The middleware of democracy. Or from knowledge to wisdom: or at least knowledge 2.0

Simon Heffer’s High Minds presents us with a portrait of the mid-Victorians in which they consciously set about building the world which became ours. A liberal democratic world.

To do so they recognised the need for all sorts of public goods. Those of education and health surely enough, an honest public service chosen on merit too (an idea they nicked from the Chinese who’d been at it for a millenium or so) and also civic virtue. It’s a stirring and a sobering story reflecting an age which I think had a more balanced understanding of the necessary ecology of public and private goods each reinforcing each other in building the Good Life.

Today for all manner of reasons – intellectual, sociological and economic – our contemporary vision is profoundly skewed toward private good and private endeavour as the paradigmatic category. That’s why I regard it as a happy hunting ground for low hanging policy fruit – a panoply of ways to drive productivity and economic growth that don’t even cost any serious government money.

But as Heffer makes clear, this Victorian quest was not just economic. It was a political project. As he argued in an interview with Geraldine Doogue – which I quote from memory because I can’t find it on the ABC website – they knew that democracy was coming, so they needed to get The People a decent education before they used their vote to wreck the place. As I explained in a post a week or so ago:

One of the most important things I know about political discourse I got from a few lines in Schumpeter’s Capitalism, Socialism and Democracy. The ‘engine’ behind democracy – what makes us engage – is not reason . . . but affect – our emotional and expressive selves. And this governs what gets covered by the media – what takes off as a meme and what doesn’t.

Note as an aside (but we’ll come back to it) that if Philip Tetlock’s work is anything to go by in many ways experts seem to be in the same boat. Certainly when one measures the extent to which experts are able to produce reliably better predictions of likely future outcomes, or even to judge the reliability of their own guesses, it’s a pretty depressing picture with the ‘hedgehogs’ focused on one big thing (I’m thinking there’s an emotional commitment here!) being worse than monkeys in front of a dart board) though prediction is just hard. Foxes (who know many things) being only a little better than an interested amateur.

The other concept Schumpeter brings to the table is the division of labour. Any social formation of even the slightest sophistication requires a division of labour. Even the local football club has a division of labour, not just out on the ground, but also as an organisation as reflected in the differentiated roles of president, secretary and treasurer. And these things pose the big problems for democracy, because democracy is rule by the people – who must ultimately understand and support whatever division of labour we work out to make all the decisions that must be made to make a government function.

We have a range of ways of handling these issues. The Victorians rightly spent a lot of their time worrying about the tyranny of the majority and so championed things like the independence of the judiciary. We also have various lower levels of independence for institutions like statutory authorities, the central bank, the bureaucracy and and so on.

Modern democracy was also built on a class system. ¹ But the class basis of democratic capitalism – in which the middle class and its preoccupation with respectability defended various abstract principles like ministerial responsibility – is being broken down by the bread and circuses of vox pop democracy and the politico/infotainment complex.

A more radically democratic way of handling the division of labour problem involves institutions of deliberative democracy – bodies chosen by sortition. The jury is a special purpose cognitive elite relating to the specific case but one that is radically and demonstrably democratic. This addresses the division of labour problem. Not everyone has to be an expert on the case – just those representing ‘the people’ on the jury in the particular case. As I and others have argued, I think such bodies could play a similar role in our democracy – either on an advisory basis or by being given some specific role as a jury is given.

One of the issues of the day is how one might be able to use the internet to scale such things. It’s not an easy question. As I’ve argued, in some ways the internet makes things worse, as it speeds the pace of deliberation and people refine their clickbaiting capabilities. But the Melbourne based website YourView tries to scale democratic deliberation by hosting discussions with the back end of the website helping to work out who is contributing to the conversation in the way that others find most constructive. Thus it rewards what it calls “epistemic virtues” by giving those that exhibit them higher credibility scores. It’s more complicated than this, but it will do for discussion to say that those with high credibility are those who make comments on issues which are rated most highly by those who disagree with them.

One interesting direction in which Philip Tetlock has taken his work is the Good Judgement Project. Here good judgement is bred first from teaching people how to make better estimates of how confident they should be in various predictions they make and also by allowing such people to compete. The project appears to be successfully cultivating a cadre of people who can recognise the amount of faith that should be put in their views and puts them in an environment that encourages them to make unbiased assessments of those matters (as does the Gruen Tender! But I digress).

But here’s the thing. It seems much easier to make progress on the ‘is’ than the ‘ought’ front. I guess that’s not surprising. Science is the miraculous engine we evolved to help us figure out how the world is. But it’s a whole lot harder to work out how the world ought to be. When I was doing the Government 2.0 Taskforce I was wondering – as were many people – how can we get a Wikipedia of government. That’s where I came up with the is/ought distinction. In figuring out how the world ‘is’ or even ‘will be’ we have made some good progress in the last decade or so with

• Wikipedia and similar informational goods on the net
• Prediction markets
• The Good Judgement project.

But while all of these new things help discipline the process by which we aggregate and judge views – including, crucially, sorting good from bad views – the fact remains that these things only measure good judgement regarding predictions of the way the world is or will work out not how it should be or work out.

That’s what YourView was built to try to do, and it seems to do a reasonable job of it. So ladies and gentlemen it’s a greatly under-appreciated public good. I think we should be developing the site – and alternative approaches to doing what it does. If I were running a large company with thousands of employees, I’d like to get all my people onto it and see how they went and see how useful credibility scores were in identifying those with good judgement and/or those who were worthwhile parts of the pluralistic pursuit of making the right decision – and who were not. That’s because such a capability would be a powerful company specific public good.

Even more so if I were running a school or university – because this kind of thing is a great training ground for those on the site – as well as a resource for better deliberation on specific issues. Indeed, I’ve lamented previously educational institutions think of themselves and their mission as synonymous with the public interest, and yet their work is hugely degraded by the crowding out of public endeavour by private competition. Thus universities preach the love of truth and all that kind of stuff, but their love of truth turns out to be somewhat less than their love of whatever it takes to get up the university rankings including publications which can’t pass the fundamental test of reproducibility.

But in an age when governments spend vast resources on education and support the ABS, the Bureau of Meteorology and any number of other public interest public good information resources, when they fund the electoral commission and things like the ABC as part of our deliberative democracy, it’s unfortunate they don’t see the merit in spending a few peanuts trying to develop the ideas and the capabilities that YourView is pioneering.

YourView is currently largely in mothballs and not being promoted by its owners any more. They’ve promoted it pretty vigorously for a good while now, seeking  support from media companies, philanthropy and various educational and intelligence organisations but so far without sufficient success to ensure viability.

So next time you hear someone banging on about how data isn’t information and information isn’t knowledge and knowledge isn’t wisdom and all that stuff, ask them if they know about YourView and if they’re helping it try to scale the solution to that rather large problem.

More here

^{1.Perhaps partly influenced by the aristocratic system of honour in which, once honour was besmirched or even called into doubt it could not be regained – something Douglas W. Allen suggests radiated out from the sovereign’s inability to surveil those putatively acting on his behalf – the middle class cleaved to respectability. Principles like that of ministerial responsibility and the need for public action to be beyond reproach were not expressions of, nor defended by, popular values or votes. (I think it is true that in the 1977 election, there was only one Liberal member who enjoyed a swing towards them – it was Phillip Lynch whose integrity was under a cloud for profiteering on land deals – though one that was largely lifted by an inquiry after the election) ?}