[EM] Trees by Proxy

[Abd ul-Rahman Lomax]

At 01:02 AM 3/30/2007, Dave Ketchum wrote:
>On Thu, 29 Mar 2007 02:52:20 -0400 Abd ul-Rahman Lomax wrote:
>>That's correct. What I'm suggesting is that voting rights 
>>immediately respond, but that deliberation rights float to some 
>>degree. You've travelled to the capital, you rented an apartment, 
>>and then somebody changes their proxy and you lose your seat?
>
>
>The proposal here is that proxies become effective (both as to floor 
>rights and voting power) some time after filing - I suggest 10 
>days.  Thus you would know your future status for the next 10 days.

I'm going to imagine that we actually make decisions here, that we 
will vote on things. It can be a useful fiction. Or we can actually 
set up polls....

Okay, so Mr. Ketchum's motion is on the floor. It has not been 
seconded. Ignoring standard deliberative rules, I go ahead and 
propose an amendment, which is that proxy rights become effective 
immediately, and that floor rights depend on the rules established by 
the assembly. Might be 10 days, might be some other number.

Because this is an online list, it is possible to have two motions on 
the floor at the same time, and to vote for both simultaneously. I'm 
not entirely comfortable with that, preferring the 
Condorcet-compatible standard deliberative process (which functions 
as a round-robin between alternate motions, through the amendment process).

Mr. Ketchum seems to think that he can define "the proposal." Sure, 
he can define *his* proposal. We know that is what he has proposed, 
it has been clear from the moment he proposed it. He tends to object 
to suggestions for change as "not *the* proposal, which he has just 
done again here. Instead of debating the suggested changes. Once 
again, he has not given a reason for delaying voting right changes, 
beyond an assumption that voting rights and participation rights are 
rigidly and immediately connected. Yet he elsewhere allows the 
assumption of voting rights upon 10 day delay at reaching a specific 
vote level, with deliberation rights being assumed at double that vote level.

In other words, he does separate voting rights from deliberation 
rights, using a fixed but different level of proxy counts for each. 
Having done that, the question is then why voting rights have any 
delay whatever. And he hasn't given one single reason for this, other 
than it is not "the proposal."

Given that neither Ketchum's proposal and my own version have been 
seconded, the whole thing is moot. Until someone chimes in and says, 
I think "this" -- whichever one is preferred -- is a good idea, the 
best raised so far.

Still, because I have another purpose here, which is to examine the 
issues around proxy democracy, something which has seen far too 
little discussion, I'm going ahead and commenting, understanding that 
all of this is dicta.

>Now, if you were on the edge of losing your seat, renting an 
>apartment is a bit dumb.  More profitable activities right now would be:
>      Campaign to round up some more proxies.
>      Concede that you do not have that much support from the voters 
> and give up.
>      Do a sideways proxy to give a legislator who shares most of 
> your goals your votes.

But the conditions of the problem are that you *have* enough votes. 
You just obtained them. And it is becoming clear that "sideways 
proxy" means, simply, that you delegate your proxy. Since you aren't 
exercising the votes yourselves, I fail to see why this is different. 
Nor why it is called "sideways." It is upward, that is, it is a 
transfer of votes from one with fewer votes to one with more votes.

Now, in Ketchum's system, proxy's don't take effect for ten days. So 
you can't delegate your votes "sideways" for ten days. Look, it's 
Rube Goldberg. There is utterly no reason to delay voting rights, 
beyond an *assumption* that they go with participation rights, and 
there are reasons to delay them. Sometimes. The necessary delay 
depends on so many variables that it is wisest to leave it to the 
assembly rules, and the assembly, by common law, already has the 
right to regulate its own debate.

Standard proxy, but delegable. Standard assembly rights. That's it 
folks. It does not need to be complicated by a priori rules that do 
nothing but gum up the works.

