Let's assume there are certain categories of government, things like "education" or "public transportation", etc.

For each of these categories, let each citizen either name another or declare him- or herself competent. You can't name someone who names you. The resulting relationship forms a "forest", a bunch of separate trees, with people in the nodes.

Record these relationships in a big, database. The relationship is transitive; if A represents B, and B represents C, A represents C.

The database could be updated in near-real-time. If I want to switch my allegiance in a field to someone else, I can do that by going to a website / making a phone call / going to a "voting booth" that is open year-round.

When decisions are made, they're made among those that don't have a representative. They speak with the weight of all the others who trust them at that particular point in time.

A representative reports to their electorate; in turn, representatives receive a running update about how many have left or joined their position.

Politicians are motivated to inform the public about impending decisions because the public actually gets to affect them. People are motivated to hear about what's going on politically because, again, they actually get to affect the outcome.

But nobody *has* to be knowledgeable about everything all of the time. You don't have to vote if you don't really care; just find someone you think is trustworthy and stick with them. (The trustworthy person doesn't need to be a career polititian, just someone who invests a little more time into making choices, maybe motivated by the additional weight.)

[Update, Nov 13 2000]

Problems with this appproach:

- [PeterSealy] It's hard to explain. -- I'm not sure that's true; I think the current system is much more complex. Granted, the things a citizen gets to do are ritualized and simple, but if one wanted to actually exert political influence, one would be faced with a baroque machinery of lobbyists and political clubs that are mostly hidden from view and difficult to traverse.

- [PeterSealy] The system would be susceptible to a country's daily mood swings. -- I think this touches on a deeper concern: most of us don't actually believe in democracy. We don't think that the people would be any good at deciding things at all, and are actually quite comfortable with things being run by a benevolent conspiracy.
I think that power and ability ultimately correlate. Only if people are exposed to the results of their mood swings, they have a chance to learn to recognize and suppress them. This opinion is clearly part of the overall liberal canon of belief that "deep down, people are good"; if you're a philosophical republican, you will find that idea ridiculous and will instead look for a better system of regulations to keep the unwashed masses in check.

- [egnor] Categories don't cleanly cover everything. -- As long as every issue people get interested in is assigned to *some* category, it would work, even if the assignment is inaccurate. (The system would still work even if there would be no categories -- one just would have to look harder for a representative.) But I'm not too happy with this myself; alternative suggestions are certainly welcome. (Maybe we can all just submit keynote expressions that dynamically pattern-match against the issue certificate.)

- [egnor] Horsetrading would be difficult. -- Secret horsetrading would be near (neigh?) impossible, as long as it relies on people remaining in power and keeping their word. Instead, it would have to be replaced with open horsetrading, where people clearly see that they support both the upside and the downside. As with [PeterSealy]'s mention of mood swings, this is part of the section where everybody gets to grow up and realize that they don't live in Fairyland.

- [Aristotle] A web-based system is subject to technologically mediated attacks. -- Yes. I think there can be technological solutions for that, but I'm also very glad I don't have to design them. The web access isn't central to this scheme; as long as there's some safe way of changing your recorded allegiance in a reasonably short amount of time. I think we're at least nearing a point where the whole thing is actually implementable; thirty years ago the thought of even trying to maintain the database would have been preposterous.

- [Aristotle] People could be spammed into voting for things. -- That's true for elections now. I'm actually reducing the effectiveness of propaganda by (a) removing the single point of decision and (b) increasing overall political literacy.

Finally, in keeping with the spirit of halfbakery, I propose that people with a sufficiently sized electorate should be allowed to wear ceremonial funny little pillbox hats with gold stripes in public.
-- jutta, Nov 12 2000

(?) Liquid Democracy Voting System http://twistedmatri...mocracyVotingSystem
Shares elements with this idea. [jutta, Dec 29 2004]

I will wear one if you will http://www.dorothea...ge.com/101_6627.jpg
Pillbox hat with gold stripes [normzone, Aug 28 2008]

Lots of people claim the current "impasse" "means" something, but it's really just random chance. Lots of people like to gripe as if uninspiring, near-"identical" candidates are some new phenomenon, but for every Nixon vs. Kennedy there are dozens of Tweedledum vs. Tweedledee. Some would argue that the production of candidates with similar platforms, driven by sophisticated polling systems and receiving very narrow margins of victory, represents the victory of modern democracy.

Jutta's proposal is very interesting. It should probably be tried in some more limited forum to see how well it works. What would be a good testbed?

One problem is to figure out how to categorize the relevant issues. Is NAFTA an "environment" issue, a "labor" issue, a "trade" issue, or what? If all of the above, how do your votes get apportioned?

