Radical changes to the research assessment exercise have been roundly rejected by grass-roots academics.
Sir Gareth Roberts, commissioned by the Higher Education Funding Council for England to propose a replacement to the RAE, said he was surprised by support for the current regime at meetings he attended across the UK.
Sir Gareth, president of Wolfson College, Oxford and author of an influential report to Treasury on the supply of scientists, sketched two possible models at this week's annual general meeting of the Save British Science lobby group. Both are based on an institution's capacity for research and require universities to submit research strategies.
The first scenario would be similar to the current RAE but would result in a research footprint rather than single rating. All institutions would submit a research strategy, explaining how their research informed teaching and financial sustainability.
Units of assessment would be larger - for example, the five modern language units would be reduced to one. Each assessment panel would have two chairs: one an academic expert, the other a professional chair who could ensure uniformity.
Submissions could be made for multi and interdisciplinary work, and all research staff - including contractors - would be submitted. The institution would be rated on, for example, pure, applied and practical-based research, strategy, interdisciplinarity, sustainability and scholarship.
The second model would allow institutions to set their own targets and to concentrate on their research strengths.
A percentage of funding would be based on research capacity defined by discipline-specific metrics as before, but the remainder would be based on research strategies. Each institution would explain its strengths, its international excellence, its multi and interdisciplinary research, its links with other universities and companies and what it would like to achieve. This might encourage smaller institutions to choose to submit only for the latter funding.
Discussing academics' responses to the RAE, Sir Gareth said: "If the 2001 exercise had been fully funded, I think there would have been far less criticism. But this RAE has reached the end of the line and people are nervous about what will replace it."
He said there had been a rejection of radical change. But methods of evaluation were agreed to be too uniform and should be tailored to the needs of individual subjects.
He added that the new system must discourage "game-playing". It should also enable institutions to build capacity in emerging areas.
Outstanding issues included how to achieve a light touch while having a rigorous system. He also suggested a rolling system, in which each year a bundle of units was assessed, and less research-intensive universities could choose to be assessed as a whole.
The models and outstanding issues will be discussed at meetings in London, Belfast, Birmingham and Cardiff, before the panel reports in May. Proposals will then be published for consultation.