We propose an unobserved components model to evaluate unemployment dat
a for evidence on hysteresis effects. Unemployment is decomposed into
a natural rate component, assumed to be nonstationary, and a cyclical
component, assumed to be stationary. Hysteresis effects are modelled b
y allowing lagged cyclical unemployment to affect the current natural
rate. The model is estimated using Canadian, German, U.K., and U.S. un
employment rate series. We find substantive hysteresis effects in the
Canadian, German, and U.K. unemployment series. Hysteresis effects in
the U.S. unemployment series are negligible.