IT feels, suddenly, as though the wheels have come off. After three years of smooth and rapid progress towards the primary-school literacy and numeracy targets, with schools at full throttle and the destination in sight, ministers have found they are no longer moving forwards.

The recently-published test results showed that 75 per cent of 11-year-olds were reaching level 4 in literacy, set against a target of 80 per cent for this year. In maths, 73 per cent of pupils were scoring level 4 in the tests when the national target was 75 per cent.

The fact that the targets were missed is bad enough, although in the case of maths the numeracy strategy has delivered most of what was demanded for both girls and boys. Far more worrying for the Department for Education and Skills is the fact that the initial year-on-year progress brought about by the literacy and numeracy strategies looks to be at an end. This year’s literacy figure shows no improvement on 2001, while the numeracy score actually declined last year before making a two percentage point improvement this time around. Privately, ministers and their advisers are scratching their heads about a block of children who seem unable or unwilling to move beyond level 3.

Nor are these ambitious key stage 2 targets the only ones that look to be out of reach for the moment. The number of 16-year-olds who leave compulsory education with no GCSEs to their name remains stuck at about 40,000 - a grave threat to ministers’ promise that 92 per cent will achieve five passes (A-G including maths and English) by the time they leave compulsory education.

Boys are taking most of the blame - as they have done for the past three years. Their apparent failure to make the grade even prompted ministers to launch an investigation into ways in which they might be persuaded to improve although, needless to say, no magic bullet has been unearthed. And slowly but surely, attention is beginning to return to a rather more traditional form of explanation and analysis - based on poverty and inequality. Ever since they came to power, ministers have been avoiding the link between family background and academic attainment: poverty, we have been repeatedly told, is no excuse for failure.

Now for the first time there are signs it is coming back on to the Government’s agenda, a development signalled right at the top when Prime Minister Tony Blair used the Redistribution word - in advance of the Trades Union Congress conference and a difficult debate on Iraq, it must be said. Chancellor Gordon Brown’s arm of policy-making, meanwhile, is nakedly redistributive through schemes such as Sure Start and the Working Families’

Tax Credit.

In a recent interview, Professor David Hopkins, the new head of the Standards and Effectiveness Unit, was questioned about the contribution of economic inequality to classroom achievement.“This is certainly high on my agenda,” he said. “Research by the Organisation for Economic Co-operation and Development is suggesting, for the first time, that the right combination of strategies and policies can begin to address poverty and social class as determinants of educational performance. That’s something I feel that there is an increasing commitment to in this Government.”

This is less boring than it sounds and, translated from the mandarin, points to something quite significant. It is an acknowledgement that, even if poverty and class are no excuse for failure, they will have to be explicitly addressed in future strategies.

It is hardly rocket science, of course, and outside Whitehall it is a widely-held view. Professor David Reynolds from Exeter University, for example, believes we are unlikely to see much more progress at KS2 until questions of disadvantage are tackled. Professor Reynolds is particularly interesting on this point, because he was in charge of the committee which devised the national numeracy strategy.

Family background and the social mix of a school population have long been established as significant determinants of a child’s future, academic and economic - a depressing truth that can be seen in most of the longitudinal short studies conducted by Professor John Rutter and Professor Peter Mortimore and others for the past 40 years. The rate of entitlement to free dinners - effectively the unemployment rate among the parents - at a school correlates strongly with poor GCSE results.

Professor Mortimore, the former director of London University’s Institute of Education, established that a good working-class school, with dedicated staff and top-level teaching, tends only to reach the level of the average middle-class school. Gender, in comparison, has only a small effect - working-class girls do worse than middle-class boys, for example. Moreover, recent evidence suggests that social mobility is actually declining, which is bad news for Gordon Brown’s already bold promise to abolish child poverty within a generation.

If, then, poverty is back on the educational agenda, the question facing ministers is how to deal with it. The answer is almost certainly that, as Professor Hopkins suggests, they will look to better teaching methods rather than wider solutions that might involve meddling with pupil intakes, limiting parental choice or engaging in expensive community regeneration.

After all, they can point out that the gloomy predictive research about working-class children tends to deal with the aggregate effects of home background rather than with individuals. Time after time individual children have proved themselves capable of defying their home circumstances and, with the right support, achieving well. So that, while GCSE results have a high correlation with home background for pupils in general, for individual pupils the correlation is very much weaker.

This is why New Labour has chosen to take the high road of optimism, more or less dismissing educational inequality as an explanation for poor results and concentrating on the improvements that can be made in the classroom. If some can do it, the argument goes, so can the rest.

Professor Reynolds is certainly confident that improved teaching is one way forward, even if it is not the only answer. For a start, he thinks that the numeracy strategy would have been more effective had it been modified for different balances of school intake. So, for example, there is plenty of research evidence from American studies in the Eighties that spending half the lesson on “whole class interactive” teaching is effective for working-class children. But there is no evidence that it helps the most affluent pupils who may well have benefited from, for example, spending only 25 per cent of the lesson in this way.(Arguing that one social class of child could do with a different style of teaching from another is, of course, politically difficult).

Professor Reynolds believes that very much more can be done to identify what elements of support turn individual working-class children into academic high-achievers when their peers continue to do badly.

But these are dangerous waters. Sooner or later any discussion of that sort is bound to run into the mass of research saying that what they most need is the help of middle-class peers. A mixed social intake has long been identified as a key element - and emerged yet again in the data from the PISA international study, according to Peter Robinson, head of research at the Institute for Public Policy Research.

“If your child goes to a school surrounded by lots of disadvantaged kids, your child is likely to do less well,” he said. “The one thing that really struck me from the PISA results in 2000 is that they found evidence for the same peer group effect across all the countries they studied. They suggested that the peer group effect was a more powerful predictor than some of the individual factors about background and class.”

No one doubts these findings, but a renewed interest in poverty does not mean that ministers will start questioning the wisdom of its policies on parental choice and open enrolment. For the time being, they would rather blame boys and pour millions into city academies.