Education outcomes for pupils attending non-selective schools in areas with grammars are lower than for similar pupils in non-selective areas, according to a new analysis.

The research by FFT Education Datalab found that students attending secondary moderns and comprehensives in selective areas did worse when it came to getting the best GCSE grades and going to top universities.

However, the chief executive of the Grammar School Heads’ Association told Tes the research failed to take into account changes made by grammars in recent years.

Exclusive: Grammar schools’ ‘insult’ to teachers as sixth-formers coach 11-plus

Profile: Chair of Grammar School Heads’ Association Jim Skinner

Read: Grammar schools ‘are revolutionary for disadvantaged pupils’

The Datalab analysis looked at the cohort which turned 16 in 2009 so it could examine outcomes up to the end of 2015-16 - the latest year for which long-term outcomes data is available.

It looked at the outcomes of pupils who go to a non-selective schools but live in areas where at least 20 per cent of pupils go to grammars, comparing their outcomes to “matched” pupils with similar prior attainment and socioeconomic and ethnic backgrounds in non-selective areas.

“Educational outcomes for pupils attending non-selective schools in areas served by grammar schools tend to be lower than for similar pupils living in non-grammar areas,” Datalab said.

“Although the percentage of pupils achieving five or more A*-C grades including GCSE English and maths was the same, the former group tend to do worse on higher-level outcomes. For instance, they were half as likely to achieve five A*-A grades at GCSE or attend a top third university.”

The analysis goes on: “It looks like outcomes for pupils who went to grammar schools are slightly higher than those of similar pupils who didn’t, but these have to be balanced against worse outcomes for those who don’t get in.

“More generally, pupil outcomes tend to be higher when pupils attend schools with higher attaining cohorts. This is the case whether pupils have low or high prior attainment.

“The grammar school debate will ultimately always come down to whether you think it is acceptable or not to segregate those with high levels of prior attainment (or measured ability) at a given age from other pupils.”

Jim Skinner, the CEO of the Grammar School Heads Association, suggested the research could be out of date because of “fair admissions” policies introduced in recent years.

“We’re looking at a cohort of pupils who took the selection tests in 2003, and a huge amount has changed since then,” he told Tes.

Datalab said they had undertaken the research in light of the Conservative leadership contest. Several candidates in the race, including Boris Johnson and Dominic Raab, have been supportive of grammars.

However, Mr Skinner played down the chances of grammar school expansion under a new prime minister, at least in the short term.

“We mustn’t forget the reason it didn’t go forward with Theresa May was when she lost the majority in the election, and that hasn’t changed,” he said. 

“I don’t think we’re anticipating any significant change in practice this side of the next general election.”