(It is arguable that standard proxy *is* delegable, unless the proxy 
assignment forbids it or limits it to direct exercise. If I'm on 
vacation, and I give a power of attorney to someone to sell my house 
-- usually it would be my attorney, but it can be anyone -- and then, 
on the day of closing, that attorney is unavailable, he or she can 
designate someone else to sign the documents, as I understand it. 
What we are doing is make delegability a clear assumption. Yes, the 
voters could provide in their assignment that proxies cannot be 
passed on. A foolish idea, but people, when it comes to voluntary, 
free-choice representation, which is the goal, should be free to be 
foolish. And the default systems set up organizationally might not 
take this restriction into account, in which case it would be up to 
the proxy to satisfy the restriction.

There are many problematic details of delegable proxy, and, though it 
seems to me that they are all soluble, I prefer to start with the 
Free Association context because proxies can do little or no harm in 
such a context. They serve to discover and measure consensus, not to 
actually implement it. Implementation, unless specific powers have 
been granted to the proxy -- typically independently of the FA -- is 
not decided or determined by the proxies, it is the prerogative of 
the members directly.

(As an example of a specific power, a member could set up a special 
bank account and give the proxy the right to transfer funds from it. 
FAs might be interested in such accounts -- they add weight to votes, 
in the free interpretation of proxy expansions that I propose -- but 
they aren't a part of the formal FA process. The FA is not going to 
collect funds and spend them, beyond nominal sums for its own 
operating costs, which should be very, very low. So low that they can 
typically be satisfied by direct payment by members, as donations. 
For example, I pay the $8.95 per year domain registration for 
beyondpolitics.org. Big whoop! I also provide the storage space and 
bandwidth, and the extra bandwidth that beyondpolitics.org consumes 
might actually be costing me $5 or $10 per month, I haven't looked at 
it. Donations accepted. But you'll notice there is no fund-raising campaign.)

>>>If a change in proxies means different delays between the old 
>>>proxy ending and the new one taking effect, the legislature will have either:
>>>     A period with no support for those voters, or
>>>     A period when those voters will have double representation.
>
>Abd disagrees, but not convincingly.

Ketchum reports the obvious. Obviously, I have not convinced him. So 
we should care?

He has not argued the point. He merely proclaims, again and again, 
that he is not convinced. Yet he has not even attempted to establish 
his proposal as having any necessity at all. What is the reason for 
delaying voting rights, given that we have already separated them 
from participation rights, i.e., floor access rights?

>>DP and other proxy assemblies can be smaller, ordinarily, than 
>>standard peer assemblies, for a given level of completeness of 
>>representation. Having ten percent more seats would mean, probably, 
>>less than ten percent more communication traffic. Not a drastic 
>>change, particularly if temporary.
>
>
>Interesting thought, and size is a topic for careful thought.

For me, this thinking entirely preceded the question of how to solve 
the problem. And a solution, it turns out, is *extremely* simple.

>   I suggest two limitations:
>      Number of seats in the legislature, filled by the candidates 
> with the most proxies.

Sure. Asset Voting does it. Then take Asset and allow the voting 
rights of those elected to float. Which makes, among other things, 
representation more accurate, it removes even the small digitization 
noise from making vote transfers in precinct blocks.

(This is proposed because it creates district-based representation, 
*mostly*. Normally, a citizen, even though the citizen has voted 
secretly, will know exactly who was elected with his or her vote, and 
this member will by local if there are sufficient local votes 
assembled to justify it. I assume that when Asset holders are 
distributing them, they will do it in precinct blocks, especially if 
the rules allow a little slop.)

However, this initial proposal, while a vast improvement over 
standard practice, still involves terms of office. Not a great deal 
of harm would be done by this, because if votes can be reassigned by 
the electors (the initial holders of assets), then in an assembly of 
reasonable size, *views* with snowball's chance of being accepted 
will rather easily find someone to incorporate them in motions and to 
discuss them. However, once we have floating voting rights, it 
becomes possible to consider floating deliberation rights. This 
essentially abolishes the fixed terms.

However, there are some obvious problems with that, in a very large 
assembly where the business of being a member is a full-time job. 
There are two responses to this: disallow the reassignment of 
deliberation rights -- the proposal above -- or delay or otherwise 
restrict such reassignment. Ketchum does both. But he *also* delays 
the movement of voting rights, which are, even in his proposal, 
distinct from participation rights. That is, you can have voting 
rights, but not participation rights.

