The merging of the Universities Funding Council and the Polytechnics and Colleges Funding Council sectors of higher education meant that the funding of universities coming from varying starting points had to be as open, transparent and, above all, as formulaic as possible.

There is, I believe, a general view that the process of integration has been carried out pretty fairly, given the constraints imposed by the lack of adequate resources overall, and that the primary task required of Graeme Davies while he has been chief executive of the Higher Education Funding Council for England has been successfully accomplished.

The time has now come, however, for a fundamental reappraisal of how public funding for universities is to be channelled in future, given the clear difficulties with the present model, which can only grow over the years ahead.

The first obvious problem with one invariant model-driven methodology for funding is that the sticks and carrots are identical for everyone. In other words, however much lip-service is paid to the ideal to which we all subscribed - that of maintaining a diverse and heterogeneous university system - all the pressures are to the contrary.

É«ºÐÖ±²¥

If there is more money for research students or part-time undergraduates or whatever, then the temptation is for all or most universities to rush down the road indicated, whether or not they are equipped for such purposes. Convergence can seem to be the only eventual outcome, except for a few favoured exceptions.

Alongside this general danger is a more specific one. A funding model which is wholly arithmetically driven is crucially dependent on the data which is fed into it. Since so many categories, whether of student type or staff member, of research contract or overhead recovery rate, are essentially fuzzy, an immediate and substantial problem arises once large sums of money depend on what are often fairly subjective judgements. Where classification is vague and yet financially very significant, advantageous interpretations will be made in borderline cases.

É«ºÐÖ±²¥

Precisely which Masters students, for example, are "taught" and which "research"? The extremes are clear - but the borderline? If funding is to be formulaic, and judgements excluded, then every category must be explicitly defined and all data checked - but is this realistically possible?

Finally, there is a further conceptual problem. Leaving aside funding for research, about which sophisticated judgements are made, is it actually plausible that, in all cases, the costs of teaching, or of the maintenance of premises, are, or should be, sufficiently similar to warrant identical funding?

I cannot for the life of me see why the teaching of astronomy, with all its attendant costs, for example, should be supposed to cost no more than mathematical physics in a laboratory of personal computers. Nor can I see how an institution is supposed to maintain the fabric of a museum or an art gallery, when the capital funding formula explicitly excludes such facilities from its purview.

The only consequence, once again, of rigid adherence to formula funding will be increased homogeneity as gradually, but increasingly quickly, the future of such special centres of excellence is called into question.

É«ºÐÖ±²¥

There is no doubt in my mind that an element of judgement must now be reintroduced into university funding. I wish Brian Fender well as he takes over as chief executive.

Martin Harris is vice chancellor of the University of Manchester.