Sources of Inequality in Earnings Growth Over the Life Cycle

Individuals that rank at the top of the lifetime earnings (LE) distribution experience almost an 8-fold increase in their annual earnings between age 30 and 55, whereas median earners only see a 50% increase. If all workers have had experienced the same earnings growth, the difference in LE between the top and the 10th percentiles of LE distribution would be 85% smaller. What explains the vast heterogeneity in lifetime earnings growth? We study both empirically and theoretically the career paths across the LE distribution. Using administrative data, we document large dispersion in job switching patterns, inci- dence of unemployment, and wage growth for stayers and switchers across the LE distri- bution. To interpret these facts, we estimate a job-ladder model featuring heterogeneity in unemployment risk, job finding rate and contact rate for employed workers, as well as returns to experience. The estimated model matches a rich set of facts including the dispersion in the career paths over the LE distribution, as well as the distribution of annual earnings changes. We use the estimated model to decompose lifetime earnings growth differences into i) ex-ante heterogeneity in unemployment risk, and offer arrival rate both on and off the job, ii) returns to experience, and iii) ex-post idiosyncratic risk.