On the face of it, PQA - post-qualification application - is the only rational way to handle university admissions. More information on an applicant is better than less; newer evidence is better than older evidence; and - if A level is the best evidence of aptitude and readiness for higher education that we can have -the conclusion follows. Taking students on the strength of what their schools and they themselves say looks like the elections in Gulliver's Travels , where politicians who had once competed to perform daring tricks on the tightrope decided it was too dangerous and now competed to make ever larger promises about the tricks they would perform if they tried.
Of course, it's not like that. Nobody offers university places on the basis of other folks' predictions; they offer them on the basis of their own best guess about an applicant's future performance; and this guess may or may not take a school's views seriously. At the so-called selective end of the market, A-level performance is a no-brainer; all 20,000 Oxbridge applicants who are going to take A level will be predicted three or four A grades and all but a few will get three or four A grades. They've been getting perfect scores for so long that it would be a surprise if they didn't.
Picking the very best isn't difficult, either - as with every other system of selection, it is drawing sharp lines between the marginal admit and the marginal reject that is the nightmare. Doing that would be easier with a full version of PQA, where you could test your intuitions against detailed A-level results.
At that level, the lukewarm version of PQA that Bill Rammell, Higher Education Minister, introduced with such fanfare is a dead loss. In the December before applicants take their A levels, Cambridge University will interview and turn down 9,000 applicants who go on to get three or four As apiece the following August; the idea that Cambridge should take, sight unseen, a couple of hundred students rather than 200 of the 9,000 they have seen and rejected is absurd. Diluted PQA is said to be the brainchild of Sir Alan Wilson; he is about to leave the Department for Education and Skills to preside over a Cambridge college, where he can see all this for himself. I suspect that he has known it all along, and that it's another piece of the Government's obsession with Oxbridge that it's been sold as a device to get bright young people from state schools into Oxford and Cambridge.
In fact, it is a useful first step, but it's only useful elsewhere, outside Oxbridge and the Oxbridge peer group. Applicants with middling grades who doubted they would get good enough scores for a particular course really would be helped if they could reopen the application process in August when they had discovered the grades were better than they'd expected them to be.
It's no good saying that they could chuck away their existing offers and take a gap year, and then apply again next year. They might be risk averse; they might find fending for themselves for a gap year daunting; and they might, quite sensibly, want to get on with their lives. Taking a gap year ought to be a freestanding decision, made for good reasons, not under duress.
There will be - there has been - some griping from universities that fear they will lose the brightest of their students to universities higher up the pecking order, but that's about as persuasive as a supermarket complaining that it is losing customers to a competitor that provides better goods at lower prices. It's the students that count, and more flexibility is self-evidently a good thing for them. The places that gripe about the competition are uninhibited in touting their own merits in their prospectuses, so they can hardly complain about living in a competitive world.
On the other hand, when 18-year-old school-leavers no longer form the majority of undergraduates - and we begin to take seriously the idea of lifelong learning -PQA, pure or diluted, drops down the agenda. What we are searching for is the Holy Grail - reliable predictors of how students will do in a new environment, regardless of A levels. One piece of information all admissions offices would find helpful is an applicant's intellectual trajectory - whether they seem to be improving or coasting or going off the boil. But how do you work that out when someone has spent the previous ten years in employment without forcing them to take endless access and foundation courses?
Tests for "deep learning", as distinct from "tick-box" learning, seem effective in picking out the very best students, but will they help across the spectrum? Who knows? The one thing that is certain is that no one solution will suit the varied system we have created over the past two decades.
Alan Ryan is warden of New College, Oxford.