This paper studies the returns to seniority, the returns to experience,
and the effects of seniority and experience at the time of a quit or layoff on
changes in the job match specific component of wages. We show that these
returns are not identified in widely used regression models that relate the wage
changes of stayers, quits, and layoffs to tenure and experience. We deal with
the identification of problems in two ways. First, we obtain theoretical bounds
on key unidentified parameters using a simple model of wages and mobility.
Second, we check the implications of assumptions about the linear tenure slope
for the estimates of the returns to tenure, experience, and the effect of tenure
on job match gains. We have three main empirical findings. First, there is a
large return to general labor market experience that is independent of job
shopping. Second, the return to tenure is probably above the Altonji and
Shakotko's (1987) estimate but far below OLS estimates. Third, quits results
in substantial job match gains for inexperienced workers. Layoffs are
associated with the substantial job match losses for workers who have been on
the job for over a year.