David Kingman looks at some new research based on the Longitudinal Educational Outcomes (LEO) dataset, which sheds more light on the complex question of whether obtaining a university degree is worth the financial cost of doing so
Tuition fees are an extremely vexed issue in contemporary British politics. In response to a perception that Jeremy Corbyn’s promise to abolish tuition fees cost the Conservative Party votes at the 2017 general election, the government launched the Augar Review earlier this year, which has been charged with exploring possible reforms to how we fund the whole system of post-18 education in Britain. Meanwhile, in a couple of weeks’ time the Office for National Statistics (ONS) is expected to announce a decision on whether the government will therafter be forced to recognise the full cost of student loans as expenditure on its balance sheet, which will draw renewed attention to the issue of whether the government is currently funding higher education in a way that delivers value for money – for either taxpayers or students.
Value for money in higher education
The latter question has become a subject of increasing controversy since the government increased the maximum tuition fees which universities are allowed to charge to £9,000 per year from September 2012, and then abolished maintenance grants (replacing them with loans that add to a student’s total debt) in 2015. (The maximum fee which English universities can charge students was subsequently increased again to £9,250 in 2016.)
Higher education tuition fees are generally justified on the grounds that the pecuniary benefits of having a degree are so large that going to university still represents a fantastic investment for the typical student. However, as we argued in a previous post on this blog, the evidence regarding the “graduate premium” is actually quite mixed, with the premium that the average student receives (which is often presented as a blanket figure) actually being highly dependent upon a range of intervening characteristics, including the student’s personal background, the type of school(s) they attended previously, prior educational attainment, the degree course which they study and the institution that they attend.
A new piece of research on this topic, The impact of undergraduate degrees on early-career earnings, jointly produced by the Institute for Fiscal Studies and the Department of Education, has now shed even more light on these variances by examining the average pay differential between graduates and similar workers who are non-graduates at age 29, controlling for a wide range of important background characteristics. This article will explain some of its key findings and attempt to distil what their implications could be for the higher education funding debate in this country.
The research was undertaken using a new resource called the Longitudinal Educational Outcomes (LEO) dataset, which demonstrates the potential that linking multiple administrative datasets held by government has for answering important social research questions. Essentially, what this dataset (which first became available last year) does is merge together several million records for individual British citizens from a number of different government databases, including the National Pupil Database held by the Department for Education, data from the Higher Education Statistics Agency (HESA) on where and when people are educated after they leave compulsory education, and data from the Department from Work and Pensions and Her Majesty’s Revenue and Customs (HMRC) on people’s tax and welfare benefits once they start working.
Linking these datasets enables researchers to have a virtually complete picture of what these individuals’ background characteristics were before they entered education, where they were educated and how they performed during each stage of their education, and then their subsequent history of employment and earnings once they entered the labour market.
Being able to take a “big data” approach to tackling questions about the long-term consequences of factors such as education and social background on future outcomes is potentially an extremely powerful tool for researchers to have at their disposal because it overcomes the problem that in the past it’s been quite difficult to get large numbers of people to participate in longitudinal social surveys over multiple waves of data collection, which meant that findings sometimes needed to be generalised on the basis of relatively small sample sizes (not to mention that it is very expensive and labour-intensive to run such longitudinal household surveys in the first place). By linking administrative datasets, data on millions of people which the government already possesses can be analysed in one go.
Show me the money
So, what did this new analysis reveal? As mentioned above, the researchers were interested in how the earnings of people who had been to university compared to those of workers with similar educational characteristics who hadn’t, by age 29. The latter part of that sentence is important, because it means that they only looked at people who obtained at least five GCSEs with grades A* to C including English and Maths, on the grounds that people whose attainment was lower than this were too dissimilar to the people who do go on to university to derive a meaningful comparison, as without good GCSE grades a person has very little chance of going on to higher education in the future.
One important point to note is that the researchers were interested in the effect of tuition fees on gross earnings; if the comparison included net earnings then it would presumably have looked somewhat different because people who graduated during this period had to make student loan repayments at 9% above the annual repayment threshold, which would have diminished their earnings in comparison with non-graduates to some extent (although net earnings would also have been influenced by a wide range of other factors which have little to do with higher education per se, such as entitlement to tax credits, so you can see why this would have made the picture look more complicated). The researchers are also at pains to emphasise that the anticipated monetary returns are far from the only motivating factor when someone chooses to study for a particular degree at a particular institution (for example, an individual student may have a passion for a particular subject, or see attending university as a rite of passage towards adulthood), so this research deliberately only takes a very narrow view of a complex decision-making process.
Their findings can be summarised thus. Firstly, attending higher education boosts female earnings by much more than male earnings: the average man who got five good GCSEs at 16 who then goes on to university is only earning 6% more on average by the time he reaches 29 than a man with the same prior educational attainment who didn’t go to university, once you account for his personal background (the raw difference is much bigger, at 25%, but most of this disappears when you control for socio-economic background); however, the average woman who’s in the same boat enjoys a much larger earnings premium (26%), mainly because female graduates are much more likely to be in the labour force and working full-time at that age than female non-graduates are (which is at least in part because female non-graduates are more likely to have children at earlier ages).
Secondly, this new report chimes with all the previous research which shows that subject choice and institution matter hugely. However, what is striking about this analysis is that it shows that there is a large group of male graduates who were actually earning less at age 29 than their non-graduate equivalents. Around one-third of the male graduates were earning less, whereas 99% of the female graduates were earning more than their non-graduate counterparts by this age. These male graduates tended to be concentrated in post-1992 universities and specialist institutions which offer very niche subjects, such as performing arts specialisms, so the researchers point out that some of these courses might have been chosen to fulfil passion projects rather than on the basis of their future earnings potential; but it still paints a concerning picture of the inequalities which the English higher education system perpetuates. Even among those who did realise an earnings premium, its size was also hugely sensitive to institution and subject choice, with a small number of courses at the most august universities – most notably maths, law, medicine, physics or economics studied at one of LSE, Oxford, Cambridge, Imperial or King’s College London – providing huge premiums of over 25% in comparison with non-graduates. By contrast, studying English, Philosophy or Creative Arts were – for men – all linked to receiving a negative graduate premium.
Worth the cost?
What implications could this new data have for the higher education funding debate? Universities are coming under increasing pressure for charging almost exactly the same (maximum possible) tuition fee for all their courses, regardless of the specific demands of the subject or the amount of teaching time, particularly when the cost of studying postgraduate qualifications in the same subjects varies enormously both between different courses and different institutions. Having such precise evidence that there is massive variation in labour market outcomes between the different courses and institutions seems likely to create further pressure towards some kind of fee differentiation, although how policy-makers should address the gender divide in returns to different degree courses is a problematic question (given that they almost certainly wouldn’t be in favour of charging different fee levels to men and women).
This research is far from the final word on the issue of trying to value the economic returns to going to university; indeed, the researchers point out that age 29 is still quite early in a graduate’s career, and you would expect the earnings of male graduates, in particular, to peak quite a lot later than this. Nevertheless, for policy-makers it does raise the key question of whether too many people are going through university courses unnecessarily when there might be more appropriate alternatives for them instead.
Photo by rawpixel on Unsplash: https://unsplash.com/@rawpixel
Help us to be able to do more
Now that you’ve reached the end of the article, we want to thank you for being interested in our work standing up for younger and future generations. We’re really proud of what we’ve achieved so far. And with your help we can do much more, so please consider helping to make IF more sustainable. You can do so by following this link: Donate