The value-in-exchange changes with distribution. The aim is for every secondary to have at least 30 per cent of pupils achieving five good GCSE grades by 2011. But the national average is well above 30 per cent, so this aim could be easily achieved by making the allocation of school places fairer, ending academic selection and other segregating characteristics such as religion, specialism and foundation status.

The Department for Children, Schools and Families says: “Many of the local authorities with the most National Challenge schools are those in selective areas where ‘secondary moderns’ face particular challenges. As part of the school improvement strategy there will be intensive engagement to raise attainment in these schools where pupils often start school with low aspirations.”

An easier way for Sir Mike Tomlinson to achieve this would be to mix grammar and secondary modern intakes. In all selective areas this would lead to no school with fewer than 30 per cent “good” GCSEs. A better mix also tends to lead to higher attainment and higher aspirations. If we could identify schools with “unacceptably low standards”, the National Challenge might still make some sense. But to use position in the bottom quartile according to contextual value-added (CVA) scores as a diagnostic reveals the scheme’s lack of logic. CVA is a zero-sum calculation. All schools could have really high standards and a quarter would still be in the lowest quartile. Eliminating unacceptably low standards using this definition is simply unachievable.

Until an individual or co-operative method of assessing school performance is developed, the flawed CVA methods will ensure that a quarter of schools remain in the lowest quartile each year. By definition. This means nothing. The challenge will fail because it is ill thought out - unless, that is, it is merely a means of meeting the target for academies by other means.

Stephen Gorard, Professor of Education Research, School of Education, University of Birmingham.