This article is more than 8 years old

Human capital: a look at IFS research on graduate earnings

IFS has published findings from their long-awaited report into graduate earnings. Here we look at the main findings, consider the response by the sector, and some of the potential policy options.
This article is more than 8 years old

Louisa Darian was Deputy Director at Wonkhe, and is now an advisor at the Department for Communities and Local Government.

Today the Institute for Fiscal Studies (IFS) has published findings of their long-awaited report into earnings of graduates in England. It represents a landmark for research into educational outcomes in the UK: it’s the first time that HMRC PAYE and Student Loans Company data has been linked for research purposes, providing real data on earnings ten years’ post-graduation by subject and, to some extent, institution.

The report contains some revealing findings with important and far-reaching implications for the sector.

Many of the headline results provide the evidence to support what many have long believed to be true, including that:

  • Graduates from richer family backgrounds earn more than other graduates: the average earnings gap between those from a higher-income background and the rest is around 30% for males and 24% females. Even after taking account of the subject studied, and the characteristic of the institution, the average student from a higher-income background earns 10 percent more than other students.
  • Graduates earn more than non-graduates: among those with significant earnings (defined as above £8,000 per year) median earnings for male graduates were 30,000 per annum compared to £22,000 for their non-graduate peers; and non-graduates were twice as likely to have no earnings compared to graduates.
  • There are big differences across, and within, universities: more than 10% of male graduates from LSE, Oxford, and Cambridge were earning more than £100,000 a year with graduates at LSE earning the most while, at 23 universities, median earnings for male graduates were less than those of non-graduates. But even within LSE, male graduates’ earnings range from £170k at the 90th percentile to around £40k at the median. The authors also note the very strong performance of some northern universities (Liverpool, Newcastle and York) whose students achieve very competitive earnings despite their local labour markets having lower earnings than in the South.
  • Subject matters: medical graduates earn the most, with men earning on average £50,000, and Creative Arts students the least: more than half earn less than £20,000, which is roughly typical of non-graduates. The earnings premium at the top end of the distribution for some subjects is good: approximately 12% of male and 9% of female economics graduates earned above £100,000 ten years after graduation. However at the lower end of some subjects and institutions, earnings are far lower than was previously assumed.


Although clearly groundbreaking in nature, there are some limitations to the methodology used in the research. The first caveat to note is that the researchers only explore institutional differences across Russell Group universities that agreed, or had sufficient sample, to analyse (19 out of the 24 total). From what we can see, three universities (Leeds, Birmingham and UCL) did not agree to be named in the analysis. And the researchers did not ask universities outside of the Russell Group if they would participate (although they say that this is something they would like to do in the future).

The research covers a sample of just 10 per cent of all borrowers since 1998 and cannot control for the location that the graduate is now located (important given local labour market variations). The research also doesn’t take account of those who drop-out of university, and excludes the 15 percent of students who don’t take out a student loan. It is also unable to capture which graduates in the study undertook further education or training after graduation.

As a result, there remain some unanswered questions including how less selective institutions, that pride themselves on the support they provide students to succeed in employment, compare against more selective ones. There is also more exploration that could be done to explain the lower earnings of Creative Arts students. Author and researcher Andrew McGettigan told us that “IFS should have given more consideration to what an early career artist or actor is doing in the first few years after graduation and what implications this has in the medium and longer term….I don’t see any recognition that earnings might be lower, because for some, work may be a means to supporting their practice and so is part-time with paid hours subordinated to time in the studio (Admittedly the current fee-loan regime is not geared for these outcomes).”


The authors identify at least three policy implications, which we set out below while also considering what the options that the Government might explore:


The Government’s income contingent loan system means that the cost to Government depends on the earnings of graduates over the first 30-years of their career. The report suggests that, over the period of time that the research took place, the proportion of students taking those subjects that deliver higher returns (e.g. Economics and Law) has been reducing while more have taken subjects like Creative Arts. It is not clear what the reasons are for this, whether it reflects student choice, or whether institutions are offering more places because they are cheaper to deliver. If the latter is true, it presents a challenge to the Government’s policy to create a more responsive market by lifting the number cap. The findings could see greater consideration given to the department level RAB charges that institutions have feared, or a move to the Australian-style fee system where loans are differentiated by subject.

