Are A levels getting easier?

Back in the mists of time when I took my A levels (1984), exactly 30% of candidates failed. The same was true in 1985 and 1986. The reason? Only 70% of candidates were allowed to pass – the system was used to rank students, not to measure standards. It didn’t matter whether everyone was good or everyone was bad – 10% would score ‘A’ grades, 15% ‘B’ grades, 10% ‘C’s, 15% ‘D’s, 20% ‘E’s and 30% would fail. This had been true every year since 1963.

In 1987, the world changed. Students needed only to gain a certain mark to achieve an ‘A’ grade or a pass. A standard was defined for an ‘A’ grade, and so long as you hit that standard, the ‘A’ grade was yours. Fast forwarding to 2008 97.2% of candidates passed, and 25.9% scored ‘A’ grades. At the same time, the numbers of students taking A levels have doubled, so that today 5 times as many students get ‘A’ grades as was the case in the 1980s.

Does this mean that A levels are easier? Ofqual, the Government’s exams watchdog, would say not – their analysis of standards over time suggests that the quality needed for an ‘A’ is the same today as it was several years ago, although the comparison is becoming increasingly difficult as A levels have become modular – can you compare the performance of a student in six discrete modules (2009) with one forced to take all the exams at the end of the two year course (1984)?

Much has changed over the last 25 years – syllabuses used to be just lists of books, whereas today they are detailed descriptions of what is and is not required, making it far easier for students to hit the required standard. As soon as the system changed from one of quotas to one where there was a defined standard, it was inevitable that more students would achieve it. If we set a given standard for a gold medal at the Olympics for the long jump, then as training techniques, equipment, diet and so on have improved, so would the number of gold medals. In 1936 Jesse Owens won the Munich Olympics with a jump of 8.06m. This would have been beaten by 7 of the finalists in the Beijing Olympics of last year. Does this mean that the long-jump is getting easier, or simply that athletes are getting better? The same analysis must surely apply to A levels.

In the end, the problem is that the A level system is being used for two mutually incompatible purposes. On the one hand we want to measure improvements in performance over time - part of the reason for the switch to setting a given standard for an ‘A’ grade. On the other hand, there is the need to use it to distinguish between students (for university application), at which point having lots of ‘A’ grades becomes a problem. It is difficult to use the system for both – either you have one gold medallist, at which point it is difficult to recognise improvements over time, or you have lots of gold medallists, but then there’s no winner.

Perhaps the solution suggests itself from the Olympic analogy – we can have lots of gold medallists, as long as we know who jumped the furthest. To do this we need simply to publish candidates’ raw scores. Then we can have the best of both worlds – we can see how many students hit the required standard for an ‘A’, but universities can see how much they beat it by. At this point, the annual hand-wringing over whether A levels have become too easy should simply go away.