The role of media for inflation forecast disagreement of households and professional forecasters

This paper investigates the effects of media coverage and macroeconomic conditions on inflation forecast disagreement of German households and professional forecasters. We adopt a Bayesian learning model in which media coverage of inflation affects forecast disagreement by influencing information sets as well as predictor choice. Our empirical results show that disagreement of households depends on the content of news stories (tone) but is unaffected by reporting intensity (volume) and by the heterogeneity of story content (information entropy). Disagreement of professionals does not depend on media coverage. With respect to the influence of macroeconomic variables we provide evidence that disagreement of households and professionals primarily depends on the current rate of inflation. -- forecast disagreement ; inflation expectations ; media coverage ; Bayesian learning