Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Inference on time-invariant variables using panel data: a pretest estimator

Abstract : For panel data models including time-invariant variables, this paper proposes a new Hausman pretest estimator of the internal instruments of Hausman-Taylor estimator. It assumes Mundlak and Krishnakumar linear specification for the endogeneity of random individual effects. Furthermore, the paper evaluates the biases of currently used estimators: repeated between, ordinary least squares, two-stage restricted between, Oaxaca-Geisler estimator, fixed effect vector decomposition, and generalized least squares. Some of these may lead to erroneous conclusions regarding the statistical significance of the estimated parameter values of time-invariant variables, especially when time-invariant variables are correlated with the individual effects.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas
Contributor : Caroline Bauer <>
Submitted on : Sunday, December 13, 2020 - 12:50:18 PM
Last modification on : Tuesday, January 19, 2021 - 11:09:09 AM

Links full text




Jean-Bernard Chatelain, Kirsten Ralf. Inference on time-invariant variables using panel data: a pretest estimator. In press. ⟨halshs-03059883⟩



Record views