...but for all its faults, the RAE might be the best way of distributing funds, says Mike Edmunds
The spectre of the research assessment exercise hangs over the UK university world. Heads of departments have submitted their last drafts, like harassed students as time is called at the end of their exam, and now await the results.
I do not envy the examining board. The RAE panels are faced with an almost impossible task. They are the unfortunate judges in a giant game, whose rules and formats have been engaging many clever minds since the last exercise. Do not even begin to think how much of this effort might have been better spent in actual research. The process was set up simply as a method of distributing resources - but then the same could be said of governments, with politics and complexity an inevitable result. The RAE's winners will get a comparatively large income per research-active staff member over the next few years, the losers very little.
A real problem is the perception that the RAE inevitably acts in an almost biblical way: "For unto every one that hath shall more be given, and he shall have abundance: but from him that hath not shall be taken away even that which he hath."
The criteria for success in the exercise include having a large research grant income over the assessment period, staff who have been invited to give lots of talks at prestigious conferences, and some excellent research publications.
That is not an unreasonable measure of success - but having large grants does not mean they have been used well or efficiently. It does imply that a committee was persuaded to part with scarce resources by an enticing research proposal, but that will not prevent the throwing of good money after bad. Invited talks may come from having lots of international buddies and lots of time and funds for travel. Review talks are, of course, an excellent device for gaining credit for work that was done by others. Good research relies on the time to think - and again, small departments with heavy teaching loads and few research funds are not conducive to quiet contemplation.
This system will inevitably drive towards fewer and larger research units. This is not necessarily a bad thing - economies of scale and breadth of experience may enhance research in these institutions.
But is it the best way to fund research? Responsive-mode grants from research councils can go to almost any department, regardless of size. As an award is usually based on a submitted proposal, the award is based on input rather than output. True output measures - of the quality of the research done - are hard to establish.
Colleagues at large institutions sometimes turn pale when I suggest a "cost-effectiveness" criterion for the RAE in which some measure of the output is actually divided by the grant awarded. The ranking of departments or individuals might then look very different. It is not a serious suggestion, but it does highlight the ambiguity of total input resources as a measure of "quality".
The inflexibility of grants may also militate against the really effective use of funds. Universities like staff to gain grants for research assistants - it allows the recovery of overhead, computing and travel costs. Some universities even have a notional target that any academic worth her or his salt will be able to wangle at least one postdoctoral research assistant out of a research council at any one time. The RAE counts them up, too. But does this macho attitude lead to optimum use of resources? A laptop, a couple of thousand pounds for travel and a few more free hours in the week might give just as much research return for many academics.
Another problem - some may say virtue - of the RAE is that it is not strategic. In an environment that is tightly constrained by the available finance, research councils might like to focus their resources on particular projects or areas. It is another debate as to whether it is possible to "direct" research effectively. But in a world where scientific experiments often require extremely expensive national or international facilities, some targeting to exploit what has already cost a lot of money seems only prudent. Any such "direction" of research would have to include a mechanism for fair competition between departments and universities. Although offering the possibility of optimum (or at least, more effective) use of resources, such planned exploitation does not seem to exist in the RAE. Perhaps this is wise. If we can judge pure "quality", then the excellent may be the best judges of their own future. But it's a big "if".
Even with all its faults, it may be difficult to evolve a significantly better mechanism than the current RAE. Inherently somewhat subjective, it is more like the academic equivalent of the Oscars ceremony than of the Olympic Games. Some awards will seem inevitable, others inexplicable. At least we will be spared the acceptance speeches.
Mike Edmunds is professor of astronomy at the University of Cardiff.