The annual debate about standards is both demoralising for students and teachers, and redundant because no comparison of past exams with today's makes sense unless it considers who actually took those papers. In the heyday of both O- and A-levels, barely a fifth of the numbers who sit public examinations today were entered. Yet many of this elite group performed astonishingly poorly – 40 per cent of O-level candidates and 30 per cent at A-level failed.

This is always taken as proof of the rigour of the exams, but a similar failure rate today, from only the brightest 20 per cent of the population at O-level, far fewer at A-level, would cause uproar, however tough the exams.

And many papers were hardly as difficult as often claimed. After two years studying just three books for English Literature O-level, I was asked to explain the source of Pip's great expectations. Some maths questions certainly were ferocious, but more than a quarter, even of the select few who passed English, failed maths.

This is hardly surprising when JMB examiners in 1956 complained bitterly about the large numbers entered who lacked "the knowledge and understanding to have the slightest chance of passing".

Yet today, results are not even included in school league tables unless students have passed both subjects at GCSE level, and at least three others.

Of course, there are other arguments about the quality of modern examinations. We hear frequent complaints from universities and employers that, regardless of their qualifications, youngsters lack essential literary and numeracy skills for courses or work.

But, even in 1961, Manchester Grammar School wrote to the Ministry of Education about "dogmatic" dons continually claiming that first-year university students were "dumb illiterates".

It is argued that recent rises in pass rates could have come about only through exams becoming easier. While this could be a factor, there are other explanations. Preparation for exams has improved immeasurably. Who had heard of Saturday or Easter holiday revision classes a decade ago? Much of the improvement over recent decades has been led by girls, many of whom under-achieved in examinations in the past. Certainly their performance in, say, maths was dire, with swathes of grammar school girls either failing O-level or even dropping the subject at an early stage.

Arguably, the improvement in girls' education after the 1970s, with the growth of comprehensive education, has impacted directly on today's youngsters. International research has shown the qualifications of mothers, particularly A-level (or equivalent) and beyond, are one of the key factors in determining a child's success at school. Recent improvements in exam rates have been wholly predictable.

Of course, there are still areas of concern. We attach too much credibility to the accuracy of public examinations and too little to teachers' assessment compared with other countries and, as a result, test our children more than anywhere in the western world. We expect testing to fulfil too many purposes, from comparing the performance of schools – and, increasingly, teachers – to diagnosing students' faults to providing colleges, universities and employers with guidance
on selection.

This level of external assessment, as demonstrated by the recent SATs debacle, is neither desirable nor sustainable. We need to free up schools to provide time to develop creativity, for extra-curricular activities and to explore issues in depth which are not examined. Most schools do a remarkably good job in producing well-rounded young adults: both creative subjects and extra-curricular activities are in better health than often appreciated.

But schools are not helped by a league table culture which both underplays the impact of social factors like poverty on pupils' performance, and leads some schools to teach to the test.

As England, in common with many countries, moves towards full-time education continuing to 18, we should query the need for so many expensive, time-consuming public examinations from seven to 16. Both primary and secondary schools now have the skills to undertake testing
in-house, backed up by random sampling of scripts to ensure consistency. We need to look honestly at the limitations of public examinations: not all subjects, for example, are equally difficult and if exam grading, especially at A-level, recognised this, it might help arrest the decline in the numbers taking languages and sciences.

In the meantime, how refreshing it would be if in late August each year we simply celebrated the achievements of the youngin England, recognising their strengths more and bemoaning their failings less.

Dr Adrian Elliott is the author of State schools since the 1950s: the good news, published by Trentham Books. He is a former York headteacher, Ofsted inspector and an A-level-examiner.