Interpretation of the phenomenon of graduate overeducation remains problematical. In an
attempt to resolve at least some of the issues this paper makes use of the panel element of
the HILDA survey, distinguishing between four possible combinations of education/skills
mismatch. For men we find a significant pay penalty only for those who are both
overskilled and overeducated, while for women there is a smaller but significant pay
penalty in all cases of mismatch. Overeducation has a weak negative effect on the job
satisfaction of women and no effect on the job satisfaction of men. Overskilling either on
its own or jointly with overeducation does so. Finally, overeducation has no significant
effect on the job mobility of either men or women, though there is a significant positive
effect on both voluntary and involuntary job loss in men who are both overskilled and
overeducated, with the results again differing for women. At least for a substantial
number of workers it appears, therefore, that overeducation represents a matter of choice
(or is possibly a consequence of low ability for that level of education), while
overskilling imposes real costs on the individuals concerned.