Forecasting the US unemployment rate

The primary interest of this paper is in out-of-sample forecasting for the U.S. monthly unemployment rate. Several linear unobserved components models are fitted and their comparative forecasting accuracy is assessed by means of and extensive rolling-origin procedure using a test period that covers the last two decades. An attempt is made to link forecasting performance to the time domain properties of the models and the evidence is that highly persistent models perform better. Deletion diagnostics and normality tests, along with documenting possible departures from linearity and Gaussianity attributable to business cycle and turning point asymmetries, foster the conclusion that these are mostly concentrated in the fit period (1948-1980). It is also argued that seasonal adjustment is not neutral with respect to these findings. A search is made for plausible non linear extensions capable of accounting for dynamic asymmetries in unemployment rates, leading to the specification of a cyclical trend model with smooth transition in the underlying parameters that improves forecast accuracy at short lead times and at the end of the sample period. Though significant, the gains are not exceptionally large, confirming our expectations. The generalised impulse response function casts some light on the interpretation of the results.