Dr Matt Dickson is Reader in Public Policy at the University of Bath Institute for Policy Research (IPR), leading the Widening Participation in Higher Education research stream, and a co-author of the latest DfE report on graduate earnings.
Our latest report, authored by the Institute for Fiscal Studies (IFS), for the Department for Education has just been released, showing how the early career earnings of young people going on to higher education (HE) in England compare to those of similar young people who choose not to go.
The headline findings are that by age 29, those students attending HE earn a lot more on average than those who don’t: for men it’s around 25% more than the average male student with 5A*-C GCSEs who didn’t go to university, for women the premium is more than 50%. However, much of this can be explained by differences in the pre-university characteristics of those who do and those who don’t choose to go on to HE.
Amongst the group who get at least 5A*-C GCSEs, the 18-20 year olds who head off to university tend to have higher attainment at GCSE and A-level and come from wealthier backgrounds and so would be expected to earn more than average even if they didn’t go to HE. Accounting for differences in background (ethnicity and a socio-economic status index based on neighbourhood measures of deprivation and free-school meal eligibility) reduces the premium a bit but it is factoring in prior attainment that really makes the big difference – once this is done, the average university earnings premium for men falls to around 6% whereas for women it also reduces a lot but still remains high at around 26%.
So the first message is that prior attainment has a dramatic effect on the difference in earnings between HE attenders and non-attenders – once we compare young people with the same attainment at age 16 and 18, the HE impacts are a lot smaller. Getting better results at GCSE and A-level not only opens the door to HE but it increases the earnings potential of young people regardless of their next step in education. This may sound obvious but it provides an important backdrop to widening participation analysis – reducing the social gradient in attainment at age 16 is the key to improving both HE participation and social mobility more generally.
Delving deeper, there is a remarkable amount of variation masked by these average figures – variation both in terms of subject and institution. For example, men studying Economics can expect a return of around 33% higher earnings, and for Medicine 23% higher than comparable individuals not attending HE. On the flip side, studying some subjects is associated with statistically significant negative returns compared to similar 29-year olds who didn’t go to university e.g. Creative Arts (-14%), English (-7%) and Philosophy (-4%). For women there are no subjects that have negative returns, even Creative Arts offering a return of more than 8% compared to those who do not go to HE. The picture is similar when it comes to HE institutions: while a handful of institutions offer men a negative return compared to not going to university, at the other end of the range Imperial and LSE are associated with 50% to 60% higher earnings. For women the returns are even higher at the top, LSE almost doubling earnings for those who go there compared to similar young people who did not go to university, with only two institutions associated with statistically significant negative returns.
Understanding what drives this variation in returns is particularly important when thinking about widening participation and social mobility more generally. As shown above, prior attainment is a large factor, and coupling this with whether or not a student has a Maths or Science (STEM) A-level provides further insight. For men with no STEM A-level, students with higher prior attainment have a much higher return that those with lower prior attainment, a pattern found also for women though to a slightly lesser extent.
Moreover, further analysis in the report shows that the differences in returns between students with different prior attainment levels is likely to be driven by HE having different impacts on the earnings of different types of students – it is not simply down to differences in subject choice or institution.
Amongst men without a STEM A-level, students with low prior attainment are more likely than those with high prior attainment to choose Business or Computing which are high returns subjects, whereas the high prior attainment students more likely to choose History and Languages which are low returns subjects. Despite this, we still see that the high prior attainment students have much higher returns (20%) compared to low prior attainment (4%) and similarly, high prior attainment students without a STEM A-level have higher returns to attending each type of HEI. This suggests it is something about the HE experience that translates into better outcomes for higher prior attainment students rather than it just being that they go to different institutions and take different subjects to their low prior attainment peers. This is clearly an area where we need more research to better understand what drives these differences.
In the meantime, importantly, students know their own prior attainment and the A-levels that they are doing, and so the information we provide in this report can be hugely useful for them in knowing the subjects and types of institution where they are likely to see higher or lower returns. For example, as shown in Figure 1, men with low prior attainment and no STEM A-level are likely to see a negative return to studying History or Philosophy but a good return to Business, Computing or Education – subjects that are taken up by a sizeable proportion of students with these prior characteristics.
Similarly, as shown in Figure 2, for low attaining women without a STEM A-level, studying Politics or Psychology is associated with a low return, yet Sociology, Communication, Education and Business degrees are all associated with substantially higher returns and are taken by large numbers of students from this grouping. Although this analysis is not claiming to provide causal effects, it does give prospective students important information about the returns on different subjects that students similar to themselves have attained in recent years.
This is especially important in the context of potential further expansion of HE in England. At the moment, at least two in three students in the middle and higher prior attainment groups go on to HE, whereas for the lower attaining group without STEM A-levels it is only one in three. If HE expands further to accommodate more of these students then it is important that they understand the subjects that are likely to lead to low or even negative returns and those that are a likely to be a better choice.
It is worth bearing in mind though that widening participation and HE expansion are not the same thing. Tackling barriers, particularly in terms of information, and ensuring that those from all backgrounds who have the ability and desire to pursue HE are able to do so is the first aim of widening participation. But this does not mean expanding HE so that increasing numbers of lower attaining students, for whom HE might not be the best option, are going to university.
What about the returns for individuals from different backgrounds once they get into HE? Ensuring fairness in access to higher education is important but it is equally important to ask whether those from non-traditional HE backgrounds are achieving the same outcomes as other students following their university experience. The size of the LEO data allows us to look at how returns vary by socio-economic background.
Positively, this evidence suggests that there is no strong social gradient in returns to attending HE for men or women. For men, controlling for the full range of background characteristics and prior attainment, those from the highest socio-economic backgrounds have a return to attending HE of 7% compared to not going, whereas men from the lowest socio-economic status group have a return of 9%. The middle groups have very similar returns, suggesting that compared to other people with similar characteristics, HE is giving the same proportionate return regardless of socio-economic background. Men who attended independent schools gain a 17% earnings premium from attending HE, higher than any other grouping, though this in part may reflect that the institutions and subjects that independent school pupils select into are particularly high return HEIs and subjects.
For women the picture is of an almost identical return to attending HE compared to not, regardless of the socio-economic background or independent school attendance, which paints a positive picture for social mobility – it is clearly not the case that socio-economic background affects the extent to which HE increases earnings compared with not going to HE.
In summary, the analysis in this report represents a huge advance in the quality of information available to prospective students on the returns to higher education in the UK, something which is particularly important for those from non-traditional HE backgrounds. Despite a large expansion of the sector over recent decades, the average return to attending HE in England remains positive and for women in particular large. Behind these averages there is a lot of variation and for men there are some courses that do not represent a good choice, with some students who currently opt to go to HE missing out on potentially more financially rewarding alternative paths – at least in terms of early career earnings. With all of this analysis however it needs to be remembered that prospective future earnings are only one reason among many why people choose to go to university and choose a particular subject and institution. The experience of HE is multifaceted with many benefits that are difficult to quantify. More research is needed to understand what drives some of the differences in returns we see for different groups, however, for all potential HE students, and in particular those from non-traditional HE backgrounds, this report provides a lot of detailed and useful evidence to inform choice.
Note: STEM A-level means the individual has at least one A- or AS-level in maths, biology, human biology, chemistry, physics, science, electronics, geology or computer studies.