The Research Assessment Exercise raises hackles everywhere. David Smith, below, argues against selectivity on efficiency and equity grounds.
University departments are gearing up for the next Research Assessment Exercise amid growing scepticism and hostility. Earlier criticism focused on technical problems of departmental rating: the perverse aspects of the funding process; peculiarities of the transfer of star performers; the harm done to what some consider to be research quality; the detrimental impact on teaching; and the rush to publish. This contribution seeks to broaden the debate, by questioning the rationale for selective research funding itself.
The argument is that, even if we could all agree on departmental research ratings, the selective distribution of resources cannot be defended on grounds of efficiency (in the sense of maximising research returns to investment), nor equity (fairness to the staff involved).
To take efficiency: how should a research budget be allocated among university departments so as to maximise the production of research? An economist might answer that it is by ensuring the marginal productivity of all departments is the same, so that there can be no net increases in output from reallocating resources.
But how is this to be achieved? There are no competitive markets for academic research, of the kind which might automatically equate productivity at the margin, and it is inconceivable that the English funding council could administer such an outcome by central planning.
One possible response would be to stimulate a market, with a purchaser/ provider split, as has been attempted in the National Health Service. But quasi-markets are subject to various distortions, in addition to the well- known imperfections of actual markets, as is amply demonstrated in the NHS.
Whether implemented through markets or administration, the efficient allocation of resources depends on prevailing production functions. Given the shortage of reliable information, we can only speculate. Suppose there are variations in efficiency among departments of geography, with the most efficient research producer being the University of Wapping. If linear production functions prevail so that increasing the resources available to this one department continues to produce more/better research than spending the money anywhere else, then aggregate research output is maximised if the University of Wapping has all the money available and other institutions none.
But if research production eventually involves diminishing returns, a point will come when it would be more efficient to invest elsewhere. Identifying this point is difficult. When the problem is expanded to require judgements among all geography departments with respect to their capacity to transform money into research at the margin, we are back to the question of whether this is capable of resolution by either market forces or calculation.
In the absence of any more persuasive theory, we might turn to intuition. It is self-evident that departments producing more/better research will continue to do this if favoured in resource allocation. There are (at least) two objections. First, high-rated departments may already be operating at levels where diminishing returns mean that it could be more efficient to favour other departments with lower (if not lowest) ratings. Second, high-rated departments may be doing more/better research simply because they were better resourced than others before the RAE was introduced. For example, if the geography department at the University of Wapping has had excellent resouces for decades, its staff might be expected to produce more/better research than those in a less well-endowed department, even if they are themselves no more talented as scholars.
It may be more efficient to build up the equivalent of the University of Wapping's infrastructure elsewhere than add to it. The onus of proof therefore rests on those who maintain that selective funding enhances efficiency (or value for money) in the production of research. What is the theory or evidence that this is true? What evidence is there that more/better research has been produced in British universities since the introduction of selective funding?
Questions of equity hardly enter consideration of funding selectivity. But they should, if only because staff in some departments feel unfairly treated. Even if everyone is convinced about the correctness of departmental ratings, a sense of injustice may prevail in departments with low ratings if there is uncertainty, even incredulity, about the general efficiency benefits of selective funding. To give the case for selectivity the best chance of success we could assume that it maximises the quantity/quality of research produced, and ask whether this outcome is equitable.
The version of utilitarianism which might be deployed in support of the efficiency argument for selectivity is indifferent to distributional inequality, and would therefore condone some participants getting little or nothing for research if this was consistent with maximisation of output in aggregate. Is this what supporters of selectivity believe?
A rights perspective might counter that everyone is entitled to some basic minimum research support by virtue of their identity as scholar. How high the minimum should be is a matter for debate, rather like the security safety net to be provided for all before the rest of society's product goes up for competition. Adequate research support for all might leave little scope for selective funding at a higher level.
There is an equity argument in favour of equal funding (per capita). Insofar as the departmental research support available to new entrants to the profession is something over which they have no control - for example, the good fortune in obtaining a post in a department rated 5 rather than 3 - selective research funding is blatantly unfair. Past and present members of staff might be held responsible for research performance; future staff cannot.
Critics could argue that government is responsible for obtaining the best value for taxpayers' money from university research. Selective funding is supposed to achieve this, but there is no evidence that it does. It could be that taxpayers, captivated by the lottery and scratch-cards, do prefer faith to evidence, but this is another issue. The second point is that the preservation of Britain's centres of excellence under resource constraints, facilitated by the RAE, might justify the losses elsewhere.
Staff with reduced research support are therefore compensated for by some others still having plenty. This argument might be persuasive, if the RAE had not replaced collective identity with self-interest, and if concentrating resources demonstrably produces more/better research. To conclude: no persuasive case can be made for selective research funding, on either efficiency or equity grounds. Those who support funding selectivity according to the RAE should explain why. If they cannot do so this time should be the last.
David M. Smith is professor of geography, Queen Mary and Westfield College, University of London. This is an abbreviated and revised version of a paper originally published in Area, 17(1) 1995. Responses by David Rhind and Paul Curran are gratefully acknowledged.