Making the grade

Is there any point in an exam which no one fails? A-level results published today show the 22nd consecutive increase in the pass rate, by 0.6 percentage points to 96%. Were the pass rate to continue to increase by this figure every year, everyone would be achieving at least a grade E by 2011. The numbers gaining A grades have risen by 0.8 points to 22.4%. In Northern Ireland, 30.4% of grades are an A. Every town in England, Wales and Northern Ireland will celebrate students with a string of top grades today; four A-grades at A-level is commonplace. And yet university academics regularly swap stories about the gaping holes in the knowledge base of even their brightest students, particularly when it comes to maths-related subjects. David Hart, general secretary of the National Association of Head Teachers, predicted yesterday that A-levels were going the way of GCSEs, where C and not E would be regarded as a real pass. The A-level, designed for a divided, two-tier system, is creaking.

David Miliband, the schools standards minister, moved early to pre-empt the (so far relatively muted) annual complaints about grade inflation. No one believed standards were slipping at Cambridge university, he said, although only 1% of students failed to achieve a degree; why should they assume the worst of A-levels? Out of an age cohort of over half a million, he argued, just 22,000 students achieved three As at A-level; that's one student out of every primary class of 30.

There is little doubt that students have got wiser at playing the A-level market, opting for subjects they judge "easier" than others. The new two-part A-level, where students take AS-level exams half-way through their course, allows them to retake papers they do badly in, or drop subjects entirely to concentrate on their favourites. Exam papers are generally more predictable, more directed and less open-ended than they were. End-of-course exams can count for just 20% of the final mark. Fewer candidates these days find that they have learned big chunks of the syllabus which don't make it on to the exam paper. Much less is left to chance.

But no one can seriously hanker for a return to 20 years ago, when candidates pursued elaborate A-level courses knowing that 30% of them would fail. Exams should be designed to measure knowledge and academic achievement, not to catch students out. A-levels, after all, are still a minority pursuit: last year just 36% of the available 18-year-old population took them. The UK continues to have one of the worst staying-on rates in the industrialised world. This lack of engagement is the most serious problem confronting Mike Tomlinson as he prepares for the October publication of his final report on the 14-19 curriculum.

Mr Tomlinson has also been set the task of finding a way to help the best universities to distinguish between the brightest students, and to stretch them more. This could include an extended essay as part of a new baccalaureate-style diploma. But Mr Tomlinson's proposals, assuming they are accepted by the government, will take a decade or so to implement. Today's results make the case for change sooner rather than later. There is no appetite for the Advanced Extension Awards, introduced to give brighter students another exam to show their worth: just 16 more students took it this year, and it remains an independent-school dominated pursuit. Ministers will not buy the idea, floated in some places, that the A grade could be divided into four subsidiary classes.

Nor are they especially keen for grades for each A-level module to be made available to university admissions tutors, to help them distinguish between able and outstanding students. That is a shame. The exam boards say they can deliver the information. It would be both helpful and more transparent.