Saturday, 26 October 2013

Why are London's schools doing so well?

London's secondary schools have been doing better than those in the rest of the country for some time now. They were already ahead in 2003. But the gap has widened over the past ten years to a chasm. Chris Cook - when he was the FT's education correspondent - was the first to analyse this trend in detail. And he realised that the official Government metrics of performance were actually hiding how far ahead London is.

Since then there's been a growing debate amongst policy wonks about why this is happening. But I think most people - even those in the education sector - still don't realise quite how far ahead London is now; especially for poorer children.

In this post I'll give some stats on how big the gap now is. I'll then question whether this is down to policy or socio-economic change. Finally I'll look at whether - if policy has driven some of this change - what that might mean for the rest of the country.

The size of the gap

The gap between London and the rest on the Government's main measure, 5A*-C including English and Maths, is negligible - 59% nationally and 62% in the capital. But that's actually quite surprising given how many more children on free school meals live in London compared to elsewhere (23% vs. 14% at secondary). When we look at just those children on free school meals we start to get a sense of the difference. In London FSM kids get 49% vs. 36% elsewhere. When you dig into LA level data the extremes are startling. Almost two thirds of FSM pupils achieved the 5A*-C benchmark in Westminster last year compared to less than a fifth in Peterborough (the lowest scoring authority).

But because 5A*-C is a simple threshold measure it obscures how much better London schools are doing for the very poorest. The graph below is taken from the recently published Milburn report on social mobility (in fact it's a variant of a graph Chris Cook did a while ago). It shows that children in London do better at GCSE and poorer children do a full grade better - across all their GCSEs - on average. The rich/poor gradient is much less steep meaning that, while London is one of the most economically unequal parts of the country, the education system is producing more equal results.

The next table is from some internal analysis we've done at Teach First. Here we've looked at only those schools that are eligible for Teach First i.e. those where at least half the pupils are in the bottom 30% of the IDACI (Income Deprivation Affecting Children Index).

Nearly all secondary schools in inner London meet this eligibility criteria. In other areas- like the South East just a small percentage do. We've compared the GCSE average points score in eligible schools in different regions (excluding all GCSE "equivalencies" and doubling the value of English and Maths). This shows eligible schools in inner London 100 points ahead of the worst performing regions (the South East and South West) which is equivalent to the difference between 6 D grades and 6 B grades.

It's also important to note that this isn't just a GCSE phenomenon down to - perhaps - better gaming by schools in London. Our initial analysis of the DfE's post-16 destination data shows the pupils from Teach First eligible schools in London have about a 9 in 10 chance of going on to further education - around the national average for all schools - compared to about 8 out of 10 in other parts of the country. Poorer students from London are also much more likely to go to university.


Why is there a gap?

For me the big question is whether this gap has been driven predominantly by policy or by socio-economic changes. If the former then it's good news because. if we can work out which policies and how best to transfer them elsewhere, we can help schools in other parts of the country improve. If the latter then it's good for London but worrying for everywhere else as the gap will continue to grow.

There have been a number of policies over the past ten years that have disproportionately benefited London:
  • The Labour Government significantly increased school funding across the country but it increased by a higher proportion in London (which was already well funded - albeit with higher salary costs).
  • The academy programme started in London and the best sponsors, while not exclusively in London, seem to be concentrated there (e.g. ARK, Harris, Mossbourne, Haberdashers).
  • Initiatives like Teach First tend to have started in London even if they were later rolled out to the rest of the country. (If you're in a challenging school in London you won't find it difficult to find a charity nearby offering tutoring or mentoring - in Clacton or Peterborough it's a different story).
  • And the one that people talk about most - the London Challenge - a programme launched by Labour as an umbrella for a number of schemes but which was, primarily, about brokering support between stronger and weaker schools. (See this Ofsted report for a good summary).

At the same time the socio-economics of London have changed massively. Essentially London has got a lot richer while much of the rest of the country hasn't. To illustrate here's some startling stats from a BBC article:

"A recent study estimated that the value of London's property had risen by 15% - or £140bn since the financial crisis began. That increase - just the increase - is more than the total value of all residential property in the north east of England. London's top ten boroughs alone are worth more, in real estate terms, than all the property of Wales, Scotland and Northern Ireland, added together."

