Abstract
Using survey forecast data, this paper documents several stylized facts about forecasters’ beliefs on income and consumption and aggregate consumption growth: (1) survey-based income forecast at consensus level is highly correlated with consumption growth; (2) consensus income and consumption growth forecast errors under-react to macro news shocks and (3) consensus income forecast error and consumption growth under-react initially and overreact subsequently in response to main business cycle shocks. Motivated by this evidence, we propose a model of equilibrium consumption determination where agents learn the exogenous latent permanent income process and extrapolate the past income realizations. Our model can generate the behavior of consumption that the rational-expectation Permanent Income Hypothesis fails to predict excess smoothness and excess sensitivity of aggregate consumption, and negatively correlated consumption growth and past income change in the medium run. This study contributes to the literature on economic belief formation and empirical consumption by showing how survey-motivated evidence can jointly explain a range of important anomalies.
Publisher
AMH International Conferences and Seminars Organizing LLC