Estimation of and testing for random effects in dynamic panel data models

Jianhong Wu, Lixing ZHU*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)

Abstract

In this article, estimation of moments up to the fourth order of random effects and errors is first investigated for dynamic panel data models. Using the QR decomposition of a matrix, the moments of random individual effects and errors are estimated without affecting each other so that the estimation procedure is simple to implement, and the asymptotic behavior of estimation is derived. On the basis of these estimations, we construct a test for the existence of individual effects. This test is asymptotically normally distributed under the null hypothesis without any distributional assumptions on the individual effects and errors other than moments. A power study shows that our test is able to detect local alternatives that are distinct from the null at a parametric rate. Monte Carlo simulations are carried out for illustration.

Original languageEnglish
Pages (from-to)477-497
Number of pages21
JournalTest
Volume21
Issue number3
DOIs
Publication statusPublished - Sep 2012

Scopus Subject Areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

User-Defined Keywords

  • Dynamic panel data
  • Estimation of moments
  • Random effects
  • Testing for random effects

Fingerprint

Dive into the research topics of 'Estimation of and testing for random effects in dynamic panel data models'. Together they form a unique fingerprint.

Cite this