[EM] DP in a legislature

[Abd ul-Rahman Lomax]

At 05:17 AM 3/3/2006, Dave Ketchum wrote:
>My topic is a legislature, which represents all the voters in the
>district.  They appoint via proxy, rather than electing, members of the
>legislature, such as a senate.
>
>Responding to Jobst:
>
>Y must know if X is Y's proxy, to be able to judge whether it is time to
>switch proxies.

Actually, if Y does not know the identity of his or her proxy, that 
person is not a proxy, as we have been using the term. A proxy is a 
personally chosen representative, as distinct from an assigned or elected one.

[For secrecy]
>Others would do elections.  Seems they can get secrecy.  They also lose
>ability to redo proxies until the next election.  Finally, the multilevel
>proxies with opportunity for communication look good to me - but hard to
>set up via election.

Secret-Ballot Delegable proxy, like the initial Asset Voting 
proposal, would be secret in the initial assignment of proxies, but 
open in the subsequent assignments. Rather obviously, for there to be 
responsible government, *somewhere* the identity of the 
representatives must become known. Delegable proxy can boil down 
representation until it reaches a manageable and efficient assembly, 
and I think we assume that the members of the assembly must be public 
figures. There seems to be little harm in what is public being larger 
than the assembly, and, indeed, it could be much larger.

Secret Ballot, in my opinion, is only necessary under conditions 
where there is reasonably possibility of voter coercion or 
intimidation. When a proxy represents a certain minimum number of 
voters, it becomes possible to devote public resources to the 
protection of these proxies, just as public representatives are now 
protected. In any case, no citizen would be forced to serve as a 
proxy who was not willing to take the risk involved. I do not think 
there is any way to protect through secrecy any level in a delegable 
proxy system higher than the primary assignment of proxies. Thus 
secret ballot proxy systems would probably lose, to some degree, one 
major advantage of delegable proxy: its ability to create efficient 
and manageable *personal* communication between the public as a whole 
and those who represent the public, made possible by a totally free 
assignment of proxies, even to the point where a proxy may only hold 
one vote other than his or her own. (But more commonly might hold, 
say, twenty, a very manageable communications burden for unpaid volunteers.)

>Seems like time for debate.

What we need is discussion, and we need something else. In the FAQ 
which is being written (and posted piecemeal here), it will become 
clear that FA/DP organizations are not merely communication fora, but 
that they also have decision-making mechanisms, or, more accurately, 
mechanisms for developing and measuring consensus, since FAs, except 
with regard to simple organizational business, don't really make decisions.

My goal is to see a much wider understanding of Delegable Proxy (and 
what I think is an inevitable precursor, at least, Free Association 
Delegable Proxy. If government were a Free Association, it would be, 
essentially, libertarian, but I'm not a Libertarian nor is Delegable 
Proxy Libertarian, in the sense of advocating libertarian government 
(which may be an oxymoron). I'm convinced that libertarian principles 
are appropriate for Free Associations, but am far from convinced that 
they would work in government, and certainly not now, if ever.

However, where freedom of thought and discussion is important, the 
mechanism which facilitates discussion must be free of coercion, 
hence libertarian.

>Do not understand your statement about forcing.  Partly because the record
>keeping should prevent loops.

Actually, all that is necessary is that loop members know that they 
are in a loop. If every member of an organization names a proxy, 
loops are inevitable. The only problem is where a loop ends up 
unrepresented in a discussion or poll because *no* member of the loop 
participates. Loops are one of the first problems that people think 
of when they actually start to consider the implications, but they 
are not really a problem, particularly if members of loops are 
notified, which may be automatic whenever a loop is formed (assuming 
some kind of automated system), or which could simply be a 
notification when all loop members are "absent" from a "meeting" or poll.

>Some thoughts on design for a senate, etc.  (here voters do proxies
>instead of elections and, as I describe it, can amend their proxies
>whenever they see need):

I.e., this is not a secret-ballot system


>Need to maintain a current list of voters.  New ones register; old ones
>die; some move into a district; some move out.
>       A district must appoint (rather than elect) multiple senators -
>likely one district to appoint the complete senate.

