Returns to on-the-job search and wage dispersion

A wide class of models with On-the-Job Search (OJS) predict that workers gradually select into better jobs. We develop a simple method based on the expected number of job offers received that can be used to measure match quality, identify the wage offer distribution and estimate the contribution of OJS to wage dispersion and the increase in wages over the lifecycle. The method uses two sources of identification: (i) time variation in job -finding rates and (ii) individual variation in the time since the last layoff. Applying this method to the NLSY 79, we find that the standard deviation of the wage-offer distribution is 13% and that OJS accounts for 8% of the total wage dispersion and 30% of the wage-increase over the lifecycle.