There are four phases to an RAE: the submission, the assessment, the outcome and then how the funding councils moderate that outcome to fit their policy objectives, such as how selective the funding allocations should be and whether a certain amount should be used to build research capacity in particular disciplines. Everyone concentrates on the first three and forgets the fourth, yet it is the most important because it has the most effect on the funding formula and thus on the QR that an institution receives.
What this means is that it is not so much what variables are used to assess quality as how they are interpreted by the funding councils that matters. There is no earthly reason why it should be more difficult for a funding council to moderate an outcome based on metrics than one based on a full, peer-review RAE. The main difference is that metrics are timely, real peer review; they change annually to reflect activity; and they carry far less administrative burden and distortion of planning inside the universities.
There are already voices saying that the metrics will distort behaviour by making academics write more grant applications, concentrate them on research councils and salami-publish to increase volume. One wonders if these individuals have been in a university as a member of staff in the recent past. As a professor in Southampton, I was an applicant on 86 grant applications in 11 years – it would simply not have been possible for me to have written any more. Especially in the sciences, our academics are writing grant applications at full speed. Furthermore, they already try to focus them on the most prestigious organisations, starting with research councils. This is partly to do with prestige but also because there is an inbuilt financial advantage to research-council income, which is based on full economic costing. Finally, the marginal financial gain in turning three large publications into four smaller ones will be so small that it is impossible to see how that will motivate an individual academic to salami-publish. What academics want is to publish in journals that will have the most impact and speak to their peer group best – these are always the highest-impact journals in their discipline.
Some themes are already emerging. Metrics work well in STEM subjects but are not the only answer in arts and humanities and social sciences, where it looks as if some light-touch, RAE-like peer review will be necessary. The main metrics will be research income, research students and some form of bibliometrics. It is also clear that their introduction will be phased over some years. This means that RAE 2008 will inform funding for a significant period of time and the funding councils will have plenty of opportunity to moderate unpredicted outcomes.
What is vital is that the sector responds to the consultation in a measured fashion, making constructive suggestions rather than a hysterical denouncement of metrics. Having complained bitterly about the current RAE for years, we are going to look particularly stupid if we fail to embrace the only real alternative in town.