[EM] Re: direct democracy / proxy system proposal

Bryan Ford baford at mit.edu
Tue Aug 10 08:36:41 PDT 2004

On Monday 09 August 2004 02:59, James Green-Armytage wrote:
> Candidate Proxy
> 	I like your idea that candidates can publish a list of whom they are
> likely to give their votes to if they don't get a majority, but I don't
> like the sort of deterministic IRV method that you propose. I prefer that
> the candidates have a space of time after the vote to organically try to
> build a majority coalition. The process of successively elimination
> candidates which you propose is, I think , unnecessary and somewhat
> limiting. Can preclude the possibility of forming a majority coalition
> around a centrist candidate with not so many first choice votes.

Yes, I can see your point, and having a separate "election completion" or 
"proxy runoff" election seems like a sensible approach.  Of course, if the 
candidates effectively get to change the way they vote with their proxy power 
(even in a constrained fashion) _after_ the popular vote has already 
happened, it will certainly draw the same objections about "taking power away 
from the people" that are commonly applied to the government-forming 
procedures of most PR-based European parliaments.  But that's not obviously a 
fatal flaw, since most of Europe lives with it OK.

> 	Aside from searching EM, you should search the web. I found some pages
> awhile ago about "Liquid Democracy" which were very close to my proposal.
> Speaking of which, my proposal is a combination of original ideas,
> repeated ideas, and ideas which were original to me at the time, but which
> had already existed elsewhere. I still don't know exactly where the lines
> are between those categories, but I know that there are some of each.

Thanks - I did some poking around starting from "Liquid Democracy", and even 
found two existing software packages implementing something like this!  
You've probably already found seen too, but in case not:

- "NetConferencePlus": http://www.usemod.com/cgi-bin/mb.pl?NetConferencePlus
- "VeniVidiVoti": http://www.usemod.com/cgi-bin/mb.pl?VeniVidiVoti

I have to look at them some more before I can form an opinion though.

> >But if depending on the
> >Internet could somehow be justified, then it could perhaps greatly reduce
> >the
> >cost and increase the richness of the resulting system.
> 	Yes. The internet would be a good medium if security could be ensured.
> Public computers could be made available to those without access, and
> those with access could use their own.
> >Based on this idea,
> >I'd like to throw out the following as an alternative, "multi-stage"
> >implementation strategy for a proxy direct democracy ("PDD"?) system:
> >- Stage 1: Experimental system, independently operated, open
> >participation by
> >anyone interested.
> 	Yes, I want to do this.
> >- Stage 2: More refined system, still independently operated, but with a
> >randomly sampled constituency for better representation.
> 	Would prefer if everyone who wants to participate, is able to, though you
> can do random invitations if you like.

What I was thinking is that the stage 2 system would be open to everyone for 
purposes of participating in discussions (e.g., reading and writing messages, 
making proposals), and anyone could even act as a proxy and thus indirectly 
wield the voting power of votes delegated to them - but only those selected 
by random invitation would actually have a "direct", baseline vote to start 
with.  That way, people (like us) who are interested and passionate enough to 
sign up on their own can serve the role of advisors in the system, while 
still preserving some level of democratic legitimacy by ensuring that these 
self-selected advisors can only wield however much voting power is delegated 
to them by the randomly-selected representative constituency.

> >- Stage 3: Realistic system, government-supported but still advisory
> >(i.e.,
> >decisions are non-binding), mass participation.
> 	Would like to stay in non-binding stage for awhile. During non-binding
> stage, security should be established to be near-absolute, and access
> should be shown to be fair. During this stage, participation should grow
> to include a substantial portion of the electorate, preferably more than
> 1/2.


> 	It's a damned good idea. We've talked a bit about setting up a proxy DD
> web site with a potential for large-scale expansion, but none of us have
> done it. I don't know if anyone else has. Even if people have made DD
> sites, it might be better to start over with a pairwise tally, with a
> capacity for all the proxy rules we like, and a capacity for expansion.
> 	So, the first thing to do is to get a web site, preferably with its own
> domain, for this purpose. The site allows people to create user accounts,
> which include a standing (ranked) proxy list. The site should be able to
> tolerate a large number of users, or at least be capable of expanding to
> tolerate more users.

For the last few days I've been looking around for some web-based software 
that might make a good starting point.

> 	There should be a discussion forum on the site.
> 	There should be a nominating and selection process for which issues
> should be voted on. To begin with, selection may not need to be limited,
> i.e. all issues could be accepted. Later on, we can introduce an STV-PR
> vote for selecting issues for a certain time period.
> 	Then, for each issue that has been selected, options for that issue
> should be nominated and selected. Issue selection process similar to
> option selection.
> 	There should be a period between issue and option selection, and the
> deadline for the actual votes... we'd try to get people interested in the
> issues during that period, recruit voters, etc. You could vote on issues
> before the deadline.
> 	Voting on the issues themselves should be ranked. (Rankings + ratings is
> possible as well, but probably unnecessary.) The tally system should be
> pairwise comparison. For each issue, it should be possible to cast a
> ranked vote or to defer to specific proxies. You should be able to search
> through the user database to find the proxies you want. If you leave stuff
> blank, the weight of your vote is carried by your standing proxy.

