The row over pupil performance and class size escalated this week as research revealed that grant-maintained schools which achieve some of the best exam results had worse pupil:teacher ratios than their local authority counterparts.
Comparisons between local authority and GM schools in London disclosed larger classes and worse PTRs in the opted-out sector.
The London Research Centre claimed that there were on average 26.3 pupils per teacher in an outer London GM school compared with 17.1 at a local authority school. It also discovered GM schools took in fewer children who were eligible for free meals and who had special educational needs.
Ian McCallum, the lead researcher, said: "It might not be unreasonable to think that GM schools can manage with large classes because they have pupils who are positively supported at home."
GM schools score well in exam league tables although Gillian Shephard, the Education and Employment Secretary, has said there is no evidence that this is due to their status. However, Robin Squire, the education junior minister, last week claimed there was a "convincing association" between self-government and higher standards. He told the Association of Grant-Maintained and Aided Schools: "There is clear evidence that, on average, GM schools have achieved better test and examination results than LEA schools."
The Government is under pressure from parents to back up its assertions that class size is not important. Last week it was forced to admit that at the beginning of this year 1.5 million youngsters were in classes that exceeded 31 - 80,000 more than in January 1994.
The Office for Standards in Education is expected to report shortly on class size and pupil performance. Chris Woodhead, its chief inspector, has been talking to the Prime Minister's policy unit and believes the new slimmed-down curriculum could be taught in the classes that currently exist.
According to the London Research Centre, 11 per cent of GM secondary classes had more than 30 pupils, compared with 6 per cent in outer London LEA secondaries and 3 per cent in inner London.
A third of outer-London GM primary classes had more than 30 pupils, compared with 24 per cent in LEA primaries in the same areas. In inner-London GM primaries 12 per cent of classes had more than 30 pupils compared with only 7 per cent in LEA schools.
However, 30 per cent of pupils in maintained primaries were on free meals this January compared with 22 per cent in GM primaries and a third of LEA secondary pupils were entitled to them compared with 14 per cent in GM schools.
In LEA primaries 1.4 per cent of pupils had statements of special educational need compared with 1 per cent in GM primaries and 2 per cent of pupils in secondary schools were statemented compared with 1.1 per cent in GM schools.
GM critics point to the high level of social advantage in GM schools in London. Local Schools Information, the local authority funded body, claimed GM schools were actually under-performing at GCSE when their pupil intake was taken into account While 55 GM schools scored above the LEA average for the percentage of pupils gaining five higher-grade GCSEs, 12 scored below the LEA's average GCSE results.
Andrew Turner, director of the Grant-Maintained Schools Foundation, said low levels of children with special needs was due to the location of GM schools. He added: "There is no way we can discriminate against pupils on free meals. "
The research also threw up startling differences between the London boroughs over pupils' eligibility for free meals. In Tower Hamlets 57 per cent of primary pupils were eligible compared to 9 per cent in Kingston. In the secondary sector 66 per cent of pupils in Hackney, Lambeth and Tower Hamlets were eligible compared to 9.6 per cent in Kingston.
Margaret Maden, a member of the National Commission on Education which is to report next month on successful schools in disadvantaged areas, said it was unacceptable that primary children were in classes of more than 30. She said the jury was still out about whether class size affects exam results but the history of a GM school had to be taken into account when considering its results - some were grammar schools while others had chosen to become more selective.