'Superstar' rating is not the issue

The Association of University Teachers, like many organisations and academics, is unable to distinguish the wood from the trees as far as the research assessment exercise is concerned ("Union puts boot into 'celebrity' culture", April 2).

The main function of the RAE (even though the bureaucracy itself seems unable to spell this out) is precisely to reallocate researchers across departments. The funding formula (and especially its change from 1996 to 2001) reveals unambiguously that the RAE's aim is to concentrate the "best" researchers in the "best" departments.

The evidence is there: whereas in 1996 departments ranked 1 to 3a obtained 29.3 per cent of overall funding for England and Scotland, in 2001 the figures dropped to 10 per cent and 11.7 per cent for each country respectively, and the share of 5 and 5* ranked departments rose from 54.3 per cent to 74 per cent and 70.2 per cent respectively, with the share of 4-ranked departments remaining essentially constant at 16 to 18 per cent.

Is this good, bad or indifferent as far as the allocation of research resources is concerned? Talk of the "superstar" culture, of the RAE being "divisive" and so on is a mere distraction. The real question is: what model of the research process underpins the RAE's very deliberate strategy towards centres of excellence?

If the research carried out by a typical department involves all the steps from basic research through intermediate applications to near-industry research, then it is easy to prove that the RAE's policy of focusing talent in a few "star" departments is indeed optimal (and therefore the superstar system, far from being a so-called quick fix, is the natural labour-market outcome).

Conversely, if departments specialise in specific stages of the research process (some focusing in basic research, others in applied), an optimal allocation of resources would involve the dispersion of talent (the opposite of the centres of excellence strategy) and thus make the RAE's induced star system highly inefficient.

The real scandal surrounds the lack of debate on which model is more appropriate for the UK and, indeed, for which UK disciplines - not the existence of "a few highly paid superstars".

Manfredi La Manna
Reader in economics
St Andrews University