摘要：Since the publication of Taylor's (2002) results supporting Purchasing Power Parity (PPP) theory using a century of data, several authors have tried to verify PPP using the same data set. While one study has rejected Taylor''s strong conclusion, others have supported it. In this paper we use yet a different unit root testing procedure to determine whether Taylor''s results are sensitive to switching the null hypothesis of stationarity with the alternative of non-stationarity. More precisely, we rely upon Kwiatkowski, et al. (1992) test and apply it to the real bilateral and real effective exchange rates of 20 countries in Taylor''s sample. The results provide support for PPP in 18 of the 20 countries that are much closer to Taylor''s findings than any other study.
