This government’s desire to tackle the roots of social exclusion and Tony Blair’s personal pledge to improve the lot of the poorest in our society is admirable.

I also applaud (both as a teacher and as a former Labour parliamentary candidate) its determination to break the cycle of deprivation and low expectations symptomatic of urban inner-city Britain.

But the reality is that, after raising benefits, the best means of reducing poverty is to raise educational aspirations and standards. Evidence strongly suggests that a comprehensive system of schooling is the best way to achieve this. My challenge to Tony Blair and Charles Clarke is to urge them to end selection in England.

Comprehensive schools work. Recent surveys have shown that there is only one factor more powerful than a pupil’s social background as a predictor of academic performance at 16 and that is the average social background of other pupils in the school.

Comprehensives help to level up. The predominantly comprehensive system in the UK produces better overall results than mainly selective Germany. It therefore would be logical to remove selection in the public sector altogether.

The prominence of highly selective schools at the top of local league tables, or in annual lists of the country’s “best” schools, proves nothing except that academically able children usually do well in exams. It is not true that selective local systems push up standards in all schools.

A value-added analysis comparing selective and non-selective local education authorities shows that where schools in an area are organised on selective lines (15 out the 150 LEAs), it depresses the educational performance as a whole.

Grammar schools may do well for the able children they select, but the evidence indicates that the performance of the other 75 per cent is lower than in non-selective systems.

As The TES recently reported (January 17), there have been concerns about the quality of provision in Kent (which has the most grammar schools in the country) where schools are “substantially” more likely to require special measures or to have serious weaknesses.

It is widely assumed that academically able children do better in selective schools while children of average and below-average ability do better in non-selective schools. In short, a trade-off may have to be made to reflect the priority of the few or the many.

Nevertheless the Department for Education and Skills’ statistical comparison of the GCSE and GNVQ results of all grammar school pupils against the top 25 per cent (equivalent to ‘grammar school ability’) in comprehensive schools indicated that the comprehensive schools had done slightly better overall. There is also evidence that fully comprehensive systems reduce the gaps in attainment between children of different abilities and between children from different social class backgrounds.

In 2002 almost half of all 15 to 16-year-olds in maintained schools achieved five or more “higher passes” at the end of compulsory schooling, compared with a mere 23 per cent who attained the equivalent five O-levels in the 1970s. Given there was no real increase in education expenditure between the mid-1970s and the late-1990s this remarkable increase in productivity as measured by qualifications is attributable, in large part, to the removal of the barrier of the 11-plus for some four-fifths of the population.

The surviving grammar schools are mostly schools for the middle classes. In England in 2002, the proportion of children eligible for free school meals (an imperfect but commonly used indicator of social disadvantage) was consistently lower in selective than in non-selective schools.

Since comprehensive education was introduced, barriers to achievement for many young people have been removed. The annual government statistics of school attainment, exam results, and participation in further and higher education all indicate a ‘levelling-up’ over the past 25 years. So come on Tony, be bold, expand the number of specialist schools, level up and get rid of selection once and for all.

Michael Ion is deputy head of Blessed Robert Johnson Catholic College in Telford. He is due to make the case for ending selection at a regional Labour party conference in Malvern today