This article is more than 6 years old

How do you solve a problem like REF 2021?

'Whole staff submission' to the next REF is reputedly a red line for the universities minister. How is this to be reconciled with a sector that believes it is unworkable and undesirable? Martin McQuillan investigates.
This article is more than 6 years old

Martin McQuillan is a former Deputy Vice-Chancellor (Research and Innovation) at Kingston University, London.

The minister certainly has his work piling up for him. His Easter is going to be spent fretting over the fate of the Higher Education and Research Bill before it returns to the Commons at the end of April. Then there is Brexit and its impact on the university sector, a big enough headache on its own. There are the TEF results to process. He also has to get his skates on as BEIS work up the new industrial strategy.

Resolving paradox

Now that the consultation on REF 2021 has closed, Jo Johnson might be forgiven for putting the responses at the bottom of his red box. However, we have been promised the outcome of the consultation by mid-2017. This is a curiously vague timescale (though not unusual for government). It is not clear whether it means early June or mid-August, when most academics have packed the Volvo and headed to the South of France. Last time the civil servants were left in the office on their own over the summer they came up with the Gold, Silver and Bronze of the TEF based upon the subliminal effects of the Rio Olympics.

Despite other pressing concerns, the future of the REF now presents a significant problem for the minister. In adopting the recommendations of Lord Stern’s independent review he has set his stall out for changes in the way we assess research in our universities. The consultation has shown that while each of Stern’s suggestions on their own may seem entirely reasonable, taken collectively they are an unworkable mess.

So, we are now in the Whitehall game of creating a solution in which the minister can claim that all his red lines have been met, while vice chancellors and mission group heads can say with a straight face, “we welcome these proposals”. Fortunately, resolving paradox is an essential skill of any senior British civil servant. The recent consultation will show sector-wide opposition to whole staff submission and the technical difficulties of submission by HESA codes. Nevertheless, the minister will have a REF based on submission of all research active staff. How are these two antimonies to be reconciled?

It is accepted by everyone that Stern’s proposal to pre-ordain a Unit of Assessment for staff through JACS code is a non-starter. It would create all sorts of perverse results and would disadvantage interdisciplinary work. We can expect this suggestion to be dropped. However, once this principle has been ceded then selectivity is immediately reintroduced into the exercise. There is nothing to stop universities creating ghost UoAs full of unfundable research to protect selectivity elsewhere. This only adds to the burden for institutions and panels, while making a mockery of a REF intended to stop game playing. It is much better simply to have a selective exercise that will protect the overall grade point average of UK research.

However, we are told whole staff submission is a red line for the minister. Stern, for what it is worth, actually recommended the submission of “all staff that have a significant responsibility to conduct research”.

Get off the road

On 15th March, a few days before the closing date for the consultation, HEFCE published a remarkable blog on its REF website. David Sweeney, Director of Research, Education and Knowledge Exchange (Mr. REF to you and I) reflected on the problems thrown up by the consultation over the definition of ‘research active’ staff. While Sweeney said there would be no re-run of REF 2014 rules, he recognised that there were issues around the use of contractual status, as notified to HESA, to determine who is research active. He suggested that the peculiarities of the Scottish system made this particularly difficult. This mystification might pull the wool over the eyes of the Westminster set, but an Aberdonian cannot kid a Glaswegian. There is no significant difference between Scottish contracts that include both teaching and research but do not determine any specific pattern of academic work, and contracts elsewhere in Britain. This is a UK-wide problem.

That being said, Sweeney went on to solicit alternative solutions to the problem of defining the research active. He offered the suggestion that “we would expect universities to develop and publish the process they used to establish agreement with their staff on their ‘research-active’ status”. He also said that the consultation might throw up better ideas on this. What this means, should it be adopted, is that universities must devise a way in which their own academic staff will declare themselves to be research inactive for REF purposes.

There would seem to be consensus across mission groups for this idea, which some are comparing to a ‘Statutory Off Road Notification’ for academics during the REF period. Staff could be asked to make a self-declaration that excludes them from consideration in the REF. Much management time would be spent identifying and persuading staff members to declare themselves research inactive. How labour-intensive this will be depends upon where you sit in the sector. The average post-92 institution submits around only 20% of its staff to the REF. It will be no small matter to harvest self-declarations of inactivity from 80% of staff in order to retain previous levels of selectivity.

However, in this way the minister can have a REF based on the submission of all research active staff and universities continue to select who is submitted. So much for the end of game playing and for reducing the burden of assessment on institutions.

It is, equally, only a sticking plaster solution for this REF cycle, allowing universities more time to dichotomise their work force between those on research designated contracts and those substantially involved in teaching. Perhaps, by REF 2027 there will be additional categories in HESA for reporting staff activity.


There are other contentious issues in the consultation but their fate is more clear-cut. We can expect the proposals for institutional level impact and environment to be quietly dropped.  Too great a percentage of QR funding would be tied up in the aggregation of mediocrity rather than supporting excellence. The impact agenda of the next REF is likely to be driven by the emerging industrial strategy and may attract a QR supplement from the hefty industrial strategy challenge fund.

It has been a close call around the portability of outputs, but the sector cannot be seen to have it all its own way. On this point the likely compromise is that staff may be able to take one output with them if they move between institutions. This will be in line with a suggested minimum of one, and maximum of four, outputs for submitted staff. How universities are to agree between them which output is to be ported, or, whether the same output can be submitted for two institutions is not clear. However, if you still think that this has anything to do with reducing administrative burdens on universities, you have not been paying attention.

All of this is a brave attempt to retain the principles of funding research excellence while keeping both the government and the sector on side. If it can be pulled off, we can look forward to that series of summer press releases welcoming the new proposals and praising the work done by the valiant men and women of HEFCE. However, in shifting the goal posts away from Stern’s recommendations, it is possible that any new set of rules will throw up their own unintended consequences. I would not be surprised to see a second technical consultation next year to test these changes.

If the prospect of a second Brexit or Scottish independence referendum fills you with dread, then spare a thought for the wonks if we have a second REF consultation.

Leave a Reply