Dynamic panel GMM using R

Peter C.B. Phillips, Chirok Han

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

GMM methods for estimating dynamic panel regression models are heavily used in applied work in many areas of economics and more widely in the social and business sciences. Software packages in STATA and GAUSS are commonly used in these applications. We provide a new R program for difference GMM, system GMM, and within-group estimation for simulation with the model we consider that is based on a standard first-order dynamic panel regression with individual- and time-specific effects. The program lacks the generality of a full package but provides a foundation for further development and is optimized for speed, making it particularly useful for large panels and simulation purposes. The program is illustrated in simulations that include both stationary and nonstationary cases. Particular attention in the simulations is given to analyzing the impact of fixed effect heterogeneity on bias in system GMM estimation compared with the other methods.

Original languageEnglish
Title of host publicationHandbook of Statistics
PublisherElsevier B.V.
DOIs
Publication statusPublished - 2019 Jan 1

Publication series

NameHandbook of Statistics
ISSN (Print)0169-7161

Fingerprint

Software packages
Simulation
Economics
Fixed Effects
Software Package
Industry
Regression Model
Regression
First-order
Model
Business
Standards

Keywords

  • Bias
  • Difference GMM
  • Dynamic panel
  • System GMM
  • Within-group estimation

ASJC Scopus subject areas

  • Statistics and Probability
  • Modelling and Simulation
  • Applied Mathematics

Cite this

Phillips, P. C. B., & Han, C. (2019). Dynamic panel GMM using R. In Handbook of Statistics (Handbook of Statistics). Elsevier B.V.. https://doi.org/10.1016/bs.host.2019.01.002

Dynamic panel GMM using R. / Phillips, Peter C.B.; Han, Chirok.

Handbook of Statistics. Elsevier B.V., 2019. (Handbook of Statistics).

Research output: Chapter in Book/Report/Conference proceedingChapter

Phillips, PCB & Han, C 2019, Dynamic panel GMM using R. in Handbook of Statistics. Handbook of Statistics, Elsevier B.V. https://doi.org/10.1016/bs.host.2019.01.002
Phillips PCB, Han C. Dynamic panel GMM using R. In Handbook of Statistics. Elsevier B.V. 2019. (Handbook of Statistics). https://doi.org/10.1016/bs.host.2019.01.002
Phillips, Peter C.B. ; Han, Chirok. / Dynamic panel GMM using R. Handbook of Statistics. Elsevier B.V., 2019. (Handbook of Statistics).
@inbook{0d2b66ca951d44cb9b91d1e87baf43b3,
title = "Dynamic panel GMM using R",
abstract = "GMM methods for estimating dynamic panel regression models are heavily used in applied work in many areas of economics and more widely in the social and business sciences. Software packages in STATA and GAUSS are commonly used in these applications. We provide a new R program for difference GMM, system GMM, and within-group estimation for simulation with the model we consider that is based on a standard first-order dynamic panel regression with individual- and time-specific effects. The program lacks the generality of a full package but provides a foundation for further development and is optimized for speed, making it particularly useful for large panels and simulation purposes. The program is illustrated in simulations that include both stationary and nonstationary cases. Particular attention in the simulations is given to analyzing the impact of fixed effect heterogeneity on bias in system GMM estimation compared with the other methods.",
keywords = "Bias, Difference GMM, Dynamic panel, System GMM, Within-group estimation",
author = "Phillips, {Peter C.B.} and Chirok Han",
year = "2019",
month = "1",
day = "1",
doi = "10.1016/bs.host.2019.01.002",
language = "English",
series = "Handbook of Statistics",
publisher = "Elsevier B.V.",
booktitle = "Handbook of Statistics",

}

TY - CHAP

T1 - Dynamic panel GMM using R

AU - Phillips, Peter C.B.

AU - Han, Chirok

PY - 2019/1/1

Y1 - 2019/1/1

N2 - GMM methods for estimating dynamic panel regression models are heavily used in applied work in many areas of economics and more widely in the social and business sciences. Software packages in STATA and GAUSS are commonly used in these applications. We provide a new R program for difference GMM, system GMM, and within-group estimation for simulation with the model we consider that is based on a standard first-order dynamic panel regression with individual- and time-specific effects. The program lacks the generality of a full package but provides a foundation for further development and is optimized for speed, making it particularly useful for large panels and simulation purposes. The program is illustrated in simulations that include both stationary and nonstationary cases. Particular attention in the simulations is given to analyzing the impact of fixed effect heterogeneity on bias in system GMM estimation compared with the other methods.

AB - GMM methods for estimating dynamic panel regression models are heavily used in applied work in many areas of economics and more widely in the social and business sciences. Software packages in STATA and GAUSS are commonly used in these applications. We provide a new R program for difference GMM, system GMM, and within-group estimation for simulation with the model we consider that is based on a standard first-order dynamic panel regression with individual- and time-specific effects. The program lacks the generality of a full package but provides a foundation for further development and is optimized for speed, making it particularly useful for large panels and simulation purposes. The program is illustrated in simulations that include both stationary and nonstationary cases. Particular attention in the simulations is given to analyzing the impact of fixed effect heterogeneity on bias in system GMM estimation compared with the other methods.

KW - Bias

KW - Difference GMM

KW - Dynamic panel

KW - System GMM

KW - Within-group estimation

UR - http://www.scopus.com/inward/record.url?scp=85061200330&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061200330&partnerID=8YFLogxK

U2 - 10.1016/bs.host.2019.01.002

DO - 10.1016/bs.host.2019.01.002

M3 - Chapter

AN - SCOPUS:85061200330

T3 - Handbook of Statistics

BT - Handbook of Statistics

PB - Elsevier B.V.

ER -