Tony Blair's criticism this week of the public sector's reluctance to face change might have been designed for the university admissions system. Neither students nor universities like it. Students suffer too much uncertainty; universities get too much paperwork about too many candidates; and the clearing scramble that ensues in summer when the results come in is awful.

The key actors - schools, universities and examination boards - have all shied away from disturbing term times (for the educators) or shortening the time it takes them to turn round exam papers (for markers and boards). They say there is no chance of a post-qualification entrance system any time soon.

But the compromise now being engineered (page 1) is not good enough and should be regarded only as a short-term one. It seems likely to bring problems of its own, although it contains some good ideas such as killing off insurance offers.

Using AS levels as a predictor of A-level performance, so that low offers can be based on them, sounds clever. But research has shown that GCSE results are a better indicator of degree class than A levels, so why not base firm or very low offers on those?

ÖйúAƬ

And the Secondary Heads Association points to Scottish experience of scholars in the second-year sixth form who gave up trying once they had easy offers from universities - not much good for the student or for school rankings.

Any new system must keep students working. It must also be open and simple. Anyone waiting for a passport might disagree, but moves to make acceptance criteria more transparent open the door for technology to take a role. If both academic and non-academic admissions criteria can be written down, a computer can see whether they have been met. And schools are getting on the internet. Interviewing could become even more of a rarity reserved for borderline cases or for courses such as medicine, where personality really is vital.

ÖйúAƬ

Among the universities with most to gain from clearer and more objective procedures are Oxford and Cambridge, whose apparent enthusiasm for the public school-educated is a matter of regular comment. They might start to address this concern by finding ways to admit students on the same timetable as everyone else, without the extra month they now find necessary. This is a barrier to potential applicants from schools unused to making Oxbridge applications.

In the longer term, moving to an automated post-qualification system must be the goal. But there are drawbacks. It could threaten one of the cornerstones of university autonomy, the right of universities to choose their students. Slow progress with automated systems may have something to do with reluctance to acknowledge that this ancient right, suited to universities admitting only a few per cent of young people, ill fits a mass system admitting a third of young people and is now a liability. It opens tutors (especially at Oxbridge) to charges of favouring the privately educated. Doing it properly requires more staff time than is available.

A post-qualification system could still protect an institution's right to set its own criteria provided these are clearly stated. But there is another unacknowledged fear: a system that was almost entirely based on certificated achievement could have social consequences. Private schools are very good at producing good exam results.

If we are to have an open, transparent and rapid post-qualification system and keep established annual timetables, universities may have to accept some loss of automomy. And the government may have to accept that admissions would no longer be available as a means of delivering policies of social inclusion.

ÖйúAƬ