Leader: Admissions war will be won only with imagination

The critics of "social engineering" in university admissions who have waded into print this week would do well to look at the government's latest youth cohort study. The annual survey of 18-year-olds charts the gulf in educational experience between children of the middle classes and those from poorer homes. But its most striking conclusion is that the higher education die is cast before teenagers reach sixth form.

In cold statistics that defy spin, it demonstrates that the government's ambitions for wider access to higher education hinge on secondary schools far more than on any campus conspiracy. Education secretary Charles Clarke admitted as much at a THES-sponsored conference, although his endorsement of attempts by universities such as Bristol and Edinburgh to broaden their intake inevitably captured the headlines. The cohort study shows that even the supposed threshold of five good passes at GCSE is no guarantee of progression beyond school: more than half of those with eight good grades are in higher education at 18, but only a quarter of those with five to seven have similarly moved on. Of those with fewer than five good passes - still almost half the age group - barely one in 20 go on to university or college. Ministers are banking on their reforms improving school performance and encouraging more teenagers to continue their education. But in the meantime, leading universities will have to wrestle with their own version of Catch-22 - accused of elitism and penalised financially if they maintain their traditional selection standards, and condemned for spineless dumbing down if they modify their criteria.

If this week's controversy is anything to go by, the debate will generate more heat than light. No amount of hard-luck cases can prove discrimination when courses are as oversubscribed as those at Bristol, but neither will parents and schools be convinced that the system is fair until it is demonstrated that selection is based on firm evidence. Part of the case against Bristol has been that highly qualified "victims" have not even been interviewed, as if they were discarded at the first opportunity. But, as in other popular universities, interviewing has long ceased to be part of the selection process other than in "exceptional cases".

Instead, admissions officers rely on schools' references and candidates'

personal statements to mark out the best prospects among hundreds per course who are predicted top grades at A level. In Edinburgh's case, selectors will gauge applicants' "commitment to higher education, their motivation to succeed, their suitability for the chosen programme and their personal resourcefulness" without seeing them or, it seems, delving beyond the application form. After that, parental background and quality of school will be used to supplement examination performance. The approach has been used, apparently successfully, in Edinburgh's medical school, but the level of interpretation required of limited and partial application information will hardly reassure doubters.

Independent schools cannot complain if universities make allowances for candidates of outstanding potential from poor schools - most of them do the same in their entrance procedures. But well-qualified applicants, whatever their background, have a right to know that they are being passed over on the basis of more than a hunch or a formula. The demand for places at the most popular universities is such that they cannot interview every applicant, but it should be possible to take a closer look at marginal candidates. This might involve interviews, US-style SATs or online questionnaires of the type used by Nottingham University's medical school.

If admissions processes are to stand up to constant scrutiny (and possibly legal challenge), they will have to be more sophisticated and imaginative than hitherto.