Rick Muir, an associate director of the IPPR, published a paper this week called ‘Not For Profit: the role of the private sector in England’s schools’ in which he argues against allowing commercial companies to play a greater role in the delivery of taxpayer-funded education. As a contributor to a recent book published by the IEA called ‘The Profit Motive in Education: Continuing the Revolution’ — which takes the opposite view — I feel duty bound to respond.
The first thing to be said is that Muir is not a rabid opponent of education reform. He’s pro-academy and pro-free school. Indeed, it’s hard to find anything of substance that Michael Gove has done since becoming Secretary of State that he disagrees with. But he believes the Conservatives will go further and include a commitment to allowing for-profits to set up taxpayer-funded schools in the next manifesto and he’s getting his shots in early.
Muir rightly notes that any talk of a ‘Chinese wall’ protecting public education from evil, profiteering capitalists (see Nick Clegg) is nonsense. For-profit companies like Nord Anglia, Edison Learning and RIM are already deeply embedded in the state education sector, operating nurseries and pupil referral units, running local education authorities, providing IT, etc. But Muir neglects to mention the most significant breach of this imaginary wall: Ofsted inspections are largely outsourced to private companies. If for-profit EMOs (Education Management Organisations) can be entrusted with inspecting taxpayer-funded schools, why not with running them?
Muir is an evangelical believer in an ‘evidence-based’ approach to public policy and the thrust of his argument is that there’s little evidence that allowing for-profits an even greater role in English public education would result in system-wide improvements. Unlike some on the Left, he doesn’t claim there’s any evidence to show that unleashing for-profits would have a negative impact on England’s public education system. Rather, his argument is that in the absence of any compelling evidence that further deregulation of the taxpayer-funded education market would have a positive effect, we shouldn’t waste our time on it.
‘Policymakers should be focusing on doing the things that we know from international evidence do the most to raise school standards: strengthening school leadership, improving the quality of teaching, giving schools autonomy while holding them robustly to account, and systematically tackling inequalities between children from different class backgrounds,’ he writes.
I don’t disagree with any of those policy objectives, but it’s odd of Muir to present this as an either/or situation. We can prioritise all of those things and still experiment with profit-making schools. Indeed, we could judge those schools on how effectively they managed to meet those targets, particularly reducing the attainment gap between rich and poor. It’s a slightly peculiar argument to say that politicians shouldn’t pursue an unproven policy because it might distract them from focusing on more tried-and-tested ones. Muir seems to be confusing an ‘evidence-based’ approach with a risk-averse approach.
The bulk of his paper is taken up with analysing the research evidence about the impact of for-profit EMOs in America, Sweden and Chile — exactly the sort of thing the ‘evidence-based’ evangelicals are keen on. But can we really draw meaningful conclusions about the likely impact of a policy in one country by analysing its impact in another? Social scientists try and get around problems like this by ‘controlling’ for variables, but in the case of two different countries there are simply too many variables to control for. Far from being a robust, scientific approach, this is precisely the kind of woolly-minded analysis that led the Nobel Prize-winning physicist Richard Feynman to dismiss all social science as ‘voodoo science’.
To take just one example, let’s look at the claim that free schools in Sweden have increased social segregation. They have — by an infinitesimally small amount. But there’s significantly less inequality in Sweden by whatever metric you choose to measure it than there is in England and there was even less when Swedish free schools were first introduced in 1992. It’s patently absurd to claim that because free schools increased social segregation in Sweden’s education system – starting from a baseline of minimal inequality – they would also increase segregation in England’s, given that England’s education system is already one of the most segregated in the developed world. How can you ‘control’ for that difference? Yet this voodoo science is the central plank of the NUT’s case against free schools.
To be fair, Muir acknowledges this weakness in his argument – sort of. He doesn’t hesitate to apply the ‘lessons’ of other countries’ education systems to England, but he does have the good grace to admit that the various research studies he relies on may themselves be flawed because of the difficulty of controlling for ‘bias’ in the results. Unfortunately, he only applies this caveat to those studies that seem to show for-profit schools doing well.
For instance, he reluctantly allows that some research evidence in Chile shows for-profit schools outperforming municipal schools. ‘Children in independent schools … perform better than those in local authority schools, controlling for income and parental education,’ he writes. But he then goes on to identify a ‘bias’ in the sample, claiming that children at the for-profit schools are more able.
In the case of Sweden, he cites a study showing that for-profit schools perform better than municipal schools and – again – he attributes this to a failure to control for all the variables. ‘Unobserved characteristics about the types of children in different types of school could affect the data significantly,’ he notes. You could say that of all the research studies he cites, but he’s selective about which ones he applies this scepticism to. Any studies that show for-profits performing at the same level as municipal schools – or worse – he accepts at face value. Any studies showing them performing better he regards as hopelessly unreliable.
Muir is selective in another way, too, which is that he only treats as significant those studies that are of whole countries rather than regions within countries.
