Table of Contents
Summary
The Research Excellence Framework (REF) is used to allocate the majority of “Quality-related Research” funding (core, flexible funding) to universities and aims to increase research quality.
The REF has been widely criticised for being bureaucratic and expensive, doubling in costs for REF 2021 to £471 million (annualised, this comes to five per cent of the budget which Research England actually distributes on the basis of the REF).
Recent evidence suggests that the REF is ineffective at improving research quality. Instead, it worsens research quality close to submission deadlines and may discourage interdisciplinary research. A majority of researchers believe that the REF discourages “blue sky research, which may yield unpredictable outcomes”.
Given a “22 billion fiscal black hole”, the Chancellor has instructed departments to find cost-savings, while universities risk closures and are making cuts to the arts and humanities amidst a funding crisis.
Adam Tickell’s Independent Review of Research Bureaucracy recommended that the government reduce the assurance burden on R&D institutions, and the new Secretary of State for Science, Innovation Technology has committed to cutting red tape for researchers.
A survey of UK researchers shows that a majority view the REF negatively because they feel the bureaucratic burden outweighs any benefit, with researchers in the arts and humanities viewing the REF most negatively. The selection process for REF submissions may also reinforce gender inequality in academia.
The new Labour government should follow through with its commitment to cut red tape for researchers, support universities and save money by scrapping REF 2029. REF should be replaced with a low-bureaucracy system where Quality-related Research funding is allocated to institutions in proportion to “external research income”, won by their researchers through grants from the private, public and philanthropic sectors.
In a scenario where REF 2029 costs do not increase further beyond REF 2021, this would save approximately £17 million for research councils and free up £430 million for universities.
This alternative system focused on “external research income” would incentivise collaboration with local industry, promoting technological diffusion and regional productivity growth, which could reduce regional inequality.
Increased collaboration with industry will also boost research financial sustainability, since private sector grants cover a greater proportion of the “full economic costs” of research than grants from the public or philanthropic sectors.
Since arts and humanities researchers are currently less able to obtain external research income, £48 million should be set aside from Research England’s £1.636 billion QR budget for a 25% increase in the budgets of the Arts and Humanities Research Council and the Economics and Social Research Council (providing an additional £17.5 million and £30.5 million respectively).
A “floor” for funding should be established, in proportion to the number of full-time equivalent research staff at the university.
Alongside allocating university funding, the REF aims to improve research quality, but is ineffective at this. Without the REF, competition for jobs and grants, alongside post-publication peer review, will continue to improve research quality. Additionally, budgets saved from scrapping the REF should be used to realise cost-savings for DSIT, and support other projects aiming to improve research quality by:
Expanding the budget of the DSIT-UKRI Metascience Unit, to identify cost-effective, evidence-based methods to improve research quality.
Supporting the UK Reproducibility Network to enable grassroots culture change, encouraging rigorous and transparent research.
A massive reduction in bureaucracy will give academics more time and freedom to pursue undirected, “blue sky” research. To further support this type of research, the Minister of State for Science, Research and Innovation should instruct UKRI to:
Adopt recommendations from UKRI’s “Review of Peer Review” to reduce the administrative burden of grant applications on researchers.
Expand schemes which provide long-term funding for “people, not projects”. This would also reduce career “precarity” and support equality, diversity and inclusion.
The Challenge
The Research Excellence Framework (REF) guides the distribution of “Quality-related Research” (QR) funding to universities (worth £1.3 billion annually in England) and aims to improve the return on investment from the government’s £9 billion R&D budget by incentivising higher research quality. First introduced in 1986, QR funding provides “core funding” to universities, providing flexible support (in contrast to grants won from UK Research and Innovation’s Research Councils).
However, the REF has been widely criticised for being overly bureaucratic and expensive. In 2006, Gordon Brown considered scrapping the predecessor system to the REF, the Research Assessment Exercise, due to similar concerns. In the 18 years since, these problems have escalated. REF 2021 doubled in cost from 2014, costing £17 million to UK Research and Innovation and £430 million to universities, which hire “impact case officers”, research assistants and external consultants specifically to prepare REF submissions. The annualised costs of the REF come to five per cent of the budget actually distributed on its basis by Research England.
Previous research on REF 2014 suggested the REF may have increased research quality at British universities in comparison to American competitors. But the latest evidence brings this into question, showing that Germany, France and Italy saw similar increases in research quality compared to Britain, despite their lack of a REF-like evaluation process. In fact, recent evidence also suggests that in some cases, the REF may reduce research quality – the REF process requires academics to submit research outputs to expert evaluation panels, and submission deadlines incentivise unrigorous and low-impact research. An evaluation of the REF commissioned by Research England suggests that it may also discourage interdisciplinary research.
