Title: A time-varying threshold star model of unemployment and the natural rate
Authors: Michael Owyang - Federal Reserve Bank of St Louis (United States) [presenting]
Abstract: Smooth-transition autoregressive (STAR) models, competitors of Markov-switching models, are limited by an assumed time-invariant threshold level. However, a STAR framework could estimate a time-varying threshold level of unemployment. One can consider this threshold a ``tipping level'' where the mean and dynamics of the natural rate of unemployment shift. If the threshold level is time-varying, one can add an error-correction term---between the lagged levels of unemployment and the threshold --- to the autoregressive terms in the STAR model. Thus, the time-varying latent threshold level serves as both a demarcation between regimes and an error-correction term.