UK Research and Innovation, Research England, Higher Education Funding Council for Wales and the Scottish Funding Council Research Excellence Framework 2028: Initial decisions and issues for further consultation

Initial decisions on the next national research assessment exercise, REF2028, have been published, following the collection of a broad body of evidence by the Future Research Assessment Programme (FRAP), on behalf of the four UK higher education funding bodies. The redesign is a bid to reshape incentives within the research system by expanding the definition of research excellence and ensuring that “appropriate recognition is given to the people, cultures and environments that underpin the research system. Submissions to REF2028 will be made in late 2027 and outcomes will be announced in December 2028. Institutions have until midday on 6 October this year to respond to the plans.

The full consultation document outlining the proposals can be found here

At-a-glance:

• The 2021 rule that staff with research responsibilities have to submit at least one research output, and up to a maximum of five outputs, has been scrapped. Universities will be required to submit 2.5 outputs per FTE of staff in each disciplinary area where they have activity. Institutions can submit outputs produced by any staff member where there is a “demonstrable and substantive link” to the submitting institution. How this link can be shown will be the subject of consultation (p21)
• Given there is no minimum output requirement for volume-contributing staff there is no need for a process to account for individual equality-related circumstances (p21)
• Rules on the submission of staff to the next REF have also changed. Institutions will no longer submit a list of staff deemed research active to the assessment exercise on a specific date. Instead, from 2024-25, universities will supply HESA with data on research-active staff. Research England will then draw this data for 2025-26 and 2026-27 to decide the volume of full-time-equivalent research staff at an institution. This will then inform the number of outputs required for submissions in late 2027 (p14)
• Staff employed on at least a 0.2 FTE basis for at least six months in this two-year REF assessment period would be eligible to submit outputs “where a link to the institution can be demonstrated”
• The three assessment elements used in 2021 to judge research - ‘outputs’, ‘impact’ and ‘environment’- have been renamed and the weightings attached to each element changed. They are now “contribution to knowledge and understanding”; “engagement and impact”; and “people, culture and environment” (p8)
• For REF2028, research outputs will make up only 50 per cent of an institution’s overall score, down from 60 per cent in 2021 and 65 per cent in 2014 (p10)
• At least 10 per cent of this will be decided on the basis of a “structured statement” in which universities will “outline their wider contribution to knowledge and understanding in the disciplinary area, supported by evidence and data”.
• Impact and engagement will maintain the 25 per cent weighting of the previous impact element (p11). People, culture and environment will constitute 25 per cent of the overall score, up from the 15 per cent allocated to environment in 2021 (p9)
• Disciplinary-level environment statements have been retained. Both institutional-level and discipline-level statements will be used to decide environment scores, using a less burdensome questionnaire-style template. The metrics used to calculate the score have not yet been decided and will be consulted on (p9)
• The funding bodies will invite the expert panels to review the ‘star ratings’ to ensure they accurately reflect an expanded understanding of excellence (p8)

Implications for governance:

As governors will know, the outcomes of the REF have a direct impact on university finances and reputation, as they inform the allocation of around £2bn of block-grant research funding each year, feed into various league table metrics, and provides an opportunity for institutions to showcase their research work to a wider audience through media coverage.

Any redesign of the exercise can have a bearing on which institutions and groups of institutions emerge as “winners and losers” of the research assessment.

The changes proposed to REF2028 build on shifts started in 2021. According to the REF2028 early decisions statement, they aim to “recognise and reward a broader range of research outputs, activities and impacts, as well as reward those institutions that strive to create a positive research culture.”

They seek to encourage a more “team-based” approach to assessing research that values the contribution of all research and “research-enabling” staff and breaks the link between the individual and the REF submission.

By scrapping the minimum and maximum output requirements for individual staff members, the value of the “star professor” appointment becomes arguably less alluring and the disincentives for institutions to hire early career researchers or recruits from industry, without track records of academic output, are reduced. The plan to include outputs produced by anyone employed by the institution on a minimum 0.2 FTE contract for at least 6 months allows staff to move between institutions and sectors with no detriment to their careers or the institution hiring them.

Another stated aim is to support and reward “a diversity of research outputs”, not just journal articles, which currently account for over 80 per cent of the outputs submitted to REF 2021, with non-text outputs making up less than 3 per cent of submissions.

According to the funding bodies, there are important output types that contribute to the wider infrastructure of research fields that are important contributions in their own right and enable the research of others, for instance review articles, meta-analyses, replication studies, datasets, software tools, reagents, video outputs, translations and critical editions.

This attempt to broaden what research outputs look like is aimed at addressing the tendency to favour short-term, measurable outputs and impacts that are likely to do well in the REF, at the expense of longer-term research or projects with less tangible outputs and impacts.

In their oversight of research, governing boards might be interested in the extent to which these diverse forms of output are a feature of their own institutions research programmes. These changes to output, coupled with elevation of the environment element of the assessment, may be seen as particularly improving post-92 institutions’ chances of being recognised for their research work.

The criteria REF2028 uses to assess research culture will be consulted on, but the document suggests they could include equality, diversity and inclusion data; quantitative or qualitative information on the career progression of current and former research staff; outcomes of staff surveys; data around open research practices; and evidence of research robustness and reproducibility.

The prominence REF2028 will give to research culture provides an opportunity for governors to look at their own research environment, the data that illuminates it, and the processes that support it, to consider if measures need to be taken to enhance them.

The funding bodies are keen to simplify and reduce the administrative burden around REF2028. For instance, the minimum requirement of two impact case studies in REF2021 has been reduced to one for units with fewer than 9.99 FTE, and the minimum 2* requirement for research underpinning impact case studies has been removed. Questionnaire-style collections of information are also planned to reduce the preparation and assessment burden.

In general, breaking the link between individuals and REF submissions should make the exercise less of an administrative burden. However, governors will know from previous iterations that any changes to REF processes and evidence requirements can be challenging and require resources and staff time.

Institutions are encouraged to take part in ongoing consultations on the outstanding practical details of REF2028, such as how to measure the research culture and environment. View are also sought on the proposed approach to the volume measure; the impacts and potential unintended consequences concerning output submissions; changes to impact case study requirements; and how to take into account any impacts on research of the Covid-19 pandemic.