[EM] democratic community, the web, implicit/explicit instant proxy

[Abd ul-Rahman Lomax]

At 10:50 AM 9/1/2006, raphfrk at netscape.net wrote:
>The trick for getting something like that to work is to keep the
>standard as general as possible while at the same time allowing people to
>do anything that they might like.

Sure. There are two problems here:

(1) How to set up DP and other filtering systems that are not 
top-down, that have thoroughly decentralized control.

(2) How to set up a web site that filters information in such a way 
as to automatically select material likely to be of interest to the 
reader, that is not excessively redundant, that is prefiltered.

The second problem is of high interest, to be sure. It's an 
opportunity awaiting the entrepreneur. And, indeed, of course, there 
are also many efforts along these lines. However, problem 2 is really 
a different problem than problem 1. The solution to problem 1 may 
*use* solutions to problem 2, but it must not be dependent on any one of them.

We do have a standard in mind, a very simple one. It's been described 
whereever we have participated in an effort to set up DP. It is the 
proxy list. It is simply this:

It is a tab-delimited (or so convertible) database, with one record 
per member. The essential fields are:

Member Handle, Designated Proxy, Proxy Acceptance, Comments.

Additional fields may be added by applications. The "Member Handle," 
in a completely open system, could be the member email address, 
likewise with the proxy, but generally we assume that privacy will be 
protected, so a system address, a user ID for the site involved, may be used.

Wikis can handle this easily, leaving a record of who edited the file 
for each edit, so we can confirm, if needed, that it was the member 
who added his or her record and designated the proxy, and it was the 
proxy who accepted.

We assume and highly recommend, however, that proxies have the email 
address of the designating client, and we likewise assume that an 
accepting proxy provide direct contact information to the accepted client.

An automated system could make all this convenient, but it can be 
done manually, it is not *necessary* to depend on systems. Which 
means that associations will not be dependent upon programmers.... 
but may use programs as appropriate.

How this list is *used* is a separate question. Current thinking is 
that the association, if large enough, will provide certain tools, 
but the data used by these tools will be generally available, and so 
to most forms of analysis can be done entirely independently of 
central control.

To use a proxy list with a vote list, which is not the only 
application, a list of votes would again be in or accessible in 
tab-delimited form. The essential fields are: Member Handle, Vote, Comment.

So an obvious tool would be a program that would take a vote list and 
bring in information from a proxy list. Where a member has not voted, 
but has named a proxy, such member votes would be expanded by 
assuming that the proxy's vote stands as the member's vote.

Many voting systems currently show votes-in-progress. It's an option 
with yahoogroups. So any member who cares can see how their proxy 
voted (or can see if their proxy has voted at all, and if not, if 
there is a higher-level proxy standing in), and, if not satisfied, 
can simply vote directly, which the member can do in any case. The 
vote mechanism is a standard one, recording the name of the voter and 
the vote. *The system does not add proxy votes, it only records raw votes.*

We assume that many votes would be reported both in raw form and 
expanded. If there is a substantial discrepancy between the raw vote 
and the expanded vote, one might want to do further analysis. For 
example, one might determine from the proxy list who are 
highly-trusted proxies, and see if the voting pattern is different 
there. From other lists, perhaps open caucus lists, one might 
determine if there are factional patterns. One may also be able to 
detect fraud. Generally, in Free Associations (FAs), it is not 
necessary to do heavy fraud detection, precisely because fraud will 
become visible in vote patterns, and because FAs are not majoritarian 
systems, ultimately, because of their ready fissioning and 
reconstitution. Basically, those who suspect that there is a large 
fraud block can simply disregard what they consider fraudulent. The 
FA does not force anyone to do *anything*, it merely creates 
opportunities for communication, coordination, and cooperation.

Worst case, the fraud creates an apparent majority and admin decides 
to accept the fraudulent vote and disable, say, the DP system. But 
the entire system exists in records held by the members. Every proxy 
has, we foresee, the email addresses of their clients. They don't 
need the central system to maintain communication, the central system 
is merely a convenience, one which can be reconstituted in a very 
short time. Proxies generally will have not only their own client's 
direct email, for example, but also that of, naturally, their own 
proxy, and many will also have the email addresses of all or most of 
their fellow direct clients of their proxy, and they may have even 
more than that. Thus even missing links in the system can be bypassed.

This is a cellular system and could have applications under dangerous 
conditions; however, secrecy creates the potential for central 
control; we do not consider it necessary to solve the difficult 
problems at this time. We are working on open systems, which we 
expect would be very difficult to corrupt.