Your functional outline sounds quite reasonable to me, and it's substantially 
along the lines I've been thinking about for a while.  I'm worried, though, 
that we're likely to fall into a trap that I suspect the authors of both of 
the pieces of software mentioned above did: implementing a system that forces 
its users to follow more or less one particular fixed, hard-coded 
"deliberative process" all the time, which, however democratic, will also 
feel sufficiently like walking in a straitjacket that only the die-hard 
democracy geeks (i.e., us) will ever actually want to use it.  The more 
"invisible" we can make the mechanics of deliberation (while ensuring that it 
remains understandable, technically sound, and democratically legitimate), 
the more readily it will be accepted.  I guess I'm just trying to say that we 
need to figure out a way to make democratic deliberation "feel" to its 
participants like they're chatting with friends around the table and drinking 
beer, rather than in a big conference room following Robert's Rules of Order.

One of the things I've been doing the past few days is reading up on and 
exploring various Wiki systems out there, particularly Wikipedia.org and the 
MediaWiki software that runs it.  It seems to me that much of the reason for 
the incredible popularity of this kind of thing is that it's so free-form: 
you can just go to a page, click "Edit", and start writing.  No particular 
"process" is required: no setting up groups or getting permissions from the 
powers that be, no making proposals and waiting for answers before checking 
in anything, etc.  And yet the sheer size of Wikipedia.org demonstrates how 
amazingly well such a thing can work with this practically non-existent level 
of security or process.  This non-security obviously creates well-known 
problems - vandalism, trolling, edit wars - which currently can only be dealt 
with by giving some users special "sysop" privileges and allowing them to 
lock down particular pages and such.  But it seems like that's precisely 
where the ideas flying around on this list could help.  Can we come up with a 
way to incorporate legitimate, secure democratic "process" into a system like 
this without making it _feel_ to users like they're being subjected to a 
deliberative process?

Here's a (still admittedly very rough) idea to play with as a starting point.  
We start with MediaWiki or some similar freeform web-based collaboration 
software, and introduce our proxy voting technology basically as a mechanism 
that only affects normal wiki-workflow when there's controversy of some kind: 
e.g., when an edit war happens.  On the zillions of pages that only a few 
people really care about and for which they find usually consensus easily, it 
works just like conventional wiki.  On the relatively far fewer pages over 
which actual controversy develops, the proxy voting system kicks in to allow 
people to judge which of several possible alternatives should be the 
"official" version of that page, while retaining all of the alternatives for 
reference and statistical purposes.  "Proposing" an issue might amount to 
creating a new page and writing both a question and a proposed an answer; 
other people can then edit it to form their own alternative pages with 
different answers (or even different questions, if they think the original 
question was ill-conceived).  When you go to that page, you can either just 
look at the "winning" majority version, or you can check out and compare 
alternative versions and vote on them yourself, or edit your proxy list so 
others can vote for you.  And if consensus is achieved on some issue, the 
corresponding page automatically goes back to being an ordinary wiki-page 
that anyone can freely work on as long as the consensus continues to hold.

In the long term, or in a true "governmental" context, I don't expect such a 
free-form mechanism could ever fully replace a more formal deliberative 
process, e.g., in which issues are nominated, chosen, debated, revised, and 
voted upon using a particular well-defined procedure.  But until we have more 
experience with how our ideas can be applied and what works and what doesn't, 
I think that trying to keep the design as free-form as possible might greatly 
improve its acceptance rate among "normal" people, and it might in turn give 
us better information to go on when designing and refining more formalized 
processes in the future.

> 	When announcing the result, we would want to show statistics on how many
> people voted directly rather than via proxy.

Right - and I also think it should be possible to look up how a particular 
proxy voted on any particular issue.  For sake of transparency, choosing to 
become a proxy should bring with it the cost of giving up one's right to vote 
anonymously; otherwise proxies would not be accountable to their 

> 	I assume that there would be plenty of logistical challenges. For
> example, how do you prevent people from signing up for multiple
> accounts?... and so on.

Yes, that's certainly one challenge.  In the Stage 1 system, I think fairly 
"soft" security is adequate: try to make sure it's possible to detect the 
most egregious abuses (e.g., when someone at a single IP address controls 
many accounts that never actually participate in anything but just delegate 
their votes on everything to one particular "active" account), but not worry 
too much beyond that.

In the Stage 2 system, it actually shouldn't be much of a problem, because 
only those randomly invited get a vote, and each randomly invited person only 
gets one account.  Anyone can have as many "non-voting" virtual accounts as 
they like.

For Stages 3 and 4, we can presume existing government-supported 
registration/identification methods could be reused.

> 	Anyway, this is something I'm interested in doing. I'm not a computer
> science expert, so I can't write the actual code. But I am interested in
> working on the project. Actually, I had thought that this could eventually
> develop into an actually nonprofit organization, eligible for 501c3
> status, foundation grants, etc. I think that it could become a very
> effective little NGO, issuing press releases and possibly getting free
> media for the results of its votes on interesting issues... meanwhile
> promoting Condorcet's method and pushing towards an official proxy DD
> system. The organization should be nonpartisan, and should provide
> discussion and education about alternative voting methods as a part of its
> mission. But first, a critical mass needs to be achieved.

Sounds good!  I _am_ a computer science geek (grad student), and I'd like to 
get into some serious code-hacking soon, but can't promise any great rate of 
progress in the short term since I have many distractions at the moment and 
will continue to at least through the rest of this year.


More information about the Election-Methods mailing list