Teacher autonomy is the most difficult issue in the education debate. Part 1

September 28, 2019

How much freedom should teachers be given to do their own thing?

Few things are more irritating to educators than knowing what needs to be done, and being stopped from doing it by those in charge of you. Whether that’s the actions necessary to keep order in the classroom, or keep students safe from each other, or the way of teaching that will make the most difference. At times, one is tempted to ask for complete autonomy. Freedom to make whatever decision one wants to. However, this principle becomes less appealing when one realises that what happens in other classrooms affects expectations of behaviour and effort in one’s own. It also becomes less appealing when one considers the consequences for children of the worst classroom practices, both educationally and with regards to their safety and well-being. A related issue is workload. An assumption that every teacher plans their own lessons, may undermine collaboration in the production of resources, leaving teachers to duplicate work their colleagues have already done. Conversely, decisions imposed on teachers by those less familiar with their classes or year group, may have adverse workload consequences for teachers who find they constantly have to make adjustments for a poorly sequenced curriculum or to prepare for badly designed assessments.

It’s easy to make rhetorical arguments in favour of teacher autonomy. Teachers should be trusted. Teachers are professionals. Teachers know their classes best. It’s easy to make rhetorical arguments against teacher autonomy. Teachers must be required to teach effectively. All teachers should have high expectations. All teaching should be based on how students actually learn. Students need consistency. Once you accept that all these arguments are sometimes true, the debate becomes about where you draw the line, and that’s tricky.

In this post I will summarise my past posts on the topic, which can be found here:

In those posts I concluded a number of things.

Managers should try to avoid giving any of the following:

  • Instructions contradicted by other instructions.
  • Completely idiotic instructions.
  • Instructions that no manager would ever subsequently admit to giving.
  • Instructions which, if followed, will be used against the teacher following them.

Some of this might seem obvious, but none of these things are uncommon. Dysfunctional management is by all reasonable accounts a problem in teaching and it is worth considering where managers should definitely leave well enough alone. However, attending to this only narrowly restricts the places where the line can be drawn.

Moving on from the day to day decisions of managers to the systems used to manage teachers, I suggested the following should be avoided in any system of holding teachers to account.

  1. Trying to achieve multiple aims simultaneously and without a clear indication of priority;
  2. Holding teachers accountable for methods and outcomes simultaneously;
  3. Enforcing, and creating paperwork for things that would happen anyway;
  4. Creating work that does not have to be done;
  5. Measuring and judging things that don’t matter;
  6. Measuring and judging things unreliably;
  7. Encouraging behaviour that is actually counter-productive;
  8. Wasting money, particularly on management salaries.

Again, this stuff might seem obvious, but it is all incredibly common. I believe almost every large school would gain from applying these principles to all of its rules and systems for holding teachers to account.

The point about the problem of trying to achieve multiple aims simultaneously is one that applies at many levels and across the public services. The philosopher Onora O’Neill, when talking about accountability described the following problem:

Traditionally, the public sector exercised control by process. We often call it bureaucratic process. The private sector allegedly exercised control by targets. When the target setting was imposed on the public sector, the process controls were not removed, hence the problem of having to be responsive to and responsible for two completely different sets of controls whose coincidence is not guaranteed.

Teachers should never be held accountable for outcomes if they were not given the freedom to affect them. In this era of workload concerns, I would add that if they can affect those outcomes, but only by taking on more work, that should also be considered unreasonable.

One helpful way of looking at restrictions on autonomy can be found in this blogpost by Doug Lemov which appeals to the concept of “positive and negative variance”.

… one of the strongest ways a school can make a difference in student achievement is to have a coherent approach to teaching, one that outlines a shared understanding of “how we do it”—things that comprise a schools core approach that everyone is expected to do. The school should name the things that are part of “how we do it” and then provide training  so predictable implementation errors are reduced. That’s a way of both aligning and implementing a philosophy but also of reducing negative variance.

But it’s super-important to balance that reduction of negative variance with an understanding of the benefits of “positive variance”… the idea that people who have achieved proficiency with a skill should have the freedom to personalize and adapt.

The example he gives is centralised lesson planning. Preparing lessons centrally will reduce negative variation in that it will make it harder for teachers to be under prepared. However, in order to encourage positive variation, teachers will need to be allowed to adapt the lessons and be progressively given decision rights that can include dropping the centrally prepared lesson entirely.

I suggested the following principles might help with ensuring there is less negative variance and more positive variance.

  1. Outcomes must be considered before processes.
  2. Schools should be upfront about what they want.
  3. If you can’t write down clearly, concisely and objectively what you want, you have no right to ask for it.
  4. The best justification for restricting autonomy is where a teacher’s behaviour will undermine colleagues. e.g. differing expectations for behaviour across the school.
  5.  Don’t take the piss. i.e. don’t have systems that can be harmful to teachers in themselves by adding to stress or encouraging bullying.

That’s about 1000 words summarising what I’ve already said on this topic. I am fully aware that everything I’ve said only suggests some constraints on where to draw the line, and doesn’t give any easy answers to the question of teacher autonomy. In my next blogpost I hope to add a few more considerations that I haven’t covered previously.


  1. I would argue that you are perhaps conflating ‘autonomy’ and ‘agency’.
    Autonomy is a comparative lack of regulation of teachers’ work—an insufficient condition for teacher professional action. Going with the flow, reproduction of habitual patterns.
    Agency, on the other hand, is about facilitating effective practice—acting with intentionality, to formulate possibilities for action and exercise choice.
    Teachers should be active professionals, not simply left alone to do things their own way.

  2. […] Teaching in British schools « Teacher autonomy is the most difficult issue in the education debate. Part 1 […]

  3. […] Teacher autonomy is the most difficult issue in the education debate. Part 1 […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: