Abstract
Applying the new panel unit root test developed in this article, we can overcome the pitfalls of old-fashioned panel unit root tests making it possible for researchers testing individual series for a unit root while taking contemporaneous cross- sectional dependence into account. The proposed test is indeed more powerful than univariate Augmented Dickey–Fuller (ADF) test in rejecting false I(1) time series. The long-run purchasing power parity (PPP) hypothesis on four OECD countries was tested between year 1950 and 1995. Evidence in favor of long-run PPP was absent when using single ADF and traditional panel data unit root test, however, when using the new test developed in this article we find strong evidence in favour of long-run purchasing power parity for three out of four OECD countries. The finite sample performance of the new test is examined though Monte Carlo Simulation, and was superior as compared to that of single ADF unit root test.
Original language | English |
---|---|
Pages (from-to) | 75-80 |
Number of pages | 6 |
Journal | Applied Economics Letters |
Volume | 16 |
Issue number | 1 |
DOIs | |
Publication status | Published - 2009 |
Externally published | Yes |