>Anyway, for something like a proxy system, you could try to
>come up with an API.  On the one hand, it could be argued to be
>centralised.  However, even in open source software, they
>still use standards.  A good standard is one that streamlines
>ensuring compatibility while at the same time doesn't limit
>the various methods of implementation.

The essential tools are not at all the problem. Those tools could be 
written by any programmer, for some in minutes. Take a proxy list and 
use it to analyze a vote list and provide an expanded vote. Or, for 
example, monitor a proxy list and periodically issue a notice to 
members of closed loops below a defined size in comparison to the organization.

The system does not prohibit loops, but loop members should know if 
there is any reasonably possibility that they will be unrepresented 
in high-level discussions and polls....

>If you did a really good job, the API could cover a wide range of
>methods.  One person might implement it with an online forum/mailing
>list and other might implement it with physical meetings and both
>systems could interact due to the API.

There are people working on the tools. See the googlegroup top-politics.

>One option here would be to have a button where a person can click
>"agree" or "disagree".  There would then be a total near the post.

I think these tools exist and are in wide use, actually. I've 
certainly seen them.

>I wonder if something like a super-wiki would work.  This could
>have multiple versions of the same article being editted and also
>some way of merging two slightly different versions of an
>article together.

Again, all these tools can and will be useful. But the basic system 
should depend on none of them.

I have seen more than once that DP has been considered a software 
problem, and when the software was lacking or buggy, DP was 
abandoned. Completely unnecessary!

It's like voting machines. The amount of labor involved in manually 
counting votes is actually small compared to the labor involved in 
*voting*. To save a small amount of labor, we have, in many 
jurisdictions, implemented automated systems which vastly increase 
the potential for fraud. Penny-wise and pound foolish, or simply 
quite clever? Might be both. The penny-pinchers cut back on expenses, 
and then the sharks move in.

> > Part of the problem was that they understood "proxy voting" to be
> > absentee voting. I quite understand this objection. Someone who is
> > not present cannot generally have followed the discussion and respond
> > to the various amendments or arguments that may have arisen.
>
>In fairness, with proxy voting there could be a (large?) group of
>people who will insist that there proxies do just that.

Whether or not this is even feasible depends on voting procedures. We 
are recommending, in fact, standard voting procedures where those 
participating vote. The voting procedure is blind to proxy voting. 
Proxy votes are *added* based on analysis. In such a system, there is 
no method in place for a proxy to cast a vote for a member which is 
different from his or her own.

Now, if a proxy is willing to vote against his or her own best 
understanding, simply to please a client, the proxy is free to do so. 
I seriously doubt, though, that such a proxy would be widely trusted! 
(Not in a DP system with an established culture expecting good 
service -- i.e., direct communication -- from proxies. In a populist 
system, early on, it might happen, and it could be quite dangerous.)

No, if the member has an opinion different from the proxy, the member 
may vote directly in the systems we contemplate. Our systems leave 
people the freedom to be foolish, and they have no opinion, 
specifically, on whether or not a member is foolish in directly 
voting. It is simply a right of the member to have his or her own 
opinion and to express it.

But we also expect that proxies will communicate regularly with 
members in a fashion similar to what has been envisioned below.... 
Proxies will be in the best possible position to move member opinion, 
and this is the reason why we expect that allowing proxy votes to 
stand in for member votes will generally converge on a reasonable 
expression of the opinion of the electorate. *After the electorate 
has had time to consider an issue, which might even be after a vote.*

That is, proxies will be, in a sense, leaders. But not governors.

>   However,
>if they were in the same room as the debate, they would be more likely
>to be reasonable.

Yes. A meeting may, perhaps, not allow those not present to vote. 
Certainly this is presently Standard Operating Procedure for 
face-to-face meetings. As a chair under Robert's Rules, I would not 
allow the entry of an absentee vote unless the established rules 
specifically permitted it. Depending on conditions, I might be more 
likely to allow a proxy vote, if the proxy was understood as having 
the freedom to alter the vote according to the proxy's best understanding.

Proxy voting is *not* absentee voting. Absentee voting does not 
require any proxy at all. A proxy who is simply casting an absentee 
vote can be replaced by a piece of paper and a stamped envelope.....

>I wonder if "asynchronous voting" could help with that.
>
>Something like:
>
>Proxies attend town meeting and discuss issue.
>
>Proxies contact the people they proxy for over the following
>week or so and explain what was discussed at meeting and how
>they will be voting

Yes. Sort of like this. Or they might have discussed an issue already 
with their clients, and then attended a public hearing, they take 
back what consensus, if any, was found, they discuss it, they take 
back further questions that they could not answer themselves, etc.

