Gains might just reflect greater success at teaching to the test, in which incremental gains should be easier to achieve at the low than at the higher end of the curve. But the gains inspire modest optimism.

Nor is the Government at fault for the smaller gains at GCSE. In 1997 it targeted primary, not secondary, schooling, and rightly so since that is the most cost-effective arena for intervention. No primary level gains would yet have affected GCSE performance. In general, commentators expect interventions to bring results unreasonably quickly.

It will take serious investigation to figure out what is responsible for the gains. A first guess would be a combination of factors. A buoyant labour market improves morale , as well as income.

The numeracy and literacy strategies have focused teachers’ efforts. The Government has improved early-years provision, and has tried, with varying degrees of success, to improve standards in the worst-achieving schools. Teachers, despite the low esteem in which they seem to be held, continue to work unreasonably hard.

Could things have been even better? The pledge to limit class sizes to 30 was a serious mistake - it has exacerbated the teacher shortage, and there is no evidence that incremental reductions in class size benefit pupils. Substantial reductions do. The Government should target the most deprived areas for primary class reductions to 20 or fewer.

The unreasonable bureaucratic burden has also worsened the teacher shortage, and undermined the energy and dignity of teachers. It must be reduced urgently.

The most efficient way to raise the achievement of the most deprived is to ease deprivation. Improving the schooling of the least advantaged is urgent.But there is still no alternative to redistribution.

Harry Brighouse is professor of the philosophy of education, Institute of Education, London University