One of the great education myths in the UK is that Scotland is better - better than England in attainment and attitudes; better than Northern Ireland in its comprehensive public sector; better than Wales in its balance of academic achievement and community.

Unintentionally, a vast experiment that tests some of these claims has been conducted since 1997 as policies in the four nations have diverged since devolution. What is little realised, and never celebrated, is that the clear winner in this is the much-disdained England.

Let us remind ourselves of the main features of comparative policy since 1997. The dominant theme in England has been fostering a diverse range of school types in the belief that the resulting competition and emulation will lead to improvement.

Some of this has been by inheritance. Unlike in Wales and Scotland, there remained important elements of selection in the public sector, and an independent sector that, while larger overall than in any other part of the UK, was particularly important in the south-east. England’s schools also inherited more religious diversity.

The Government since 1997 has done little to dilute these legacies. But it has done much through special funding for a multiplicity of school types; through its encouragement of school specialisms; and through its scepticism of the standardising effect of local authority bureaucracy.

None of this has happened in Scotland, and what little there was in Wales has been set aside by the Assembly government since 1999. Uniformity of structures in these two nations is believed to be the only route to providing equal opportunities and to managing the tensions of multiculturalism.

Northern Ireland had a different legacy, not only in its persisting selection for secondary school (likely to end soon), but also in the denominational character of nearly all its schools. That kind of diversity is not undermined by devolution because it underpins many of the province’s political structures.

This ideological context is largely irrelevant in Great Britain, but in this sense too the explanation for England’s distinctiveness is political. Labour may have remained strong in Wales and Scotland (though tenuously now), but New Labour never had the same mesmeric effect on their public debate as it had in England. If “standards, not structures” signified the end of Labour’s allegiance to a uniform comprehensive system in England, it indicated precisely the opposite in Scotland and Wales, where an obsession with structures would have meant a determination to abolish what the 1960s had bequeathed.

So what has happened? Many liberals - and I would have been among them then - were sceptical about England’s reforms. Diversity, we feared, would lead to the worst kind of competition and would allow invidious selection to be re-established. The national literacy and numeracy strategies would destroy children’s confidence so much that attainment would deteriorate. Far more preferable, we thought, was the other territories’ more emollient approach to curricular prescription and testing.

Scots in particular had superiority of attainment and high levels of participation in post-school education that had lasted for as long as records had been kept. Surely, it was believed, this would not be disturbed by the mere fashion of a decade?

Yet it has, and this brings us to the core of the matter. England’s education system may still have its problems. There may be more pressure on children and teachers than ever before. There may be types of diversity and places where competition has undermined morale. But the essential point is that attainment in England has improved much more than in the other three nations. That is true in primary, as shown last year by the Timss (Trends in International Mathematics and Science Study). It is true of attainment at GCSE or equivalent. And it is true of progression rates into higher education.

Wales and Scotland have got a bit better. Northern Ireland, with its selective system, has bested these two and remains ahead of England, although it has not improved as much. None of these comparisons is invalidated by the possibility that exam standards are declining, because that would have affected all four nations similarly.

Moreover, diversity in England has not been accompanied by any marked increase in formal selection. The selection that does take place, such as by favouring existing pupils’ siblings, is no worse (and probably more benign) than selection by mortgage into catchment areas. When measured by appropriately subtle statistical means, there has not been any large increase in social segregation among schools, nor any worsening of inequality of attainment and progression.

Of course, exam and test results are not the only relevant measures, but they are necessary. The best that has been thought and said cannot begin to be appreciated without a firm and traditional grounding.

Improvements in England cannot be attributed to any one policy, such as that of specialist schools. Although such policies have probably improved attainment slightly, the most likely effect they have had is to encourage a spirit of constructively competitive improvement.

We do not have enough data to tell us whether these rather speculative explanations of England’s impressive advances are valid. This is largely because the liberal consensus of academic research has not - with distinguished exceptions - really looked.

So here are some research hypotheses for the next decade: competition and diversity are quite compatible with high standards; equal opportunities don’t require uniformity of structures; and England has quite a lot to teach the rest of us.

Lindsay Paterson is professor of educational policy at Edinburgh University.