With grammar schools seemingly back on the agenda – it feels like 2016 all over again. Thankfully we are not at that stage yet, but both candidates in the Conservative leadership election have shown their support for grammar schools, in the hope it will get them the support of party members.
It has reignited the debate on grammar schools, with the usual arguments for those in support of selecting pupils at age 11 resurfacing.
But does the data back them up?
The argument invariably begins with the claim that grammars create a great levelling opportunity, where all pupils can thrive no matter their background.
Just over five per cent of pupils who attend grammar schools are eligible for free school meals (FSM), around a quarter of the national average. Such disparities should come as no surprise given that by the end of primary school, pupils from low income families are typically 9 months of learning behind their peers.
They kick off the 11+ a couple of goals behind, and facing a more affluent opposition who are more likely to have benefitted from private tuition when efforts to make tests ‘tutor-proof’ have been unsuccessful.
People may argue that we shouldn’t judge grammar schools by attainment gaps earlier on in the school system, or that we might expect FSM rates to be lower given the areas they serve.
But after allowing for this, the FSM rate in grammars is still about a third of what we would expect and those that do make it in are not typical of the disadvantaged pupil population as a whole.
A commitment to expand academic selection can feel like the default policy when politicians have no sense of how they are going to tackle the real school issues
They tend to have been eligible for free school meals for a shorter time than average and this is particularly important within the context of the disadvantage gap where the gap for the persistently disadvantaged is wider than for the disadvantage group as a whole.
There has been no progress in closing that gap in over a decade, and concerningly the persistently disadvantaged are a growing group.
The argument then moves to what grammar schools achieve, not judging them by who they take in. Though it’s easy to argue that their outcomes are achieved precisely because of who they take in.
Our own research did find a small positive attainment effect for grammar school pupils that was equivalent to at most a third of a grade in each subject at GCSE. But in the most selective areas we also found a negative effect for those that didn’t get in, equivalent to around a tenth of a grade in each subject.
In essence, selection does not increase overall attainment, it redistributes who gets the grades. These differences can persist, with the wage distribution for those who grew up in selective areas more unequal than those in comprehensive areas.
But what about if you want to argue that this isn’t about outcomes for all, but for those at the high end of prior attainment? Well, we found that those pupils do just as well in high performing comprehensives (and there are a lot more of them).
And if you want to argue that more grammars are what people want, then recent polling found just 30 per cent of adults in favour of more selection, with nearly half saying either that there should be no increase, or that selection should end completely.
If ministers do want to expand selection – perhaps in a few limited areas – the question becomes where? When we examined this we found very few areas where there was likely to be demand, and where the opening of a grammar school would not be damaging to existing schools.
Ultimately, a commitment to expand academic selection can feel like the default policy when politicians have no sense of how they are going to tackle the real issues facing schools.
When they have no clear policies to address the fact that successive governments have failed to attract enough teachers into the profession nor do enough to make them want to stay, or where they are unwilling to find money to address the near fifteen-year school funding squeeze.
And saying you will open a grammar school to give children from poorer backgrounds more opportunities in the future is a lot easier than saying how you will make sure they are not poor right now.
Cognitive Ability Test (CATs) were carried out at my school on naive year 7 pupils and again two years later, when the same cohort was in year 9. Divide the year 7 scores into the top 40% for selection (the exact percentage does not matter for this argument – it could be anything from 20% to 60%). Now look at the scores in year 9; of those in the higher selection category, approximately 30% would no longer be considered high achievers. Similarly, for those in the lower band, there was a churning of cognitive ability; some who in year 7 were considered special needs for low attainment were well above average two years later.
One boy who score in the top 2% in year 7 left school with virtually no qualifications, and a girl with scores of straight 100 points in all three categories of CAT attained a degree and is now teaching.
I spoke to some academics at Cambridge who told me it was not worth publishing as it was well known that cognitive ability is as variable as any other growth factor. For example, can we predict growth? Will the smallest person in year 7 remain the smallest in all years? is the smartest person in year 7 likely to get all top grades?