The peer-review system for allocating research grants is on the brink of breakdown, it was warned this week, as research councils struggle to find enough suitable academics willing to review applications.
Research councils say they have to approach many more scholars than they did in the past to find the number and calibre of reviewers needed because the experts are saying they are too busy to undertake the work.
The research councils' grant system relies on experts agreeing to take time away from their own research and teaching, usually without payment, to review research proposals in return for others doing the same.
But evidence gathered by Times Higher Education reveals serious difficulties.
"We are teetering on the brink of the system breaking down," said Nigel Brown, former director of science and technology at the Biotechnology and Biological Sciences Research Council and now vice-principal of the University of Edinburgh. "It is not impossible, but it is getting quite difficult. People are under a lot of pressure in their day jobs, and it is just easier to say 'no'."
The rise in the number of grant applications has been met with a fall in the number of academics willing to participate in the system.
Both the BBSRC and the Engineering and Physical Sciences Research Council report that they have experienced a rise in the number of academics refusing to take part in peer review. Similar reports have come from the Medical Research Council.
This is despite the fact that most research councils have some form of "peer-review college" to provide a ready pool of experts.
"The system is creaking under the demand people are putting on it," said Alf Game, the BBSRC's deputy director of research, innovation and skills.
Dr Game said a BBSRC proposal needed at least two referees - ideally three - before it went before a panel. He said that whereas a few years ago this was achieved by approaching four people, it now took six or seven.
"It is not unknown to go to 12 people to get two responses," he said. "Very occasionally, we can't even get two."
The EPSRC said that its success rate in getting academics to review work was about 50 per cent - a drop from the 65 or 70 per cent it enjoyed three years ago.
It also said that the speed of responses to peer-review requests was a problem even though there was a monetary incentive for undertaking and returning peer reviews on time.
The BBSRC also reported that it was facing a problem with the quality of reviews it received.
"We simply can't get reviewers to be critical enough to give proposals enough differentiation," Dr Game said.
He said the research council may move in future to introduce sanctions for academics who are in receipt of BBSRC grants but decline to take part in peer review. "Sanctions are quite likely if it continues to become more difficult."
Not every council seems to be struggling in the same way. The peer-review college introduced in 2003 by the Natural Environment Research Council to improve falling response rates delivers an almost 100 per cent hit rate, said Anne McFarlane, its research funding manager.
However, Nerc's system is straining under other problems. A review published earlier this year showed that Nerc researchers were not confident that members of the peer-review college were qualified enough to identify the highest-quality proposals. Biases in the system were also identified.
Figures compiled last year by Times Higher Education showed a 13 per cent rise in the number of funding applications to all research councils in the 2007-08 financial year compared with the year before. Last year, the EPSRC received 4,758 applications for 1,442 awards. The BBSRC got 1,983 applications for 581 awards.
It can be hard to find reviewers for large grants to multiple participants because everyone who could examine the proposal is involved in it, Professor Brown said.
How two academics deal with the deluge of requests
Research councils may bemoan a lack of peer reviewers, but overworked academics struggle to balance their "community service" duties with a growing burden of teaching, research and administration tasks.
One is Claudio Stern, a professor of anatomy and developmental biology at University College London who is also an editor of the journal Mechanism of Development.
"The whole system is suffering as a result of overstretching," he said. "I try not to say 'no', but I can't accept everything. It depends on when the requests come ... there are times when I get dozens a day."
There is no alternative to peer review, he said. He would like to see academics better rewarded, but he is aware that such a move would mean cuts to research.
One of his big complaints about the system is that commercial publishers are making huge profits by using academics "free for everything". Another is that some scientists are not as committed to the system as others. "It is rather unfair," he said.
More than ten years ago, when he received a bundle of 12 grant proposals from a research council for a single funding round, John Gray made a stand.
Already overburdened, the head of the department of plant science at the University of Cambridge refused the requests and instituted a new personal rule: if the Biotechnology and Biological Sciences Research Council wanted his help, it would simply have to pick up the phone and ask him. Should it continue to send out proposals electronically, he would do them if he could; but if he could not, he would not, and the BBSRC would not get a reply.
"They refused point blank and that is how it has been ever since," he said. "I guess I have reviewed only two or three applications since."
He believes that the BBSRC has lost its personal touch. "We just get things fired off to us and are expected to drop everything and do them. They say they have not got time to telephone, but I suspect that if they phoned reviewers first, they would get a better response ... it would save them a lot of hassle."