How should Government consult? There is – obviously – a long piece of guidance on the matter – owned, bizarrely, by the business and skills bit of government, but it doesn’t go nearly far enough.
The Code of Practice on Consultation is pretty strict – twelve weeks to respond, don’t do it during an election period, set out a “consultation stage impact assessment”, and so on. It even reminds officials not to use consultation when what they really mean to say is “here’s the policy, it won’t change” – which still happens more than it should.
Yet despite the guidance, the real experience of being consulted is pretty dispiriting. In most cases, there is a list of questions that you are invited to answer, either by ranking items or filling in freeform text. It’s like a really dull opinion poll, or perhaps the sort of quiz show you quickly switch over from. There aren’t opportunities to stand up and say, “but you’re assuming…” or “but why aren’t you asking about…”.
The Government’s consultation rules aren’t wrong per se, they work pretty well for situations where you are asking a group of reasonably knowledgable people with an opinion to give their comments on a technical proposal. However, that’s only a small part of the interaction that Government needs to do around policy, and on many of the other occasions standard consultations produce doubly-negative reactions: the citizen gets frustrated trying to fit what he wants to say in to a narrow question-and-answer format, and the policy officers get frustrated reading all these stupid responses that JUST AREN’T ANSWERING THE BLOODY QUESTION.
It’s no wonder that lobby groups, who know exactly what they want and employ people to answer consultations effectively, seem like such allies when you’re a policy official – they can give you answers you can use, and they understand the constraints of the process.
Government could do consultation much better, and I hope that with the work on engagement being done in the GDS, there’s a better chance of improvement than there was a few years ago. We can’t just digitise the existing narrow process, and we should aim for consultation throughout the development of a policy, not just right at the end.
Representative democracy means that Ministers have the right to set the framework for decisions – but policy makers and politcians could get more out of the public, and develop better policies, by engaging in different ways, appropriate to different situations. They should think more about the audience, the expectations framework within which they are setting up the discussion, and what data they really want to get out at the end.
If we can get it right, Government at national level can support and model better democratic conversations around issues – and provide an infrastructure that local-level public services can use as well.
What sort of situations should we consider? As a start, I’ve come up with five broad types of consultation, different enough to need different tools, but all usable in developing a single policy.
Type one: Mastermind – a list of questions on a particular topic, looking for a set of simple answers. This is most like the traditional model. Government is looking for discussion on the detail, and the audience is potentially general, but realistically restricted to people who are interested either because of personal involvement or political belief. The expectations framework ought to be what the underlying policy and boundaries are (such as EU law), and it should be clear that this is a chance to comment on the detail. Preparatory information can assume a reasonable level of interest in and knowledge of the issue. At the end of it, the civil servant gets a set of answers to inform the final decisions on detail. Most of the answers are probably fairly sensible, or at least predictable, and a few will be
completely barking independent-minded. It’s not a representative sample, but it’s probably got good coverage of those who take a personal or professional interest in the issue.
Type two: Crystal Maze – running people through a simulation of something and seeing what decisions they make. This is pretty common at local level in the form of council budget simulators, and some are quite game-like. Government is trying to entice people who are probably not interested in a dry policy area, and take them through some structured choices and options in an entertaining way, so they can see where people end up. Expectations framework: we’ve written the game with a set of rules that reflect reality and the policy framework, that’s how it’s going to have to be played. Preparatory information: ideally none, but the game has to start from an expectation of zero knowledge. What you get: a set of outcomes and the pathways that people took through the game, with (if you’ve asked) basic demographic information on who has said what. A lot of data that you can fiddle with in spreadsheets, but not information provided on the basis of much deliberation.
Type three: X Factor – lots of people saying or suggesting something on an issue, not considering the comments of others. This is the “crowdsourcing” model of Your Freedom – and also the “fire and forget” model of much online commenting. I didn’t love Your Freedom when it was out, but there’s a time for gathering in random ideas or comments from the public – you just have to be credible and honest about how you’re going to follow through. Expectations framework: what Government is asking for, and credible promises on how it will be openly used. Preparatory information: depends on the issue, but ideally a lot more than Your Freedom provided, and not so much that no-one gets past the second page of the website. What you get: the aggregated jerking of knees – but a lot of knees. You won’t get comments on detail of policy, you won’t even be sure that the policy is what people are reacting to, but you will have a huge sample to wade through and analyse.
Type four: Open University – deeply technical requests to deeply technical people. Lots of Government consultations are like this, and they are so different in nature and purpose from everything else in this list that they really need a different term. What you get: seven really, really detailed comments on aspects of the policy you’d never thought of before.
Type five: Review Show – throwing out some creative things and getting interesting people to talk about them. This is where the most interesting potential lies, and it’s a model that’s farthest away from the traditional approach. It’s not quite deliberative democracy on the Fishkin model, and it’s not quite the networking community activists that we are working on in our We Live Here project, but it’s the potential for connection between the two – a sort of conversation-in-the-network or (to steal Patrizia Nanz‘s phrase) an Internet of Citizens. The expectations framework is simply that you’re participating in a conversation that will be listened to, and that contributions that are evidenced and take account of others’ views will be listened to most readily. The preparatory information needs to be both a good introduction to the topic, and something that anticipates (and responds to) questions or possibilities raised during the discussion. In this model, Government can’t stand at the back of the room with its arms folded, as it does with traditional consultations, it has to respond live – not necessarily pitching into the discussion, but presenting useful information and data that’s informed by what people are talking about. In return, there is the potential for rich information not just about preferences and discussions during this debate, but also where people are coming from (networks and places) with those opinions, and how opinions are changing when people are exposed to new data and information.
There is also perhaps a sixth type, not really feasible at the moment, and not even really a consultation, that you could call “reality TV”. By that I mean consultation on the model of “smart cities” – watching the aggregate of action or interaction as people in a place go about their lives, without trying to influence them, just connecting that lived reality into the policy-making process.
I’m sure these categorisations are wrong, or incomplete, and any comments are welcome below, but I would suggest GDS start with this sort of broad view of interaction between state and citizen when thinking about how they handle consultation in new ways.