Research produced by the bulk of the elite Russell Group of large research-intensive universities is no better than that generated by their smaller counterparts, according to a new analysis.
Data from the UK research evaluation firm Evidence show that, outside an Oxford-Cambridge-London "golden triangle", smaller research-intensive universities are performing slightly better than their larger research-intensive rivals.
The findings challenge the dominance of the Russell Group of 20 large research-intensive institutions, turn on its head the mantra that bigger institutions are automatically better at research, and also raise questions about how research funding should be concentrated.
Evidence, part of Thomson Reuters, analysed citation counts - the number of times academics' work is cited by their peers - in the 1994 Group of smaller research-intensive institutions and the Russell Group.
The study compared the percentages of all cited research papers published by the two groups between 2002 and 2006, taking into account subject field and year.
The findings reveal the full extent of the dominance of the golden triangle: papers from the universities of Oxford, Cambridge, Imperial College London, University College London and the London School of Economics were cited far more often than the world average.
Taken as a whole, the Russell Group only just produced a greater proportion of highly cited papers than the 1994 Group.
But when the five golden-triangle institutions were removed from the equation, the 19-strong 1994 Group beat off competition from the rest of the Russell Group, producing a greater percentage of highly cited papers.
Jonathan Adams, director of Evidence, said the key message was that there was an "elite group" in the system producing an "exceptional percentage of high-quality papers", but that outside the golden triangle it was difficult to distinguish between the rest.
"The 1994 Group produced only 58,000 papers, compared with the Russell Group's 200,000 papers," he noted.
About 7 per cent of papers from golden-triangle universities were cited at least four times as often as the world average. This compares with about 5 per cent of papers for the rest of the Russell Group, and more than 5.5 per cent for the 1994 Group, against a UK average of less than 4.5 per cent.
"The Russell Group has a relatively higher proportion of papers around the world average, but actually it is pretty difficult to distinguish between the 1994 Group and the Russell Group at the top end," he said.
In the wake of the proposal by Michael Arthur, vice-chancellor of the University of Leeds, that the UK should focus research funding on 25-30 institutions, Dr Adams said the study raised the question of which ones should benefit.
"Based on this analysis, the (remaining) Russell Group institutions would not stand out significantly," he said, adding that Professor Arthur's proposition fell down because it did not reflect "how the UK research base was distributed".
Dr Adams said there were two challenges facing the Higher Education Funding Council for England when distributing future quality-related (QR) research funding.
The first was to recognise that golden-triangle institutions were "disproportionately contributing" to the UK's overall research performance, so it was imperative that their funding was sustained. The second was to balance this with recognising and supporting the diversity that existed below them.
"It is not enough for Hefce to say we can just support good departments and not worry about institutions as a whole," he said. "If Hefce sets up a system that does not sufficiently sustain those institutions at the peak of excellence, which has grown with more concentrated funding under successive research assessment exercises, then it risks undermining that peak."
Commenting on the data, Wendy Piatt, director-general of the Russell Group, said citations were only "one of many" indicators which, when used in combination, help to compare the performance of universities across the world.
She added that a comparison of the Russell Group without a quarter of its members was "very curious", and pointed out that Professor Arthur had suggested that the top-five institutions should receive a "larger share" of cash.
"It is essential that the UK continues to support our leading research-intensive universities through the tough economic climate to keep the UK competitive," she said.
Paul Wellings, chair of the 1994 Group and vice-chancellor of Lancaster University, said the data "show that 1994 Group universities are second to none".
FEEL THE QUALITY: BIRMINGHAM V-C STANDS UP FOR QR
Block grant funding gives institutions the freedom to undertake risky and ground-breaking research that may otherwise go unsupported.
That is the message from Universities UK in a report released this week, which argues for the importance of quality-related (QR) research funding, provided to universities on the basis of the former research assessment exercises.
The report, Securing world-class research in UK universities: exploring the impact of block grant funding, follows a warning last month from David Eastwood, former chief executive of the Higher Education Funding Council for England, and vice-chancellor of the University of Birmingham.
He said "constant vigilance" was needed to protect QR, arguing QR had already been rebalanced in favour of research council funding and further losses would be inappropriate.
"Constant vigilance is necessary on behalf of all those who fight the funding cause, because QR from the point of view of the Treasury ... looks like a very particular form of investment," he told a conference at the Royal Society run by the Higher Education Policy Institute.
Times Higher Education has partnered with Thomson Reuters to produce its annual World University Rankings from 2010 and beyond. To help us create a new, more rigorous and transparent methodology, visit http://bit.ly/ErAag to fill in a survey and leave your comments.