A survey commissioned by Research England found that a majority of researchers consider the REF to have a negative impact on their work because the bureaucratic burden (in terms of workload, time and money) outweighs purported benefits. Researchers in the arts and humanities were most likely to have a negative view of the REF. Respondents felt that the main impact of the REF was to promote “game playing”, where universities optimise for REF performance with unintended negative consequences for research quality and staff development. The selection of research for evaluation under the REF process may reinforce gender inequality in academia, and aspects of the REF process have led to perceived bias against the humanities. The University and College Union is in favour of scrapping the REF.
Due to financial pressures, multiple universities have made cuts to the arts and humanities and some may face the risk of closure - polling indicates the public would blame the government if closures do come to pass. Meanwhile, the Treasury’s “22 billion fiscal black hole” means that the Department for Science, Innovation and Technology must urgently find cost-savings. The REF poses a significant expense to universities and the government, without commensurate benefits.
In March 2024, New Zealand abruptly cancelled the “Performance-Based Research Fund”, its REF-like evaluation process of universities, recognising that marginal benefits were small relative to the large bureaucratic burden. In 2022, Australia cancelled the 2023 round of its similar “Excellence in Research in Australia” process, promising a move to a “modern data driven approach”. In the UK however, the REF is set to continue – the next exercise will require universities to provide their submissions in 2027, with results published in late 2029. Initial plans for REF 2029 came under severe criticism for potential negative effects on the humanities.
A majority of researchers believe the REF discourages “blue sky research, which may not yield predictable outcomes”. This problem is exacerbated by UKRI’s dominant grant-making system, which primarily offers short-term funding for specific research projects. Compared to approaches where individuals and teams receive long-term funding, UKRI’s “project-based” approach can increase the assurance burden on researchers, reduce autonomy and discourage undirected, blue sky research. A lack of long-term funding also contributes to “precarity” in academic careers, with negative effects on equality, diversity and inclusion (EDI). Time-consuming grant applications with low acceptance rates also contribute to red tape. A famous Australian study found that 550 years of researcher-time could go into a single grant round, and the Economist has argued that scientists spend too much time “looking for money”.
Labour’s Industrial Strategy highlights the importance of collaboration between universities and businesses, and Labour’s Start-Up Review explored how university spin outs could be supported to promote regional productivity growth. In 2017, the Conservative government’s interest in the role of universities in industrial strategy led to the creation of the “Knowledge Exchange Framework”. This comprises metrics which aim to measure the contribution of universities to society, including income from collaboration with businesses and the rate at which students and graduates create start-ups. However, the Knowledge Exchange Framework does not play a role in funding allocation, so universities have little incentive to improve their performance on these metrics.
UKRI supports research via grants which officially cover 80% of the “full economic costs” of research, which includes both direct and indirect costs. Universities must find the remaining 20% from other sources, such as their QR funding or teaching income. Philanthropic grants cover an even smaller proportion of the full economic costs of research, which came to an average of 58% in 2022/23. This system has led to concerns about the sustainability of research funding, since universities are currently using income surpluses from teaching to cover a £5.3 billion “research deficit”.
For large-scale systemic reform, such as reforms to the university funding system, it is not feasible to generate perfect, real-world evidence before implementation. The size and complexity of the REF makes it difficult to conduct pilot studies or controlled experiments for alternative systems without significant disruption. In this case, reforms must be based on theoretical reasoning, expert analysis, practical considerations and evidence with a narrower scope, focused on specific aspects of university funding.
The Opportunity
Adam Tickell’s Independent Review of Research Bureaucracy recommended that the government reduce the assurance burden on R&D institutions, and the new Secretary of State for Science, Innovation Technology has committed to cutting red tape for researchers. In opposition, Labour refused to commit to retaining the REF in its current form, and in government, Labour has indicated openness to reform.
Building on these ideas, REF 2029 should be scrapped. To minimise bureaucracy, the funding allocation function of the REF should be served by a new, simple system where universities declare their external research income (total grant winnings from the public, philanthropic and private sectors) and are awarded QR funding in proportion to this.
