Is there overinvestment in overschooling?

Start date: 01 januari 2010

Edwin Leuven, Hessel Oosterbeek


Indirect/long term policy relevant


Research question: In this project we survey the economics literature on overschooling. The original motivation to study this topic was the suspicion that the strong increase in the number of college graduates in the early 1970s in the US caused a decrease in the returns to college education. We discuss the relation between this original motivation and the celebrated ORU specification of the wage regression. Next we discuss how over/underschooling have been measured at the level of individual workers, and what the incidence of over/underschooling is. We then analyze in more detail the ORU specification of the wage equation, the potential problems with it due to endogeneity and measurement error, and review the results previous studies have obtained using this specification. Results: While many studies document that over/underschooling is difficult to measure, we identified not more than one study that attempts to correct the wage returns to over/underschooling for measurement error. Unfortunately this study overlooked that measurement error in over/underschooling cannot be treated as classical measurement error. Addressing measurement error in a more rigorous way using the PSID, we find a substantial reduction of the returns to overschooling and a substantial increase of the penalty to underschooling. Status: this paper is prepared for the Handbook of Economics of Education and will be finished May 2010.

Related information

Share this page:   Deel op Twitter Deel op LinkedIn Deel op Facebook Deel via e-mail