And here's a scary infographic from the ONS:

Not only has London got richer. It's got richer in precisely those areas where schools have improved the most. Look at this map discovered by Economist journalist Daniel Knowles. The red areas are those which have gentrified the most over the past ten years; the blue areas have gone "downmarket".

The areas that have seen the biggest change in socio-economic status - Westminster, Southwark, Newham, Tower Hamlets, Hackney - are also the areas which have seen the greatest gains in school performance.

This certainly isn't the whole story behind London's improvement but it's an interesting correlation. The secondary schools in these areas haven't suddenly filled with yuppies' kids, they're nearly all still eligible for Teach First, but many have become more mixed. Moreover, the children in those schools are exposed to very mixed communities, unlike some of the worst performing parts of the country where communities are much more uniform. And those areas have become a lot more attractive to teachers - especially young teachers.

So has policy or economics driven change? Given the lack of in-depth research on the topic all we can is it's a complex mix of factors. The extra money available in London probably helped but it wouldn't have been worth as much without the increased supply of bright young teachers heading towards the capital. Yes lots of charities like Teach First started in London but they did so because that's where recruitment is easiest.

London Challenge is perhaps the most interesting. Many headteachers who were involved will argue it was a hugely valuable experience - though there is little in the way of hard statistical evidence (e.g. comparing improvement in London Challenge schools with other schools in London that weren't part of the programme). But if it did work it was in large part because there was already a supply of excellent headteachers and advisers to provide mentoring and support.

What might it mean for the rest of the country?

The complex relationship between policy and non-policy factors in London mean we have to be very careful about drawing simple conclusions about what might work elsewhere.

It would be easy to say - and many have - let's roll out "challenges" in other regions. But they were tried in Manchester and the Black Country and weren't nearly as successful - perhaps because the necessary concentration of existing expertise wasn't there to make it work. And if it isn't there in Manchester it certainly won't be in coastal towns or dispersed rural areas where we have the greatest problems of underperformance.

Likewise there'll be little benefit to changing the funding system to switch funds from London to elsewhere if other schools can't get access to a ready supply of talented new teachers and the types of mentoring/tutoring initiatives that are so prevalent in London.

That doesn't mean regional placed based strategies aren't worth trying - I believe they are - but they need to be bespoke rather than just attempts to recreate London Challenge.

I get the impression that Nick Clegg's unfortunately named Champions' League of Headteachers is the beginning of an attempt to do this; but busing in new heads won't be enough. If regional strategies are going to be tried in the places that need them most there will have to be a concerted effort to create a sense of collective mission. Existing heads and teachers in the areas of focus will have to buy in; successful heads and teachers from other areas (and especially London) will have to recruited to provide support; charities will have to be encouraged to venture out of their urban hubs and, ideally, educational change will be linked to wider community projects and economic regeneration.

Change happened in London but it already had a lot going for it. Making it happen elsewhere will be a lot harder.

Sunday, 20 October 2013

What do the opinion polls tell us about support for free schools?

Opinion polling is a valuable but dangerous tool. Dangerous because superficial analysis can lead to badly misjudging the public mood. And because so few people understand how polls actually work. This weekend I saw people tweeting puzzlement at two polls; one showing Labour with a 3% lead and one with a 11%. But with a margin of error of 4% and very different methodologies between the two polls it's not surprising at all. (The trick to reading voter polls is to look at trends across multiple polls over time).

These two polls happened to both contain questions about free schools after an eventful week for the policy. There isn't really much polling on free schools for such an important policy (politically at least) so it's worth unpicking in some detail. 

First ComRes for the Independent on Sunday asked:

Parents, teachers and charities should be encouraged to set up new state schools, even if there are already schools in the local area:

Agree 27%
Disagree 36%
Don’t know 37%

So free schools are unpopular? Not so fast... This is a badly designed, and so misleading, question. The problem is it asks two different things in one question. First should non-state providers be allowed to set up schools and then should they be allowed to do so even if provision already exists. It doesn't tell us how many like the idea of non-state providers but not the "waste" of surplus capacity. And it's confusing- as can be seen by the high number of "don't knows".

Meanwhile Opinium for the Observer asked some much more detailed questions. First they looked at general support for the policy:
"Like state / comprehensive schools "free schools" are schools that are funded by the government but, unlike state / comprehensive schools, they are not under the control of the local authority. They can be set up by parents, teachers, charities or businesses and are free to attend."