Another issue common to most direct-democracy schemes is that everything is focused on the short term. Under a representational system like our current one, politicians are free to make decisions they know are unpopular now, in the expectation that the end result will be popular in a couple years. This is especially relevant to "horsetrading", where candidates can trade support for issues of varying importance. I'm not expressing this very well, but I'm trying to say that it's important sometimes to vote for a leader you believe in, rather than a platform you agree with. I realize that this system allows some of that, but there may be something to be said for only being allowed to vote every few years.

Finally, you might allow people to split their vote between multiple "authorities" on a topic.
-- egnor, Nov 13 2000


Egnor: regarding a good testbed for this idea... here at Halfbakery we have categories, voting, and a big database -- three things mentioned by Jutta in introducing her idea. So maybe Halfbakery would make a good testbed!
-- PotatoStew, Nov 13 2000


PotatoStew: What, "I designate PotatoStew to vote for me on all 'Car' ideas"? Mmmaybe, except that the votes don't actually do anything important.

Collaborative filtering could be an alternative (you vote on a sampling of issues, and eventually the System correlates you with others), but it's probably too mysterious to adopt for Important Political Matters, and its anonymity fails to encourage people to actively "campaign".

Aristotle: You're being awfully vague. Describe how these "spam based attempts at manipulation"? Make some effort to explain why this reasonably well-thought-out form of variable representation is not suitable to escape the "anarchistic" bogeyman you invoke? It hardly seems like anarchy to me.
-- egnor, Nov 13 2000


This sounds like a system of Proxies and Plebiscites.

Categories, as egnor pointed out, sound problematic. One simplified alternative is this: Create the same sort of system of computerized plebiscites. Everyone who votes on these has the option of making their position public. Anyone who wishes to automatically cast their vote in with someone else's can choose to do so before each vote. This does away with categories, but need not eliminate the forest -- I may give my vote to someone I merely trust, they may know someone who they believe to be competent in the relevant area. (This requires a voter to choose someone else for each issue, unless they are devoted to a single person, but in the original idea they were obliged to choose someone for each category anyway; people could now maintain their own lists of people whose judgement they trusted in different matters and judge which was best for the decision at hand).

The desciption leaves me a little uncertain, but I take it that there are still elected officials of the usual kinds (as there must be for a national government, but not for some things this system might be well-suited to). Perhaps the thing to do is to elect representatives in the usual way (or in a more direct way, which might be nicer) and then oblige them to bring their stated policies before the mob to be vetted or replaced. Decisions they make thereafter can be challenged by a sort of gigantic no-confidence vote, or overturned by judges, if they appear to be inconsistent with the policies (these would, in fact, comprise a sort of temporary constitution). It might be difficult otherwise to decide which decisions require public backing and which do not. Another advantage is that this gets things over with, something else engnor mentioned, and avoids the possibility of a state of perpetual warfare over, say, to pick a random example, abortion.

This has all the usual disadvantages of direct democracy (the 'mobocracy' ones), and the proxy system seems like it would be open to abuse of all kinds, but it still sounds like a good (and very neat) idea.
-- Monkfish, Nov 22 2000


It's an interesting idea and one I like - but I see one or two dangerous possibilities.

Since these executives would have to remain in place for a period of time, and as such become knowledgable in their subject matter, it's not inconcievable that they themselves might become representatives themselves.

A process of lobbying the 'tops' of disparate trees could rapidly swing the public vote in wildly unexpected and unwanted directions. It would be a matter of knowing and being able to influence the people at the tops of those trees in a fluid enough manner that would then become the basis of power. In short, there is potential for massive power to unnacountable individuals.

A curb on this fluidity might be the enforcement of a rule that limits the number of people in any tree (or have some degree of locality). This way, you couldn't have the whole population of the country suddenly switching their vote because a celebrity sold their responsibility on to someone on the opposite side of the fence (as if that doesn't already happen)

A 'buck stops here' flag, or timing device might be another method of curbing the fluidity; Where people, in addition to not nominating anyone else, also have their intention not to do so at a later date enforced by the system. Or perhaps one's ability to nominate someone else is slowed in proportion to the weight of the people one is representing.
-- zen_tom, Dec 29 2004


Awwww....

I thought it was a Vernon idea.

Still, +.
-- DesertFox, Dec 29 2004


//I thought it was a Vernon idea.//

I did too for some reason.
-- supercat, Dec 29 2004


To make sudden power concentrations or vote swings less likely, you could have a delay when someone changes their vote or representative. The new vote takes a short time, maybe a month, to become valid, but the voter may have the old one revoked immediately. This wouldn't work for quick decisions.

Another compromise is to have a second house elected for longer term, which has veto power, but have the veto overridable either by sufficient votes, or expiring after a limited time.

