Two Applications of wild bootstrap methods to Improve Inference in cluster-IV models

Leandro M Magnusson, Keith Finlay

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Microeconomic data often have within-cluster dependence, which affects standard error estimation and inference. When the number of clusters is small, asymptotic tests can be severely oversized. In the instrumental variables model, the potential presence of weak instruments further complicates hypothesis testing. We use wild bootstrap methods to improve inference in two empirical applications with these characteristics. Building from estimating equations and residual bootstraps, we identify variants robust to the presence of weak instruments and a small number of clusters. They reduce absolute size bias significantly and demonstrate that the wild bootstrap should join the standard toolkit in IV and cluster-dependent models.
Original languageEnglish
JournalJournal of Applied Econometrics
DOIs
Publication statusE-pub ahead of print - 6 Apr 2019

Fingerprint

hypothesis testing
microeconomics
trend
Wild bootstrap
Bootstrap method
Inference
Weak instruments
Standard error
Microeconomics
Toolkit
Hypothesis testing
Instrumental variables
Bootstrap
Join

Cite this

@article{288eee88933d4d08b974278593efcab8,
title = "Two Applications of wild bootstrap methods to Improve Inference in cluster-IV models",
abstract = "Microeconomic data often have within-cluster dependence, which affects standard error estimation and inference. When the number of clusters is small, asymptotic tests can be severely oversized. In the instrumental variables model, the potential presence of weak instruments further complicates hypothesis testing. We use wild bootstrap methods to improve inference in two empirical applications with these characteristics. Building from estimating equations and residual bootstraps, we identify variants robust to the presence of weak instruments and a small number of clusters. They reduce absolute size bias significantly and demonstrate that the wild bootstrap should join the standard toolkit in IV and cluster-dependent models.",
keywords = "instrumental variables, wild bootstrap, hypothesis testing, weak instruments, clustered errors",
author = "Magnusson, {Leandro M} and Keith Finlay",
year = "2019",
month = "4",
day = "6",
doi = "10.1002/jae.2710",
language = "English",
journal = "Journal of Applied Econometrics",
issn = "0883-7252",
publisher = "John Wiley & Sons",

}

TY - JOUR

T1 - Two Applications of wild bootstrap methods to Improve Inference in cluster-IV models

AU - Magnusson, Leandro M

AU - Finlay, Keith

PY - 2019/4/6

Y1 - 2019/4/6

N2 - Microeconomic data often have within-cluster dependence, which affects standard error estimation and inference. When the number of clusters is small, asymptotic tests can be severely oversized. In the instrumental variables model, the potential presence of weak instruments further complicates hypothesis testing. We use wild bootstrap methods to improve inference in two empirical applications with these characteristics. Building from estimating equations and residual bootstraps, we identify variants robust to the presence of weak instruments and a small number of clusters. They reduce absolute size bias significantly and demonstrate that the wild bootstrap should join the standard toolkit in IV and cluster-dependent models.

AB - Microeconomic data often have within-cluster dependence, which affects standard error estimation and inference. When the number of clusters is small, asymptotic tests can be severely oversized. In the instrumental variables model, the potential presence of weak instruments further complicates hypothesis testing. We use wild bootstrap methods to improve inference in two empirical applications with these characteristics. Building from estimating equations and residual bootstraps, we identify variants robust to the presence of weak instruments and a small number of clusters. They reduce absolute size bias significantly and demonstrate that the wild bootstrap should join the standard toolkit in IV and cluster-dependent models.

KW - instrumental variables

KW - wild bootstrap

KW - hypothesis testing

KW - weak instruments

KW - clustered errors

U2 - 10.1002/jae.2710

DO - 10.1002/jae.2710

M3 - Article

JO - Journal of Applied Econometrics

JF - Journal of Applied Econometrics

SN - 0883-7252

ER -