Government not planning to reintroduce 11+, says former Education SoS Michael Gove in U-turn over selection

Janet Downs's picture

The Government isn’t planning to reintroduce the 11+, claimed former Education Secretary Michael Gove on The World at One*.  This would be ‘totally wrong…a retrograde step’.

Restoring the 11+ isn’t what the Government is trying to do, he claimed.  There’s no plan to reintroduce 11+ tests across the country.   Education Secretary Justine Greening has made the same claim.  But it’s disingenuous to imply there won’t be some kind of screening.  It might have a new name - Centre of Excellence Aptitude Test (CEAT) perhaps - but it would still be the 11+.  To paraphrase Shakespeare, 'A cabbage by any other name would smell as foul'.

What the Government is doing, Gove said, was taking a ‘detailed, evidence-based approach’ to increase the number of good schools.  But all the evidence is against selection at 11 (see here, here and here).  It may (and I say ‘may') help those who are selected but it has a negative effect on the rest.  And any positive effect doesn’t appear to last: the Sutton Trust found pupils from comprehensive schools outperformed their equally qualified peers from both private schools and state grammars at university.

Nevertheless, Michael Gove says the Government is taking a ‘detailed, evidence-based approach’.  But as we know, the attitude of the Department for Education to evidence is inconsistent and slippery**.

It is wrong to be guided by ‘pristine ideology’, Gove said.  He had introduced new selective schools at age 16 so being ‘utterly opposed’ to selection was ‘silly’.  But supporting selection at 16 is not the same as supporting selection at 11.  Schools which offer A level courses select by ability because the exams require a certain prior level of achievement.  Secondary schools, however, all offer the same exams – GCSEs. 

Gove claimed the approach by Prime Minister Theresa May and Education Secretary Justine Greening was ‘unimpeachable’.  It is right, he said, to be motivated by what works.  But as noted above, a selective education system does not work for the majority of pupils

A former Conservative Education Secretary Edward Boyle said the 11+ system was wasteful.  Mounting evidence showed selection at such an early age was flawed.  That was over 50 years ago.  But half-a-century later a Tory Prime Minister is still calling for grammar schools.  It would be laughable if it were not so serious.  Boyle had the courage to speak out against selection – it cost him his career.  But today we see another former education secretary doing a complete about-turn from his previous emphatic opposition to grammar schools.  A cynic might say such a change of heart is a calculated career move.

REMINDER: You have until 12 December to complete the consultation about the ‘garbage’ grammar proposals.  Be warned: the questions imply respondents already agree with the proposals.  If you disagree with them, say so.

*Full interview here (education views from 18 minutes).

**See my comment to the Education Select Committee on the DfE’s use of evidence here (scroll down). 





Share on Twitter Share on Facebook

Be notified by email of each new post.


jane's picture
Thu, 27/10/2016 - 12:15

Thanks for posting this, Janet. I have done my best on the questionnaire but like so much of what the Government does it presupposes agreement with its policies! Just as evidence, my reference was ANON-EKAK-FWR5-Z

ian thompson's picture
Wed, 02/11/2016 - 18:26

Grammar schools do not have negative effects on those who aren't selected, and they have consistently been shown to benefit those who are, particularly poorer children . Pupils on the other hand in comprehensive schools on average fail to achieve their potential and experience wide variations in progression to selective universities that are determined by their parents' socio-economic circumstances.

Janet Downs's picture
Thu, 03/11/2016 - 08:06

Ian - follow links in paragraph three in main article to evidence that grammars DO have negative effects on the unselected particularly disadvantaged children.   Also OECD: ' When schools are allowed to apply selective academic and income criteria, this aggravates school composition segmentation, as oversubscribed schools tend to hand-pick their students, crowding out disadvantaged students and students with low performance.'  (p38 School Choice and Equity: Current policies in OECD countries and a literature review.  Feb 2012). 

ian thompson's picture
Thu, 03/11/2016 - 13:48

There is so much ridiculous misinformation there that it is difficult to know where to start.

Studies from Smith and Naylor a decade or so ago to Crawford more recently show that for a given level of A level attainment the worse the school that a student came from, the better they perform at university; under-performing comprehensives are worse than better comprehensives which are worse than independent schools and so on. This is frequently summarised as 'comprehensive pupils do better than grammar school pupils or independent school pupils at university '(even by the Sutton Trust) but the grade difference that they actually achieve in more facilitating subjects, more than makes up for it and similar independent school pupils end up with better degrees because degree and A level achievement correlate too strongly for that not to be the case. The consequence of this is seen at Cambridge where students from independent and grammar schools are more likely than students from comprehensive schools to achieve firsts. Twice as many firsts are awarded to independent school students as comprehensive school students and despite the even smaller proportion pupils who in grammar schools they account for nearly as many firsts as comprehensive school students. I won't trouble you with the value added data from EPI and CEM which speaks for itself.

