While the provision of higher education has expanded around the world in recent decades, not everyone has benefitted equally from doing a degree. In particular, women and students from lower socio-economic backgrounds often earn less after graduating from university than comparable men or high socio-economic status students. Where do these post-graduation disparities come from?

One potential source of inequality is differences in the types of universities and degrees attended by men and women, or students from low and high socio-economic status (SES) backgrounds. This might be a particular issue in countries such as the UK and the US, where there is a wide range of options in university education, and substantial variation in both the type and the quality of degrees which are available.

What is ‘undermatch’?

‘Undermatch’ is the phenomenon of students attending universities or doing particular degrees for which they are over-qualified in some way. Undermatched students are relatively highly qualified in comparison to their university classmates, and could potentially be at a more selective university or on a more financially rewarding degree. Undermatched students are less likely to complete their degrees, and tend to earn less on graduation than those who are better matched to their university degrees. If undermatch is more concentrated in women or lower SES students, we might be concerned that undermatch is perpetuating gender and SES income inequalities.

Research examining the extent of undermatch and the types of students affected has mostly come from the USA, and has mostly examined the phenomenon at the university level. While this evidence is informative, we and our co-authors felt we could contribute in this area by examining undermatch in England – where large-scale data are available covering students on every course at every university, as well as information about the school grades and the average earnings of graduates from each course. The linked data allowed us to track a cohort of students from school into higher education.

We examine two types of undermatch. The first is based on how a student compares to others on the university course in terms of their prior academic attainment. For example – we might consider a student undermatched if they got three As at A-level, and are doing a university course on which the median student got three Cs. The second type of undermatch we examine is based on how a student’s relative academic attainment compares to relative expected earnings from the course. For example, are the highest attaining students in the country on the courses which lead to the best paying jobs? And are students with average attainment on courses which lead to average paying jobs?

What sorts of undermatch are there?

We find that students from low SES backgrounds are much more likely to undermatch than those from high SES backgrounds based on both definitions. This is true across the attainment distribution: from the lowest to the highest attaining students, lower SES students attend less academically prestigious courses on average, and attend courses with lower earnings potential. This finding suggests that the university system is not contributing to the reduction of socio-economic inequality as well as it might.

We also find that, although female students attend courses that are just as academically selective as those taken by male students, women tend to enrol in courses which are lower-earning than those taken by men. Again, this suggests that the university system could be doing more to alleviate inequality, in particular the gender pay gap.

We go on to consider the factors which could be producing these inequalities. Examining the SES gap in undermatch, we find that even within the same degree subjects, low SES students are more likely to be undermatched than their high SES peers. This suggests that the choice of university, rather than the choice of subject, is the important driver of undermatch among low SES students. We also find that even comparing only those low and high SES students who go to a local university, low SES students are more likely to attend an institution for which they are overqualified, suggesting location is not driving undermatch.

Unlike with the SES gap, university subject choice does seem to be the crucial driver of the gender gap in match. This suggests that women are not attending less prestigious institutions than similarly qualified men, but within those universities, they are studying subjects which tend to lead to lower-paying jobs.

What should we do?

We suggest the best policy response to these inequalities is to improve the level and quality of information available to school students about different universities and course options. Students might be provided with information on the typical entry requirements and later earnings of graduates for different courses at key decision-making ages. However, careful targeting is important, since in general those who could gain the most are the least likely to consume such information.