapt to the strength of identification of the parameters not under test. This test has correct size and strictly higher power than the subvector Anderson and Rubin test by Guggenberger et al. (2012). We provide tables with conditional critical values so that the new test is quick and easy to use. Application of our method to a model of risk preferences in development economics shows that it can strengthen empirical conclusions in practice.

UR - http://www.scopus.com/inward/record.url?scp=85065446978&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065446978&partnerID=8YFLogxK

U2 - 10.3982/QE1116

DO - 10.3982/QE1116

M3 - Article

AN - SCOPUS:85065446978

VL - 10

SP - 487

EP - 526

JO - Quantitative Economics

JF - Quantitative Economics

SN - 1759-7323

IS - 2

ER -