Authors: Jan J. J. Groen, Richard Paap, and Francesco Ravazzolo

This paper revisits inflation forecasting using reduced-form Phillips curve forecasts, that is, inflation forecasts using activity and expectations variables. We propose a model that results from averaging across different regression specifications selected from a set of potential predictors that includes inflation lags, different real activity data, term structure data, nominal data, and surveys. In each of the individual specifications, we allow for occasional shocks of random magnitude in the regression parameters. As such, our framework simultaneously addresses structural change and model uncertainty that unavoidably affect Phillips curve forecasts. Over the full 1960-2008 sample the framework indicates several structural breaks across different combinations of activity measures, which often coincide with, among others, policy regime changes and oil price shocks. We find less evidence than elsewhere in the literature for variance breaks and inflation gap persistence. In a real-time out-of-sample forecasting exercise our model specification generally provides superior one-quarter- and one-year-ahead forecasts for quarterly inflation relative to a range of regularly used forecasting models.