The original thinking that led to my own formulation of delegable 
proxy went through a stage where the idea was that voters would meet 
in small groups and elect one of their number to represent them to 
another small group composed of those similarly elected, etc. The 
initial idea was to hold a Presidential election in 10 days, starting 
from scratch, where every "round" in this election involved 
face-to-face meetings of perhaps 10 people. The first and most 
obvious problem was in this election of a single representative. 
Though there is a possible solution, the meetings could be 
essentially preselected for strong affinity, I eventually came to the 
idea that a meeting could send on more than one representative, and 
then, even simpler, that people could assign proxies; the structure 
becomes chaotic, a fractal, which at first seemed a disadvantage 
until I realized that nervous systems are organized as fractals. It 
may be hard to visualize, but it would be self-organizing, like 
biological structures, and the functioning would be relatively simple.

Forcing proxies to be assigned within districts may not be harmful 
when the districts are large, but I don't see any reason for it 
except where the district is the district of the entire assembly, 
i.e., is its jurisdication. What if the person I most trust in the 
world lives across town, in a different precinct. Why should I not be 
able to give my proxy to him, to represent me in the city council?

Indeed, my most active current project is for a Parent Free 
Association at a private school. The Parent Council currently 
consists of Class Representatives, which are theoretically elected by 
class. It's an obvious way to do it, but the fact is that Parent 
Council at this school and at other affiliated schools are famous for 
passing in and out of existence every few years. Class representation 
is *very* spotty. Many parents don't even know who their class rep 
is, for they did not attend the meeting where the rep was elected. 
Communication is quite poor.

Class reps are like district representatives with a small district. 
You may or many not have *anyone* in the class with whom you have 
good rapport, and good rapport is an essential element in building a 
strong organization.

It is not that class representation is wrong, it is just that it is 
neither complete nor necessarily efficient. What I'm proposing is 
that parents be encouraged to name a *personal* representative, on a 
public list, someone they would most trust to represent them and to 
inform them of what they might need to know of Parent Council 
activities, if they are not themselves present.

>Any voter offering to be a proxy holder needs to offer a "platform" to
>identify what kind of action they offer.

This is political thinking, and it really assumes quite a gap between 
proxies and those they represent. Delegable Proxy opens the 
possibility for choosing on the basis of something much better than 
platform, which is all too often something that a politician tells 
you he supports in order to get you to vote for him, and *maybe* he 
will follow it. And maybe we are all better off if he does not, for 
politicians will promise the moon.

I'd suggest, instead, choosing proxies based on a simple 
characteristic: you trust the person. And, more than that, you have 
access to the person. This is why, once people realize what a proxy 
is, they would no more choose Clint Eastwood as their proxy than they 
would choose him to handle their affairs when they were on vacation 
or incapacitated; i.e., only if he was actually a personal friend and 
available for the task.

It is this possibility of easy access that is what sets delegable 
proxy apart from all other representational systems that I've been 
able to imagine.

And, I'm sure, that is why biological systems are organized in a 
similar way. You don't have one neuron linked to a million sensors.

>       Both lone voters, and those who already hold proxies, can offer
>their collection to this holder.

I'd never do it without a personal conversation.

Many people, presented today with such a system for public 
governance, would not know what to do. Fortunately, proxy democracy 
will almost certainly see broad adoption in non-governmental 
organizations, so, by the time it hits public life, they will know what to do.

This aspect of the discussion is largely moot: if a proxy system is 
set up, proxies and the public will do what they will do, the rules 
would not state anything, I'm sure, about public statements of 
platform. Personally, were I willing to serve as a proxy, I would 
*not* issue a platform. Platforms are either useless fluff intended 
to attract votes, or they actually bind a sincere politician who 
wants to do what he promised and who therefore is inflexible when 
faced with information which would lead him to something contrary.

Platforms and the debate around them are useful under present 
conditions because they *can* give us a glimpse into the character of 
those who debate them, and some idea of how these people might 
function. But, we might note, politicians often disappoint, perhaps 
more often than not, when measured against their platforms.

With DP, sure, you can change your proxy at any time, but this also 
makes platform not nearly as important. Indeed, if what we are 
talking about is direct democracy with delegable proxy, i.e., 
citizens may vote directly on any issue if they so choose, and 
especially if citizens may see how their proxy voted and remove their 
vote from his by directly casting a vote, platform becomes *much* 
less important. What matters much more is that the proxy serves well 
as a communicator and analyst.

