Second-order corrected likelihood for nonlinear panel models with fixed effects

Geert Dhaene; Yutao Sun

doi:10.1016/j.jeconom.2020.04.001

Second-order corrected likelihood for nonlinear panel models with fixed effects

Geert Dhaene, Yutao Sun^*

^*Corresponding author for this work

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

7 Citations (Scopus)

55 Downloads (Pure)

Abstract

We propose a second-order correction for nonlinear fixed-effect panel models. The correction is made via the log-likelihood function. It removes the two leading terms of the bias of the log-likelihood that arises from estimating the fixed effects. Maximizing the corrected likelihood gives a second-order bias-corrected estimator, with bias OT⁻³, where T is the number of time periods. The corrected likelihood also gives second-order corrected test statistics. The correction applies to general nonlinear fixed-effect models with independent observations. The bias correction properties are confirmed in simulations for binary-choice models.

Original language	English
Pages (from-to)	227-252
Number of pages	26
Journal	Journal of Econometrics
Volume	220
Issue number	2
DOIs	https://doi.org/10.1016/j.jeconom.2020.04.001
Publication status	Published - Feb 2021

Bibliographical note

JEL classification: C23

Funding Information:
We thank Stéphane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11 ].

Funding Information:
We thank St?phane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11].

Publisher Copyright:
© 2020 Elsevier B.V.

Access to Document

10.1016/j.jeconom.2020.04.001

1_s2.0_S0304407620301196_mainFinal published version, 845 KBLicence: Article 25fa Dutch Copyright Act

Cite this

@article{5c076cd1205543bf8b86530b90817b4e,

title = "Second-order corrected likelihood for nonlinear panel models with fixed effects",

abstract = "We propose a second-order correction for nonlinear fixed-effect panel models. The correction is made via the log-likelihood function. It removes the two leading terms of the bias of the log-likelihood that arises from estimating the fixed effects. Maximizing the corrected likelihood gives a second-order bias-corrected estimator, with bias OT−3, where T is the number of time periods. The corrected likelihood also gives second-order corrected test statistics. The correction applies to general nonlinear fixed-effect models with independent observations. The bias correction properties are confirmed in simulations for binary-choice models.",

author = "Geert Dhaene and Yutao Sun",

note = "JEL classification: C23 Funding Information: We thank St{\'e}phane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11 ]. Funding Information: We thank St?phane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11]. Publisher Copyright: {\textcopyright} 2020 Elsevier B.V.",

year = "2021",

month = feb,

doi = "10.1016/j.jeconom.2020.04.001",

language = "English",

volume = "220",

pages = "227--252",

journal = "Journal of Econometrics",

issn = "0304-4076",

publisher = "Elsevier",

number = "2",

}

TY - JOUR

T1 - Second-order corrected likelihood for nonlinear panel models with fixed effects

AU - Dhaene, Geert

AU - Sun, Yutao

N1 - JEL classification: C23 Funding Information: We thank Stéphane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11 ]. Funding Information: We thank St?phane Bonhomme, Roger Moon, Cavit Pakel, participants of the 2017 Tinbergen Institute Conference, the 2017 International Panel Data Conference, and the 2018 International Association for Applied Econometrics Annual Conference, two referees, and the editors for helpful comments. Geert Dhaene acknowledges financial support from the Flemish Science Foundation [grant number G.0505.11]. Publisher Copyright: © 2020 Elsevier B.V.

PY - 2021/2

Y1 - 2021/2

N2 - We propose a second-order correction for nonlinear fixed-effect panel models. The correction is made via the log-likelihood function. It removes the two leading terms of the bias of the log-likelihood that arises from estimating the fixed effects. Maximizing the corrected likelihood gives a second-order bias-corrected estimator, with bias OT−3, where T is the number of time periods. The corrected likelihood also gives second-order corrected test statistics. The correction applies to general nonlinear fixed-effect models with independent observations. The bias correction properties are confirmed in simulations for binary-choice models.

AB - We propose a second-order correction for nonlinear fixed-effect panel models. The correction is made via the log-likelihood function. It removes the two leading terms of the bias of the log-likelihood that arises from estimating the fixed effects. Maximizing the corrected likelihood gives a second-order bias-corrected estimator, with bias OT−3, where T is the number of time periods. The corrected likelihood also gives second-order corrected test statistics. The correction applies to general nonlinear fixed-effect models with independent observations. The bias correction properties are confirmed in simulations for binary-choice models.

UR - http://www.scopus.com/inward/record.url?scp=85084233850&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2020.04.001

DO - 10.1016/j.jeconom.2020.04.001

M3 - Article

AN - SCOPUS:85084233850

SN - 0304-4076

VL - 220

SP - 227

EP - 252

JO - Journal of Econometrics

JF - Journal of Econometrics

IS - 2

ER -

Second-order corrected likelihood for nonlinear panel models with fixed effects

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this