Most but not all QR funding is distributed based on the REF. In England, the £114 million “QR business research element” is distributed based on income from private sector grants, and the £219 million “QR charity support fund” is distributed based on income from philanthropic grants. Allocations via the QR business research element (Figure 1) and the QR charity support fund (Figure 2) are strongly correlated with funding distributed based on the REF (also known as “Mainstream QR Funds”). The Scottish Funding Council, the Higher Education Funding Council for Wales and the Department for the Economy (Northern Ireland) also already consider income from philanthropic grants when allocating QR funding. External research income is associated with greater impact and higher quality of research environment in the REF, and current plans for REF 2029 will increase the weighting given to “people, culture and environment”. While universities may respond differently to the incentives under a new system, and any reforms to the system will create winners and losers, the available evidence suggests that a system focused on external research income will result in a QR funding allocation similar to existing plans for REF 2029, but without the associated bureaucracy.
Figure 1: The associated between funding distributed based on the REF and funding distributed based on private grants (England)
Data for 2023-24, obtained from UKRI. Statistical testing using Shapiro-Wilk normality test and Pearson product-moment correlation. Correlation is statistically significant using a significance level of p = 0.05.
Figure 2: The associated between funding distributed based on the REF and funding distributed based on philanthropic grants (England)
Data for 2023-24, obtained from UKRI. Statistical testing using Shapiro-Wilk normality test and Pearson product-moment correlation. Correlation is statistically significant using a significance level of p = 0.05.
As part of these reforms, budgets currently distributed by Research England through the QR charity support fund and QR business research element should also be distributed based on total external research income. This means Research England would distribute close to £1.636 billion (including the £1.303 billion previously distributed based on the REF) based on external research income.
Incentivising universities to focus on increasing external research income will encourage universities to support spin outs, while driving greater collaboration with local industry in line with Labour’s Industrial Strategy, due to a strengthened incentive to obtain grants from the private sector. Increased collaboration with local industry aligns with recommendations from the Paul Nurse review of the research landscape, and Richard Jones’s report for the Productivity Institute, which highlights the need for technological diffusion to drive regional productivity growth. This could lead to a reduction in regional inequality. The role of this new system in promoting spin outs and regional growth means that it should also replace the “Knowledge Exchange Framework”, further reducing costs for DSIT.
Increased collaboration with industry will also boost research financial sustainability, since private sector grants cover a greater proportion of the full economic costs of research (76% in 2022/23) than grants from UKRI’s research councils (69% in 2022/23) and the philanthropic sector (58% in 2022/23).
Researchers in the arts and humanities are more reliant on QR funding and under the status quo may be less able to obtain external research income than STEM researchers. To ensure arts and humanities research is supported, £48 million of Research England's £1.636 billion QR budget should be set aside to enable a 25% increase in the annual budgets of the Arts and Humanities Research Council and the Economics and Social Research Council (providing an additional £17.5 million and £30.5 million respectively). Compared to the REF, increasing the budgets of relevant research councils is a far more efficient approach to supporting the arts and humanities, since approximately £471 million of wastage is avoided. Researchers in the arts and humanities may also favour this approach compared to the REF, since they have a more negative view of the REF than researchers in other disciplines. This change would also offer universities in Scotland, Wales and Northern Ireland an opportunity to win this funding, further supporting regional equality.
Under the new system, a “floor” for funding to universities should be established, in proportion to the number of full-time equivalent research staff, a factor which the Department for the Economy (Northern Ireland) already considers in its QR allocation formula.
Even without the REF, competition for grants and jobs alongside post-publication peer review will continue to drive greater research quality across all disciplines. However, to further improve research quality, some savings for Research Councils from scrapping the REF should be allocated to the DSIT-UKRI Metascience Unit and the UK Reproducibility Network, with the rest realised as cost savings for DSIT.
In contrast to the REF, the DSIT-UKRI Metascience Unit takes an evidence-based approach to improving the quality of R&D funded by the taxpayer, generating evidence on the most effective ways to improve research integrity, reproducibility and impact. Coupling this evidence based approach with cost-effectiveness analyses will lead to interventions which improve research quality in a more effective manner than the REF, without imposing the same bureaucratic burdens on universities.
The UK Reproducibility Network, supported by Research England (the body which distributes QR funding to English universities), is a grassroots initiative at UK universities promoting higher-quality research. Their schemes include “ReproducibiliTea” and “RIOT Club”, networks of local clubs at universities where early-career researchers discuss efforts to improve research quality and provide training on key research quality issues. This grassroots approach to improving research quality minimises bureaucratic burden.
