In this paper we suggest panel data unit root tests which allow for a structural breaks in the individual effects or linear trends of panel data models. This is done under the assumption that the disturbance terms of the panel are heterogenous and serially correlated. The limiting distributions of the suggested test statistics are derived under the assumption that the time-dimension of the panel (T) is fixed, while the cross-section (N) grows large. Thus, they are appropriate for short panels, where T is small. The tests consider the cases of a known and unknown date break. For the latter case, the paper gives the analytic form of the distribution of the test statistics. Monte Carlo evidence suggest that our tests have size which is very close to its nominal level and satisfactory power in small-T panels. This is true even for cases where the degree of serial correlation is large and negative, where single time series unit root tests are found to be critically oversized.
Download the paper in PDF format
Yiannis Karavias and Elias Tzavalis
View all Granger Centre discussion papers | View all School of Economics featured discussion papers
School of EconomicsUniversity of Nottingham
Nottingham, NG7 2RD
Connect with the University of Nottingham through social media and our blogs.
Campus maps | More contact information | Jobs