Student choice

While information on earnings is not an indicator of teaching quality, it is nevertheless important information for students – particularly in the context of higher fees and given the low returns of some subjects. Research finds that four in ten graduates say their earnings fell below what they were expecting. While there would be risks associated with including this information in the TEF, it could be included in the Key Information Set alongside contextual information to inform student decision-making. The Government’s recent Budget suggested that the delivery of this information is in development. The research findings also suggest that students need access to information on how earnings compare to non-graduates, to help inform a decision about whether university is the right route for them, but also access to early advice so that they can consider the best subject choices to support their chosen career path.

Social mobility

The report raises implications for social mobility given the impact that family background has on earnings. The authors note that there may be a number of reasons for this: wealthier students may have access to better financial and cultural capital; and students from lower-income backgrounds may face discrimination in the labour market. Either way, it demonstrates the importance of looking at measures to retain and support students when they get to university. In their response to the report, the University and College Union call for reforms to admissions processes and NUS point out that the Government’s decision to scrap maintenance grants will do nothing to support social mobility.

Across the different sector bodies, the response has been fairly consistent: the research highlights the benefits of a university education while demonstrating that there is more to do, particularly for students from disadvantaged backgrounds following this fresh evidence. The key question now is what ‘more’ will look like.

The above is an initial look at the research, but its far-reaching findings will be digested and debated for some time yet. We’ll be adding further analysis as it emerges.

Download the report How English domiciled graduate earnings vary with gender, institution attended, subject and socio-economic background in full here.

5 responses to “Human capital: a look at IFS research on graduate earnings

  1. I think it is important not to overlook this conclusion: “These findings have implications for our understanding of the nature of subsidy of higher education. Given the relatively low earnings of graduates with degrees in some subjects, the level of public subsidy for these graduates is likely to be greater than for other graduates in other subjects, such as economics, even given the lower costs of provision for some subjects as compared to others. Making this explicit when considering the shape of higher education and in particular where any further expansion might take place would seem important.” p 55

  2. Ten years ago economics and medicine were mainly taught by more selective universities. The conclusion about subject is actually more due to average tariff score and not subject chosen.

    Economics and medicine students of ten years ago had better A levels than their peers and went on to earn more, that isn’t a surprise but it has been dressed up as if students should now be choosing economics regardless of their level 3 qualifications, this is dangerous as another subject may suit them much better in terms of future wellbeing than economics at a less selective institution.

    1. Interesting point. On p26 of the report the authors control for other factors that affect earnings, incl prior achievement, and find that at the lower end of earnings distribution in particular there is less variation in earnings by subject. Different at the 90th percentile.

  3. This report has been long awaited and now that we’re finally able to read it, it tells us a lot that we expected, not much that we didn’t and raises as many questions as it answers. Even so, it’s useful to have data to support the positions so many people have held without good evidence to support them.

    However, the most profound limitation of this study – and it’s not a methodological flaw, just simply a fact of what the data can possibly show – is that it encourages us to equate graduate success with money earned.

    For nurses, for teachers, for social workers – even for the creative artists – success is about fulfilment, satisfaction and contribution. Many of these graduates make at least as big an economic contribution to society as those who studied law or economics. Indeed, many make a far larger contribution, but take less for themselves. I would not want to I’ve in a world with more economists that teachers, not simply because our education system would suffer, but because our society – our humanity, even – would be poorer for it.

    We need to explore ways of recognising positive graduate outcomes that are not purely based on earnings. Young people rarely make choices about their future based on unguaranteed salaries and nor should they.

Leave a Reply