Do Wages Rise With Job Seniority?

Abstract

Many previous studies have found a strong positive effect of job seniority (tenure) on wages. This paper re-examines the evidence using a simple instrumental variables scheme to deal with the fact that tenure is likely to be related to unobserved individual and job characteristics that affect the wage. We use the variation of tenure over a given job match as the principal instrumental variable for tenure. The variation in tenure over the job is uncorrelated by construction with the fixed individual and job match specific components of the error term of the wage equation. Our main finding is that the partial effect of tenure on wages is small, and that general labour market experience and job shopping account for most wage growth over a career. The strong cross section relationship between tenure and wages is due primarily to heterogeneity bias. Prior researchers (see Mincer and Jovanovic (1981), Bartel and Borjas (1981), Borjas (1981), Cline (1979) and Mellow (1981) and many others) have established that job seniority has a strong positive relationship with wage rates in a cross section or a cross section-time series of individuals. In this paper, we re-evaluate the evidence concerning the partial effects on wages of job seniority (tenure) and total labour market experience. The extent to which wages rise with tenure is important for several reasons. First and foremost, the wage-tenure profile is a fundamental question about the structure of earnings over careers. Second, the wage-tenure profile is a key determinant of the extent to which the earnings power of individuals is tied to specific jobs, and it is important for assessment of the losses suffered by displaced workers (Hamermesh (1984)). Third, evidence that wages rise with job tenure has been used to explain the decline in quits with tenure, since the wage growth on the current job lowers the probability that the worker will locate a superior alternative. A growing theoretical literature (surveyed in Parsons (1984)) has taken the evidence of a strong wage-tenure profile at face value and sought to provide an explanation for the relationship. The most prominent explanation is the theory of specific human capital (Oi (1962), Becker (1962), Mincer (1974), Parsons (1972), Kuratani (1973), and Hashimoto (1980)), in which the growth of wages with tenure (holding experience constant) is attributed to the worker's share of investments in firm-specific skills. More recently, Lazear (1981) has presented a supervision model of wage growth in which firms defer compensation as a means of inducing workers not to shirk. Freeman (1977) and Harris and Holmstrom (1982) provide an explanation for wage growth based on an insurance motive. Salop and Salop (1976), Nickell (1976), and Guasch and Weiss (1982) present adverse selection models of wage growth in which firms use a wage-tenure profile as a