How biased are the estimated wage impacts of overeducation? A propensity score matching approach

Abstract

This article uses Propensity Score Matching (PSM) techniques to assess the extent to which the costs of overeducation are likely to have been over-estimated as a result of unobserved skill heterogeneity in studies adopting the standard ordinary least squares (OLS) wage equation framework. It was found that the PSM estimates were very much in line with those generated by OLS, suggesting that the overeducation phenomenon is likely to be imposing real and significant wage and productivity costs on individuals and the economy more generally and therefore cannot be dismissed as merely arising as a result of an omitted variables problem.