Belief Formation In Macroeconomics

I use forecasts of macroeconomic variables from the monthly Wall Street Journal Economic survey to explore the belief formation process across individuals. I find that respondent’s forecasts tend to underreact to news, in contrast to previous findings. In addition, by tracking individuals over the sample I find that the degree of under or overreaction depends on the experience of the forecaster. Furthermore, I infer forecaster’s perception of the Federal Reserve reaction function from their joint revisions to macroeconomic variables. Again I find heterogeneity across individuals in how they perceive the Federal Reserve’s reaction function depending on what type of firm they are employed by. Together these findings highlight the importance of panel composition when testing models of belief formation