Forecaster inattention : measurement, determinants, and policy implications

Using individual forecasts from the U.S. Survey of Professional Forecasters during 1968-2014 and Consensus Economics for G7 countries during 1990-2014, we provide direct econometric estimates of time varying inattention, defined as the common component in forecasters’ inattentiveness when predicting many economic variables. Based on this measure, we find that professional forecasters update their information sets every four months on average, and they are less inattentive in periods of recession and high economic uncertainty. Through the time varying structural vector autoregression model, we show that the same sized monetary shock has more persistent real effects when the degree of inattention is high. Our findings contribute to the literature on the transmission of monetary policy shocks and suggest inattention as an additional explanation why policy might become less effective during recessions.