Non-grammars are most socially selective state schools
1 October 2008
The vast majority of England's most socially selective state secondary schools are non-grammar schools, a major review carried out for the Sutton Trust reveals today. However, England's remaining grammar schools are currently enrolling half as many academically able children from disadvantaged backgrounds as they could do.
The research, by the Centre for Evaluation and Monitoring at Durham University, found that only 17 of the 100 most socially selective state schools in England are grammars. 54 of the 100 most socially selective schools are faith schools.
Comparing the national test results of pupils at age 11, it found that 50 non-grammar schools are also more academically selective than the least selective grammar school in England. However, when the research considered the top 25% of academic achievers at age 11, it found that just under 2% of those in grammar schools are on Free School Meals, compared with 5.5% in non-selective schools.
The Durham University study of 125,000 grammar school children between 11 and 16 is the most comprehensive review ever produced on the academic characteristics and social impacts of the remaining 164 grammar schools in England. The researchers led by Dr Rob Coe have developed new methods to determine the social and academic 'selectivity' of schools by comparing the characteristics of their pupils with other children in the same electoral wards. It is the first time that the geographical spread of grammar school intakes has been mapped in this way.
The research also concludes that the impact on the academic results of non-grammar state schools' due to the 'creaming off' of pupils to grammar schools is negligible. They conclude: "Grammars have a widespread, low-level, impact on pupil enrolments across the sector. A relatively small number of non-selective schools do see a significant proportion of pupils 'lost' to nearby grammars, but the research suggests that this does not damage such schools, at least in terms of academic achievement."
They also found that on average pupils in grammar schools achieve between zero and three-quarters of a GCSE grade per subject more than similar pupils in non-selective schools.
Dr Lee Elliot Major, Director of Research at the Trust, said:
"One of the most unexpected findings from this study is that the most socially – and in some cases academically – selective state schools in the country are not grammar schools at all, but 'comprehensive' schools. Top state schools – whether non-grammar or grammar – are steeped in traditions of educating pupils from all backgrounds, yet the stark reality is that their pupil intakes have little in common with the local communities in which the schools are based.
"The figures also suggest that grammars are not enrolling as many academically able pupils from disadvantaged backgrounds as they could do. While ideological debates will continue to polarise opinion over whether England's remaining 164 grammars should exist, we believe more spotlight needs to be focused on ensuring grammars do all they can to reach out to all potential pupils."
"We must monitor effectively the impact of the Government's latest secondary school admission reforms to see if they improve the situation. But we should also carry out a review of 'eleven plus' selection tests to see whether they deter bright pupils from disadvantaged backgrounds applying to grammar schools and urge more grammar schools to adopt outreach programmes such as the Trust supported scheme at the Pate's Grammar School in Cheltenham."
Dr Robert Coe, leader of the research team at Durham University's CEM, said:
"Some of the commonly held beliefs about grammar schools are just not true. For example, it turns out that there are no self-contained selective areas whose performance can be compared with comprehensive areas. Many pupils travel large distances to attend grammar schools, often crossing Local Authority boundaries. A large number of non-selective schools each lose a small number of pupils to grammar schools so the creaming effect is very widespread but low-level.
"Another common view is that grammar schools raise achievement, but secondary moderns depress it. We tried lots of different ways to test the first part and many of the models did suggest that those who get in to grammar schools achieve better GCSEs than similar pupils in other schools, though other models showed no difference – it just depends what assumptions you make. Also, grammar school pupils appear to be already making better progress between KS1 and KS2, before they even set foot in a grammar school. On the second part, we found no evidence that the performance of secondary moderns, or any schools creamed by grammars, was different from other schools.
"Grammar schools as a whole do not seem to take their fair share of economically disadvantaged pupils, although there are other groups of schools for whom this is just as much true. The ways in which school choice and selection operate to reinforce disadvantage within the system as a whole are complex and subtle. Abolishing grammar schools, without addressing these systemic problems, might do little or nothing to increase fairness overall."