A massive reduction in bureaucracy will give researchers more time and space to pursue undirected, blue sky research. Research demonstrates that long-term funding for “people, not projects” can give researchers further time and space to pursue this type of research, and the Royal Society already runs the University Research Fellowship, which provides long-term flexible funding for promising researchers. Labour’s commitment to ten-year funding settlements for UKRI could help it run similar schemes, providing long-term funding for ten or more years to further cut red tape for researchers and support undirected, blue sky research.
To cut red tape further still, the Minister for Science, Research and Innovation should instruct UKRI to adopt recommendations from the “Review of Peer Review” report, commissioned by UKRI itself, which would reduce the administrative burden on grant applicants. This includes reduced grant application form lengths, the use of “partial randomisation” or funding lotteries and two-stage application processes. The British Academy and the Natural Environment Research Council are already trialling partial randomisation, and Nesta has highlighted the potential benefit of this approach in reducing the time researchers must spend filling out grant applications. A reduction in grant application lengths could be delivered by a 10 A4 page cap for most grant applications – ARIA currently accepts grant proposals at this length. These reforms will further reduce bureaucracy, and create further time and space for blue sky research.
Plan of Action
In line with his stated commitment to cut red tape for researchers, the Secretary of State for Science, Innovation and Technology should immediately announce the cancellation of REF 2029 and the “Knowledge Exchange Framework”, and their replacement with a system where QR funding is distributed in proportion to external research income.
£48 million of Research England’s £1.636 billion budget to be distributed based on external research income should be set aside for a 25% increase in the budgets of the Arts and Humanities Research Council and the Economics and Social Research Council (£17.5 million and £30.5 million respectively).
A floor for QR funding should be established, in proportion to the number of full-time equivalent research staff at the university.
Cancelling the REF 2029 will save money for Research Councils. While the amount is difficult to predict, in a scenario where REF 2029 costs would not have increased further beyond REF 2021, £17 million would be saved. Some of this funding should be reallocated to the DSIT-UKRI Metascience Unit and the UK Reproducibility Network, with the remainder realised as cost-savings for DSIT.
The Minister of State for Science, Research and Innovation should direct UKRI to:
Build on promised 10-year funding settlements by expanding programs which fund “people, not projects”, to cut red tape even further, giving researchers time and space to conduct undirected, blue sky research.
Adopt recommendations from UKRI’s “Review of Peer Review” to reduce the administrative burden of grant applications on researchers.
Cost
Scrapping the REF would be cost-free and save money. In a scenario where REF 2029 costs would not have increased further beyond REF 2021, the new system focused on external research income would save approximately £17 million for research councils, some of which should be realised as cost savings for DSIT. Approximately £430 million would be saved by universities.
FAQs
How will these reforms affect different types of research?
The total volume of research will increase because:
Taking the costs of REF 2021 as a lower bound, approximately £471 million will be saved by cancelling REF 2029, which could be used towards more research.
Reformed incentives will cause universities to obtain additional research funding from the private sector.
Time and money will be saved by cutting bureaucracy in grant applications, which could be used towards more research.
Table 1: Effects of proposed reforms on the volume of different types of research
What about non-STEM research, such as in the arts and humanities?
Recent changes to REF rules led to reduced funding for the humanities, and researchers in the arts and humanities have the most negative view of the REF.
Currently, stretched university finances are leading to cuts in arts and humanities departments. A reduced bureaucratic burden on universities from replacing the REF will enable approximately £430 million of cost-savings. Instead of hiring “impact case officers”, research assistants and external consultants to prepare REF submissions, this money could be used to protect jobs in arts and humanities departments.
Arts and humanities researchers are currently more reliant on QR funding and are less able to obtain external research income than STEM researchers. To correct this and strengthen the incentives of universities to pursue arts and humanities research, £48 million of Research England’s £1.636 billion budget to be distributed based on external research income should be set aside, enabling a 25% increase in the budgets of the Arts and Humanities Research Council and the Economics and Social Research Council (providing an additional £17.5 million and £30.5 million respectively).
What about blue sky and basic research?
A majority of researchers believe that the REF discourages “blue sky research, which may yield unpredictable outcomes”. A massive reduction in bureaucracy from scrapping the REF and reforming UKRI’s grant application processes will give researchers more time and space to pursue undirected, blue sky research. To further support this type of research, new 10-year funding settlements for UKRI should be used to provide long-term funding to the most promising individual researchers and research groups.
Basic, curiosity-driven research will also continue to receive support via grants from the public and philanthropic sectors. External research income will include grants won from the public and philanthropic sectors, so under the new system, universities will still have incentive to conduct basic research.
How will we improve research quality without the REF?
