What shape is a feedback loop?
This is a question prompted by some slides published by the Alphagov team which look at ideas for "closing the feedback loop" by improving online consultation and policy engagement across government.
Co-author of the slides Steph Gray has been seeking "some more first-principles thinking on consultation", and many of the important issues have been covered in the explanatory post from Neil Williams and in follow-ups by Simon Burall of Involve here and here (the comments on these posts are worth reading).
This Andrea Di Maio post on where to use technology in the policy-making process is also worth reading in the Alphagov context and in light of the specification for the new DirectGov ePetitions system which tends to suggest the costs of moderating comments on government websites is a problem.
Anyway, there is not much I could add to what has already been covered elsewhere, so instead this post is heading off at a tangent to deconstruct the metaphor of the feedback loop.
This is excessively pedantic, I realise. In principle, obviously the concept of feedback and corrective action is great. But perhaps there are some distinct senses in which this can be analysed and applied.
A traditional policy cycle
In its Green Book, the Treasury explicitly spells out its view of how the policy cycle begins with a rationale which leads to objectives, appraisal, monitoring, evaluation and feedback.
Whether policy-making can actually be done in neat cycles like this is an issue discussed in the Institute for Government (IfG) report Policy-Making in the Real World.
Interestingly, the IfG comes to a similar set of conclusions to the Alphagov team, suggesting that "we need to bring the policy process closer to the real world, and bring the real world closer to the policy process".
While policy design is a broader topic of which consulting and seeking feedback is a subset, the desired outcomes are essentially the same.
Still, if there is an underlying assumption to both the Alphagov and IfG thinking, it seems to be that through feedback and iteration, policies can be improved over time by taking them in a particular direction.
To me, though, that idea seems fairly top-down and Whitehall-centric.
I think there is a parallel here with Simon Burall's post which argued that the idea there is a "real, knowable Public Opinion waiting to be discovered" should be abandoned.
The idea that there is a platonic policy on any issue which the government can slowly move towards is equally fanciful.
So perhaps a feedback loop really looks something like this (as you can tell, I'm not a graphic artist).
Here each loop may be some localist microinnovation broadly heading in the same direction as the thrust of government policy.
But there is still something missing, I think, around understanding the complexity of a landscape which has more localism and diverse service providers and greater personalisation.
Maybe a better image, then, would be this.
Here the trunk of the tree is the original policy intention.
In the case of, say, health service or school reforms then the trunk might be the plan to create new GP consortiums or more academies, but there would also be a pretty significant branches near the base of the trunk given that the opt-in nature of these bodies means that existing structures can continue almost in parallel.
Feedback on one set of policies or organisations might be entirely different to feedback on another. This points to why the idea of a single loop is problematic.
The Alphagov team also suggests using the website to gather feedback on both the delivery and policy at the point at which citizens interact with a government service.
However, the tree image can prompt us think about the blurred line between policy and delivery.
If the trunk is the policy, by the time we reach one of the many individual leaves at the end of the branches then we may well be talking about operational issues.
This might range from the seating arrangements in a Jobcentre Plus in Norfolk to the visiting hours at a hospital in Swansea.
But it can be unclear at which point along the branches the policy becomes an operational judgement. To take a current example, are recent criticisms of the Care Quality Commission due to management misjudgements or policy confusion? Will citizens know enough to be able to make a meaningful distinction?
And if the policy is fundamentally flawed in some way, no amount of feedback and tinkering with delivery is going to improve it (what impact would feedback on the Child Support Agency have had?).
So having deconstructed to death the word 'loop', my conclusion seems to be that even circles can be too linear.
A question of leadership
One other thing is missing so far from both the loop and tree metaphors though: Where is the beginning?
In part this is an issue about co-creation and how early in the process citizens get to have their say (even to the extent of defining what the problems are that need to be addressed).
But it is also a question of leadership and not ignoring the role of politics and politicians in this process.
One purpose of consultation and engagement in the policy sense is to help leaders make better decisions with a broader range of information and a wider spread of views.
Another purpose is to educate the participants in the difficulties of making choices in government.
A minister, or indeed a civil servant, needs to lead debate online as they do in other media if they are to do their job effectively. This is particularly the case when ministers have manifestos to implement and political agendas which they are elected to follow.
Public feedback is never going to be the only driver of innovation or policy, so the explanations of choices which might be provided back to citizens as part of the loop may be intensely political. But that is not a negative thing.
What makes policy trees bear fruit?
Let's stick with the tree imagery a moment longer and consider that co-creation may perhaps be viewed as the roots, the trunk is policy, branches are innovation and leaves are delivery.
In the terminology of Sir Michael Barber (head of the Number 10 Delivery Unit under Tony Blair), a line drawn from the trunk to any given leaf would constitute the delivery chain.
The current government's approach to structural reform and incentivisation, however, would seem to suggest that feedback at the level of each individual leaf is what matters.
This is stretching the metaphor of the tree beyond breaking point, but it might just point to one more lesson about these contrasting approaches.
Under Labour it was the flow up through the trunk and out to the leaves which was important.
The coalition believes that it's the sunlight of transparency and competition at the leaf level which matters most.
But actually both are needed if the tree is to remain healthy.