*Communication.*

Thinking of proxies as voting agents is not more than a small part of 
the concept. In FA/DP applications respecting Town Meeting 
governments, there really isn't any proxy voting. There is just the 
communication and generation of consensus. When there is sufficient 
consensus, there is practically no need for voting; that is, the 
existing voting methods will simply ratify it. (yes, a faction could 
attempt to pack the Town Meeting. Which, I can tell you, would go 
over like a lead balloon; even if successful, that vote would not 
last long! Town Meeting towns already operate mostly on consensus. 
Many Town officers run unopposed, simply because they have such broad 
acceptance that a contest isn't really needed. What I'm trying to 
introduce is something that will improve communication, so that we 
don't see the disconnect represented by Town Meeting voting, for 
example, to approve and borrow the funds for a Public Safety complex, 
which is then voted down through the mandatory submission to the 
voters at the next secret-ballot election. I know *exactly* why this 
happened. It isn't intimidation at Town Meeting. It is simply that 
the voters weren't sufficiently informed, nor was the Town Meeting 
proposal broadly enough discussed, there remained unresolved issues.

At lot of planning effort was wasted because of poor communication. 
Bringing everyone in through DP, *before* holding votes, will, I 
think, work much better. That DP could revolutionize politics is a 
side-effect, here.... but, of course, I'm quite aware of it!

>Anyone can withdraw/transfer their proxy at this stage

At any stage. But it is simpler, if we are only talking about one 
issue, to vote directly.

>2nd meeting held for the actual vote

This is common process, actually, many issues require two meetings to 
be resolved, if I'm correct.

>This means that nobody can vote unless they have at least
>received a report from their proxy about what happened
>at the meeting.

No, people can vote whenever they want.... but, with such a system, 
they will have the *opportunity* to get a personal report. My comment 
on the Public Safety complex vote was that nobody from the town 
called me up to discuss it with me. When I saw it on the ballot, I 
thought "I haven't the vaguest idea whether this is a great proposal 
the expenditure necessary, or it is a boondoggle to provide a palace 
for the fire and police." And I did ask one person active in town 
government, and she happened to be one who thought the proposal 
premature. So I abstained from voting on it. Many voters, however, 
will instinctively vote against a tax override, which is what this 
was, unless they are convinced of the necessity. It is simply not 
amazing that it failed. And it shows how there can be a gap between 
Town Meeting and the voters. Town Meeting is the collection of those 
willing and able to be *active* in Town government. It is a skewed 
sample, skewed by the selection criteria, and further skewed by being 
those most knowledgeable about the actual proposal.

DP attempts to deskew the representation, at the same time as it 
broadens the discussion and amplifies the trust of citizens in 
general in the town government process.

It is *very* important for the introduction of an FA/DP organization 
like this to not be seen as an attack on the existing government. 
Changes often are made at times of polarization. This is one best 
made when things are *not* so polarized.

>This would allow compromising of one issue against another,
>the proxy could say "well I know this isn't what we wanted,
>but if we agree, then we get this other thing that we wanted."

Under present conditions, proxies don't have binding votes to cast, 
except their own. A proxy, under present conditions, becomes only a 
communications link; but this is actually the most important function 
and voting is not crucial. Yes, when there is conflict and major 
division, voting becomes important, but I do expect that such 
conflict will not be as common where we have FA/DP organizations in place.

Once there *are* such organizations, they will be much better able, 
not only to judge if legal changes should be made (such as changing 
state law to allow proxy voting at Town Meeting), but also to collect 
the political power to actually get the law changed.

>It could very well come down to organisational "culture".  I am
>not so sure that there won't be alot of people who won't pick
>proxies who are unmovable by debate.  Picking a proxy you trust
>is hard.  Picking one who will vote a well defined way on every
>issue is alot easier.

And a lot more frustrating. Politicians are famous for making 
promises that they don't keep. And, in fact, if politicians always 
kept their promises (the kind of promises that they currently make), 
the whole system would break down.

We are not trying to change human nature. We are simply trying to 
*use* human nature to create more effective social intelligence. 
Proxies who are ideologues are, generally, *not* trustworthy, and, 
further, they will not be good communicators. I don't think that 
proxy systems will amplify their power, quite the reverse. DP systems 
will dilute their power. They will still be there and they will still 
have followers. But they won't be gatekeepers, people will have other 
choices....

In the FA/DP environment, the ideologues are welcome to their own 
opinions. The FA is not going to contradict them. But caucuses can 
and will form to counter special interests (and ideology is a kind of 
special interest). FA/DP organizations don't link issues. There are 
people who are deeply opposed to abortion. And to war. One might 
think these positions quite compatible! However, present public 
politics pretty much requires a Hobson's choice of these people. They 
may make a rough calculation of what they see as lives lost. Hmmm. a 
hundred thousands lives in this war vs millions of lives by abortion. 
I guess I have to support that pro-war candidate because he opposes abortion.