Given that the two are separated, it then becomes logically necessary 
to consider them separately. We know there are reasons to delay 
participation rights. But what is the reason for delaying the 
movement of voting rights? At all, beyond the necessities of 
registration, i.e., notice.

And what is the reason for not allowing *anyone* holding votes, that 
is, any elector, from voting directly? The technical problem is trivial.

Further, if direct voting is allowed (remember, this is not direct 
voting by citizens in general, but by chosen electors who received 
votes from citizens in a secret ballot election), there is much less 
need to change proxies, so the composition of the assembly, in 
theory, becomes more stable. (If you disagree strongly with your 
proxy on one issue but you generally trust the proxy, you can reserve 
your direct voting for the rare occasions that this is the issue 
coming up. Otherwise you may feel compelled to change your proxy to 
someone else, simply because on one important issue you can't support 
the position of the proxy. This, by the way, solves the dilemma of 
the person who is socially liberal but is religiously convinced that 
abortion is murder. I know a fair number of such people....)

>     Minimum proxies to occupy s seat.  I suggest 1% to vote; 2% to have
>floor rights and thus full membership

This limits the assembly to fifty people. Not bad, but I'd prefer to 
*allow the assembly* to determine its own operating size. If direct 
voting is allowed, votes on assembly rules would require majority 
support, and would not affect votes on subsequent issues. The 
incentive, then, is for the assembly to make its effective size, for 
deliberation, optimum for efficiency: efficiency requires that all 
significant factions on an issue be represented in debate, and that 
the size be small enough to be operationally manageable. I find no 
way to predict this, and optimum size would vary greatly not only 
with the context but also from time to time.

Note that if, for example, an assembly majority decided to restrict 
the assembly size such that it was disproportionally represented, the 
minority could form its own unofficial assembly and use it to 
deliberate publicly. I'm sure such would be newsworthy, and it would 
be insane for members of the official assembly to not follow the 
debate in the minority assembly. Their constituents, generally, will 
expect them to be informed....

I don't see proxy democracy as nearly as prone to the divisive and 
polarized party politics of the present, because much of the driving 
force for this is removed. There is no *contest* involved in proxy 
democacy. There is only a process of determining, through free 
choice, a representative assembly of manageable size, and we have 
shown how this is possible while still preserving privacy at the primary level.

>>The confusion arises because we think it best to assign seats based 
>>on votes. But that is just an *indication* of whether or not 
>>someone should have a seat. I've thought that legislatures might 
>>give some people seats who don't have any votes other than their own.
>
>
>Interesting thought.  I do not propose such, but do not object to 
>legislatures managing such affairs themselves.

Since Ketchum now acknowledges the reasonableness of legislatures 
setting their own rules, why then prejudge what size it will choose? 
It is reasonable, in discussing how a proxy legislature would work, 
to suggest that it might have such and such a size, but most people, 
in seeing that comment without it being qualified, will assume that 
the size is fixed. I'm just suggesting that, if a formal proposal is 
being made, references to sizes be explicitly examples of what an 
assembly might choose.

>>>Direct voting would be a complication that would make the basic 
>>>proposal harder to evaluate.  Such comparatively minor changes 
>>>could be considered by themselves later.
>Abd suggests that direct voting is more important than electing via proxy.
>
>I disagree.

Instead of extracting my proposal by quotation, something that he 
seems to avoid, quoting only himself, he paraphrases me and then 
disagrees with the paraphrase. Nice. So do I. I do not suggest that 
direct voting is more important than electing via proxy. I suggest 
that direct voting, otherwise impossible, becomes possible if there 
is a proxy assembly where direct voting is allowed, but participation 
rights beyond that are restricted according to assembly rules.

And if there is direct voting, assembly rules will respect the rights 
of all parties -- or there are easy ways around an attempt to 
monopolize debate.

What I'm suggesting is that the foundations of democracy are about 
the right of direct participation, and that scale and efficiency 
demand that representation be the norm. Representation by proxy is an 
extension of the right of direct participation.

