The seven research councils responsible for a £2.8 billion annual UK research budget have rejected plans for a new system to replace the research assessment exercise as "not acceptable in their current form".
Research Councils UK, which represents the seven subject-specific research councils, has taken the unusual step of making public criticisms of the Higher Education Funding Council for England's plans to replace the RAE with a new research excellence framework.
The research councils and Hefce are responsible for administering the two strands of the so-called dual-support research funding system. Research councils provide grants to individual researchers while Hefce allocates £1.4 billion in block grants to universities using the results of the RAE, which the funding council plans to replace with the REF after this year.
The formation by Gordon Brown of the Department for Innovation, Universities and Skills had already led to speculation over the future of dual support. Bringing control of both strands under a single Whitehall department could make it easier to create a single funding system.
The current divergence over Hefce's proposals has fuelled speculation further.
"The coffee-room chat is that if Hefce can't evolve the RAE, then what future do they have in terms of funding research?" one pro vice-chancellor said. "They have to come up with a simplified RAE or the funding risks going to the other arm - the research councils."
Bahram Bekhradnia, director of the Higher Education Policy Institute and a former head of research at Hefce, said he could not think of a previous case where one agency of a government department had so publicly criticised the proposals of another.
"I find that extraordinary, and it makes me wonder if this may not be part of some wider jostling for position within DIUS."
Under its plans for the REF, Hefce is proposing to measure research quality in the sciences using "metrics", such how often an academic's published research is cited by peers, instead of using peer review. Non-science subjects would be judged primarily on the basis of "light-touch peer review".
The RCUK submission to the consultation is at odds with Hefce in many respects. It suggests that there should be a single system for science and non-science subjects, it says all subjects should include some form of "light-touch peer review", and it argues that research into benefit to the economy and society is being underplayed.
It also recommends, contrary to Hefce's proposals, that all staff should have their work counted under the citations system, rather than allowing institutions to enter the work of only some academics for consideration.
"This will allow a fairer overall assessment of research quality," the submission says.
"RCUK would welcome the opportunity to work more closely with Hefce in developing revised proposals ... in order to ensure that the two arms of the dual-support system operate together efficiently and effectively," it says.
A Hefce spokesman said he was pleased to note the significant areas where the RCUK supported Hefce's proposals. "A number of issues raised by RCUK are matters where we are undertaking further work," he added.