Selective education: some evidence

Does selective education work? The CMPO has gathered some evidence on this (it'll appear here shortly).
First, David Jesson claims that the 164 grammar schools that currently exist in England "do not offer a ladder of opportunity to any but a very small number of disadvantaged pupils". This is because they just don't go to such schools. Only 2% of grammar school pupils are eligible for free school meals. Seven times as many were at private school before entering the grammar.
Second, Damon Clark found a neat natural experiment to compare how grammar schools compared with non-grammar. He looked (pdf) at those pupils in the East Ridings of Yorkshire in the early 70s who had 11+ scores on the borderline for grammar school entry. Some attended grammars, and some didn't. And he found that those who attended grammars did no better on maths and English tests in their fourth year than those who attended other schools. Grammars did, however, enable their pupils to take more O levels - especially pupils from poorer backgrounds - which might have increased their chances of going to university.
A more positive take on grammar schools comes from Eric Maurin and Sandra McNally, who looked at the effect of expanding Northern Ireland's grammar school intake in 1989.  They found that this significantly improved GCSE scores, with the impact similar for both poorer and  wealthier students. Wider access to grammar schools, it seems, improves standards but not equality of opportunity. Or put less kindly, excluding pupils from grammar schools damages them.
These results seem rather inconclusive. As this paper (pdf) concludes, "we probably do not know very much about  the effect of comprehensive schooling."
Which makes one wonder why everyone thinks they know.

Comments

[Dipper]

some counties have retained selection and some haven't. Surely there must now be a load of data to analyse selection at 11?

[Bob B]

Another inconvenient truth:

The London Borough of Sutton has a cluster of outstanding maintained selective schools - of which no less than five feature in the top 40 (forty) state secondary schools in England according to this recent league table from The Sunday Times:
http://www.timesonline.co.uk/parentpower/league_tables.php?t=state_secondary_schools

The continuing consequence is that the London Borough of Sutton rates at the top or close to the top of the regular annual league table of Local Education Authorities because the effect of the outstanding selective schools is to boost the AVERAGE attainment across all Sutton schools in the school leaving exams:
http://news.bbc.co.uk/1/hi/education/6250433.stm

There is a regular myth that Sutton is unusually affluent. It isn't. Data on household income distributions in the London boroughs, posted on the official website of the London Councils, shows that the parameters for income distribution in Sutton are very close to the AVERAGE for all London councils - namely, 21% of Sutton households have a household income less than £15k (compared with 22% for London); 53% have a household income less than £30k (53% for London); and 85% have a household income less than £60k (85% for London).
http://www.londoncouncils.gov.uk/doc.asp?doc=19944

By these official DfES statistics, the percentage of Sutton residents with graduate (level 4) qualifications is above the average for England but below the average for London:
http://www.dfes.gov.uk/rsgateway/DB/STA/t000747/index.shtml

[Nik Khat]

@Dipper Yes there is a ton of data. If it "proved" the case for "social mobility" and a "better" overall education don't you think they pro 11plus crowd would be shouting it from the rooftops? In fact in Buckinghamshire the opposite is true - all the data shows ethnic minorities & white children from deprived areas do badly - that is they don't get in to Grammar schools.To compound the disparity only 1 or 2 of the many "upper" schools is above the national average.

Remember by definition in a selective system the vast majority 70-80% of children will go to Upper schools.