Survey Data and Subjective Beliefs in Business Cycle Models.

This article develops a theory of subjective beliefs that systematically deviate from rational expectations by incorporating time-varying pessimism and optimism, as evidenced by household survey data on inflation and unemployment forecasts. Using a calibrated structural business cycle model with nominal rigidities and a frictional labor market, the authors demonstrate that these belief biases have quantitatively significant effects on macroeconomic aggregates, particularly amplifying unemployment volatility and generating upward biases in inflation and unemployment forecasts during recessions. The model links subjective belief distortions to agents overweighting adverse future states associated with low continuation utilities, and it is disciplined by extensive empirical survey evidence, including cross-sectional and time-series patterns. The framework also introduces a novel solution technique for dynamic stochastic general equilibrium models with subjective beliefs and highlights the importance of firms' beliefs in matching inflation forecast biases. Empirical validations using local projections and forecast error regressions support the model's predictions, and the study connects belief fluctuations to measures of consumer sentiment and idiosyncratic risk, suggesting that subjective beliefs play a crucial role in business cycle dynamics and macroeconomic fluctuations.