We study the revision of survey expectations in response to macroeconomic shocks, which we identify in vector autoregressive models with sign restrictions. We find that survey respondents distinguish between movements along the Phillips curve and shifts of the Phillips curve, depending on the type of shock that hits the economy. In addition, interest rate expectations are revised broadly in line with a Taylor rule. Consistent with models of rational inattention, macroeconomic shocks account for a small share of the forecast error variance of survey measures elicited from consumers, while they are more relevant for the expectations of professional forecasters.

No abstract

We study the effects of macroeconomic shocks on several measures of economic inequality obtained from U.S. survey data. To identify aggregate supply, aggregate demand, and monetary policy shocks, we estimate structural vector autoregressions and impose sign and zero restrictions on impulse response functions. Our results show that the effects of the macroeconomic shocks on economic inequality depend on the type of shock as well as on the measure of inequality considered. Contractionary monetary policy shocks increase expenditure and consumption inequality, whereas income and earnings inequality are less affected. Adverse aggregate supply and aggregate demand shocks increase income and earnings inequality, but reduce expenditure and consumption inequality. Our results suggest that different channels dominate in the transmission of the shocks. The earnings heterogeneity channel is consistent with the inequality dynamics in the aftermath of monetary policy shocks, but it appears to be less crucial when the economy is hit by either aggregate supply or aggregate demand shocks. In the aftermath of aggregate supply and aggregate demand shocks, inflation and the real interest rate appear to drive inequality dynamics to a larger degree. Using variance decompositions, we also find that although the macroeconomic shocks account for large shares of the variation in the macroeconomic variables, their contributions to the dynamics of the inequality measures are limited.