Heterogeneous Information and the Labor Market ∗

I introduce dispersed information in a search and matching model of the labor market, where firms are hit by aggregate and idiosyncratic productivity shocks. The latter induce larger responses in recruiting activity than the former because aggregate shocks are relatively less persistent and have general equilibrium effects which partially offset the change in fundamentals. Informational frictions prevent firms from disentangling aggregate and idiosyncratic shocks. With Bayesian updating, firms attribute aggregate shocks largely to idiosyncratic factors, because the latter are significantly larger. Since the optimal response to an idiosyncratic shock is to change hiring more aggressively, this misattribution translates into an increased responsiveness of employment to aggregate shocks, relative to an economy with full information. I show that in a calibrated model, this channel has quantitatively significant effects and offers a potential solution to a well-known puzzle the fact that standard search and matching models fail to generate sufficient volatility in labor market variables. In particular, the model with dispersed information brings the relative volatilities of employment and market tightness very close to those observed in the data. It also generates a negative comovement of the labor wedge with aggregate employment.