The REF is ineffective at improving research quality. The REF may even reduce research quality by incentivising unrigorous and low-impact research close to deadlines. Without the REF, competition for jobs and grants, alongside post-publication peer review, will continue to raise research quality across disciplines. Research quality can also be improved via specific interventions identified as effective by the DSIT-UKRI Metascience Unit, and by grassroots culture change driven by the UK Reproducibility Network. Budgets saved from scrapping the REF can be reallocated to these initiatives to improve research quality.
The DSIT-UKRI Metascience Unit aims to generate evidence to support a more effective innovation system. The scope of its £5 million grants programme includes projects which could “inform practice within research-performing organisations”, such as projects aimed at improving research training and skills, research integrity, research reproducibility and research impact.
The UK Reproducibility Network, supported by Research England (the body which distributes QR funding), is a grassroots initiative at UK universities promoting higher-quality research. Their schemes include “ReproducibiliTea” and “RIOT Club”, networks of local clubs at universities where early-career researchers discuss efforts to improve research quality and provide training on key research quality issues. This grassroots approach to improving research quality minimises bureaucratic burden.
What about research financial sustainability?
Currently, under a system focused on the REF, QR funding allocations are not directly influenced by grant income from UKRI’s Research Councils in any of the four nations, while incentives to win philanthropic grants are stronger than incentives to win grants from the private sector.
However, in 2022/23, a greater proportion of the “full economic costs” of research were covered by grants from the private sector (76%) and UKRI’s research councils (69%) than grants from the philanthropic sector (58%).
Replacing the current system for QR funding allocation with a system focused on external research income, which includes private, public and philanthropic grants, will strengthen universities’ incentives to obtain grants from the private and public sectors, boosting research financial sustainability.
Would a system based on external research income lead to ‘game-playing’?
Respondents to a survey commissioned by Research England felt that the main impact of the REF was to promote “game playing”, where universities optimise for REF performance with unintended negative consequences for research quality and staff development. There is evidence that game playing in the REF leads to worse research quality close to REF submission deadlines.
Currently, most but not all QR funding is distributed based on the REF. For example, in England, the £219 million “QR charity support fund” is distributed based on income from philanthropic grants, and the £114 million “QR business research element” is distributed based on income from private sector grants. Income from philanthropic grants also influences QR funding allocations in Scotland, Wales and Northern Ireland. This means incentives to obtain external research income already exist. Universities will have incentives to ensure that external research income consists of genuine grants which support research, since failing to do so would harm their reputations and open them up to accusations of fraud.
How do academics feel about the REF?
A survey commissioned by Research England found that a majority of researchers consider the REF to have a negative impact because the bureaucratic burden (in terms of workload, time and money) outweighs any benefits. The University and College Union is in favour of scrapping the REF.
What do academics have to do for the REF?
As part of the REF, academics have to: select and prepare their research outputs for evaluation, develop impact case studies which demonstrate how their research has benefited society beyond academia, gather and present evidence to demonstrate that their research is high quality and high impact, engage in internal reviews to prepare for REF submission, participate in internal “mock REF” exercises, attend internal training events on the REF process and keep up-to-date with changes to the REF process. This bureaucracy leads to the REF being viewed negatively by a majority of researchers.
What are other countries doing in this area?
In March 2024, New Zealand abruptly cancelled the “Performance-Based Research Fund”, its REF-like evaluation process of universities, recognising that marginal benefits were small relative to the large bureaucratic burden. In 2022, Australia cancelled the 2023 round of its similar “Excellence in Research in Australia” process, promising a move to a “modern data driven approach”.
Acknowledgements
Thank you to Tym Syrytczyk (UK Day One) for data visualisation and statistical analysis.
Sanjush Dalmia
Sanjush Dalmia is the Science Policy Lead at UK Day One. Previously, he was a Policy Advisor to the Shadow Minister for Science, Research and Innovation. He holds a Master of Research in Medicine degree which focused on ageing. He is in the process of completing his medical training in Yorkshire.
Richard Carr
Richard Carr is an Associate Professor in Public Policy and Strategy at Anglia Ruskin University. Previously, he worked in the think tank and public policy sphere. He is the author of several books including Alice in Westminster: The Political Life of Alice Bacon (2016, co-authored with Rachel Reeves MP), March of the Moderates: Bill Clinton, Tony Blair, and the Rebirth of Progressive Politics (2019) and Britain and Ireland from the Treaty to the Troubles: Independence and Interdependence, c.1921-1973 (2025). He writes in a purely personal capacity.
Contact Us
For more information about our initiative, partnerships, or support, get in touch with us at hello@ukdayone.org