Children who go to grammar schools in England achieve better grades than those of similar ability who are not in selective areas, researchers claim.
A Bristol University study suggested pupils from poorer backgrounds do particularly well.
But relatively few children from the poorest families go to grammar schools.
Probably because bright children from poor families have already suffered from particularly poor state education by the time they reach their 11+. On a larger scale, the left has always been against grammars, so traditionally Labour areas are more likely to have abolished them altogether. And these areas are, on average, poorer than Tory-controlled areas, which are more likely to have kept selection, to one degree or another. So, across the country as a whole, children from poor families are simply less likely to have access to a grammar school.
Overall there is little or no impact on attainment, but the quarter of children educated in grammar schools do substantially better than their peers in similar non-selective areas.
By substantially, they mean 3.5 grade points - the equivalent of raising 3.5 GCSEs from grade C to B.
The other 75%, who did not go to grammar schools, were disadvantaged by 0.5 grade point.
The Labour government promised not to extend selection but did not abolish the existing grammars, where pupil numbers have in fact increased.
Instead it set up a system of parental ballots. Only one was held, in Ripon, where parents decided to keep the selective system.
Pro-comprehensive campaigners say the ballot process is biased against those who want to get rid of grammars.
Why only have ballots about abolishing grammar schools, in those areas that are still lucky enough to have them? Why not let people who don't have the opportunity to send their children to grammar schools vote on whether they would like that opportunity?
And they're not pro-comprehensive campaigners. They're anti-grammar campaigners.