We propose a nonlinear framework that exploits both macro and micro data to study the dynamic transmission of aggregate and idiosyncratic shocks to household income. We develop identification arguments and estimation methods that combine macro time series and time series of panels. Using U.S....