In the UK, people from all ethnic minority groups are now more likely to go to university than white British people. But does university education pay off when it comes to their future earnings?
I looked at this question in a recent report, co-authored with Jack Britton at the Institute for Fiscal Studies and Lorraine Dearden at UCL. We found that the financial benefits from university are positive on average for all ethnic groups even after accounting for taxes and student loans. Gains are highest for South Asian students, middling for white students, and mostly lower for black students.
The benefits are especially large for Pakistani students, with an estimated boost to average earnings of more than a third by age 30. Adding up predicted gains over the whole life cycle and taking into account taxes and student loans, we found that doing a degree is worth around £200,000 for Pakistani students – around twice the average return for all students we calculated in previous work.
This is not because Pakistani graduates have especially high earnings. In fact, the opposite is true: Pakistani graduates have the lowest graduate earnings of all ethnic groups, with typical earnings at age 30 of £23,000 for men and £19,000 for women.
Instead the reason is that – based on comparing similar people who did and didn’t go to university – Pakistani graduates would have earned much less had they not gone to university. Typical earnings at age 30 of Pakistani men and women who did not go to university are only £13,000 for men and £11,000 for women.
An important factor explaining the large earnings gains for Pakistani graduates (compared to not attending university) appears to be that Pakistani students are more likely than White British students to choose subjects with good job prospects at university, such as business, law, or pharmacology. They are also less likely to choose degrees with low or negative financial returns, such as creative arts.
This reflects a more general pattern. All Asian groups are more likely to study “high-return” subjects at university, which appears to be a major factor behind the comparatively large gains for these groups.
These findings appear to contradict a claim in the government’s recent race commission report. According to the report, an explanation for the low graduate earnings of many ethnic minority groups is that “ethnic minority students, and especially Black students, from lower social status backgrounds are not being well advised on which courses to take at university”.
Our findings suggest that the opposite is true for South Asian students, as they tend to study more lucrative subjects than white students. We also find no evidence that black students choose lower-return subjects than white students. This does not mean that poor career advice is not a problem – but it doesn’t seem to affect ethnic minorities disproportionately.
The government’s report also suggests that ethnic minorities have low graduate earnings because they attend less selective universities. It is true that students from ethnic minorities – especially black students – are more likely to attend lower tariff universities, and that graduates of these institutions earn less than other graduates.
But importantly, this does not mean that these universities offer low returns. Many graduates of these institutions would have had much lower earnings still if they had not gone to university at all. Overall, we found no evidence that ethnic minorities’ institution choices lower their gains from attending university.
Differences by socio-economic background
We also looked at differences in gains from having a degree by socio-economic background. We found that those who went to private schools – especially men – benefit much more from university than state school graduates. Among the state-educated, returns vary relatively little by socio-economic background, but if anything, those from the poorest families benefit the most.
Again, this is not because graduates from poor families earn a lot: typical earnings at age 30 for this group are relatively low at £25,000 for men and £21,000 for women. Instead, it’s because they would likely have earned much less had they not gone to university. For state-educated 30-year-olds from these families who did not go to university, typical earnings are only £20,000 for men and £11,000 for women.
What explains differences between groups?
All of the analysis above compares students who did or did not go to university within each ethnic or socio-economic group. When looking across groups, we found that some but not all of the differences in earnings are explained by differences in school attainment and other background characteristics such as special educational needs or English as an additional language. Unexplained earnings differences are smaller among graduates than among non-graduates, suggesting that university can to some extent level the playing field between different groups.
Strikingly though, even among graduates, white British men earn significantly more on average than men from all non-white ethnicities after controlling for all observable background characteristics. Similarly, among the state-educated, those from wealthier families earn more than those from poorer families even after controlling for the full range of factors.
These unexplained differences could reflect the effects of discrimination in hiring, promotion and pay. But they may also point to other differences between groups, such as differences in geographical mobility or access to professional networks.
The race commission report rightly highlights that a large share of racial disparities is explained by economic disparities, leading to gaps in school attainment. But our research shows that there are also unexplained gaps between ethnic groups, as well as gaps between socio-economic groups that can’t be explained by attainment gaps.
Even if some of the race commission’s particular conclusions won’t ultimately stand up to scrutiny, its distinction between explained and unexplained disparities is helpful for pinpointing the sources of earnings differences between ethnicities. Getting to the bottom of what causes unexplained disparities will be an important task for further research.