Teaching to the test. Is it a good or a bad thing? I believe this is just about the most fundamental question facing education. Yet those in charge of our schools have repeatedly failed to answer it, in detail or conclusively. They need to do so urgently.

A strong case can be made that preparing pupils for their next assessment dominates the educational experience of most pupils for most of their school careers, to a greater extent than occurs anywhere else in the world, even though many teachers try to avoid its worst excesses.

A survey of hundreds of primary schools by the Qualifications and Curriculum Authority made this clear. Primaries were asked directly how much time they spend on exam coaching during the four months leading up to the key stage 2 tests in May.

Ten hours a week on average was the reply - or nearly half the teaching time available. Though the picture is not uniform, few working in primaries would dispute that this is happening in many schools, or that extensive test preparation now also takes place with younger children. In secondaries, of course, the practice continues. English, maths and science lessons in Year 9 are dominated by building up to the tests. Pupils taking GCSEs and then A-levels find that four of their last five years in school are geared to exam preparation.

The replacement of education by test revision is likely to be accentuated in the future: the Making Good Progress trials of new national tests - which pupils can take at up to six-monthly intervals when their teachers believe they are ready - are poised to emphasise assessment-driven teaching.

This should come as no surprise to TES readers. But is it a positive or negative development? Does teaching to the test represent the best way of ensuring children get the education the public wants for them, or is it simply an unanticipated side effect of the pressure on schools for results? Startlingly, the problem has arisen without the Government having offered a detailed analysis of its merits or downsides.

Repeatedly, ministers have rejected accusations of cramming with the suggestion that pupils spend very little time actually taking national tests. Sue Hackman, chief adviser on school standards, used this defence again last week when she appeared before a parliamentary inquiry on assessment. She said only 0.14 per cent of teaching time in KS2 and 0.2 per cent in KS3 was spent on test preparation. She later had to inform the committee that the figures actually related to the time pupils spend sitting the tests.

But it is the preparation, rather than the time spent in the exam hall, that is contentious. Ms Hackman’s statement also obscures the fact that the Government appears to believe that extensive coaching is to be welcomed.

Advice sent to primaries in 2003 by the Primary National Strategy urged schools to start test-geared teaching in the January of Year 6, making sure that plenty of test questions are integrated into lessons, discussing mark schemes with pupils and, from February, doing past papers. The trials for the new tests also enshrine official assumptions that pupils need to be confronted regularly by external tests to have their achievements in class validated.

Again, ministers and officials appear to be facing both ways on teaching to the test. David Bell, the permanent secretary at the Department for Children, Schools and Families, told the parliamentary assessment inquiry that he did not see much evidence of test coaching during his visits to schools. But Ms Hackman said that some preparation for tests was important, that the Making Good Progress assessments could motivate pupils by giving them an exam to aim at, and implied that national tests helped to inject a sense of “urgency” into Year 6 teaching that was lacking in Year 7.

The previous week, Andrew Bird, deputy director general of the AQA exam board, told the same committee of the Government’s thinking on the new functional skills tests in English, maths and ICT. “The policy position is that making (the tested skills) explicit will lead to more discrete teaching of those skills, and hence to a rise in performance,” he said.

In other words, ministers want more teaching to these tests.

Mr Bird and the heads of three other examination boards were quizzed by MPs. But there was little discussion of the extent to which Years 9-13 are now dominated by exam-driven teaching, and what this involves. A science GCSE, for example, features 155 learning outcomes, including that pupils should know that “it is more cost-effective, in terms of energy, to produce a field of wheat rather than a field of beef cows”.

Does teaching to this exam describe an ideal education for the pupils experiencing it? Or is it, in its specificity and predictability, designed simply to provide the shortest route to good results for those pupils and, crucially, their teachers?

It may be that ministers, civil servants, and even exam boards believe teaching to tests - over which the Government and, through it, the electorate has control - is indeed the best way to improve our schools. But if they truly think that, they should say so clearly, and explain why. Failure to do that would represent a missed opportunity, in two senses. First, an honest answer would enable a more realistic appreciation that if teaching to the test is going to happen, then it is important to mitigate what many people outside schools would see as the downsides, such as the tests becoming too predictable.

Second, silence on this issue will invite political cynicism: the belief that ministers are happy to see teaching to the test continue, no matter what its side effects - just so long as it delivers improvements in the statistics by which the outside world judges them. That would be a great shame.

Warwick Mansell, TES curriculum and assessment specialist and author of ‘Education by Numbers: the Tyranny of Testing’, published by Politico’s.