Who should work how much?

A production efficiency perspective naturally leads to the prescription that more productive individuals should work more than less productive individuals. Yet, systematic differences in actual hours worked across high- and low-wage individuals are barely noticeable. We highlight that the insurance available to households is an important determinant behind this fact. Using a dynamic heterogeneous-agent model with insurance frictions, income effects calibrated to match aggregate hours across time and space, and financial frictions that deliver realistic wealth dispersion, we report stark effects of insurance: perfect insurance would raise aggregate labor productivity by 9.6 percent and decrease hours worked by 7.7 percent.