Alan Ryan

Government policies are confused. We must decide what different institutions are really supposed to be doing.

Talk of back to basics summons up images of former prime minister John Major's unfortunate campaigns for the moral revitalisation of Britain.

Still, the risk must be taken. The government's policies on higher education are so confused in so many dimensions that it is time to get back to basics.

Here are three basic thoughts.

First, stop talking about "top universities". A university is good to the extent it does its job and bad to the extent it doesn't. Good means good for students, faculty, staff and the larger public - good at teaching, research, industrial cooperation and enhancing community life.

Second, we really must decide what different institutions are supposed to be doing. Loose talk of "world-class", "access" and "diversity" confuses the issue, as does lumping together everything from reading Heidegger to running a team of 30 as "research". Finding money for the first is easy; paying for the second is not.

Third, although there is some consensus on what institutions are for, there isn't agreement on how to pay for them and how to balance competing claims.

The consensus is this: some places exist to do blue-skies research, they are open to commercial demand and supply commercially useful ideas, but not at the expense of blue-skies research; they can also give a demanding and expensive undergraduate and postgraduate education that will produce the researchers of tomorrow, along with the leading cadres in the professions and public administration.

Others institutions exist to do more immediately useful research and provide an undergraduate education that leads to professional training, teaching, public service, or postgraduate work and blue-skies research.

Others still provide the first port of call for school-leavers, mature students and mid-life learners who are not certain of their futures, who want to change gears and careers, who may want to get a rigorously academic education but aren't sure and aren't sure of their skills. And they can do a lot for their local economy, through training courses and research work.

Do uniform tuition fees make sense in such a context? No. No fees anywhere would make sense, given a rationing system such as the California master plan. In any other universe, the "first-taste" institutions must be subsidised to get the punters in by all means possible. The first taste is high risk, and first-tasters must be allowed to move to vocational training, or to find out they really want the most academic courses available; so we don't want them to be deterred from that first taste, and we do want them to be helped to move around. Since the public purse isn't bottomless, another obvious move is to bribe first-tasters to stay at home by giving them a grant to stay where they are rather than loans to go somewhere else.

Research-heavy institutions cost a lot of money. Britain could do without them and buy its pure research from America, but we might then find that anyone with any gumption headed for the US on graduation. Still, it's not economically essential to compete with the US at the blue-skies level - Japan's growth rate was a compound 7 per cent a year when its advantage lay in using other people's pure research in manufacturing and electronics. If we want blue-skies research, we will have to pay for it.

That washes over into the cost of teaching, since there is a global market in high-grade research faculty; and they aren't going to do a great deal of teaching anyway, so your staff-to-student ratio will be high and pricey.

Even at the current low levels of pay, Oxbridge, Imperial College London and University College London cost between £15,000 and £18,000 a year per undergraduate because they are science-heavy serious-research universities. The London School of Economics is cheaper because it is spared the financial horrors of the sciences. No university anywhere in the world expects students to pay the full cost - in the Ivy League it is well over £30,000 a year and tuition fees are £15,000.

The middle range of institutions seem most likely to suffer at present with their teaching budgets cut to divert funds towards the first-taste institutions and their research budgets cut to divert funds to the blue-skies institutions. But they are in principle the easiest places to fund and the ones where modest but genuine tuition fees make most sense.

They could get a long way on £10,000 a student - £3,000 from the student, £4,000 from teaching funding and £3,000 in research funding - especially if they steered clear of biomedical research.

It's all arithmetic, but politicians prefer to grandstand than read spreadsheets. They want uniform, one-size-fits-all solutions to problems whose intellectual interest lies in the fact that they are complicated.

Alan Ryan is warden of New College, Oxford.