At the heart of much of what passes for debate about education is one false dichotomy or another. Subjects or skills? Breadth or depth? Rigour or well-being? The only rational response to such polarised debates is to say "both matter, obviously".
In the accountability debate, the polarisation can be extreme. On one side, there are those who argue passionately about the distorting effects of the accountability system, saying that excessive focus on the grade C borderline, for example, targets ever-narrower groups of children. On the opposite side, others argue strongly that getting children to a C is rightly a major focus, because it unlocks doors to progression in learning and life.
As ever, there is right on both sides. Our accountability system both creates a focus on outcomes and risks distorting behaviour. Any such system carries that risk. But what I think should not be in question is the merit of accountability.
It is not a necessary evil but positively a good thing. We are all at our best when we are properly accountable for our actions, when there is proper scrutiny of the things we have done and we bear the consequences.
All recent scandals of public and corporate life occurred when people did not feel accountable for their actions. The MPs' expenses debacle, the excesses of banking and the phone hacking incidents all happened in a culture where people ceased to feel that they would be held to account by society at large.
Yet it is also true that any system of accountability, particularly one based on targets and measures, can carry the risk of perverse consequences. This begins to happen as soon as people stop seeing the target as the proxy for what is to be achieved (in our case, educating all children well, to enable them to be good citizens, get decent jobs, have a successful life and so on) and instead as what is to be achieved in itself (get as many children as possible across the CD borderline).
Where the focus on targets is pronounced and the consequences of not meeting them grave, the risks of system degradation grow. Behaviour may focus on the targets rather than the real aims.
We have seen this in the GCSE performance tables. Equivalences led some schools to focus on qualifications that counted heavily in performance tables but could be taught in a single option block, irrespective of their value to young people. The introduction of the English Baccalaureate measure into performance tables led some schools to move young people off a course they had started in order to make sure they met the new standard.
This example and others like it suggest that, in the face of a very powerful accountability system, many schools have lost the confidence to determine for themselves what is right for their pupils and to stick to it. Many feel under such pressure to work towards centrally-set measures that instantly adapting the curriculum in response to them feels natural.
What is the answer to this? In part, bravery: the determination never to do the wrong thing in order to look good in an accountability system. Another part of the answer lies in the system itself. The government is looking again at that system and perhaps asking some fundamental questions. How to incentivise high standards without placing so much weight on a single grade boundary? How to place more emphasis on the progress of every child?
I would argue that far from there being too much accountability, there is too little. As a result, too much weight falls on a few indicators. Schools should be accountable for all their actions: do they welcome the vulnerable or do they surreptitiously try to remove from their roll those who might bring down their league table position? Do they maintain a broad curriculum throughout key stage 3 or do they begin to game the system in Year 9 by teaching GCSE courses that give them league table points but give pupils no chance to progress into further learning?
Part of the new accountability system should be to put out into the public domain much more comprehensive and connected information. It should set out more fully what schools are doing for children, as well as financial and staffing information.
Equally, however, we need to go beyond the "arm's length" approach to accountability. Data can tell us a great deal but not everything. Ofsted's approach to focusing more sharply on observational evidence, quality of teaching and progress in lessons may be challenging but it is right. Even that, though, cannot tell us everything about a school's behaviour.
Education is fundamentally a local service and there need to be local mechanisms for accountability. As we move from 2,000 to 20,000 academies across the country, intervention from the Department for Education in the event of serious problems will become increasingly unrealistic. Even more unrealistic would be expecting the DfE to have the ability to tackle coasting schools or spot poor admissions or exclusions practice.
It could be argued that market mechanisms could meet some of this need. However, the availability of capital, the pressure of providing places for growing numbers and the costs of providing schools mean that we remain a long way from that point.
So, I believe that in the long run (as all schools in an area become academies and the local authority ceases to be a provider), local authorities will need to be given the responsibility and powers to challenge providers that are not serving communities well. If schools do not meet minimum standards under one provider, local authorities should be able to intervene to allow other providers to take charge. Such local powers will mean that providers can be more fully accountable for all their actions. In the long run, it may be that removing local authority responsibilities for provision liberates them to focus more on the progress and success of children - whoever provides the schools.
Jon Coles is the former director general of the DfE and the chief executive of United Learning, the academy chain formerly known as the United Learning Trust.