Job Loss, Consumption and Unemployment Insurance∗

It is well known that job loss is associated with both pre- and post-job loss declines in hourly wages and earnings. Using recent data from the Panel Study of Income Dynamics, I show that consumption dynamics mirror these wage dynamics. To account for the consumption dynamics in the data I introduce a correlation between individual hourly wages and job loss into a life-cycle model with self insurance (through savings), social insurance, and endogenous unemployment durations. I find that this model is able to replicate the joint dynamics of wages, job loss and consumption that we observe in the data. I then show that accounting for the correlation between wages and job loss has important implications for the optimal design of unemployment insurance (UI). The consumption smoothing benefits of unemployment insurance are larger, and the cost of insurance lower, than suggested when this correlation is absent. Thus, while a model that assumes away these correlations yields optimal UI replacement rates close to zero, a model that incorporates the correlations predicts optimal rates of 0.54, slightly higher than the current US level. ∗This paper is based on the first chapter of my dissertation. I am deeply indebted to my advisers Nick Bloom and Luig