Increasing dispersion in the returns to graduate education is found, using quantile regression. This trend is related to rising overqualification. We distinguish between and validate measures of "Real" and "Formal' overqualification, according to whether it is or is not accompanied by underutilisation of skill; and using a unique data series in Britain we report the trend in overqualification types between 1992 and 2006. The distinction between types is relevant because employees in the Real Overqualification group experience greater, and more sharply rising, pay penalties than those in the Formal Overqualification group. Real Overqualification, but not Formal Overqualification is associated with job dissatisfaction. Formal Overqualification has been increasing over time, and in 2006 characterised nearly one in four graduates. Real Overqualification has been steady or rising only slowly; in 2006 it affected less than one in ten graduates. Conditioning on graduates being matched to graduate jobs, it is found that there is no significant increase in the dispersion of returns to graduate education. The normative implication drawn is that the state should provide regular public information on the distribution of the returns to graduate education.
Stored only when both papers are in the corpus — 490 citations exist upstream.