LEE responds to government green paper

Local Equal Excellent have submitted their response to the government’s green paper ‘Schools that work for everyone’. The response sets out the case against the government’s plans to expand grammar schools, as well as the extensive evidence showing that selection suppresses attainment, damages schools and causes social division.

The response states:

‘The government should continue the current prohibition on new wholly or partially selective schools.

The ‘evidence’ presented by the government in the green paper is extremely thin and narrated in a partial way. In the entire section on selection, the government only cites one study (CEM 2008) but calls it something different in each footnote (see footnotes 9, 10, 11 and 13). This gives the impression that the government has sought to represent one evidence source as four different evidence sources.

It is also unfortunate that the government has sought to rely so heavily on a report by CEM, which is an organisation that makes more than £1 million each year from providing 11-plus test papers, and is therefore not impartial.

The absence of any other references in the selection section of the green paper suggests that the government have not looked at any of the evidence on selection since 2008; for instance, Boliver and Swift 2011, OECD 2012, Harris and Rose 2013, IFS 2013, Sutton Trust 2013, Burgess at al 2014, Sullivan et al 2014, Strand 2014, and EPI 2016. Each of these studies presents evidence that is problematic for the government’s central contention that selective education produces a net benefit.

The government also makes various statements to support the case for greater selection which are not referenced, and for which there does not appear to be evidence. For instance, the government says, ‘Many selective schools are employing much smarter tests that seek to see past coaching and assess the true potential of every child.’

There is no evidence that grammar schools have succeeded in using ‘much smarter tests’. In Buckinghamshire in 2013, the thirteen grammar schools sought to introduce a new test that they claimed would be resistant to coaching. But outcomes data from the first four years have shown quite clearly that results for children from poorer homes have got no better, and that private school children have made gains while state school children have done worse (Local Equal Excellent 2015; see also Buckinghamshire County Council website ‘Grammar School Statistics’ section).

The government’s own Chief Scientific Adviser, Dr Tim Leunig, has conceded this failure of Buckinghamshire’s new ‘tutor proof’ test. In November 2016, he told the House of Commons Education Select Committee, “I don’t want to cast any aspersion on their reasons for doing it or on the people who designed the test [CEM], but it didn’t work.”

A further statement which could potentially mislead the public is, ‘Those children that attend selective schools enjoy a far greater chance of academic success.’ Any school that only takes the top 20-30% of most able children will naturally have better exam results than schools admitting children of all abilities. The government should supply the evidence showing that pupils in selective schools do significantly better than the top 20% of children in comprehensive schools. In terms of the suggestion in paragraph 8, that children from disadvantaged backgrounds make better progress at grammar schools, the government could have also quoted this finding from the CEM study:

‘Although these analyses indicate that grammar school pupils appear to make greater progress from KS2 to KS4 than other pupils, we also find that these same pupils were already making more progress from KS1 to KS2. This suggests that there may be important but unmeasured differences between grammar and non-grammar school pupils and somewhat undermines our confidence in these estimates of a ‘grammar school effect’.’ (CEM 2008)

Perhaps more importantly, this point is undermined by the government’s own concession in paragraph 8 that the majority of children in a selective system – the 70-80% who attend secondary moderns – achieve worse results than they would have in a comprehensive system. EPI’s recent study found:

‘There is also a large penalty for pupils living near selective schools but attending non-selective schools, when compared with comprehensive schools elsewhere in the country in the top 25 per cent of schools. If these pupils had attended a high-quality comprehensive school, they could have achieved an additional 4.2 grades spread across eight GCSEs (equivalent to 0.5 grades per subject). For FSM pupils the potential benefits of attending a high-quality comprehensive instead of non-selective schools in areas with grammar schools are larger still. We estimate a 6.4 grade difference spread across eight GCSEs (equivalent to 0.8 grades per subject) for these pupils.’ (EPI 2016)

Given that the government’s stated concern is for boosting the achievement of children from disadvantaged homes, the weight of evidence showing the overall detriment suffered by children from poorer homes in selective systems seems to seriously undermine their case.