The public process won't immediately change if we manage to create 
FA/DP political organizations. But it will become much more possible 
for *true* pro-life candidates -- that is anti-abortion *and* 
anti-war -- to gain traction, to find a support community.

(I've just chosen these issues as familiar ones, not to express any 
personal opinion. Sure, I've got those opinions, and strong ones! But 
that's not what is going on here.)

>The issue is that for any kind of negotiation, there must be a
>penalty for not coming to an agreement.  Time is often used as
>the penalty.

Consider this. There is an FA/DP political organization, perhaps it 
is Metaparty (http://metaparty.beyondpolitics.org). There is a group 
which thinks that we should immediately withdraw from Iraq, and 
therefore we should create a campaign fund for candidates who agree 
to support this. Another group thinks that this is surrender to 
terrorism, and that we should support candidates who disagree with 
withdrawal, who will push through toward victory. Let's say that the 
numbers are 55% withdrewal and 45% stay-the-course. Okay, if they 
form the actual PAC that collect money -- the FA is not going to do 
this, and every member of each faction contributes to their 
respective cause, and, let's say, there are 10 million so-active 
members and each one contributes an average $20, we are going to 
spend $110 million politicking in one direction and $90 million 
politicking in the other.

What a huge waste! If we can find some position that would enjoy a 
broader consensus, we can work for it, and not only will our efforts 
not mostly cancel each other out, but we will, if we have broad 
representation within the organization, i.e., our organization 
represents to some degree society as a whole, also not have such a 
difficult time in the political market. We will already be proposing 
something broadly acceptable, and it may not be necessary to spend so 
much money.

($20 each? How much would *you* spend in a reasonable attempt to 
resolve this issue? It is costing each of us, on the average, 
thousands of dollars in taxes, just from a financial point of view 
alone, not to mention the lost lives. If we continue uselessly, what 
a waste! and if we withdraw prematurely, what a waste! How much 
effort is it worth, to find agreement? The existing political system 
thrives on disagreement, it fires up voters. Where is the "Let's Find 
Agreement" Lobby?)

FA/DP, because it leaves power (i.e., money and votes) decentralized, 
strongly favors, I'd suggest, consensus solutions. And the DP 
structure makes it much more practical to find consensus, by breaking 
down the discussion into relatively small groups.

Why won't it work? There are lots of reasons. I'd suggest they are 
worth looking at. I think there is a way.

It's happening, the discussion has been broadening. Over the last 
year, I'd say, far more people are aware of the DP option, and some 
of them are recognizing that there just might be something to this FA 
stuff. (People with extensive experience in existing FAs may already 
know that, and, for years, I've known people who said, "If only the 
rest of the world could function like this...." But what was not 
known was how to scale it. DP is an attempt at solving the problem of 
scale, while still preserving the rigorous anarchist/libertarian 
concepts behind FAs.)

>If a person/group is willing to sit at a meeting for 2-3 hours
>and not budge an inch, then it can be implied that they really
>care about the issue (or as you say have alot of time on their hands).
>This also applies in buisness negotiations,  if a person spends all
>day and isn't willing to lower their price, then maybe it really
>would bankrupt them.

Yes. However, as the meeting gets smaller, in general, the process 
also gets easier. The time expands rapidly as meeting size grows. 
Smaller equals better, as long as the meeting does not get so small 
that it no longer broadly represents the diversity of position and 
opinion in the association. That's the trick, and DP is designed just 
to jump through that hoop.

>The Senate Filibustering system is the same kind of thing.  If you
>are willing to meet the filibuster consditions, then you likely
>care more about the issue.

Yes. And then there are those who simply care about winning, and who 
threaten to simply disregard the rules by interpreting that they 
don't apply, even though they patently do. This is the nuclear 
option, and we already experienced it, essentially, in the 2000 
election, when the Supreme Court utterly disregarded its own 
precedent and the Constitutional process for resolving disputed 
elections, to issue a ruling designed to produce a desired outcome. 
There is no other explanation, and apologists for the court have 
essentially argued that the court is a political institution, get over it.

(The chair can rule pretty much anything if a majority will sustain 
it. Of course, some people have ethics and would not rule contrary to 
precendent and clearly stated standing rules, but others.... Of 
course, such an august official as the Vice-President of the United 
States would never do that, would he?)

