REF 2029 downgrades research culture, restores five output limit

Radical plans to allow researchers to submit an unlimited number of outputs to the next Research Excellence Framework (REF) have been abandoned, with the exercise moving closer to rules and policies used in the 2021 exercise.

In a reversal of planned reforms to the exercise, by which the cap on submissions would have been abolished, scholars will now be limited to submitting a maximum of five outputs.

Proposals to remove the requirement for REF-eligible researchers to submit at least one output in the seven-year assessment period have, however, been retained, with institutions required to submit 2.5 outputs for every full-time equivalent researcher employed during 2025-26 and 2026-27.

In another significant move, the REF will downgrade the weighting attached to “people, culture and environment” (PCE), now renamed “strategy, people and research environment”, from a proposed 25 per cent to 20 per cent.

The weighting of the main “quality” component of REF 2029, known as “contributions to knowledge and understanding” (CKU), will be increased to 55 per cent, having been reduced to 50 per cent from a 60 per cent weighting in 2021. Research impact, known as “engagement and impact”, will remain at 25 per cent, according to an update announced on 10 December.

Rules on portability will also change following criticism from researchers in the arts and humanities, who objected to their outputs remaining with their employer even though they may have left the role during the two-year census window. Long-form outputs, such as monographs, will be granted an exception to allow them to remain attached to a researcher for up to five years.

The overhaul of REF policies follows a three-month pause to the exercise announced by science minister Patrick Vallance in early September amid growing criticism over what some universities believed was the excessive weighting attached to PCE. Some university leaders argued that the diminished focus on academic excellence in the form of a reduced weighting for quality could be politically damaging for research.

The pause was announced less than a month after Ottoline Leyser, a major champion of REF reform, stepped down as chief executive of UK Research and Innovation.

Concerns were also raised over the feasibility of using metrics to assess PCE. Results of a 18-month pilot to test how PCE could be measured have been published alongside the updated guidance.

Vivenne Stern, chief executive of Universities UK, said the organisation “strongly welcomes this pragmatic approach, which shows that the four UK funding bodies have listened carefully to researchers and universities”.

“This matters because university research shapes our daily lives, and the REF helps funders target investment where it will drive the next generation of discoveries,” said Stern.

REF 2029 director Rebecca Fairbairn said the “changes confirmed today represent an effective way to incentivise excellence and align with government ambitions for research and innovation across the four nations”.

“This collaborative process has ensured the REF is grounded in operational realities and responsive to the rapidly changing context for higher education,” she added.

Panels for REF 2029 will begin meeting in early 2026 to set criteria and an institutional-level working group for strategy, people, and research environment will be established. According to the REF organisers, weightings for this component will be split 60:40 between institutional-level and unit-level statements.

Panels will be asked to develop the final criteria and template – the same approach used in REF 2021 – by drawing on the indicators developed in the PCE pilot.

To reduce burden, the assessment of impact has been broadly retained from REF 2021 and unit-level statements have been removed from both impact and CKU.

In line with the initial decisions from June 2023, outputs will remain decoupled from individuals and impact case study requirements are reduced to one for the smallest units. The 2* qualifying threshold for underpinning research has also been removed.

To maintain the original REF 2029 timetable, there will be no formal consultation on the guidance or panel criteria and working methods.

In an opinion piece published on 10 December, Helen Cross, deputy chair of the REF Steering Group and director of research and innovation at the Scottish Funding Council, said the “changes [are] designed to bring clarity, reduce burden, and create space for what matters most: enabling excellent research that drives innovation, supports economic growth, and works to address the many challenges that the world is facing”.

On reverting to 2021 rules on maximum output levels, Cross said: “The sector told us clearly that, in a decoupled framework, guard rails were needed to support inclusive and equitable practice. In response, we’re reintroducing a recommended maximum of five outputs per researcher, consistent with REF 2021.”

“This restores clarity and fairness, while reducing pressure on individuals by removing the minimum requirement...we want to strike a balance: encouraging representativeness, supporting equity and maintaining confidence in the process,” she said.

Science minister Patrick Vallance welcomed the decision to increase the weighting for academic excellence to above 50 per cent.

Speaking at a Universities UK conference shortly after the changes were announced, Vallance said he was “very pleased that the percentage [attached to quality] has gone up”, noting the research excellence element was essential for public and political trust in the REF which is used to allocate some £2 billion in research funding annually.

“That is a fundamental thing that REF has to do,” said Vallance, explaining this reassurance is particularly important for fundamental research.

“The only question that a science minister should ask [of a piece of research] is: ‘Is this world-class?’.”