In our last blog post, we looked into a new and deeply flawed ‘research report’ by former Department of Education official Iain Mansfield. It was written in support of the Prime Minister’s argument that expanding grammar schools across England would help ‘ordinary working families’.
The pro-grammar lobby know they’ve lost the social mobility argument in terms of grammar schools and children on free school meals (FSM). Less than 3% of children at grammar school are listed as FSM, and most of those only spend a short time on FSM. That is why Theresa May has shifted the argument to what she calls ‘ordinary working families’ i.e. above free meal entitlement but low income.
Faulty figures have been used to suggest that 45% of grammar school pupils are below median income – an unlikely calculation. Even if that were true, the data in Mansfield’s report still ends up revealing that grammar schools are letting down low income families.
The report (page 53) draws on data from Cambridge University. It divides the population into five: the lowest quintile Q1 are largely FSM. The second quintile Q2 fit Theresa May’s notion of ‘ordinary working families’. Cambridge admitted 572 grammar school pupils in 2017, but only 22 from Q1 and 32 from Q2, compared with 297 from Q5. Even this might not tell the whole story: the quintiles are based on the areas where children live, and not every Cambridge entrant living in Q1 or Q2 is from a low income family.
The argument for grammar school expansion falls flat on its face. If grammar schools were set up across the rest of England, i.e. if another 500 or so grammar schools were opened, only an extra 20 or so young people would reach Cambridge – hardly the basis for such major policy change.
Of course Cambridge University itself must carry some responsibility: adding together comprehensives, grammars and private schools, its Q1 (bottom fifth) students make up only 3% of Cambridge entrants and Q2 (second lowest fifth) another 6%, with 56% of places going to Q5 (the top fifth).
Other high-status universities
The report also looks at grammar and comprehensive school pupils going to the “top third” of universities, according to Government definitions. Here again, the pro-grammar school argument collapses.
We learn that 53% of grammar school students move on to these ‘top third’ universities, compared with 23% of comprehensive school students. This is hardly surprising, since grammar schools only select the most academic at age 11.
Given that grammar schools admit roughly the highest achieving quarter of their local population, this 53% should be divided by four to provide a fair comparison. In other words, we should be comparing 23% from comprehensive schools with 13% of the population of areas served by grammar schools – hardly data which supports the case for expanding grammar schools across England.
Mobility for minority ethnic students
The most worrying data in Mansfield’s report concerns BME pupils. He argues that BME pupils are far more likely to progress to Cambridge if they go to grammar schools:
486 from grammar schools vs 362 from comprehensives
This is worrying because grammar schools are located in only a quarter of local authorities, albeit quite affluent areas. However even this data needs close scrutiny.
Firstly, Mansfield’s report has combined three years figures without any explanation. For 2017 the annual figure is:
160 from grammar schools vs 133 from comprehensives.
Secondly, a quarter of grammar school pupils come from outside their own local authority, mostly from non-selective local authorities. Allowing for this, the adjusted figure would be something like:
120 from local authorities with grammar schools vs
133 from comprehensive local authorities.
Still worrying, but what conclusions should we draw from this? These are small numbers, spread across the whole of England. No experienced researcher would reach big conclusions on the basis of such small numbers.
We know that the ‘proportion of pupils from non-White backgrounds going to grammar schools is higher than in other schools’. It would only take one Indian hospital consultant, perhaps, in each non-selective local authority to send a highly academic daughter or son to a grammar school in a neighbouring authority to cause this imbalance. Perhaps just for the sixth form… and maybe with GCSE results that would have won her a place anyway… Given the rest of the data from Cambridge (see earlier), very few of these BME students can be from ‘ordinary working families’.
This it not to suggest that grammar school teachers are doing a bad job… simply that policy decisions should be based on clear reasoning and solid evidence.