Changing Income Risk across the US Skill Distribution: Evidence from a Generalized Kalman Filter—Braxton 2021

Whose profits risk has changed? To answer these problems, we develop a filtering mechanism that recovers persistent and transitory earnings for every individual over time. Our estimation allows first and second moments of shocks to depend on observables and zero incomes (i.e., unemployment) and readily integrates into theoretical models. We filter 23.5m SSA-CPS records. First, we show that our earnings-based filter catches SSA-CPS shocks such job switching and layoffs. Despite a drop in overall earnings risk since the 1980s, persistent risk has risen for employed and jobless employees while transitory risk has declined. Persistent earnings losses from full-year unemployment are up 50%. Using geography, education, and occupation data from SSA-CPS records, we disprove ideas about routine and low-skill workers' employment chances and geographic theories about the Rust Belt's collapse. Rising persistent earnings risk is concentrated among high-skill workers and linked to technology adoption. Rising persistent earnings risk while working (jobless) leads to 1.8% (0.7%) welfare loss, and bigger persistent earnings losses while unemployed lead to 3.3% welfare loss.

Braxton, J. C., Herkenhoff, K. F., Rothbaum, J. L., & Schmidt, L. (2021). Changing income risk across the US skill distribution: Evidence from a generalized Kalman filter (No. w29567). National Bureau of Economic Research.