Education: Comment - 'The reputations of universities ride, or stumble, on the results of the RAE'

IS THE research assessment exercise ruining higher education? As the fourth RAE looms, this thought is crossing more people's minds - and not only those who have always believed that measuring research was a gross intrusion into the so-called "donnish dominion", but also perhaps the higher education funding councils and even the Department of Trade and Industry.

Every five years the RAE measures the quality of research in colleges and universities on a sliding scale from 5* (research of international excellence) to 1 (what research?). The next one is due in 2001 - which, given publishers' schedules and queues of journal articles, is just round the corner. Lecturers and professors are divided into the elect and the damned, research-active or inactive. Institutional reputations ride, or stumble, on the results of the RAE.

But, just when the RAE has come to be accepted as a routine matter of life (and death) by the academic community, tough new questions are being asked. The higher education funding councils have commissioned a raft of studies into different aspects of the RAE - for example, its impact on inter-disciplinary research and its relevance to regionalism - to inform their "fundamental review" of the RAE (the next one, that is, RAE 2001 is happening regardless).

Meanwhile, officials in the DTI (which, through the Office for Science and Technology runs the research councils) are reported to be sceptical about the RAE.

No prizes for guessing what they would like to see in its place - a wholesale transfer of the pounds 1bn worth of research funding, which is at present being distributed by the funding councils to itself, and the establishment of targeted research programmes directed by Blairite businessmen.

The funding councils, of course, are trying to improve the RAE; the DTI to bust the so-called "dual-support system" for research which means that universities receive research funding as part of their core grant from the funding councils and also bid for project and programme funding from the research councils. The "fundamental review" is intended to address the weaknesses in the RAE. These weaknesses are both technical (for example, the bias against research outside mainstream disciplines) and behavioural (such as the mad rush to sign up research "stars" in time for the next RAE).

Such weaknesses can be remedied but not removed entirely. Any measurement system needs categories to measure, and those at the edge of (or in between) the chosen categories will always be disadvantaged. The problem is not so much with the implementation of the RAE; it is with its intentions. Basically there are two - to increase the quality of research; and defend the "dual-support system". Both, the funding councils argue, are served by selectivity - scarce resources will not be wasted on indifferent research; and the DTI will not be able to accuse them of failing to "manage" their part of the research budget.

The first is difficult to prove. Britain has a very healthy share of global research outputs, as measured by crunching citations. But most science policy experts attribute our performance to deep-rooted - and, maybe, historical - factors. Recent policy initiatives such as the 13- year-old RAE, do not figure much in their analyses.

The second is also a doubtful success. The core justification for the dual-support system was that the two funding streams were different; the "free" money from the funding councils supported fundamental science, while research council grants were targeted on specific projects. But, if their intentions have converged, why maintain the dual-support system?

Maybe it is too late to rewind the RAE completely. But a lighter-touch and less selective RAE would have two advantages.

First, too much research selectivity not only tends to make the research- light universities less exciting places because it erodes their self-worth; it also makes the Darwinian struggle to succeed in the research race even more intense which, paradoxically, detracts from teaching.

Second, too much selectivity is bad for research because it reduces its variety. In the bad old days before the RAE, research choices were made by tens of thousands of individuals; in the RAE era that number has been cut to thousands; if the DTI gets its way, it would be further reduced to a few hundreds.

But in the real, as opposed to bureaucratic world, research is becoming ever more widely distributed throughout the "knowledge society"; the number of researchers (even if they are not immediately recognised as such) is proliferating; good science is no longer just good laboratory science, but must be socially robust (as the BSE controversy has so graphically illustrated). The risk is that our present, and future, funding arrangements will produce excellent 20th-century science but indifferent 21st-century research.

The writer is vice-chancellor of Kingston University