The claim
“[Three academic studies] all show the more free schools you have in a municipality the higher standards are. They all show standards are higher in free schools than other schools.”
Michael Gove, BBC2 Daily Politics education debate, 3 May 2010
The background
It’s one of the Conservatives’ most eye-catching policies – that parents unhappy with what’s on offer for their children locally can set up their own school. Over the weekend, David Cameron said this could be a reality by September if his party forms the next government.
The idea is that increased competition drives up standards – both in the new schools, and by putting the pressure on existing schools to attract parents who could go elsewhere.
The Tory manifesto contains a page praising the Swedish free schools movement, with which the policy has many similarities. Since the early 90s, anyone has been allowed to get public funding for a non-selective, non-fee paying school – before this, the only option was a municipal school, run by the local authority. The Conservatives say this has led to higher standards across the board.
Yesterday shadow schools secretary Michael Gove was challenged on the evidence behind the policy during a TV debate. He said that several academic studies in Sweden showed that more free schools meant higher standards, both in the free schools themselves, and in other schools in the area.
So back to the books – what does the Swedish experience show?
The analysis
The Tories pointed us towards a summary of the evidence by free school champions New Schools Network.
We went back to the original Swedish papers, and there is some evidence for Gove’s claims.
The first major analysis, originally written in 2002 by Fredrik Bergström and Mikael Sandström (the latter is an economist who now works for the government) looked at exam results from 28,000 ninth graders (the last year of compulsory school).
The researchers also took into account things like parents’ education and income, the size of the school, and spending per pupil.
They found competition had positive effects, and no real negative effects. According to the study, results in state schools were also better when a larger share of pupils went to free schools.
Another study cited by the New Schools Network, from 2003, found a 10 per cent increase in the number of children attending free schools led to a 6 per cent increase in maths test results.
Not bad – but then, the same piece of research also found increased school competition had “no significant effects” in English and Swedish performance. That doesn’t torpedo Gove’s point, but it does suggest improvements weren’t across the board.
Another report re-analysed the data and found a positive association between the share of independent schools and Swedish students’ average performance, but that there was little effect on foreign-born or students from less-educated backgrounds.
The authors also said there was some evidence that independent schools increased segregation.
The most recent study (2008) seems to provide the greatest pause for thought. Anders Bohlmark and Mikael Lindahl of Stockholm University found that an increase in free schools had a positive effect in the short term – backing up Gove.
However, they concluded these effects were pretty short-lived. They didn’t find any effect when they looked at subsequent achievement in optional further education, or at university, and concluded that increased competition did not have a lasting effect. Sandra McNally, director of the LSE Centre for Economic Performance’s education and skills programme, says this study is the most credible piece of research.
The Conservatives said there is a difference of opinion about which study is best. They also pointed us to a Policy Exchange report which questions some of the methodology of the Bohlmark and Lindahl report (pages 25 and 26), and they added that there is further “overwhelming” evidence from similar reforms in the USA and Canada.
“The fundamental point is that even those who are against the system admit that there have been positive effects,” a spokesman told FactCheck.
The verdict
There are studies backing up Gove’s claim about improved results in Sweden. But other research also casts doubt on how long lasting those effects are.
It’s perhaps worth pointing out the differences between the English and Swedish education systems – when free schools were introduced in Sweden, there was virtually no alternative to non-selective state schools.
Britain already has a range of fee-paying schools, selective grammar schools, technology colleges and around 200 academies, which like free schools, can be set up by independent bodies.
Researchers at the Institute for Fiscal Studies looked at a range of international research, and concluded there is no definitive evidence for the effects of school competition.
Given these uncertainties, we feel unable to give Michael Gove’s claim a firm rating.
