University funders have been warned not to allow the research excellence framework results to “impoverish the humanities and social sciences” after medicine and life sciences scored particularly well in the exercise.
Main panel A, which covers medical and life sciences, had the highest overall proportion of 4* research compared with the other three main panels, which cover physical sciences, social sciences, and the arts and humanities.
The result has led to concerns that the performance of medical and life sciences could divert quality-related research funding towards universities that carry out a large amount of work in these areas – although such institutions could still decide to invest their money across the board.
The uplift was driven largely by assessors rating 60 per cent of submissions to main panel A as 4* for impact and environment, which account for 20 and 15 per cent of scores respectively. This compares with 37 to 39 per cent rated as 4* for impact, and 38 to 41 per cent rated as 4* for environment on the other three main panels. According to analysis by one unnamed university, the amount of research in main panel A that was rated 4* grew by 140 per cent since the 2008 research assessment exercise. For the other panels combined, this rise was 51 per cent.
£6 billion shot in the arm
It has also been suggested that additional investment in the medical and life sciences of about £6 billion from a national research funder during the REF period is likely to have contributed to the rise in research quality.
Richard Black, pro director of research and enterprise at Soas, University of London, said he was concerned about whether a “narrow” view of impact could lead to a concentration of funding.
“The high rating of impact in main panel A does appear, on the face of it, to have driven up the assessment of overall research quality in life sciences,” he said.
“It would be to the detriment of all if the result were to impoverish the humanities and social sciences, both of which are critical to societal well-being, and have major, if indirect, impacts on the public good.”
How much universities with a significant presence in higher REF-performing subjects stand to benefit when the annual £1.7 billion of QR funding is next distributed remains to be seen.
David Sweeney, director of research, education and knowledge exchange at Hefce, said the size of the pot of money awarded to each subject area was predetermined. But that could change, he added: “It is fair to say the funding will follow the results.”
Stephen Holgate, chair of REF main panel A, said that “if funding matches the profiles panel by panel, then the medical and life sciences will proportionately do well”.
Professor Holgate, who is also clinical professor of immunopharmacology at the University of Southampton, said that the uplift in research quality in the subjects was down to a change in the way that the government distributed funding to medical schools just before the REF period began.
In 2006, Dame Sally Davies, who was director of research and development in the NHS at the time, stopped a research funding stream to medical schools over fears that it was being used to cover patient costs not science, said Professor Holgate.
That year, she created the National Institute for Health Research, which awards the £1 billion a year research funding from the Department of Health to English NHS organisations that are mostly affiliated with a university. In 2013-14, 22 per cent of this money went on research programmes assessed by international peer review, 66 per cent on research infrastructure and 10 per cent on research training, according to the NIHR.
This “produced a big kick” in the quality of clinical and applied research that coincided with the 2008-14 research assessment period, said Professor Holgate.
“The reason our impacts are so high is because this [NIHR] money is about transforming research into patient benefits,” he added. Other changes that occurred around the same time also had an effect, he said.
The Medical Research Council focused on translational research, and two cancer research charities merged in 2002 to create Cancer Research UK, which began to invest more in the quality of research, he added.
Strength in numbers: market share of South-East regional alliances grows
The performance of regional university alliances in the research excellence framework has raised questions about the possible concentration of funding in the golden triangle of London, Cambridge and Oxford.
All regional research alliances boosted their overall grade point average scores in the 2014 REF compared with the research assessment exercise in 2008, according to an analysis by Times Higher Education. But only those in the South East managed to increase their market share of research power (calculated by multiplying GPA by the number of people submitted) at the same time.
There are now five regional alliances: the N8 group of northern universities, the Midlands-based M5, the GW4 in the South West, the Science and Engineering South Consortium, and the Eastern ARC.
Calculations by THE using the REF data reveal that all five alliances increased their overall GPA score by between 0.37 and 0.49 over the past six years. The N8 group had the most marked drop in market share, 1.9 percentage points, while SES universities saw the biggest boost in market share, 1.6 percentage points.
Data on the proportion of eligible staff included in the REF show that overall the N8 universities submitted the lowest proportion, 76 per cent. SES, by contrast, submitted the highest, 91 per cent.
David Price, vice-provost for research at University College London, said it was a “huge management error” for institutions to exclude staff, referring to the fact that funding is in part determined by the number of staff submitted to the REF.
Kieron Flanagan, senior lecturer in science and technology policy at the University of Manchester’s Business School, suggested that the golden triangle was becoming a “dominant scientific labour market” and that the South East’s “magnetic pull” would only grow if the funding formula is unchanged.