I had an idea for the categories, I don't think I'd trust it but anyway: each representative chooses the category for each issue. Disputes could arise though. For example, my Public Spectacle representative says some unusual fireworks display is a public spectacle issue and my Military representative says it's a military issue, and they disagree about the vote. If I've ranked them, that determines my vote, otherwise I abstain since it's unresolved. My ranking wouldn't just be determined by importance of the issues, but also how much I trust them to stick things in the right categories. If I noticed the issue I could make my own decision of course.
-- caspian, Jan 04 2005


My Pork Barrel representative votes in favour of a subsidy in my direction, my Tax System representative votes against it, and my Foreign Affairs representative wants to get the subsidy proposed, but then withdrawn in return for a concession from a foreign government competing with my industry.

Hang on - I've read about something like this; scenes from Paris in 1870, when the French were trying to conduct the Franco-Prussian war directly from the floor of the parliament, without the use of a cabinet. Hmm...
-- pertinax, Jun 21 2006


Updating this idea circa 2006, the real experts on any issue are the lobbyists and the associated bloggers. The lobbyists already influence the votes, and in recent years, so do the bloggers.
-- theircompetitor, Jun 21 2006


I like the idea of people representing different government functions rather than regions. The current system rewards pork, as opposed to working together to solve national problems. I agree with pertinax that jurisdiction issues would be common and with egnor that this and other variations should be tested on a small scale before risking widespread use.
-- Ford, Jan 17 2008


Yes this is similar to my idea, however yours is more a system of dynamic representation whereas mine is more a ongoing poll of experts in a field.
-- johnbakersmon, Aug 28 2008


I can't believe I've never seen this before. Nice one jutta.

It would be informative to know what shape the trees were - they could be very bushy, if many people choose the same person as their proxy (such as a known politician or celebrity), or very branched (if for example people tended to assign their representative from someone they know well and personally).

Idealistically, I'd like the latter, but realistically, I'd expect the former.

What happens if your representative doesn't vote, and doesn't assign a representative?
This isn't a show-stopper, it could be a legal requirement that representatives either vote or have their own representative assigned. Or you're informed if your representative doesn't have their own. Or various other possibilities...

How do you avoid generating circular references? Suppose someone is a reference for a great many people. They will therefore find it hard to pick their own proxy, because anyone whose views they like will already have picked them, perhaps indirectly.

An alternative would be to allow circular references, but to allow anyone to 'override' their reference by voting. If members of a circular chain were informed of being such, yet noone votes, we can safely infer that they don't give a damn.

It should also be the case that people who have assigned proxies be told (or be able to find out) how their votes were cast (although perhaps not who cast them if cast indirectly). That way people can avoid those who say they support one thing but vote differently.
-- Loris, Aug 29 2008


I was thinking about this idea yesterday or today. I had forgotten that you had included the categories feature; I was planning to post a version that added categories.

The solution I thought of for categories being imprecise was to allow each proposal to be in multiple categories, and have a separate category for voting on which category/ies a proposal should be in. (Maybe: Each proposal gets assigned categories by its proposer. A proposal is automatically created in the category category to approve this categorization. Each proposal whose proposed categorization fails gets several proposals in the category category. Each one of these proposals is for the main proposal's membership in one of the non-category categories.)

I'm not sure how to solve the issue of delegates in different categories voting differently on the same proposal. Maybe you could set an order of precedence for categories or delegates, or maybe in that case the system could prompt you to vote for yourself.

The other improvement I thought of was to implement it as a distributed system backed by cryptography (like how cryptocurrencies and DAOs work). It would still need some form of central control, probably, by the government, to ensure that each citizen has exactly one account, but it could use technologies like tokenization (widely used with online credit card payments) to add security and anonymity if necessary. (Because nymity might be desired for delegation, you can of course connect and verify your identity with the token, and if you want to be anonymous again, just generate a new token, which automatically invalidates your old one.)

(One possible way this could fit into a wider national identity system: The government gives you an ID number at birth registration or immigration/naturalization, and then you can generate some kind of one-way-linked identity based on that main ID. You can generate as many as you want, to use for different systems like voting and taxes, so your votes can't be linked to your taxes (e.g.). But they can all be linked back to your primary identity record only enough to ensure that you're not committing some kind of identity fraud. Perhaps the same kind of technology SQRL uses to generate a new identity for each website you log into while regenerating them on-the-fly from your main identity each time you log in could be applied in some form.)

Any online form of this idea could also display your history of voting and delegation, and how your delegates voted (though I guess that's an obvious feature to have).
-- notexactly, Jul 24 2017


The issue of mood swings is handled by precedence in Westminster based systems of government --- that being pretty much all of the first world.

Perhaps a rolling average of voting can be put in place and a decision made when the rolling average settles at a definite value for a period of some years. This will allow the Judiciary time to hear cases on a matter for several years before a public decision is made.

Having said that --- this is the way that new laws are enacted. If the courts keep having to pass law on an area, and the judgments are on the whole the same, then Government can be persuaded to pass a law.
-- madness, Aug 02 2017



random, halfbakery