>       Note that there are many possible platforms - positions with few
>backers end up with no one both offering a platform AND getting enough
>proxies to get seated in the senate (access to abortion easier or harder
>and access to drugs easier or harder are reasons for 4 platforms for the
>4 basic combinations on these two issues).

In a DP system as I foresee it, there would be multiple ways to be 
heard in a "Senate." The proxy system does three things: it allows 
decisions to be made without the direct participation of the voter, 
which is especially valuable when the voter has no opportunity or 
inclination to become properly informed; it is a means whereby 
citizen concerns can be communicated to the assembly of the whole; 
and it is a means whereby the whole can communicate and explain 
decisions to the citizens.

However, suppose the citizen has a critical concern and the citizen's 
trusted proxy, on that particular issue, refuses to carry the concern 
to the assembly. If this causes the citizen to lose trust in the 
proxy, then the remedy is simple: change the proxy assignment. But 
what if this is a simple disagreement and the citizen does not 
consider that it renders the proxy otherwise untrustworthy? There is 
nothing to prevent the citizen from approaching any other person with 
the concern, and if this person is willing to pass it up, this person 
passes it (and explains it) to *her* proxy.

The fractal structure of DP, if the organizational habits are right, 
allows personal communication to become general communication. While 
some might worry about the "telephone game" problem, the loss of 
information in transmission, that problem only exists when there is 
no parallel redundancy, such as written communication passed up -- or 
group edited on a wiki.

>       Perhaps this holder offers communication with those voters whose
>proxies are held directly.

It think that is essential. Sure "distant" DP might be a vast 
improvement over elected representation, but it will be a pale 
imitation of what it might be if communication is open.

(It is for this reason that I've considered that proxies would not be 
effective until *accepted*. A proxy who is too busy to accept another 
"client" might suggest one of his existing clients to the 
applicant.... This is very different from what we see in the 
political sphere, where we expect politicians to seek us and to sell 
us on voting for them. That, indeed, is a formula for creating all 
the mess we see today.... When a corporation wants to hire an 
executive, they search for him or her, and I think they might be very 
suspicious of someone who spent millions of dollars to get them to 
hire a particular candidate. After all, if I want to pay *you* to 
allow me to handle your affairs, what would you think?)

>   If so, there almost HAS to be a limit as to
>how many voters this offer extends to - not necessarily all voters whose
>proxies are held.

The offer can and should extent to all givers of *direct* proxies. 
Not to indirect. Those who have named a lower level proxy communicate 
with the higher level through their lower level proxy, the one whom they chose.

*However,* I do expect some broader communication than that. A 
high-level proxy might, for example, operate a mailing list for all 
of those represented, and this list might extend and include those at 
lower levels.

Where a high-level proxy represents *many* people, such as millions, 
that list could not be unmoderated. But proxies will work all this 
out themselves, it does not at all have to be part of the structural 
rules. The structural rules, in fact, can and should be *very* simple.

>Secrecy of proxies:  NEEDS to be doable, yet parts of my proposal require
>openness.

Secrecy is doable at the primary level and even there it does inhibit 
realizing the full value of the system.

>The senate needs rules.  Some should be standard - likely based on
>Robert's Rules.  Some need to be tailorable - some easier than others.
>Basic coding I use below:
>       E - easy to modify, for little problem with manipulating.
>       H - hard to modify, where there can be excessive temptation to
>modify for wrong reasons.
>       HH - REALLY dangerous to change.
>
>Size of senate:
>       E - to increase - when present membership sees value in letting in
>other viewpoints, and that bigger is practical.
>       H - to decrease - can be temptation to lock out annoying members
>holding few proxies.

The problem is not as difficult as it might seem, *if* voting in the 
senate remains completely open, i.e., if it is a direct democracy 
with respect to voting, if not with respect to generating universal 
traffic (i.e., speaking to or entering motions before the whole 
assembly). The rules are subject to vote as in Robert's Rules; the 
rules restricting *full* membership actually benefit everyone, so 
there should be no problem in getting the membership to accept them. 
After all, Town Meeting governments devolve into City Councils, 
resulting in a loss of citizen power, because citizens realize what a 
pain it is to deal with tedious meetings. As long as the rules 
changes, and any decision at all, really, must receive the approval 
of the required majority of all who vote, personally or by proxy, 
there is no problem with restricting full membership. It will be 
self-regulating.

