Infrequent Large Shocks to Unemployment: New Evidence on Alternative Persistence Perspectives

This paper tests whether the observed high persistence of unemployment rates in most OECD countries is due to (full) hysteresis against the alternative that it is caused by adjustment towards an increased natural rate. The analysis relies on standard univariate unit root tests. Usually such tests cannot reject the presence of a unit root in the unemployment rate, pointing to full hysteresis. This paper shows that the unit root hypothesis can clearly be rejected once infrequent level-shifts are allowed for.