Universities are being told to improve their social mix, but what can they do when early years exert such influence on life chances, Chris Bunting asks.

Education secretary Charles Clarke has had many nicknames in his rumbustious political career - "Biggles" because of his boundless enthusiasm and "two pizzas" because of his boundless appetite, to name but two - but the shaken-looking civil servant knocking back shots at the bar had a new appellation for his boss: "the rhinoceros".

"Estelle Morris was all very nice and fluffy. She knew her education stuff incredibly well. But you get the feeling with this guy that when he starts charging nothing much is going to change his course," said the mandarin, fresh from a "moderately terrifying" Clarke briefing.

It is a description those in ÖйúAƬ would do well to note, because the bull rhinoceros's ears are pricked and he is peering in their direction. The issue is widening access to ÖйúAƬ.

ÖйúAƬ

Clarke told a meeting of headteachers in November that it was "absolutely ridiculous" that up to 80 per cent of middle-class children went to university while only 5 to 10 per cent of poor children did so. Few university leaders, he growled, were "committed to ensuring that people from all backgrounds have an equal crack at getting into university".

A month later, he confirmed his deadly seriousness about the issue in a hugely significant policy change indicating that the government's previous top-priority target of getting 50 per cent of 18 to 30-year-olds into university was less important to him than ensuring "a much better class basis" in the 43 per cent currently attending.

ÖйúAƬ

His determination is understandable in light of the appalling figures. A National Audit Office report at the start of last year showed that only one student in ten at elite universities such as Cambridge came from working-class families. Even at the least socially exclusive institutions, less than half the students came from such backgrounds.

Less than a quarter of students in the sciences, social studies, law, medicine, languages and the humanities had a skilled, semi-skilled or unskilled manual worker as a parent. The most socially diverse subject area, education, drew only one-third of its students from such families.

Overall, working-class families accounted for more than 40 per cent of the UK population but produced only 25 per cent of university entrants in 2000-01 and, according to the NAO report, the figure was not rising. For Clarke and his new Labour cronies, who have discarded Labour's traditional emphasis on income redistribution in favour of social reform based on ensuring "equality of opportunity", the unthinkable possibility is that widening participation in ÖйúAƬ may be benefiting dull middle-class children more than the underprivileged.

In a society in which the lack of an undergraduate degree means exclusion from most professional jobs, a university sector that is now equipping most middle-class youngsters with BAs and BScs but excluding most of their less privileged peers may be functioning as a brake on social mobility.

But is ÖйúAƬ entirely to blame? A large body of evidence attests to deep divides in educational attainment between youngsters from different social backgrounds long before they think of university.

In an article for a Department for Education and Skills magazine in November, David Miliband, the minister of state for school standards, admitted: "We continue to have one of the greatest class divides in education in the industrialised world, with a socioeconomic attainment gap evident in children as young as 22 months."

He said the class divide, once established, became wider and more rigid as children progressed through school. Research by Leon Feinstein, an economist at London's Institute of Education, showed that 22-month-old toddlers from middle-class families were significantly better at tests in putting on their shoes, noting different facial features, stacking bricks and drawing shapes than children from working-class families.

ÖйúAƬ

Children of parents with A levels scored 14 per cent higher in Feinstein's tests than those without. A difference in weekly family income of ?100 correlated with a 3 per cent improvement in scores.

ÖйúAƬ

These differences are partly explained by variations in child-rearing between parents from different social backgrounds. A study by academics at the universities of Alaska and Kansas in the mid-1990s found that professional parents were much more talkative than working-class parents. Children of the talkative parents had correspondingly higher scores in simple intelligence tests at age three. Reading to children, which varies with class, strongly influences cognitive development.

But there are also systematic differences in access to childcare and education from a very early age. While the government has invested millions in expanding pre-school education, low-income families are still much less likely to access high-quality childcare than professional parents, according to the Daycare Trust. This is significant because good childcare significantly boosts later educational attainment. The DFES's Effective Provision of Pre-School Education Project study reports that children who enter childcare settings at an early age and attend regularly show higher cognitive attainments than those who do not, even when controlling for the influence of child, family and home environment.

Some progress has been made in reducing differences between social classes'

attainment at primary schools. Since Labour gained power in 1997, the gap between the scores of junior pupils in England's poorest ten wards and the national average has narrowed by one-third. But they still score an average of almost a grade a subject lower than their peers.

Not only does a child's family background have a powerful effect, the background of those they rub shoulders with at school also relates with their achievement. The Organisation for Economic Cooperation and Development's Programme for International Student Assessment reported in 2000 that, across the 32 countries it studied, pupils with classmates from poor backgrounds did less well at school, even allowing for their own social disadvantages. Truancy, low self-esteem and mental illness are disproportionately high among underprivileged children.

At secondary level, the situation is worse. Achievement falls significantly among working-class pupils as they begin to adopt the adult expectations and norms of their social environment. Here, too, there has been some levelling-off in recent years. The number of pupils getting five or more GCSEs at grades A to C rose more than twice as quickly among children of skilled, semi-skilled and unskilled manual workers than among children of professionals. But a huge gap remains - 69 per cent of pupils from professional families get five good GCSEs, compared with 61 per cent of the offspring of other non-manual workers, 45 per cent of skilled manual workers, 37 per cent of semi-skilled manual workers and 30 per cent of the unskilled. Pupils from an unclassified social grouping, including large numbers of children from families with no wage-earner, had just a per cent chance of getting five good grades.

Higher education's social exclusiveness is therefore perhaps not as "absolutely ridiculous" as Clarke would have us believe. There is a philosophical question about the nature of education systems. Are they, as new Labour believes, a tool for social change? (As Peter Mandelson put it in a recent article, the education system is the key to "transforming life chances, giving young people in deprived areas the chance to escape the limits of their birth".) Or are they, as the great French sociologist Pierre Bourdieu and others have argued, really about recording, justifying and reproducing existing class differences?

An interesting philosophical puzzle but perhaps it is best not to get too philosophical when a rhinoceros is charging your way.

ÖйúAƬ