Research Excellence Framework: next REF delayed to 2029

The UK’s next Research Excellence Framework has been delayed until 2029 amid concerns over plans to fully break the link between individual staff and university submissions.

The exercise, whose results were due to be published in December 2028, will now conclude in the final month of 2029, it was announced on 7 December.

That decision was made “in recognition of the complexities” for universities in “the preparation for using Hesa [Higher Education Statistics Agency] data to determine REF volume measures” and “fully breaking the link between individual staff and institutional submissions,” said the team behind the REF, which is run by Research England and its funding council counterparts in Scotland, Wales and Northern Ireland.

Under new proposals for REF 2028 – now REF 2029 – research-active staff will not be required to submit a minimum number of outputs (which was a stipulation in REF 2021) nor will they be limited to entering a maximum number of outputs (five in 2021).

While units of assessment would still be required to submit 2.5 outputs for every researcher for REF 2029, the total volume of submissions would be based on that unit’s submission of research-active staff to Hesa.

Institutions would have been allowed to submit outputs from anyone employed on at least a 0.2 full-time equivalent basis for at least six months in the two-year REF assessment period up to the submission deadline.

Research England believed the proposals would have led to about £100 million in efficiency savings on the £471 million cost of the REF 2021– thanks to a reduction in costly internal selection and efficiencies related to special circumstance decisions – but the plans have faced criticism in recent months.

Earlier this week the Royal Economic Society said submissions based on a definition of “volume-contributing” staff “provides greater scope for institutions to misrepresent research power by inflating volume-contributing staff numbers by including staff with a minimal workload attributed to research and innovation”.

With staff not restricted by a maximum number of outputs, “institutions may choose to focus their financial resources more on recruitment/retention of ‘superstars’”, it added, with the society’s chair Sir Anton Muscatelli, vice-chancellor of the University of Glasgow, adding “the proposed revisions for REF2028 risk making REF less inclusive”.

Announcing the delay, which follows the conclusion of a consultation on the REF’s initial decisions announced in the summer, the REF team said it would also review the proposed weighting of research culture and environment, which was due to rise to 25 per cent in 2028, the same weight as impact. An extended consultation on this issue has just closed.

Steven Hill, director of research at Research England, said the delay reflected concerns raised by institutions regarding the Hesa data-led submission of researchers.

“We have learned from the consultation that there was more complexity than we thought,” he said. “We thought [our plans] were feasible in the summer but we’ve learned from the experts on submitting Hesa data.”

Asked if the worries voiced in the consultation reflected growing opposition to the REF’s new structure, Dr Hill insisted that these were a “normal part of the process to get feedback from the sector”.

“We make proposals and listen to what people have to say,” he said.

On the people, culture and environment plans, the REF team added that further work is being commissioned on indicators that will be delivered during 2024.

On the changes to how output volumes are decided, “further work is needed to ensure the balance of benefit and the minimisation of unintended consequences”, the REF team added, saying that the delay would “allow additional time for institutions to resolve challenges with using Hesa data for REF volume purposes, and to allow testing of the approach the REF team will take when using this data to calculate average full-time equivalent staffing levels.

“Moves to break the link between individual staff members and unit submissions were welcomed by the community and this principle will be maintained,” it continued. However, it was “working to find a balanced approach” that would ensure outputs represented the institution’s research staff, which might include the reintroduction of maximum output requirements, or the use of sample testing to check if research-active staff are included in output submissions.

The minimum number of impact case studies that an institution can submit per disciplinary submission will also be reduced to one, it said, while impact case studies would no longer be required to be linked to 2* quality research.

Tim Bradshaw, chief executive of the Russell Group of research-intensive universities, said that the delay to the next REF was “welcome”.

“This is a sensible approach to allow the necessary preparation to take place in universities and ensure the robustness of the assessment. It is also positive that Research England is exploring options to test and pilot the expanded people, culture, and environment element and will review the proposed weighting based on the evidence,” Dr Bradshaw said.

“We look forward to working closely with Research England and other stakeholders on the design of the next REF, including to help develop a set of indicators to accurately assess excellence in research culture.”

jack.grove@timeshighereducation.com