Specific Human Capital Accumulation and Job Match Quality - Implications for Measuring Returns to Tenure

This paper uses the theoretical argument presented by Stevens (2003) that suggests that the measured returns to tenure will unambiguously be biased downwards. We illustrate this effect for data from a UK internal labour market using the counterfactual methodology outlined by DiNardo, Fortin and Lemieux (1996). Finally we argue that Stevens's theory offers a possible explanation for the apparent puzzle presented by Medoff and Abraham (1980) who find that their estimated coefficient on tenure did not fall when direct measures of productivity were introduced into the wage equation