I find it quite interesting that many organizations do allow direct 
participation, but prohibit participation by proxy. Why? I do 
understand it, quite well, I've seen organizations in action and see 
how this combination of rules works.

The annual meeting is held in the home town of those who control the 
organization, and thus a majority of those who show up at the annual 
meeting are the managers or those close to the managers. So existing 
management, if proxy representation is not allowed, have a huge 
advantage, often insuperable, even where the majority of members, if 
they could be represented, would make different decisions.

An example of this, in theory, would be Citizens for Approval Voting 
or Americans for Approval Voting, I forget which way the corporation 
reads. Annual meeting located conveniently for the founder, 
dues-paying members can vote, proxy voting not allowed. I'm not 
singling them out, because this is quite standard practice. Why is it 
standard practice? Well, it benefits those who are the initial 
organizers, and they are the ones who write, usually, the bylaws.

There are two reasons why corporations can't get away with the same 
thing. First of all, if shareholders became aware of the restriction, 
they would be unlikely to hold shares, and, secondly, such a 
restriction is generally unlawful by statute. I think that there is 
some movement in this direction when classes of stock are set up, 
with some classes being non-voting....

But nonprofits aren't regulated in the same way. The state is 
concerned that there is a body of people responsible for ensuring 
that the corporation acts within the law, and it cares little or 
nothing about how these people are chosen, it only wants to know if 
they have assumed the responsibility. So nonprofits and political 
action groups can pretend to be democratic by having "members," but 
can effectively shut out the members from control. Standard practice.

However, if the members really wanted to change the situation, and 
were aware of FA/DP, they could do it, rather easily. They simply 
form an independent organization of members, and, among other things, 
allow a superproxy, or as close to that as possible, to be found, and 
then it is arranged for enough members to attend the annual meeting, 
one time or as necessary, to change the rules to allow proxy voting. 
And from then only, only *one* member need attend....

The evils of proxy voting come when it is allowed, but unused. What 
happens then is that someone notices the proxy provisions and 
secretly collects a lot of proxies. Perhaps this person even gets 
people to join and assign him or her their proxies. Then he shows up 
at the organization's meeting and suddenly can vote a majority on 
everything. Does tend to get people riled up. And so they outlaw, if 
they can still manage it, proxy voting. Jan knows this story, it 
happened with the Libertarian Party.

>>If Ketchum is saying that those who vote should understand what 
>>they are voting on, great. But who decides who is competent, who 
>>understands enough? My claim is that the proper one to make this 
>>evaluation is the voter himself or herself. Direct democracy by DP, 
>>I expect, will *increase* the participation power of exactly the 
>>right people, those who are widely trusted by those who know them *closely*.
>
>Anyone who is TRULY competent should be able to convince enough 
>voters to provide proxies as backing.

This argument is incorrect. Someone may be truly competent and 
knowledgable, but may be holding a view which is, shall we say, ahead 
of its time.

I find value in having a legislative vote, expanded by proxy, of 
2,374,089 to 1. Wouldn't it be interesting if, ten years later, that 
one vote was seen as being the only correct one? And, with 
Asset/Proxy, the one vote would be of an identified person. Perhaps 
next time more people would listen to him or her?

 From my point of view, the argument does not hinge on how it can 
work without allowing direct voting. It can. The question is whether 
or not it works *better* with direct voting, and there is another 
question, the question of rights.

If I hold a divergent view on a topic, such that to agree on it with 
any member of the legislature (100 people in Ketchums' proposal have 
the right to vote), I may be forced to not assign a proxy at all, 
because I would consider it repellant for my vote to be recorded in 
favor of a measure. I am forced, by Ketchum's proposal, to continue 
the odious practice of present electoral democracies, that I must 
compromise, not merely on issues, but on *representation*.

Again, compromise on representation is necessary as to deliberative 
rights, because we cannot allow a million points of view to be 
represented in deliberation, nor even much smaller numbers than that. 
If my view is sufficiently divergent, it is not going to be heard at 
the top level, though it may be heard at lower levels than that, and 
I may be able to gain support, but not enough to reach the assembly level.

But compromise on representation in *voting* is odious. I'm *not* 
represented if it is not my free choice.

