I shall start, since most things seem so to do this month, with the late Diana, Princess of Wales. What exactly was the secret of her appeal? I suggest that it lay in her embodiment of the message that you can have it all ways. Here was a woman who sailed the Mediterranean, dined at the Ritz and shopped at Harvey Nicks. Yet she could combine this extravagant, diamond-encrusted lifestyle with a caring, compassionate image. She held the hand of an African land mine victim one day, the hand of a millionaire playboy the next.

She thus legitimised the central delusion of our age: that we can somehow improve the lot of the poor and hungry, both here and abroad, without any diminution in our own living standards. We can carry on making and spending money, we just need to be nicer about it. Diana represented a kind of designer compassion; she turned charity (which previously had a rather dowdy, down-at-heel image) into something chic and fashionable. Meantime, talk of taxation, income redistribution and social equality remains the political equivalent of flared trousers.

No wonder, then, that Diana was so readily embraced by the Blair and Clinton administrations. They, too, offer compassion without pain: not just a free lunch, but a free Oxfam lunch, with champagne and caviar. They do not challenge the middle-classes and what J K Galbraith has called their “culture of contentment”. Rather, they hold out the prospect of yet more contentment, by promising, in effect, to remove any of the guilt associated with a high-income, high-spending, resource-guzzling lifestyle. Worried about unemployed miners, homeless Londoners, hungry Liverpudlian children? Relax. Tony Blair has task forces at work; he is “including” these unfortunate folk. Meantime, you can go shopping with your Labour party credit card or attend one of their premium-price fund-raising dinners, and so combine your caring and your consuming into a seamless whole.

Education, as we all know, plays a central role in this project. It is the prime instrument of “inclusion”, the most important (perhaps the only) component in.. what remains of the social democratic strategy for tackling inequality. It too holds out the prospect of social justice without any disturbance to the present lifestyles of the comfortable. Poverty, the argument goes, is the result of people being uncompetitive in the labour market. Give them higher levels of education and skills and they will command better jobs and higher wages. Then they can live in semi-detached houses, drive round the M25 and eat mange tout like the rest of us.

I do not deny the truth of these propositions (whether Britain can support more suburban housing and the world’s climate more pollution are matters for others to determine). I have no doubts about the enormous potential of education to improve a country not just economically but also socially and culturally.

But I do not think Blair’s “education, education, education” lets him or us off the hook so easily. Here, I turn to the recent report on “Literacy, Numeracy and Economic Performance” by Peter Robinson of the London School of Economics (to which, I think, my fellow columnist, Michael Barber, did scant justice last week).

Robinson shows that the evidence for and against particular types of school organisation, curriculum or teaching methods is simply inconclusive. Mixed ability against setting or streaming, progressive against traditional teaching styles, primary school homework against no homework - none of these issues, which so dominate debates about standards, makes a measurable difference either way.

Three things, however, do make a real and lasting difference. The first is social and economic background: put simply, poverty impedes learning. Second, schools that have a high intake from middle-class homes do better for all their children, including those who come from poor homes. Third, parental interest and involvement in their children’s education have positive effects that can override class background.

Robinson’s conclusions are based on two large-scale studies of children born in 1958 and 1970. But there is a mountain of other research that reaches similar conclusions: for example, school intake, the “peer-group effect”, was highlighted by Professor Michael Rutter and others in Fifteen Thousand Hours in 1979.

Robinson argues that “a serious programme to alleviate child poverty might do far more for boosting attainment I than any modest intervention in schooling”. That seems to me incontrovertible. But such a programme would involve, for example, a high minimum wage, improved child benefits, better social support services, lower taxes for families on below-average incomes. In other words, we would have to revive quaint, half-forgotten ideas such as redistribution, progressive taxation and, if you like, socialism.

Fat chance. The same applies to the second sure way of improving attainment: trying to achieve a socially balanced intake for each school. If ministers wanted to do this, they would have to remove financial privileges from the private sector, restrict parental choice, abolish the remaining grammar schools.

But ministers will do nothing that upsets those who are already advantaged. That, in effect, was what they promised during the election campaign. Voters were promised easy consciences at nil cost to their purses and personal interests - a package so attractive that, in retrospect, a landslide should never have been in doubt. In education, they were promised, Labour would produce excellent schools for children of all classes without affecting the freedom of the better-off to manipulate the system to their own advantage.

Diana’s death is supposed to have revealed a more egalitarian mood in the country; but, as a Frenchman once observed (and the French, it must be admitted, have some experience of these matters), “what makes equality such a difficult business is that we only want it with our superiors”. All the same, we should seize the moment.

Those of us who argue for the overwhelming influence of social background on school achievement are accused of setting low expectations. Not so. Remove poverty and there is no limit to what children may achieve. I do not think it a coincidence that the two most unequal societies in the Western world, Britain and the United States, are those with the biggest educational problems.

So let us, first, have some high expectations from Labour party leaders. Expectations that social justice can be sold once more to the British people; expectations that we can tackle child poverty directly; expectations that the better-off will pay more tax for the sake of a fairer society and better schools. And expectations that compassion can come to mean more than laying flowers at the gates of Kensington Palace.