Open any volume of official statistics that covers more than 10 years and you are likely to find nearly all the tables are intersected by dotted lines. "Change in basis of calculation," you will be informed, is a nightmare for historians and an irritation for journalists, no doubt. But nobody seriously expects a modern government to be collecting figures on spending and taxation in the same way as Gladstone's administrations. The components of the Retail Price Index are reviewed annually.
The Conservatives, famously, have changed the basis of the unemployment figures on numerous occasions. Politicians and economists argue endlessly about the definition of poverty but agree (at least I hope they do) that it has changed since the Middle Ages.
So why do people expect examinations to be immutable? And why is there so much fuss about whether "standards", in GCSEs, A-levels and degrees, are comparable with those of 20 or 30 years ago?
Biology, for example, has changed beyond recognition. Maths has been transformed by the pocket calculator: when did you last employ long division or long multiplication on paper? Examining techniques in some subjects have, beyond any doubt, changed for the better.
Thirty years ago, I sailed through O-level French without acquiring the smallest understanding of what any French person might be saying to me. Nowadays, people complain about "open book" exams. Yet, almost incredibly, my Latin O-level required us to render into English a passage of which we had a year's prior notice. Naturally, the teacher made us chant the English aloud until we knew it off by heart.
My history O-level was not much better. "Daniel O'Connell, three Is, three-fold aims," the master kept telling us, as though our lives depended on it. I cannot now remember who O'Connell was, still less what his aims were.
There is no sensible way of ensuring that the standards of 1966, for example, are comparable to those of 1996. A maths exam taken with a calculator is simply different from one taken without. A French exam that requires oral fluency is different from one that tests written translation.
Once, the leading universities required a classical language for entry to many courses. Now most would prefer some understanding of word-processing. In that context, any talk about standards is just irrelevant poppycock.
The failure to accept that we can't cling for ever to the same standards is behind much of what is wrong with English education. Take the GCSE. This was supposed to be the start of a brave new era in schools - a modern exam system that would reward children for what they knew rather than penalising them for what they did not. It was supposed to encourage the development of learning skills rather than rote memory; to use varied methods of assessment; and to rid us of the divisions between O-level sheep and CSE goats.
The old system should have been buried and the public informed there was no relationship whatever between the old grades and the new. Instead, the Government insisted the top three grades of GCSE had to be equivalent to O-level standards, putting the new exam in a straitjacket from the start.
The lower grades of GCSE have no more esteem (perhaps less) than the CSE. We still manage to brand roughly half our children as failures at 16, while simultaneously creating new doubts about whether the other half are really successful at all.
Examining at 18 faces the same problem. We have tried and failed for 30 years to tackle the notorious over-specialisation of the English sixth-form curriculum. We shall never succeed as long as governments insist on preserving the A-level "gold standard". A curriculum that covers five subjects and straddles both arts and sciences is, again, different from one that concentrates on, say, physics, chemistry and maths.
Fiddling around with half A-levels, AS-levels, N-levels, or whatever you choose to call them, while still preserving A-levels, just muddies the water. Universities, too, though they are far more flexible than they used to be, are still hamstrung by the traditional English idea of a degree: a three-year, full-time course, based on largely theoretical learning. The problem is not so much the lack of alternatives as the belief that these must be second-best, that the traditional degree is the standard to which all the best people should aspire.
And that is the nub of it. We cannot get it out of our heads that a qualification is only worth having if it can be clearly demonstrated that most people would fail it. As Groucho Marx might have said, I don't want to take an exam that would allow me to pass.
If 15 per cent achieve a standard once achieved by 12 per cent, we are determined that something must have gone wrong. Rigour must have been sacrificed, depth lost, knowledge diluted. Yet we do not react similarly to improvements in other walks of life.
Until the 1950s, nobody had climbed Everest or run a four-minute mile. Now, hundreds have done both. We do not therefore appoint steering groups and engage consultants to tell us whether Everest has shrunk or the mile become shorter.
The O-levels, A-levels and degrees of my youth were deliberately designed to produce an elite. As R H Tawney put it, thousands of tadpoles had to die so that a few might become frogs. A modern economy, however, demands a far greater spread of educated talent. Why should we expect a system that turns a third of the population into graduates to embody the same standards as one that gave degrees to barely one in 20 young people?
The point is not whether standards are higher or lower - these things are unknowable - but whether the standards are the right ones for the world we live in.
I am not suggesting we jettison entirely the idea of following standards from year to year. We should, indeed, do it more rigorously and methodically than we do now. But, at ten-yearly intervals, we should pause and consult employers, teachers and students on how the demands we place on young people should change.
Then we should draw a dotted line through any tables of examination grades and passes and announce: "Basis of calculation changed".