Abstract
This paper provides a new way of analyzing tenure profiles in wages, by modelling simultaneously the evolution of wages and the distribution of tenures. We develop a theoretical model based on efficient bargaining, where both log outside wage and log wage in the current job follow a random walk, as found empirically. This setting allows the application of real option theory. We derive the efficient separation rule. The model fits the observed distribution of job tenures well. Since we observe outside wages only at job start and job separation, our empirical analysis of within job wage growth is based on expected wage growth conditional on the outside wages at both dates. Our modelling allows testing of the efficient bargaining hypothesis. The model is estimated on the PSID.
Highlights
A large empirical literature has looked at wage returns to job seniority, using a whole arsenal of econometric techniques, see Farber (1999) for a survey
This paper addresses explicitly whether the existing evidence is consistent with efficient separations by modelling simultaneously the evolution of wages and the distribution of job tenures
We have proven the remarkable result that in this model the evolution of log wages in completed job spells does not provide any information whatsoever on wage-tenure profiles, since this evolution is independent of the drift in log wages
Summary
A large empirical literature has looked at wage returns to job seniority, using a whole arsenal of econometric techniques, see Farber (1999) for a survey. The evolution of an individual’s within-job log wage is reasonably described by a random walk with transitory shocks, as previously found by Abowd and Card (1989), Topel (1991) and Topel and Ward (1992), hypothesis that we verify on our PSID estimation sample Whereas this observation received little attention among labor economists, we take it as cornerstone of our modelling. From the distribution of job tenures we are able to estimate the surplus of the job’s productivity above its reservation value and a (linear) drift of this surplus, up to a normalizing constant (the variance of the random walk). The estimated tenure profile is on the high end of the spectrum, 5% per year, though more than five sixths of the return take the form of a declining outside productivity instead of a rising inside productivity. The paper is structured as follows: the model is discussed in Section 2, the empirical analysis in Section 3 and Section 4 concludes
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
