Academics take aim at ‘flawed’ grammar school study

Eight leading education academics have hit back at a “flawed” study that claimed poor pupils at grammar schools were twice as likely to attend Oxbridge.

The Higher Education Policy Institute has published a collection of essays today, titled ‘Social Mobility and Higher Education: Are grammar schools the answer? ‘

It was published partly in response to a controversial study last year from Iain Mansfield that claimed poor pupils living in regions with grammar schools were twice as likely to attend Oxbridge than those in non-selective regions.

But Matt Dickson and Lindsey Macmillan, from the University of Bath and UCL Institute of Education respectively, said the most “reliable conclusion that can be drawn is that social mobility – as measured by progression to elite higher education – is unequivocally damaged by the selective schooling system”.

They added there were “significant data issues” in Mansfield’s study which mean the proportion of disadvantaged pupils attending selecting schools is “overstated”.

The pair’s essay also claims Mansfield’s statistical analysis was “flawed” as it conflates correlation with causality – “therefore severely overestimating the selective school advantage”.

Finally, they say the report’s conceptual methodology is “limited” as it ignores the impact of selective schooling systems on those pupils not attending grammar schools.

Mansfield also previously claimed academic researchers could be ignoring the benefits of grammar schools because they are politically biased against them.

Professor Vikki Boliver, from Durham University, and Dr Queralt Capsada-Munsech, from the University of Glasgow, added: “To dismiss a virtual consensus among academics as merely the “views” of experts is to fail to appreciate the difference between subjective beliefs and objective evidence.”

They state most empirical studies find selective education does not boost educational attainment or foster social mobility.

The collection ends with vice-chancellor of the Open University Professor Tim Blackman calling to reduce selection in higher education, rather than increase it in secondary education.

Blackman states that high entry grades are used to “signal prestige” and there is no expectation on universities “to add value to students’ prior attainment”.

He states less selection would create “more diverse and inclusive” student communities which would enable peer learning and “likely improve educational outcomes”.

Mansfield, who was special adviser to former universities minister Jo Johnson, said the paper raises “some interesting points”. But he added it “contains a number of significant issues that call into question its value as a definitive statement on this subject”.

“Notably, at no point is the paper able to challenge the most striking finding of my report last year – the five-fold increase in propensity for BME children in selective areas to progress to Oxbridge.”

Former prime minister Theresa May’s attempts to end the ban on new grammar schools were unsuccessful. The current government is focused instead on expanding existing grammar provision, with a £200 million fund available to selective schools that agree to improve access for poorer pupils over four years.

However a Schools Week investigation found despite the promises to open up access from the first 16 grammar schools to get cash, they still lagged behind comprehensives.

A Department for Education spokesperson said: “Selective schools are some of the highest performing schools in the country.. They can only receive expansion funding if they meet the high bar we have set and clearly demonstrate how they will admit more pupils from disadvantaged backgrounds.”


Your thoughts

Leave a Reply to Janet Downs Cancel reply

Your email address will not be published.


  1. The HEPI report used DfE provisional analysis. The DfE warned caution should be used when interpreting the stats. The data had several limitations.

    But HEPI did not use caution.

    HEPI defined ‘disadvantaged’ as being below median income which scoops up many more children than the accepted (if crude) pupil premium definition. The DfE provisional data showed 45% of grammar school pupils were from below median income households but the figure was much higher in non-selective schools: 67%. This was not mentioned by HEPI.

    The provisional data, used with such confidence by HEPI shows grammar schools still took far fewer ‘disadvantaged’ pupils than non-selective schools.

    • CLARIFICATION: The HEPI report in my comment above was the one written by Mansfield a year ago. Apologies.

      Although Nick Hillman says in the preface to the latest paper that the first one was to stimulate debate, this was not clear in Mansfield’s paper. Hillman wrote a foreword which said:

      ‘The debate on grammar schools has become very one sided. Researchers line up to condemn them for inhibiting social mobility, and the schools do not perform well on every single measure. But the full evidence is more nuanced and shows some pupils benefit a great deal.’

      This sounds like endorsement to me.

  2. Re BME students which Mansfield says his critics ‘notably’ didn’t address. Mansfield wrote:
    ‘Astonishingly, 163 grammar schools sent over 30 per cent more BME entrants to Cambridge (486) than the nearly 2,000 non-selective schools combined (362). With more than threequarters of the country having no grammar schools, these fgures (sic) represent a shocking indictment of the comprehensive system.’
    But there were 3,436 state-funded secondary schools in England according to data available at the time. It seems Mansfield mislaid a large proportion of English state secondary schools.