Inference for Unit Roots in Dynamic Panels

Paper number: 96/04

Paper date: February 1996

Year: 1996

Paper Category: Discussion Paper

Authors

Richard Harris
University of Exeter

Elias Tzavalis
University of Exeter

Abstract

This paper proposes similar unit root testing procedures for both homogenous and hetrogeneous dynamic panel data models, based on Least Squares estimates and assuming that the time dimension of the panel is fixed. It is shown that the limiting distributi9on of the tests is standard normal. Similarity with regard to the initial conditions is achieved by including fixed effect dummy variables in the autoregression, while similarity with respect to the fixed effects is achieved by including a linear trend for each individual unit of the panel. When fixed effects or individual trends are included as regressors the Least Squares estimate of the autoregressive parameter is inconsistent and thus the test statistics must be corrected for the asymptotic bias. Simulation evidence suggests that the proposed tests have empirical size that is very close to the nominal 5% level and considerably more power than other panel unit root tests and the corresponding unit root tests for the single time series case.

JEL Classification Nos: C22, C23, F43
Keywords: Panel data, Unit roots, Fixed effects, Central limit theorem

Corresponding Authors: Richard Harris and Elias Tzavalis, Department of Economics, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ, Great Britain, tel:(44) 1392 263215, fax: (44) 1392 263242, email: R.D.F.Harris@exeter.ac.uk and E.Tzavalis@exeter.ac.uk

 


The authors thank K. Aabdir, K. Hadri, C. Hsiao, M. Magdalinos, G.D.A. Phillips, Z. Psaradakis and R. Smith for helpful discussions and comments.