The government also needs to clarify the following statement: ‘We believe that there is a case for relaxing restrictions on selective education, in order to provide more good school places within the system.’ Firstly, the government is proposing providing more places in schools that by definition will exclude the majority of children. Secondly, the net effect could be the creation of more poor school places. This is because of the direct impacts of selective schools on neighbouring schools.

In fully selective Buckinghamshire, 10 out of 24 non-selective secondary schools are currently rated ‘requires improvement’, and only one is rated ‘outstanding’. This means that nearly one quarter of secondary-aged children across the county attend secondary schools that ‘require improvement’. This is despite the fact that Buckinghamshire is a relatively prosperous area with below average numbers of children on free school meals. Latest school census data for 2016 shows that a child in Buckinghamshire is more likely to attend a school that requires improvement than a child in the fully comprehensive Royal Borough of Windsor and Maidenhead (which is a ‘statistical neighbour’ in terms of demography).

For any consultation to be meaningful, the government needs to inform the public of all the evidence pertaining to selection, including the extensive evidence that does not support their case.

In particular, there is considerable evidence to show that creating wholly or partially new selective schools will:

  1. cause detriment to other local schools and to the attainment of children attending these schools (Boliver and Swift 2011, Harris and Rose 2013, OECD 2012, Centre for Evaluation and Monitoring [CEM] 2008, Education Policy Institute [EPI] 2016);
  2. cause intensified social and racial segregation in schools and communities (Institute for Fiscal Studies [IFS] 2013, Sutton Trust 2013, Local Equal Excellent 2016, Burgess et al 2014);
  3. deepen the problem of selective areas having much larger attainment gaps than non-selective areas (Strand 2014, EPI 2016).

Furthermore, there is substantial evidence showing that the children who would take up additional places in selective schools would do as well at a good comprehensive school (CEM 2008, EPI 2016, OECD 2012, Sullivan et al 2013).

The consultation questions concede a problem that is inherent in selection – that children at non-selective schools close to selective schools suffer from inferior educational opportunities and outcomes. We are deeply sceptical whether any of the measures proposed by the government in this section are capable of altering this fact. The only way to ensure that non-selective schools close to existing selective schools are not disadvantaged by selection is to abolish selection.

While the government is keen to avoid using the term ‘secondary moderns’ because of perceived negative connotations, there is no escaping the fact that the net effect of this green paper will be to create a new generation of secondary modern schools.

To understand this, it is instructive to consider this question through the eyes of the headteacher of a non-selective school. From their perspective, what benefit can a neighbouring selective school possibly bring to their school which will outweigh the problems that the very existence of that selective school creates? As a direct consequence of the selective school, the non-selective school will be admitting disproportionate numbers of children from poor homes, children with special educational needs and children from certain ethnic backgrounds. Conversely, the selective school will be admitting very small numbers of these children and so will be least well placed to support the non-selective school in addressing the associated challenges, which will range from pedagogy and behaviour management, to inferior resource levels and problems with teacher recruitment and retention.

In terms of teaching and learning, perhaps the most significant challenge for non-selective schools that are close to selective schools is trying to create an optimal learning experience in the absence of the most able children. Modern pedagogy understands children and young people to be co-producers of learning in the classroom. This understanding tells us that it is impossible for any teacher to deliver the same kind of learning experience that is achievable when the most able children are part of the learning group. It is deeply worrying that this crucial and uncontroversial understanding of how teaching and learning work in our most successful schools seems to elude the government.

Meaningful visits to some of the many excellent comprehensive schools that are thriving in challenging circumstances would perhaps help politicians to gain greater insights into how excellent education can be achieved for all children in the same setting. These schools do not have to assist struggling local schools that their very existence has helped to create.

The attempt at defining conditions for either new or existing selective schools serves to highlight the fundamental flaws with selection. Each condition is a signpost to one of the problems inherent in selective systems. These problems are not preventable through national policy-making because they are a function of the very system the government wishes to expand.

Every school should be required to adhere to fair admissions standards which do not result in bias against any particular group of children – whether those from poorer homes, those with special educational needs or those from certain racial groups. The evidence is clear that it is impossible to meet such standards (in other words to ensure non-discrimination) with a selection test at age 11, and the narrative and questions in this green paper, presumably unintentionally, confirm that this is the case.’

LEE’s full response can be found on the ‘Our briefings’ page of this website.