We examine dynamic asymmetries in U.S. unemployment using nonlinear time se
ries models and Bayesian methods. We find strong statistical evidence in fa
vor of a two-regime threshold autoregressive model. Empirical results indic
ate that, once we take into account both parameter and model uncertainty, t
here are economically interesting asymmetries in the unemployment rate. One
finding of particular interest is that shocks that lower the unemployment
rate tend to have a smaller effect than shocks that raise the unemployment
rate. This finding is consistent with unemployment rises being sudden and f
alls gradual.