In 2016 the Government launched a 12-week public consultation, Schools that Work for Everyone, setting out its intention to allow selective schools to expand, for new ones to be set up and for non-selective schools to become selective.

The Education Policy Institute published new, rigorous data analysis and evidence to inform this important debate. This is the most detailed study of Government education data on selective schools for almost a decade.

You can read the full report here.

Key findings


1. Once prior attainment and pupil background is taken into consideration, we find no overall attainment impact of grammar schools, either positive or negative.

  • At school level, grammar school pupils perform highly in raw attainment terms, with 96.7% of their pupils achieving five A*-C GCSEs, versus the national average of just over 57% in all state-funded schools.
  • This high performance is driven however by the very high prior attainment and demographics of pupils in grammar schools.

2. Pupils who are eligible for free school meals (FSM), a proxy for disadvantage, are under-represented in grammar schools. Only 2.5 per cent of grammar school pupils are entitled to FSM, compared with an average of 13.2 per cent in all state funded secondary schools.

  • A main cause of this significant under-representation of disadvantaged pupils in grammar schools is that, by the time the ‘11 Plus’ entry exam (or equivalent) is taken, 60 per cent of the disadvantaged attainment gap – equivalent to 10 months of learning by this stage – has emerged.

3. We do not find a significant positive impact on social mobility. The gap between children on FSM (attaining five A*-C GCSEs, including English and Maths) and all other children is actually wider in selective areas than in non-selective areas – at around 34.1 per cent compared with 27.8 per cent. Our analysis indicates the reason for this is:

  • grammar schools attract a larger number of high attaining, non-FSM pupils from other areas and so, in selective areas, there is a disproportionately large number of high attaining, non-disadvantaged children. Indeed, pupils travel, on average, twice as far to attend a selective school as a non-selective school.
  • pupils eligible for Free School Meals in wholly selective areas that don’t attend a grammar schools perform worse than the national average.

4. An expansion of grammar schools in areas which already have a large number of selective schools could lead to lower gains for grammar school pupils and small attainment losses for those not attending selective schools – losses which will be greatest amongst poor children.

  • In the most selective areas, the positive effect of attending a grammar school is 2.3 GCSE grades spread over 8 subjects (0.3 grades per subject).
  • Within those highly selective areas, that gain falls to 0.8 of a grade overall (or 0.1 of a grade in each of eight GCSEs), in areas where grammar school places outnumber the proportion of high attaining pupils.
  • In the most selective areas there is a small negative effect of not attending grammar schools – an average of 0.6 grades lower per pupil across all GCSE subjects (or just below 0.1 grade per subject).
  • But that impact is greater for pupils eligible for free school meals who do not attend grammar schools, they achieve 1.2 grades lower on average across all GCSE subjects (or just below 0.2 of a grade lower in each of eight GCSEs).

5. If you compare high attaining pupils in grammar schools with similar pupils who attend high quality non selective schools, there are five times as many high quality non selective schools as there are grammar schools.

  • This means high attaining pupils perform just as well in high quality non-selective schools as selective schools.
  • These are schools which are in the top 25 per cent based on value added progress measures and represent good quality schools operating at a large scale
  • These schools are much more socially representative than grammar schools, admitting close to the national rate of FSM pupils (12.6% versus 13.2% nationally, and just 2.5% in grammar schools). They also admit close to the national share of children with special educational needs.

6. Other interventions to raise school standards and attainment have proven to be more effective than grammar schools in raising the attainment of disadvantaged pupils. The Labour sponsored academies programme has had a more positive impact on the attainment of disadvantaged pupils compared with the present grammar school system. This finding is based on:

  • Research commissioned by the Education Policy Institute from the LSE, which showed that the Labour sponsored academies demonstrated average attainment gains of one grade in each of five subjects (or 0.6 of a grade over eight subjects). The pupil intakes of grammar schools and sponsored academies are clearly very different in terms of prior attainment, but it is notable that those early sponsored academies educate around 50,000 FSM entitled pupils compared to around 4,000 such pupils in grammar schools.


In December 2016 the Education Policy Institute published its second major report in this area  – Grammar schools and social mobility: Further analysis of policy optionsThis report looks at what the effects would be of putting new grammar schools in different parts of the country, and assesses the impact of introducing quotas for free school meal children. You can read it here.