Sunday, October 21, 2007

Theory and Practice of Netarchy

[warning -- way too long]

Netarchy is a word I came up with to describe an imaginaged system of governance that relies, not on top-down hierarchies of power, nor on the supposedly bottom-up collective decision-making of markets, but on power distributed, transmitted, and exercised through social networks.

There are two different versions of netarchy. There's the purely descriptive version, which is just taking note of the very common observation that in the real existing world, individuals exercise influence by means of their social networks. Having spent some small time with high-powered business types, I have observed that their Rolodex is their most important asset and tool. The same is true in government, journalism, academia, and just about any other field of human endeavor. It's not a particularly original observation, of course. Everybody and there brother is trying to get social networks online or otherwise cash in on them. However, nobody seems to have quite nailed down in words the fact that these networks rule the world. So, perhaps the word netarchy can be of some service in foregrounding that fact.

The second version of netarchy is more interesting -- it's the idea that we can craft systems of governance that make these networks explicit and bring more people into them, and replace traditional party structures with something more dynamic and flexible.

A political party or organization is essentially a network or coalition that connects the leadership with the rank-and-file, along with various middlemen (organizers, consultants, etc). Parties interface with various interest groups (businesses, labor, lobbying groups like the AARP) via contributions and socia connections. The goal of a party is to build a coalition that can take and hold power.

A party or coalition should ideally express the common values and goals of its members. The US 2-party system is rather poor at this, since it requires widely disparate groups and values to be lumped together. The Republican party, for example, has to hold together religious fundamentalists, business interests, quasi-libertarian anti-government types, and hawkish neoconservatives. There is no particular reason that all these factions should be pulling in the same direction.

Parliamentary systems have greater latitude for forming coherent coalitions. Typically countries with such systems have a 2-4 major parties and a host of smaller ones. The parties themselves come together in coalitions to form governments, but the existence of the separate parties serves to coalesce group values in a way the US system doesn't.

Direct democracy

All the parliamentary systems were designed for eras before electronic communication and certainly before the internet. In that era, it was necessary for elections to be infrequent, for representatives to debate and decide things on behalf of the larger populace. While that might still be desireable, it is no longer necessary. It would be perfectly possible nowadays to have direct democracy, where debate and voting is handled over the web. The web has in fact evolved numerous debating societies, but has no real power. It's slowly replacing journalism, but not actual government.

Of course, most people don't have the time or inclination to participate directly in government. Nor should they have to. On the other hand, pretty much everybody hates the political system and the very few choices it presents. So, can we design a system that works better? Let's not worry about the fact that changing the fundamental political system, or even modifying it slightly, is next to impossible. Maybe the coming global-warming-induced collapse of society will create an opportunity for new systems to arise.


Here's my loose proposal for how a fully networked governance should work.

Anybody anywhere can start an interest group, with roughly the same effort it takes to start a blog.

Interest groups can have members, and this process is recursive -- that is, an interest group can join another, larger interest group. This is the coalition/party forming mechanism.

Groups can form their own internal governance mechanisms to make decisions (like which larger groups to join). There will be standardized models available.

Groups that are big enough get to be part of the government. "Big enough" is defined by some threshold of membership, or by taking the biggest n, or something like that.

Assume there's an issue to be voted on. The vote is called, and all the organizations that are part of the government get to express their vote. Every group has a certain amount of voting power. How is this power determined?

Here's where it gets a little bit interesting. This general structure can support a number of different schemes, separately or together. In the simplest model, ever person is allowed to join a single group, and a group's voting power is determined by the (recursive) sum of their membership. The vote of a group is determined in a winner-take-all vote. Issues that come up for a vote flow down the tree, and votes travel back up.

If we let all votes flow up the tree without a winner-take-all step, the result is equivalent to direct democracy, with one imporant difference, which is that belonging to a group absolves individuals of the need to actually vote on every issue. That is, if you belong to a group (say, the equivalent in this new world of Planned Parenthood), and a vote comes up on an antiabortion measure, you have the option of transmitting your vote through the organization, or you can just trust them to exercise a vote for you.

Group membership could be changed at any time, although some groups might try to get longer-term commitments from their members, which would increase their barganing strength. Presumably the large-scale groups would have to engage in the kind of poltical bargaining that goes on now, where one faction trades votes with another -- this is hard to do if the coalitions are too dynamic, so it might be desireable to build some friction into the system.

Complicating things

The above assumes that each individual is a member of exactly one low-level group, which is then bundled up into coalitions. But in reality, people might want to join multiple groups that express their values in different areas. For instance, say I'm an anti-choice environmentalist. Under the current system I'm stuck, since my values don't place me solidly in either party. In netarchy 1.0, above, I'm also stuck. But suppose you can join multiple groups and somehow split your individual voting power between them?

There's a number of different ways this could work. One obvious way is to split your single vote up into fractional powers that get distributed to different groups, but that's not very satisfactory, since it means the more things you care about the less pull you have with each. A more elaborate scheme would be to create a series of rules or filters, that basically assigns a different group to each vote depending on the content of the bill (if bill contains "abortion", count me with The Catholic League, if it contains "environment", count me with the Sierra Club, otherwise, count me with the Libertarian Party).

The computational and network mechanisms to accompish all this remain to be designed. They may not be practical -- after all, designing systems to support our ordinary, simpleminded voting system has many nontrivial security issues to deal with.

What this looks like

Changing the focus to user experience -- what does this mean in terms of web media? I think the idea is that we transform the current sprawl of online forums, blogs, and chatter into a network of debating societies, but debating societies that can actually make decisions and send their collective wisdom upstream. Like any group, there will be more and less active members, leaders and followers -- groups might have internal governance structures with elected officers -- but anybody is free at any time to start their own group. It's still pretty hierarchical, but that's basically a requirement of any system that collects votes and funnels them up to a decision point. If it's a hierarchy, it's at least one that can be dynamically reconfigured at any time, on any level, if the participants feel like it.

Essentially we are replacing the legislature with a more dynamic networked structure for collecting and conveying people's opinions and votes.

Well, this will never happen, barring a major revolution. Still, it's interesting to imagine what could happen if our 225-year old structures of governance could be given a modern technological makeover.

Hm, on reflection, what I think is more likely to happen is the government devolving and getting slowly replaced by networked organizations that are more efficient and responsive. Maybe they'll work as described above. Maybe, as they start small and grow and interconnect to form something bigger that can actually manage the planet, or what's left of it.

1 comment:

goatchowder said...

"It's not a particularly original observation"

But a very useful one. My old man used to drill it into my head when I was a kid: "It's not WHAT you know, but WHO you know, that matters."

He was, and still is, a salesman, evangelist, and politician (and now a Ph.D. and college professor too), so the sum total of his life's work has been the plying of social networks.

Humans are social animals. Ain't no way around that.

Excellent, thoughtful post.