44% believed that "free schools" are generally a good thing for education in the UK while 22% said they are a bad thing; with a further 22% saying neither good nor bad and 12% don't knows.

Here the focus is on the "free" part not the surplus places and that bit of the policy seems popular; even among Labour voters who were 39/30 in favour (Lib Dems were even more pro - 50/18).

On the specific issue of whether such schools should be allowed to hire unqualified teachers respondents were asked:

"While teachers hired by state / comprehensive schools have to have a PGCE or equivalent teaching qualification, "free schools" are free to hire whoever they choose, as private schools do. The advantage of this is that they can employ people who have more experience of the world outside teaching while the disadvantage is that it may risk children being taught by under qualified teachers."

A resounding 60% were concerned that "free schools" are able to hire teachers who may not have a PGCE or equivalent teaching qualification while 30% are not. Amongst all three parties' voters there were more "concerned" than "not concerned". So that bit of the policy doesn't seem very popular.

But perhaps most interesting Opinium asked about what respondents wanted a future Government to do about free schools. Here responses were divided. 23% wanted the policy to continue as is (this was the most popular for Tory voters).

27% wanted what is - in effect - Labour's current policy- to allow them to continue but only in areas where there aren't enough places and if all teachers are qualified. This was the most popular option for Liberal Democrats - just beating leaving the policy as it is - and the equal most popular amongst Labour voters.

Just 12% wanted to prevent any further free schools from opening while leaving the existing ones alone and 20% thought that all free schools should be taken under local authority control (this was equal most popular with Labour voters).

What does all this tell us? It seems the majority of voters like the idea of non-state providers being allowed to set up their own schools but that they really don't like the idea of unqualified teachers. Most people want the policy to continue in some form but those that do are pretty split between allowing them in areas that already have enough places. So Labour's current policy is probably most in tune with the public (if they could explain it clearly...)

I expect, as we approach the next election, more polling companies will look in detail at this issue so we should be able to see if these hypotheses hold as people learn more about the policy.

Thanks to Daniel Boffey at the Observer for sending me the Opinium numbers. When these appear on the Opinium website I'll add a link.

Friday, 11 October 2013

Thoughts on the OECD survey of adult skills

The first OECD survey of adult skills (PIAAC) was published this week. It's a fascinating - and, at almost 500 pages, immense - report. As England comes close to the bottom of the league tables - especially amongst 16-24 year olds - it has been presented as a condemnation of our education system. And you'd have to be remarkably complacent to read the report and not be concerned about our apparent global weakness in basic skills but it's worth highlighting some things that a superficial reading might miss:

1) The report isn't really telling us anything we didn't already know from PISA - the OECD's existing international comparative test for 15 year olds. It looks worse than PISA because we're at the bottom of the league table rather than somewhere in the middle, and our scores are below average rather than average, but that's a bit misleading. Far fewer countries participated in this survey than in PISA and most of those that did also do better than us in that test.

2) Those countries that did better than us in PIAAC but not PISA (e.g. France, Sweden, Czech/Slovak Republic) all have compulsory literacy and numeracy for those in post-16 education. As do pretty much all the other countries that participated. As a majority of our 16-18 year olds don't study English or Maths it's not really surprising that even countries we're marginally ahead of in PISA do better than us in PIAAC. If we were to increase the numbers of 16-18 year olds doing these subjects (as both the Government and Labour want to do) I'd expect to see us rise up the tables a bit.

3) The questions PIAAC uses (like PISA) are very applied - i.e. are focused on using literacy and numeracy in "everyday" situations. Our education system doesn't focus on applied skills in the way that others do; especially post-16. By contrast the other big international comparative test for 15 year olds, TIMSS, tests whether pupils have mastered the specific knowledge and skills outlined in curriculum content common amongst participating countries. We do much better in TIMSS than PISA so it's fair to assume our post-16 issue is also with applied skills rather than curriculum skills. All of which seems to support mathematician Tim Gowers' proposal for post-16 courses in real-world maths.

4) This difference between applied and curriculum skills may go someway to explaining how it can feel to so many people that our system has improved over the last 15 years despite our PISA results flatlining and our PIAAC results showing our 16-24 year olds doing so badly. We may have improved a lot in things that PISA/PIAAC don't measure; while making little progress in things they do. Another factor may be that the politics/media world tends to focus on London where the system almost certainly has improved while ignoring continued underperformance in other parts of the country.