Chris Cook uses a compost of deprivation data that would do no credit to a sixth form geography project. He rides roughshod over considerations of subject difficulty, cross boundary movements, non-educational factors that shift his curves to the right or left, and data dredging. There is a very good reason why it was confined to news sites and never appeared in peer reviewed journals. A proper analysis for the Sutton Trust showed no disadvantage to pupils who did not get into grammar schools in grammar school areas, but it did show as other studies do, that clever FSM eligible children do disproportionately poorly in comprehensive schools.

As for the OECD, England does substantially worse for basic literacy and numeracy than Northern Ireland which retained the eleven-plus and ends up bottom amongst comparable countries. The OECD also singles out England for the parental gaming that goes on in its nominally non-selective system.

Janet Downs's picture
Fri, 04/11/2016 - 08:41

Ian -   Re Northern Ireland v England scores In reading in OECD PISA tests 2012:

'The score for reading in Scotland was 506 points, slightly higher than that in England (500), and Northern Ireland (498). This compares with scores of 500 in Scotland, 495 in England and 499 in Northern Ireland in PISA 2009'.

Again, 498 is lower than 500 although it would be ridiculous to say this ever-so-slight lower score was the results of England having few grammars.


Janet Downs's picture
Fri, 04/11/2016 - 08:47

Ian - Re Northern Ireland v England scores In science inOECD PISA tests 2012:

'Scores for science were 516 in England, 513 in Scotland and 507 in Northern Ireland'

Again, scores for England are higher than in Northern Ireland.

(Apologies - my comments seem to have been posted in the wrong order.   My comments would make more sense if the one posted at  08:35 was read first.

Janet Downs's picture
Fri, 04/11/2016 - 09:18

Ian - re Crawford.  Again, you provided no link.  But I found this Guardian article which summarised Crawford's research.  She recommended that unis lower their expected grades for pupils attending state non-selective schools or low-performing state schools.  This could be interpreted as (a) these schools are don't provide a good enough education to ensure their pupils get high A level grades or (b) private and grammars are better at ensuring their pupils get good A level grades but the higher performance doesn't seem to last at uni.  

The latter intepretation seems to be upheld by research from Oxford:

'...the sociological evidence is now clearer: the school type effect at Oxford University is likely to be driven by short-term teaching effects upon the secondary school grades of private school students.'  ie, private schools pupils are more likely to gain a place at Oxford but the teaching that got them there has a short-term effect.  However, this tells us nothing about the relative degree performance of pupils from comps or grammars.  

Janet Downs's picture
Fri, 04/11/2016 - 09:05

Ian:  Re - the research you cited by Smith and Naylor.  I had to search for this because you didn't provide a link.   I found this which summarises Smith and Naylor.  It compares private schools with state schools and doesn't differentiate between state selective schools and state comprehensives.   As such, it neither proves nor disproves the Sutton Trust finding that comprehensive pupils outperform their equally-qualified peers at university.

Janet Downs's picture
Fri, 04/11/2016 - 08:35

Ian - in no particular order.  Re Northern Ireland v England scores In mathematics in OECD PISA tests 2012:

'Scores for mathematics were similar in England (495) and Scotland (498). In both cases this showed little change from PISA 2009, where the scores were 493 and 499 respectively. The score in Northern Ireland was 487, compared with a score of 492 in PISA 2009. As in PISA 2009, mathematics performance in Wales was lower than the rest of the United Kingdom, with a score of 468 points compared with 472 in PISA 2009.'

It's a long time since I took O level maths, but in my reckoning 487 is lower than 495.

That said, there is actually very little difference in the scores.  In any case, it would be wrong to say any higher score was the result of the presence (or not) of selective schools.  As noted above, Scotland's score was higher than England's.   Scotland has no grammar schools.  But it would be wrong to say that this higher performance was the result of having no grammars.  Equally, it would be wrong to say Wales's low score was the result of the principality having no selective schools.


ian thompson's picture
Fri, 04/11/2016 - 11:31

Sorry, I should have cited my sources properly.

These are a justification for the use of contextual data in university admissions, but they are frequently used as an endorsement of poorly performing schools:

For how comprehensive school pupils, grammar school pupils and independent school pupils achieve at GCSE, A level and university you can simply but crudely compare the numbers in each group achieving the right grades, getting into universities and their subsequent performance. Unfortunately the best quality university figures are more or less limited to Oxford and Cambridge, and Oxford doesn't split degree class by school type although their admission data is more detailed. You need to constantly bear in mind that the number of pupils dropping out of the academic figures because of selection at sixteen in comprehensive schools is substantially greater than the number lost from the other types of school.

The OECD source related to basic literacy and numeracy, you can also compare exam results in the two systems which do not favour England.