(And I thought that the Supreme Court was likewise far more worthy of 
respect. Boy, was I mistaken! Tbhis is not about having voted 
Democratic. I've experienced a lot of lost elections, and I *never* 
before felt that it was an injustice. It was simply not having 
convinced the American people.... but, ahem, I digress!)

>Hmm, maybe proxy voting could be assumed to be a method to remove
>the cost to the individual of being a hold out?  (and that would
>be a bad thing)

The cost is still there for the proxy. Proxy representation 
establishes a far more effective freedom. Yes, freedom can be abused, 
but that proxy still has to slog through.

The *individual* may not have an incentive not to hold out, but the 
individual is not sitting in that meeting, the proxy is, and, I don't 
know how many times I have to repeat it, the proxy is not a 
rubber-stamp, unless he is willing to so serve....

I think the practices will work themselves out. Having simple votes 
that are *not* proxy votes and then expanding them in analysis means 
that the proxies are *not* casting their client's votes. They are 
casting their own votes. The analysts then make of it what they will. 
In an FA, the goal of voting is the measurement of consensus. Votes 
in FAs, except for pure process issued, don't move money and power. 
It is a *different* environment, so a lot of conventional political 
thinking and habit will not apply.

> > Town Meeting allows all citizens of the town to vote directly on Town
> > issues by attending Town Meeting. However, there is a huge group of
> > Town citizens who are effectively locked out by this. For starters,
> > single mothers.... Town meeting is always held in the evening, and it
> > can go on late on occasion. There is no absentee voting at Town
> > Meeting. And, of course, proxy voting is not allowed.
>
>Systematic biases are obviously a bad thing.  What about something like
>this:
>
>The total amount of time for a town meeting is divided equally between
>all potential voters.
>
>When you set someone as a proxy, you are granting them the right
>to use your time allocation.

That might work. If you are there, you have your own time, which you 
can voluntarily yield to anyone. That is, in fact, how the Senate 
often functions.....

>Time not used in one meeting can be carried forward to the next (with
>probably a limit to how much can be stored).  A proxy uses up the same
>amount of time for all people he is proxy for, so they all bear the
>burden equally.

It can be done. If it is worth the effort, it will be done.... When 
the meetings are relatively small, it isn't necessary. Meetings can 
vote to close debate, and they can vote to censure members for 
whatever reason they choose, if someone takes up too much time and 
refuses to yield. Relatively small meetings can work all this out, 
they generally will go overboard to ensure that everyone who wants to 
speak gets to. It is when they get too large that the problems become 
intractable. Proxy representation will allow meetings to get smaller, 
I think. The large meeting will be replaced by a smaller central 
meeting and *lots* of informal meetings taking place throughout the 
town, phone calls, etc.

>No vote on an issue is final until nobody wants to spend any
>more of their time trying to convince others to change their mind.

It's called standard Robert's Rules. Except that debate can be 
closed, normally, with a two-thirds vote. Unless the chair decides 
that the decision of the Majority Whip is final, and a majority 
dutifully agrees that, yes, this is what the rules really mean! Aagghh!

>No meeting may run for longer than the planned duration.

Give that one up! Meeting majorities have the absolute right to 
extend or shorten meetings.... "Motion to Adjourn!" Undebatable, 
simple majority carries. "Motion to set aside the Orders of the Day" 
Also undebatable, simple majority carries. (The Orders of the Day 
would include a preset time for adjournment.)

>A
>meeting may end early if nobody else wishes to speak.  Votes
>which haven't been finalised by then, are tabled until the next
>meeting.

Tabling unresolved issues is quite common. Sometimes it is to allow 
time, sometimes it is an attempt to kill a proposal without debate.

>This means that everyone has equal time to speak.  If you really
>don't want an issue passed, you can have your proxy use up all
>of your time holding out and preventing the issue from being
>finalised.  However, this means that you lose power for other
>issues as you have spent all your time on this one issue.

The idea needs work. But, yes, sometimes time-allocation systems can 
work. It is very difficult to get people to keep to them, though.

>Also, there would probably be a rule that some time (say 25-33%) is
>"free" and speakers are picked at random.  This would given non-proxies
>(who probably have only 30-40 seconds "in the bank") a chance to speak.

The fact is that if you can't convince any of the proxies who are 
present to speak for you, you have little chance of accomplishing 
much at that meeting. But, yes, it could be done.

>In practice, there might even be a rule where a proxy can say
>"... and I speak for 15 more minutes".  He wouldn't have to actually
>speak, it would just remove 15 minutes from the maximum time the
>meeting can run.

That is a truly interesting idea. But I don't think that meeting 
maximums would fly. Indeed, most of my work is toward standing 
meetings, as by mailing list or other forum.