• Local Labor Markets: Job Ladder, Wage Growth and Unemployment

Abstract: I document that within location job-to-job wage growth is on average 40% greater in high unemployment MSAs than low unemployment MSAs in the U.S.. This heterogeneity contrasts with the nearly flat job-to-job rate profile against local unemployment rate. Along the job ladder, the more job-to-job switches a worker has made post-unemployment, the less wage growth they yield in their job switch. Interpreting these facts jointly using a job ladder model with location specific job type creation choice, I highlight that a larger job-to-job wage growth signifies greater mismatch in the original job. The calibrated model offers a novel channel to speak to the empirical finding that productivity in high unemployment area drops more than low unemployment area during the 2007-09 recession. The same decrease in aggregate matching efficiency drags down productivity in high unemployment area more, since the lost job-to-job switches there have larger productivity gains.

• Job Recalls and Worker Flows over the Life Cycle  (with Xincheng Qiu)

Abstract: The share of jobless spells that end with recalls, i.e., returning to the previous job, as opposed to finding a new job, is strongly increasing in age. The fact is robust to an extensive set of controls and various alternative sample selections, and is confirmed in administrative data. Recalls lead to different wage outcomes than exit to new jobs or job to job transitions. Throughout the life cycle, wages are barely changed after recalls, and the distribution of wage changes is very concentrated. We find that a job-ladder search model with recall options can successfully account for all documented facts. The introduction of recall options provides a novel mechanism that reconciles the puzzle of a positive comovement between separation and job finding rate over the life cycle. The model highlights that the deterioration of matching efficiency in bad times hurts young workers more than old workers, because the young rely more on the matching function of the labor market to find a new job, whereas the old rely less and get more recalls.