SFR37/2015: Provisional GCSE and equivalent results in England, 2014 to 2015, 15 October 2015

Janet Downs's picture
Fri, 04/11/2016 - 13:57

Ian - thanks for the links.  I've just started reading the doc.  It says:

'...those from independent and selective state schools, those from state schools with a low proportion of FSM-eligible pupils and those from high-value-added state schools are now significantly more likely to drop out, significantly less likely to complete their degree and significantly less likely to graduate with a first or a 2:1 than their counterparts in non-selective state schools, state schools with a high proportion of FSMeligible pupils and low-value-added state schools respectively.'

This surprised me.  I thought it would be the other way round with pupils from independent schools and grammars LESS likely to drop out and MORE likely to complete their degree than their peers from comps etc.   I've actually read this several times to make sure I've understood it properly.  The report also says this picture doesn't change if the researchers factored in 'type of university attended and subject studied (or when we restrict attention to pupils attending high-status universities only)'.

The authors warn against concluding a 'causal' effect between school types but went on to suggest lower offers to pupils from state comps or low-performing schools.

Leaving aside the 'lower offer' argument, what are possible causes for the increased rate of drop out etc by independent or grammar pupils?  Could it be, as Oxford suggested (see my comment above) that the 'school effects' of a private education didn't last into university?  Could it be that pupils at private/grammar schools have been taught-to-the-test and their higher A level results don't accurately show their true ability (hence the higher drop out rate, lower degree etc)?  Or, on the other hand, are comps not providing the type of education which gets high A level results for their high achievers?

I don't know the answer, but any dropping out is a waste both for the drop-out and for the potential student who could have taken the place.


Janet Downs's picture
Fri, 04/11/2016 - 14:12

Ian - I'll try to keep my comments short or they run the risk of being too long to be easily digested and/or straying too far from the original article.  You say the document is 'frequently used as an endorsement of poorly performing schools'.   Your reply implies all 'poorly performing schools' are also poor schools.  This isn't necessarily the case.   When league tables were introduced, we knew immediately that our school would be judged 'poorly performing'.  Why?  Because we were a non-selective school in a selective county.  We were, therefore, creamed of a large number of previously high-attaining pupils.  We also had a 'unit' for statemented children.  This resulted in our school's intake being skewed to the bottom end.  It was obvious, without even looking at the league tables, we would be 'poorly performing' when judged only by results. 

And that applies to most non-selective schools in selective areas.   



Janet Downs's picture
Fri, 04/11/2016 - 14:27

Ian - re the Adult Skills Survey you cited towards the end of your comment.  The OECD urged caution when using these survey's results because most of the countries did not meet the response criteria.  England and Northern Ireland were the worst.   I argued with the OECD that the figures were so flawed as to be useless but the stood by their data (see here).  

That's not a sign of complacency, however.  The performance of young people was dire, being bottom in literacy and second to bottom in numeracy.   But the researchers pointed out that 30% of England's 16-19 year-olds were not in full time education or training leading to formal qualification whereas there was near universal participation in other countries.  This is a problem which has still not been sorted despite the raising of the participation age.  The education of 16-19 year-olds is the poor relation of the education system: even more inadequately funded that 5-16, fragmented and not compulsory.

That said, the Adult Skills survey had important insights.  These were mainly ignored (see here and here).  

Janet Downs's picture
Fri, 04/11/2016 - 14:37

Ian - following on from the Adult Skills Survey (to be used with caution), the poor performance of England's young people wasn't confined to those who attended comps.  If the sample were properly weighted, it would have contained young people who had attended selective and private schools.   I pointed that out at the time (see here). 

ian thompson's picture
Fri, 04/11/2016 - 18:22

In many ways the Crawford study is the less satisfactory of the two. By lumping firsts and upper seconds together, given that a minority of degrees are graded below that, they were in effect focussing on the smaller under-performing tail of university students. They also found that the type of university attended affected their main conclusions in that differences in degree outcomes for different types of school were weaker in more selective universities but they curiously opted not to publish the relevant details. Taking both points together their approach was designed to overlook the higher end of ability where selective and independent schools might be expected to have an advantage. Smith and Naylor covered that much better.

Students from independent schools grammars and comprehensive schools with low FSM rates are least likely to drop out, it was only in their statistical model having corrected for student background and attainment at KS4 that the reverse appears to happen.

Absent from discussions on the possible causes of differential degree attainment was any consideration of progression rates to university. Students from the lower half of the independent or grammar school ability distribution curves might be predicted to compete relatively poorly with pupils who see themselves as high flyers from the top ten percent of pupils in comprehensive schools.

The bottom line of course is that school type makes a 6-9% difference in probability of achieving a good degree, not 100% or even 50%, but it might make a pupil 400% more likely to get into a good university in the first place.

ian thompson's picture
Fri, 04/11/2016 - 18:30

I agree with your comments on the OECD figures.

Add new comment

Already a member? Click here to log in before you comment. Or register with us.