Unit Root Tests and the Burden of Proof

Simulation evidence is presented on the finite sample properties of two tests for stationarity recently proposed by Kwiatkowski, Phillips and Schmidt (1991) and Park (1990). Unlike earlier unit-root tests, these test the null of stationarity against the alternative of a unit root, thereby reversing the usual burden of proof. We also examine the consequenes of using the Kwiatkowski, Phillips and Schmidt test in conjuction with a standard unit-root test. These results suggest that the frequency of incorrect conclusions may be decreased relative to the application of only standard unit-root tests. Also, such a joint testing procedure may in some cases permit researchers to be more confident about their tests' results.