The 1991 wave of the British Household Panel Survey is used to examine
the extent of, and the returns to overeducation in the UK. About 11%
of the workers are overeducated, while another 9% are undereducated fo
r their job. The results show that the allocation of female workers is
more efficient than the allocation of males. The probability of being
overeducated decreases with work experience, but increases with tenur
e. Overeducated workers earn less, while undereducated workers earn mo
re than correctly allocated workers. Both the hypothesis that producti
vity is fully embodied and the hypothesis that productivity is complet
ely job determined are rejected by the data. It is found that there ar
e substantial wage gains obtainable from a more efficient allocation o
f skills over jobs.