This paper explores the importance of shocks to consumer misperceptions, or "noise shocks", in a quantitative business cycle model. I embed imperfect information as in Lorenzoni (2009) into a new Keynesian model with price and wage rigidities. Agents learn about the components of labor productivity by only observing aggregate productivity and a noisy signal. Noise shocks lead to expectational errors about the true fundamentals triggering aggregate fluctuations. Estimating the model with Bayesian methods on US data shows that noise shocks contribute to 20 percent of consumption fluctuations at short horizons. Wage rigidity is pivotal for the importance of noise shocks.
Imperfect Information Noise Shocks Aggregate Fluctuations Bayesian Estimation