One of the pieces of research that proponents of profit-making schools often seize upon is a study carried out in Philadelphia in 2009 by Paul Peterson and Matthew Chingos. In 2002, the School District of Philadelphia asked both for-profit EMOs and not-for-profit EMOs to take over the running of its lowest-performing schools, awarding 30 contracts to the for-profits and 16 to the not-for-profits. Bingo! A perfect ‘controlled’ experiment. Muir summarises Peterson and Chingos’ conclusions as follows: ‘Using individual level data, [they] found that in this case the for-profit providers outperformed both the old public schools and the not-for-profit organisations in terms of pupil attainment.’
Quite an exciting result for disciples of Milton Friedman like me, but Muir dismisses this research because the sample size is … wait for it … too small. ‘It would be unwise to reach general conclusions on the basis of a single case,’ he writes.
Unwise? Or just inconvenient?
Muir’s decision to only treat as significant those studies that examine the system-wide impact of profit-making schools seems arbitrary. Rather than lump all for-profit EMOs in a country together, why not compare the performance of different ones? Even if we accept that, taken as a whole, for-profit schools do little or nothing to raise attainment across an entire country, groups of them do raise attainment in particular regions and individual profit-making chains raise attainment across regions. For instance, Sweden’s largest for-profit chain, IES, outperforms all other types of Swedish school, including not-for-profit free schools.
If Muir is sincere about wanting to take an ‘evidence-based’ approach, it makes sense to focus on smaller sample sizes where there are fewer variables to control for. Having identified those for-profit EMOs that perform best – the Benjamin Franklyn Charter Schools in Arizona, for instance – we could then go about trying to duplicate those successes. Why not repeat the Philadelphia experiment in Hull (England’s worst performing city when it comes to state schools) and allow a handful of reputable for-profit EMOs to set up, own and operate two dozen taxpayer-funded secondary schools and see how they get on? To rule out any profit-making state schools on the grounds that, taken as a whole, for-profit EMOs haven’t produced system-wide improvements in other countries is ridiculously dogmatic. It’s using ‘evidence’ to justify what is transparently an ideological bias.
It also seems dogmatic to rule out the possibility that the for-profit education sector will gradually improve as the market punishes failure and rewards success. The reason IES is the largest free school chain in Sweden is because it has expanded in response to demand, a reflection of the quality of the education it provides. Again, it seems arbitrary to judge the for-profit sector as a whole on the basis of its current performance — like taking a snapshot of any evolving industry and assuming it’s never going to improve.
‘The whole point of charters is that you can close the ones that fail,’ says Geoffrey Canada, founder of the Harlem Children’s Zone. ‘I’m all for it! You close them and constantly innovate, and things get better.’
Muir says he approves of innovation, so his dismissal of the for-profit sector is particularly odd. ‘Opening up the system to new actors can have beneficial effects,’ he writes. If he’s pro- innovation, why rule out a whole category of new actors in advance? Muir’s position is: ‘I’m in favour of innovation, just so long as it’s not commercial companies doing the innovating.’ Thank God that’s not the attitude of the German government to the automotive industry.’
In the final section of his paper, Muir is less circumspect about trying to conceal his ideological bias. He argues that schools are particularly unsuited to being run by commercial companies because they’re responsible for ‘imbuing norms and values’. ‘Schools have a special status because they play a critical role in shaping the kind of people we want children to become,’ he writes. ‘Schools inculcate values and send out important messages to children.’
Tellingly, Muir doesn’t think for-profit EMOs can be trusted to disseminate the right ’norms’ because, in his view, there’s something fundamentally unsavoury about ‘unfettered capitalism.’
I accept that schools do more than simply transmit knowledge, but rather than see that as a reason to resist the entrance of commercial actors into the taxpayer-funded sector, I regard it as a reason to break up the state’s monopoly. As a classical liberal, I want to reduce the state’s opportunities for disseminating state-approved ‘messages’. I don’t want to preserve it or — God forbid — extend it.
Muir isn’t troubled by this because the ‘values’ he wants state schools to ‘inculcate’ are shared by him. There’s a revealing passage in which he tries to articulate what these ‘values’should be and ends up producing something that sounds suspiciously like the preamble to a Labour Party manifesto: ‘live a healthy lifestyle … social and civic participation … work together to achieve shared goals … relate to people from different backgrounds … understand the rights and responsibilities of citizenship … etc.’
Muir regards these bland, New Labour platitudes as apolitical, as if the ‘values’ contained therein are universally shared. Maybe among members of the Fabian Society, but not outside North London. It is precisely because I don’t want the progressive Left to be able to use schools as instruments of social engineering that I want to break the state’s monopoly over taxpayer-funded education. Not-for-profit academies and free schools are a good start, but we need to go further if we’re to make this revolution permanent. And notwithstanding Muir’s review of the research studies, there are reasons to be optimistic that profit-making schools will outperform local authority-run comprehensives in cities like Hull, provided we take a genuinely ‘evidence-based’ approach to selecting them.