(If for some reason membership is restricted too far, a majority of 
those speaking before the assembly might be promoting a certain 
agenda, yet the agenda would fail to receive majority approval. In 
order to get majority approval, you need to allow broad participation.)

>Puzzle:  Size of senate has to be kept practical.  Yet potential number of
>platforms approaches number of voters.  My best thought is that the small
>parties have to develop a platform that enough of them can share to have e
>holder of enough proxies to get on the floor of the senate.

Again, I think platform secondary to character. It is quite possible 
for a representative of character to fairly represent the views of 
people with whom  she disagrees.

DP, by the way, makes political parties rather unnecessary. I'm not 
sure they would continue. They would not be needed to fund elections, 
because there would be no elections as we understand them. Everyone 
who wants to be is represented, there is no need -- or means -- to 
"beat" the other party. Rather, controversy will devolve into issue 
controversy rather than platform, party controversy.

>       Size of senate has to be kept manageable.
>       Do not see a voter having a voice or vote except via a holder with
>enough proxies to qualify for a seat.

Exactly. The exact number I would leave to the vote of the assembly, 
assuming that this is a *full* vote, not merely the vote of the members.

(I've assumed, generally, direct democracy, membership restrictions 
are only about those things which must be restricted to avoid 
communications overload, i.e., addressing the whole and entering 
motions. It is these things which require either the action of a 
qualified proxy or the permission of the assembly (presumably upon 
the motion of a qualified proxy).

>[...]
>Puzzle:  Power of individual senators has to be limited.  If one senate
>held and voted a majority of the proxies, that senator would have a
>monopoly on voting.  Make the limit a percentage of votable proxies
>votable by one senator:
>       H - to make limit over 25% or under 10% (my offer).  Over 25%
>encourages concentrating voting power too close to a monopoly; under 10%
>discourages reasonable concentrations.
>       E - to vary limit between those points.

I've thought about this. The problem is moot in Free Associations, 
but it could be a serious one in government. However, if votes are 
held in stages, where all qualified proxies cast their votes first, 
and then there is a period in which those not qualified (on down to 
individual citizens) may cast their vote independently, the problem 
disappears. There is no power problem even if there is a superproxy 
(which might be defined as one who represents everyone, or, perhaps 
more usefully, as above, i.e., one who represents a majority of votes 
and thus theoretically able to make decisions directly. But only if 
those represented continue to consent!)

So I don't recommend a limit. *As long as the proxies are revocable 
at any time and as long as direct voting by a citizen (or lower level 
proxy) removes that citizen's vote from the high-level proxy's total.*

>Write-in votes?  I sympathize but, when I try to think farther my head
>aches.

Fractals do that to ya.

Fortunately, we can try all this out in NGOs, including discussion 
and advocacy groups, where failure will not cause serious harm.

>   I do say that a voter can retract his proxy, but this leaves him
>mute until s/he finds a more acceptable platform to back (this voter can
>even do a new platform, but remains mute until attracting enough proxies
>to demand a seat).

This problem only exists if direct voting is not allowed. I see no 
reason to prohibit it, and plenty of reason to facilitate it. You 
visit the Senate, and a motion is being considered about which you 
have an opinion. Perhaps you listen to the debate. And when it comes 
time to vote, you go to a terminal which is set up for members of the 
public to vote. You enter some identifying information and a 
password, and you cast your vote. And that vote is recorded and, 
indeed, reported. Imagine this: The vote in the Senate today on the 
bill to establish a fund for the support of frumindillys passed today 
by a vote of 603,507,276 to 1, the only dissenter being Mr. Contrary 
from Hokokomo.

And when it turns out that frumindillys are really bad for the 
environment, won't Mr. Contary look good?

Seriously, the vast majority of votes cast in the Senate, even if 
remote voting is allowed (i.e., internet voting), will be cast by 
proxies, and probably most of them by qualified proxies, i.e., those 
with debate rights. But the very fact that everyone can vote if they 
so choose solves the problem of exclusion. There is very little harm 
in excluding some from debate if they are *not* simultaneously 
excluded from the decisions themselves.

Today, we are both excluded from debate, the vast majority of us, but 
also from decision. The former is necessary, but the realization that 
decision and debate could be separated has been one of my better moments....