With direct voting allowed, it becomes much easier to compromise on 
representation in deliberation, because no longer is it necessary 
that I have no serious disagreements with my proxy. I can simply 
follow the business of the legislature, its agenda, sufficiently that 
I will know when something important to me, and where I fear that my 
proxy will vote in a manner offensive to me, and then I can follow 
the debate and process on that and vote as I see fit.

There is no reason *at all* to prohibit this. I see no practical harm 
from it. It does not interrupt or make more difficult the legislative 
process, and if it introduces delay, such delay would be quite small 
and could be bypassed specifically in emergencies. (In which case 
votes would be provisional, as to full expression, but still 
effective as to immediate effect. I would take away the proxy (my 
proxy) of any representative who abused the emergency powers -- some 
city councils have emergency rules and use them routinely simply as 
an exercise in political power.)

 From my point of view, direct voting is the basic right, and the 
removal of that right must be justified. What is the justification? 
That is the question I'm asking.

So far, Ketchum has advanced these arguments:
         1. Direct Voting would be unpopular, so it would delay implementation
         2. Direct Voting would be irresponsible, uninformed people 
would vote. I presume that this is what is "destructive."

The first of these objections is a path to implementation question, 
not one concerned with conceiving of an ideal legislature. I've 
proposed a specific path which does not involve direct voting 
initially, but which makes it an easy step to achieve: this would be 
Asset Voting, which can elect a traditional-looking peer legislature, 
except with very exact proportional representation, so proportional 
that it can be considered *almost* totally free choice. Asset creates 
a class of electors, who effectively hold primary proxies assigned by 
secret ballot. These electors then elect the actual assembly by using 
or reassigning votes. (Any elector with enough votes may, but is not 
required to, assume a seat and then use remaining votes to create 
other seats, either independently, if the elector has enough votes, 
or in negotiated combination with others.)

So the first argument, while not entirely wrong, is not actually 
relevant when it comes to implementation strategy. Asset creates the 
class of people who could be direct voters: the electors.

And then this brings us to the second argument. Given that those 
voting directly would not be single citizens, in general, but persons 
who registered as candidates in the election, they would be, usually, 
relatively well informed. The second argument is actually a standard 
argument against democracy itself, it is considered that the "people" 
are ignorant and irresponsible. It has to be both, because I can be 
ignorant and simply abstain from making decisions where I'm ignorant.

Obviously, some people are ignorant and irresponsible. But if that is 
a majority, the society is in deep trouble no matter what style of 
government you set up. Normally, this is *not* true of the people, 
they are collectively intelligent and responsible, given institutions 
that allow this to be expressed. Far too often, the existing 
institutions don't allow it.

I seriously doubt that the Hutu majority in Ruanda would have decided 
through free process to attempt to exterminate their Tutsi neighbors. 
Rather, the institutions of majoritarian democracy set up conditions 
where a rather small minority of active fanatics could take power. 
Our system here is vulnerable to the same thing, potentially. 
Hopefully not so extreme!

Proxy democracy, if it is based on maximized freedom of choice, I 
expect to select for wisdom and intelligence. The vast majority of 
electors under Asset Voting would be quite capable of voting intelligently.

Ketchum has not responded, really, to the Asset proposals as being a 
means of electing the legislature that would not involve a complex 
local structure to exist first. But his proposals and the Asset 
proposal both create potential electors that would be reasonably 
qualified to vote.

So where is the beef?

I have alleged harm from not allowing direct voting, and the only 
reason we don't recoil in horror from this harm is that we already 
tolerate it, such offensive compromises (even worse, indeed) are 
routine in American politics. You hate abortion. Fine, I'll represent 
you. And, of course, I'll vote for the war, that's my party position.

People offended in this way often simply don't vote. That way, they 
avoid, maximally, having blood on their hands. I see no reason to 
exclude these people from the political process, indeed, I'd want 
them to be a part of it. These are people who *think*.

(What happens in fact, is that one side becomes so destructive that 
they *must* intervene, they hold their noses, pray for forgiveness, and vote.)

One of them is one of the most intelligent women I know. Why do we 
shut these people out?