Byoungchan Lee, University of California - Berkeley
This paper studies dynamics of earnings inequality and its macroeconomic implications. I construct a quarterly measure of earnings inequality based on the Quarterly Census of Employment and Wages. Using the new measure, I investigate how earnings inequality responds to macroeconomic shocks. Although technology and fiscal policy shocks explain a significant fraction of the cyclical dynamics of earnings inequality, most of the short-run movements are not predictable. I define an inequality shock as the unanticipated component of the inequality index, which reduces real GDP and interest rate similar to a negative demand shock. It contributes to about 30% of the forecast error variance of real GDP at the peak. To understand the mechanism, I build two agent new Keynesian models. Using a simple model, I analytically illustrate why an inequality shock is related to a demand shock in a representative agent model. Then a medium-sized DSGE model is proposed for quantitative evaluation. Finally, I study a non-linear prediction of the model that higher inequality may lead to a more volatile business cycle. I explain why this is the case in the model and suggest empirical evidence supporting the view.