More Uncertainty about the Unit Root in U.S. Real GNP

Abstract

The new KPSS test is used to test the null hypothesis that U.S. real GNP is a trend stationary process. When appropriately sized, the test fails to reject the trend stationary null. This provides an important counter-example to the generic inability to reject the difference stationary null hypothesis for output by classical hypothesis testing. The evidence in favor of the trend stationary representation is weakened, however, by showing that such size correction dramatically reduces power. Reversing the null and alternative hypotheses, this paper complements Rudebusch’s (1993) recent analysis.