What would Nelson and Plosser find had they used panel unit root tests?

Abstract

In this study, we systemically apply nine recent panel unit root tests to the same 14 macroeconomic and financial series as those considered in the seminal paper by Nelson and Plosser (1982). The data covers OECD countries from 1950 to 2003. Our results clearly point out the difficulty that applied econometricians would face when they want to get a simple and clear-cut diagnosis with panel unit root tests. We confirm the fact that panel methods must be very carefully used for testing unit roots in macroeconomic or financial panels. More precisely, we find mitigated results under the cross-sectional independence assumption, since the unit root hypothesis is rejected for many macroeconomic variables. When international cross-correlations are taken into account, conclusions depend on the specification of these cross-sectional dependencies. Two groups of tests can be distinguished. The first group tests are based on a dynamic factor structure or an error component model. In this case, the nonstationarity of common factors (international business cycles or growth trends) is not rejected, but the results are less clear with respect to idiosyncratic components. The second group tests are based on more general specifications. Their results are globally more favourable to the unit root assumption.

Acknowledgement

I am grateful to Benoit Perron, Rafal Kierzenkowski and Regis Breton for helpful comments and suggestions. I am grateful to Faouzi Boujedra for research assistance. All the tests used have been programmed under Matlab. Codes and data are available on the website: http://www.univ-orleans.fr/deg/masters/ESA/CH/churlin_R.htm

Notes

¹It is for instance the case, when the cross-section dependences are specified as standard spatial error processes (Baltagi et al., 2005) and nonlinear IV tests (Chang, 2002) or bootstrap based tests (Chang, 2004) are used.

²Based on standard time-series unit root tests, the seminal paper by Nelson and Plosser pointed out that American macroeconomic series feature, quasi-systematically, stochastic tendencies and unit root properties.

³The second slight difference is that we consider GDP (and GDP per capita, real GDP) rather than GNP for data availability.

⁴The corresponding estimated numbers of factors are exactly the same except for employment and velocity. This slight difference is due to the fact that in Bai and Ng (2004) the information criteria are computed from demeaned first differences whereas in Moon and Perron the residuals are used.