Education Policy in the Election 3: Autonomy, Regulation and Academy Chains

By:  Simon Burgess

Over half of all secondary schools now have Academy status, and the number will continue to grow rapidly in the next parliament. The case for academy schools is largely the case for school autonomy, and there is robust evidence from US Charter schools that autonomy can be helpful in raising pupil attainment. While the constraints on schools might have been overstated after Local Management of Schools, there is really no going back, and schools will almost certainly retain most or all of their ‘freedoms’.

But there are some important issues about the governance and autonomy of academies at a different level. These have been hardly discussed at all and, given the sheer number of academies, may well become one of the most serious issues for the next Secretary of State to deal with.

Academy chains are now a big part of the education scene, reportedly covering about half of all academies. They have also come to bear the burden of quite a lot of policy hopes for school turnaround.  But policy-makers need to think more about the governance by and of chains.

There is also a paradox here – if academies were born from a desire for autonomy, how much autonomy do the schools have within a chain? And if the most effective chains are the ones keeping the tightest grip on their schools, did we get wrong the level at which to increase autonomy? Should it perhaps be groups of schools that have autonomy (chains and of course, ironically, local authorities), rather than individual schools?  While leading commentators champion autonomy, this is usually meant as autonomy from government rather than autonomy from Chain HQ.

There are three main issues: the formation of school chains or groups, the activity and performance of groups, and the setting up of new schools.

Formation of school groups

One of the most important things we want policy to achieve with school chains is to mimic the operation of a market, in the following sense: we want highly successful schools and leaders to provide education for more pupils and unsuccessful ones for fewer. The formation of school groups can help a lot to achieve this. We want existing high performing schools to form a group with (in slightly less polite language, take over) a nearby low-performing school. Or an existing chain of high-performing schools to do that.

There seems to be little point in encouraging just random assemblages of schools to join together beyond pooling back-office functions and bulk-buying stuff. More clearly, we would surely want to rule out a group of all local low-performing schools banding together. Why? Two reasons – first it removes effective choice from parents if all local schools are of the same ‘brand’. Secondly, by removing choice it also removes the pressure to improve. While all this seems ‘top-down’ in fact the idea is simply to mimic a well-functioning market in which the successful performers grow and either take-over or out-compete low-performers.

As the number of school take-overs and chains and ‘brands’ increases, there will be issues of market power to think about. This is a major issue in hospitals, but they are an order of magnitude bigger so not a major worry yet.

What do school groups do?

There is a tendency (until very recently, basically a blind hope) to see all chains as good things – highly effective or at worst, benign. Recent research, however, including from the path-breaking Chris Cook, suggests that perhaps just a couple of chains are transformative (ARK and Harris), and the others are pretty average; one chain, AET, has been strongly criticised by OFSTED. (You can now see for yourself the stats on the secondary school chains, having first read the health warnings about the data).

The widespread current view is that what school groups do is aid collaboration. Who could be against collaboration? Two points to make. First it is not clear that collaboration in the everyday sense is what effective school chains do. Some do no doubt. But in some cases it is more that the over-taking chain does some possibly rather un-collaborative turning around of the taken-over school. This is surely what policy makers and local parents want. That may not always work, but I think it is a mistake to see chains as synonymous with collaboration.

Second, the London Challenge is seen as the epitome of school collaboration, and the London effect is cited as evidence that this works. I have written about that here, here and here, so no more now; just to say that while the London Challenge may have done many wonderful things for London schools, there is little evidence that it has impacted dramatically on pupil attainment.

In summary, collaboration may not be what the few effective chains do, and there is little evidence that collaboration works to raise attainment.

Setting up new schools

The organisations in prime place to set up new schools should be existing high performing schools, or existing groups that have shown that they can deliver a high quality education. Whilst there should definitely be an open and transparent competition to set up new capacity, we could use a deliberately loaded question: is there any reason why this group should not provide a new school?

What about Free Schools? These are essentially academies but set up following proposals from groups other than the LAs. In principle the idea is to make the education market ‘contestable’ and to provide the competitive pressure to keep schools on their toes and performing well. However, the evidence does not suggest that this is having much impact.

The big issue associated with free schools is this: there is only limited capital to spend on setting up new schools. So there is a strategic choice to be made between provision of extra capacity in a place where schools are ineffective but places are plentiful, and neighbourhoods where there are insufficient places. And of course, where schools are ineffective, places are almost bound to be plentiful.

Two conclusions: contrary to Cameron’s announcement of an expanding free school programme, we should abandon the current policy of building free schools in the hope that they will raise standards all around.  However, where new school places are needed, we should retain the principle that outsiders are able to propose to set up a new school as part of an open competition, but with a strong presumption in favour of existing high performing schools or chains.