Graduate earnings: what will politicians do with the data?

“Worthless degrees: graduates earn less than school-leavers,” ran the headline in The Times. “Questions raised over value of a degree,” reported the Financial Times.

The long-awaited graduate earnings research published last week by the Institute for Fiscal Studies actually said in its executive summary that its “main finding” was that graduates from wealthy backgrounds out-earn their less wealthy counterparts even when they attend similar courses at similar institutions.

But the two newspapers focused on another eye-catching finding from the research: that at 23 institutions (unnamed by the researchers, who only approached Russell Group institutions for permission to identify them in the paper), earnings for men 10 years after graduation were lower than those for non-graduates. As well as particular institutions, particular courses came under the spotlight. The researchers found that “disappointingly, creative arts delivers earnings which are roughly typical of non-graduates”.

People in government will be “crawling all over” the research, according to Gordon McKenzie, GuildHE chief executive and former deputy director for higher education strategy and policy in the Department for Business, Innovation and Skills. BIS is working on its own, separate project to gather and publish graduate earnings data by institution.

In terms of the political impact of the data, could such figures be the catalyst for transformative change in tuition fee caps and public subsidy for courses in English higher education? And would such changes be beneficial or destructive?

David Palfreyman, director of the Oxford Centre for Higher Education Policy Studies, believes that we should be told the identities of the 23 institutions where male graduate earnings were below those of non-graduates.

He said: “The student-debtor punter deserves to have access to this data, especially if thinking of doing a degree that does nothing for his/her employment chances [or] doing so at a university that adds nothing through its brand value to his/her employability.”

And, he continued, taxpayers “should also see the data and decide if spending on higher education [is] really worth it compared to spending on schools, further education, [the] NHS, etc”. 

‘Easy-option blanket-gorging’

Facing up to the potential challenges brought by graduate earnings data would be a healthy process, argued Mr Palfreyman. “If punters and/or taxpayers decide it is not worth it, then the higher education industry needs to work on delivering a cheaper product than its current egregious easy-option blanket-gorging” on £9,000 annual fees over three or four years, he said.

This should be “easily done since the for-profits already deliver at circa £6,000, there are massive efficiency savings to be made, there is cost-reduction though massive open online courses use, courses could be speeded up”, he added.

“This would be the normal application of market forces to any economic activity”, Mr Palfreyman said, adding that higher education has until now been “sheltered by being a (quasi) public sector monopoly and one where consumer information is (sometimes deliberately) missing or distorted”. 

Perhaps it is not surprising that different sections of the media drew different messages from the IFS paper – as the researchers who worked on the project seem to have taken different messages as well.

While the IFS press release, executive summary and full paper all emphasised the key role of students’ pre-university social background in shaping graduate earnings, Neil Shephard, professor of economics and statistics at Harvard University, instead emphasised variability in earnings by university and course.

“This impacts on which parts of the higher education sector the UK government funds through the subsidy inherent within income-contingent student loans,” he said. “The next step in the research is to quantify that variation in funding, building on today’s paper.”

That refers to the calculation of more detailed figures on the resource accounting and budgeting (RAB) charge – the public subsidy in writing off the portion of loans never repaid – by university and course.

Such calculations would show, in exact detail, how the level of subsidy is highest in those courses – such as creative arts – where earnings appear to be lowest.

Bahram Bekhradnia, president of the Higher Education Policy Institute, said of the graduate earnings research: “I have always thought that while it might be thought useful to have this information, potentially the implications of what a government that is ill-intentioned might do with it are worrying.”

He said the direction of the IFS research was “going far down the road of treating higher education in a very atomised way, as millions of individual transactions rather than as a whole”.

One solution to reported low graduate earnings in creative arts subjects could be to “close down all creative arts courses and make it appear that you have a better rate of return [on investment]”, he said. “Is that what we want as a country?”

Mr Bekhradnia cautioned policymakers to be “very careful what policy conclusions you draw” from earnings data.

He added that the logic of Professor Shephard’s position is that those with lower loan repayment rates should be charged more for their borrowing, which is likely to mean that women, students from working-class backgrounds, or students on courses such as creative arts would pay “a higher rate of interest” on their loans. This would be “intolerable”, Mr Bekhradnia said.

What the IFS research does is to link anonymised individual Student Loans Company data with HMRC tax records for 260,000 students to examine variation in graduate earnings by university attended and subject studied, looking at earnings up to 10 years after graduation.

In addition, it looks at average entry tariffs of courses and socio-economic background of students to create a measure of the “value added” by degrees.

Andrew McGettigan, author of The Great University Gamble, said it was important to separate the dataset created by the HRMC/SLC data and the value added measure.

“It is hard to argue against publishing the raw data by institution and subject as it is will inform prospective applicants (albeit about how students from over 15 years ago fared subsequently)," he said. “How should that dataset further inform policy? That's hard to say and depends on how far you think the government should think like an investor looking for a return on its HE funding through loan repayments and higher tax revenues.

"Regarding the proposed 'value add' - that looks perfectly acceptable for economics degrees but I doubt we want a diverse HE sector to be reforged in the image of that one discipline."

Jo Johnson, the universities and science minister, said in his response to the research that it "reveals the worrying gaps that still exist in graduate outcomes. We want to see this information used to improve the experience students are getting across the higher education sector, and it’s why our reforms will continue to encourage universities to focus on teaching quality and helping their students progress into fulfilling careers.”

No mention there of what some see as the hidden agenda behind the government’s interest in graduate earnings data – as a means to reduce public subsidy for certain courses or universities deemed to be low performing by reducing their fee caps, or a means to allow fee caps to rise for those deemed to be high performers.

Mr McKenzie said: “Right now, I think ministers see it [graduate earnings data] primarily in terms of better student information, potential performance metrics…and as a challenge to social mobility policies.

“That said, someone, somewhere in government is always interested in questions like where, and by how much to provide public subsidy; how best to incentivise universities or how to cut the public cost of the student loan book. I imagine people will be crawling all over this and generating ideas, but they won’t necessarily translate into politically viable policies.” 

john.morgan@tesglobal.com