Linear Regression for Panel With Unknown Number of Factors as Interactive Fixed Effects

Hyungsik Roger Moon, Martin Weidner

Research output: Contribution to journalComment/debate

53 Citations (Scopus)

Abstract

In this paper, we study the least squares (LS) estimator in a linear panel regression model with unknown number of factors appearing as interactive fixed effects. Assuming that the number of factors used in estimation is larger than the true number of factors in the data, we establish the limiting distribution of the LS estimator for the regression coefficients as the number of time periods and the number of cross-sectional units jointly go to infinity. The main result of the paper is that under certain assumptions, the limiting distribution of the LS estimator is independent of the number of factors used in the estimation as long as this number is not underestimated. The important practical implication of this result is that for inference on the regression coefficients, one does not necessarily need to estimate the number of interactive fixed effects consistently.

Original languageEnglish
Pages (from-to)1543-1579
Number of pages37
JournalEconometrica
Volume83
Issue number4
DOIs
Publication statusPublished - 2015 Jul 1

Fingerprint

Factors
Fixed effects
Linear regression
Least squares estimator
Limiting distribution
Coefficients
Regression model
Panel regression
Inference

All Science Journal Classification (ASJC) codes

  • Economics and Econometrics

Cite this

@article{d7a08e2effe8466c9e3bf214e6a3bf9e,
title = "Linear Regression for Panel With Unknown Number of Factors as Interactive Fixed Effects",
abstract = "In this paper, we study the least squares (LS) estimator in a linear panel regression model with unknown number of factors appearing as interactive fixed effects. Assuming that the number of factors used in estimation is larger than the true number of factors in the data, we establish the limiting distribution of the LS estimator for the regression coefficients as the number of time periods and the number of cross-sectional units jointly go to infinity. The main result of the paper is that under certain assumptions, the limiting distribution of the LS estimator is independent of the number of factors used in the estimation as long as this number is not underestimated. The important practical implication of this result is that for inference on the regression coefficients, one does not necessarily need to estimate the number of interactive fixed effects consistently.",
author = "Moon, {Hyungsik Roger} and Martin Weidner",
year = "2015",
month = "7",
day = "1",
doi = "10.3982/ECTA9382",
language = "English",
volume = "83",
pages = "1543--1579",
journal = "Econometrica",
issn = "0012-9682",
publisher = "Wiley-Blackwell",
number = "4",

}

Linear Regression for Panel With Unknown Number of Factors as Interactive Fixed Effects. / Moon, Hyungsik Roger; Weidner, Martin.

In: Econometrica, Vol. 83, No. 4, 01.07.2015, p. 1543-1579.

Research output: Contribution to journalComment/debate

TY - JOUR

T1 - Linear Regression for Panel With Unknown Number of Factors as Interactive Fixed Effects

AU - Moon, Hyungsik Roger

AU - Weidner, Martin

PY - 2015/7/1

Y1 - 2015/7/1

N2 - In this paper, we study the least squares (LS) estimator in a linear panel regression model with unknown number of factors appearing as interactive fixed effects. Assuming that the number of factors used in estimation is larger than the true number of factors in the data, we establish the limiting distribution of the LS estimator for the regression coefficients as the number of time periods and the number of cross-sectional units jointly go to infinity. The main result of the paper is that under certain assumptions, the limiting distribution of the LS estimator is independent of the number of factors used in the estimation as long as this number is not underestimated. The important practical implication of this result is that for inference on the regression coefficients, one does not necessarily need to estimate the number of interactive fixed effects consistently.

AB - In this paper, we study the least squares (LS) estimator in a linear panel regression model with unknown number of factors appearing as interactive fixed effects. Assuming that the number of factors used in estimation is larger than the true number of factors in the data, we establish the limiting distribution of the LS estimator for the regression coefficients as the number of time periods and the number of cross-sectional units jointly go to infinity. The main result of the paper is that under certain assumptions, the limiting distribution of the LS estimator is independent of the number of factors used in the estimation as long as this number is not underestimated. The important practical implication of this result is that for inference on the regression coefficients, one does not necessarily need to estimate the number of interactive fixed effects consistently.

UR - http://www.scopus.com/inward/record.url?scp=84938149024&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938149024&partnerID=8YFLogxK

U2 - 10.3982/ECTA9382

DO - 10.3982/ECTA9382

M3 - Comment/debate

AN - SCOPUS:84938149024

VL - 83

SP - 1543

EP - 1579

JO - Econometrica

JF - Econometrica

SN - 0012-9682

IS - 4

ER -