Over the last couple of months I've been asked several times for my thoughts about deliberative democracy on the internet. The subtext has been clear in each case, "What should we be doing?". The timing of these requests has not been coincidental either; most people are asking these questions for the same few reasons:
1. Government faces even lower levels of trust than even the habitual norm - is there anything that can be done?
2. The increasing consumerisation of society suggests that people want government to listen to them more intently, more often.
3. The scope for politically realistic innovation in both direct democracy and normal elections seems limited and unexciting.
So, the idea that deliberative democracy is the most rewarding channel for investment in public engagement has popped simultaneously into the minds of government officials, pollsters, entrepreneurs, techno pundits, and academics.
Which leaves the original question: what is to be done? My answer is as follows.
Before building anything, we need to see that deliberative democracy is not a one-size-fits-all problem. Only once we have a map of how the terrain undulates can we start building robust democratic bridges.
One simple way of making this map is to identify which factors will most heavily influence the type of approaches which can be used. My guess here is that these are 1) the number of people consulted, and 2) the complexity of the issue discussed.
On the first issue, the main cleavage which must exist within any analysis of deliberation is the issue of scale. Quite simply it will take very different tools and approaches to consult with large numbers of people, and just a few dozen.
The second issue is the complexity of the subject of the consultation. Is the issue simple and highly tangible (speed bumps in your road) or huge and abstruse (an act of parliament)? Totally different approaches will be needed for different ends of this scale.
Now that we have two factors, they can be mapped on one of those predictable two by two grids that social scientists love so much. As always, the key question here is whether the axes themselves are truly the significant delineators for different design approaches. I can't really defend my axes as empirical absolutes, but lets call this a starter for ten. I've put letters in each corner so that I can discuss the types of problems and software which might tackle them without actually having to think up names.
A) Complex issues & few people. This is currently classic government consultation stuff, and the main challenge here is how to make the task as pleasant as possible for those people who are most likely to be involved anyway. Paragraph by paragraph annotation seems like the most likely immediate innovation here.
B) Complex issues & many people. This is stepping into newer territory, and is undoubtedly the most difficult of the options. On the bright side, these types of issues are normally the highest profile media stories. One blessing here is that the scale of the issues can be used to dilute the raw amount of traffic, although there will be greater need here for Slashdot and ebay style reputation and rating systems to help the most valuable contributions to bubble up.
C) Simple issues & few people. Most probably local, revolving around differences of opinion about the desirability of clear outcomes (i.e. road humps). In these situations the users will need tools that let them ‘build’ the issues around which they are arguing themselves – setting the terms of the debate themselves, rather than responding to some kind of official document.
D) Simple issues & many people. These are the issues most likely to become screaming matches, and clever moderation will be the must here.
I know that these are hardly incredibly detailed software specifications, but that isn't the point of this post. Rather, this is my opening gambit in what I hope is a discussion which will still be live many years from now.