This paper proposes a test statistic for the null hypothesis of panel stationarity that allows for the presence of multiple structural breaks. Two different specications are considered depending on the structural breaks affecting the individual effects and/or the time trend. The model is exible enough to allow the number of breaks and their position to differ across individuals. The test is shown to have an exact limit distribution with a good nite sample performance. Its application to a typical panel data set of real per capita GDP gives support to the trend stationarity of these series.Keywords: multiple structural changes, panel data, stationarity test, GDP per capita JEL codes: C12, C22 ResumAquest article proposa un estadístic de prova per contrastar la hipòtesi nul¢la d'estacionarietat en panell permetent la presència de múltiples canvis estructurals. Es consideren dues especicacions diferents en funció de si els canvis estructurals afecten els efectes individuals i/o la tendència temporal. El model és el sucientment exible com per permetre que tant el nombre de canvis com la seva posició puguin diferir entre els individus. El treball mostra que la distribució asimptòtica de l'estadístic és exacta. Experiments de simulació indiquen que el comportament del contrast en mides mostrals nites és bo. La seva aplicació a un panell típic de PIB per capita real proporciona evidència a favor de l'estacionarietat de les sèries.

It is well known that (seasonal) unit root tests can be seriously affected by the presence of weak dependence in the driving shocks when this is not accounted for. In the non-seasonal case both parametric (based around augmentation of the test regression with lagged dependent variables) and semi-parametric (based around an estimator of the long run variance of the shocks) unit root tests have been proposed. Of these, the M class of unit root tests introduced by Stock (1999), Perron and Ng (1996) and Ng and Perron (2001), appear to be particularly successful, showing good finite sample size control even in the most problematic (near-cancellation) case where the shocks contain a strong negative moving average component. The aim of this paper is threefold. First we show the implications that neglected weak dependence in the shocks has on lag un-augmented versions of the well known regression-based seasonal unit root tests of Hylleberg et al. (1990). Second, in order to complement extant parametrically augmented versions of the tests of Hylleberg et al. (1990), we develop semi-parametric seasonal unit root test procedures, generalising the methods developed in the non-seasonal case to our setting. Third, we compare the finite sample size and power properties of the parametric and semi-parametric seasonal unit root tests considered. Our results suggest that the superior size/power trade-off offered by the M approach in the non-seasonal case carries over to the seasonal case.

No abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Abstract: This paper proposes a set of tools for analysing the regional distribution of unemployment. As we were interested in the characteristics of the distribution as a whole, results from a traditional regression analysis were complemented with those obtained by estimating its external shape before and after being conditioned to factors underlying regional unemployment. In addition, the paper specifically considers the spatial characteristics of the distribution, and the empirical model developed in order to determine explanatory factors includes spatial effects. This framework is applied to the study of the provincial distribution of unemployment rates in Spain. Results point to increasing spatial dependence in the distribution of regional unemployment rates, and a change in the factors causing regional differentials over the last decade. Terms of use: Documents in EconStor may