This sounds quite interesting, Abd ul-Rahman. Where can I learn about
your FA/DP idea? Your discussion here is helpful, but I feel like I
am missing out the important prerequisite pieces in order to make
sense of it. (I know about delegable proxy, but haven't heard about
FA/DP specifically).
Thanks,
Duane Johnson
On Apr 21, 2010, at 9:57 AM, Abd ul-Rahman Lomax wrote:
At 04:42 PM 4/19/2010, Duane Johnson wrote:
Hi Everyone,
I am new to this forum, thanks to James Green-Armytage who sent me
the address. I am a software engineer in Chicago who also happens
to be interested in voting methods.
I'd like to propose a voting method that may be of interest here.
It has also been cross-posted to the ideas group at forums.e-
democracy.org. This system seems almost too simple when you
understand it, but the implications are deep and, I believe,
profound. I am interested in your feedback.
I'm glad to see more people thinking about process and voting as
involving communication. Which leads to considering communication as
the foundation of democracy, not voting per se. Functional democracy
is deliberative democracy; democracy without communication is easily
manipulated and if power is directly exercised, there is a tendency
to mob rule, where the intelligence of crowds is dumbed down instead
of amplified. Wikipedia, for those who study it, is a great example
of how not to do it.
However, I strongly urge people who attempt to analyze the situation
and to propose reforms to:
1. Keep it simple. An extraordinarily powerful system for fully
proportional representation consisting of a seemingly-simple tweak
on Single Transferable Vote was proposed in 1883 or so by Charles
Dodgson (Lewis Carroll). If a simple system that is *obviously* far
more democratic doesn't attract notice for more than a hundred
years, what chance does something more complicated and dodgier
(i.e., involving lots of unknowns) have?
2. Don't propose chickens without eggs or vice versa. Imagine that
there is some ideal political system, that allows good ideas to be
efficiently considered, with the necessary depth. If such a system
existed, it would be easy to suggest and propose and agree upon it,
and if people agree on just about anything, they can do it, as long
as the actual already-existing system is reasonably democratic. I
invented FA/DP (Free Associations with Delegable Proxy) as a method
for considering and forming consensus on ideas like FA/DP.
3. FA/DP is terminally simple, but, in reality, it's like pulling
teeth to even get people to consider it. Sure, lots of people will
say, "What a great idea," if they don't get stuck in the knee-jerk
objections, like, "*They* will corrupt anything." or, more
sophisticated, "Iron law of oligarchy (see the Wikipedia article),"
etc.
4. Notice: the method by which one would develop consensus and
implement better political systems is a political system.
Revolutions tend to empower the revolutionaries, or those who
inherit power from them. If we want a true democratic revolution, we
must want something different from the norm of revolution, which
tends to follow the same traditional power structures, thus, in
effect, simply changing faces. Traditional power structures boil
down to two kinds: oligarchical and distributed. Oligarchical power
developed and prospered because it was more efficient when the scale
became large -- even though it is, from an ideal perspective, very
inefficient -- due to the involvement in process normally required
for distributed power to function, which expands exponentially with
the number of active participants.
5. FA principles are natural for humans, most peer organizations, in
their infancy, are roughly FAs. But if the FA principles aren't
understood and solidly maintained, and as organizations grow, they
naturally develop oligarchical structure, it is what people know how
to do, and they are not aware that there are alternatives. When the
organization is small, implementing something like DP seems too
complicated. Can't we just discuss things like we always have? When
the scale becomes large enough that DP is truly needed, it's too
late. De-facto oligarchies have already developed, and the Iron Law
of Oligarchy begins to function and resist change back to
distributed power. The oligarchy believes that it knows best, and,
indeed, it often does. It's the exceptions that are killers, that
reduce long-term efficiency and support, that allow originally
wonderful nonprofit organizations, for example, to become divided
and weakened, to be co-opted by corruption, to become no longer
truly representative of the aspirations of their members, but
because the organization has been "successful," and comes to
dominate its field, it is very difficult to start anew and such
efforts will be considered "divisive" and "disruptive."
6. So: consider delegable proxy, how simple it can be when applied
within a Free Association, which does not concentrate power *at
all*. In the Montesqueuian sense, it is pure judgment, which I think
of as advice. In theory, if the executive and judicial power are
fully separated, the judicial system has no direct power, it only
advises but cannot coerce the executive system. A wise executive,
though, wants good advice! In an FA/DP system, the system functions
solely to advise its members as well as anyone else interested. The
only coercive power it might have is the distributed power of the
members, if they act in a coordinated manner, and it cannot, without
blatant abandonment of FA principles, coerce them. The FA itself
does not collect power except for the power to communicate, and it
cannot control that, for the communication structure is easily
replicable, if a central structure is corrupted in some way, the
delegable proxy structure can re-establish it, network by network,
reconnecting outside the original structure, because most of the
communication in an FA/DP structure is not through the central
structure, it's under the control of the proxies themselves, who
cannot compel any client to participate, who do not "own" their
clients, nor do the clients own them.
7. Often overlooked, because of the habits of thinking only in terms
of hierarchies that concentrate power or representation, is that, if
everyone names a proxy, all paths in the directed graph produce form
loops. There is nobody at the "top," or, more accurately, there must
be at least two. I.e., if there is what I call a superproxy, someone
who, if nobody else participates in some "discussion," would
represent every member, there must be at least two, because the
proxy named by this superproxy would have the same representative
power.
8. Lots of people, thinking about this stuff, try to eliminate
loops, because a loop can mean that some are not represented in some
high-level process. However, when that "defect" happens, it's
visible. It's easy to notify those who are "absent," and to do so
efficiently, through identification of unrepresented loops and the
"proxy rank" of those involved. ("Proxy rank" looks at each member
and considers how many clients are represented if the member
participates and nobody else does, it's a bit complicated because it
involves recursion to avoid the appearance of equal representation
due to a superproxy of some natural caucus naming a proxy within the
caucus.) Loops represent natural caucuses, and they only cause loss
of representation if no member of the caucus, in the loop,
participates in a process. Attempting to avoid loops involves
restricting the rights of members to name people in their own
trusted group, forcing the naming of someone outside that group.
That's coercive, in the name of "more democratic." Bad idea.
Instead, let people decide if they want to be represented or not. If
they do, and if it is not practical for them to be represented
directly through a loop member, all it takes is one member of a loop
naming a proxy outside the loop to break it.
9. Alternate proxies are another solution to the loop issue. An
"alternate proxy" is a backup proxy, considered to represent a
member if the primary proxy is not present or represented. But,
remember, every complication represents risk. Single-proxy DP
encourages members to think about whom they most trust, from among
those who are available to them. A single proxy has a clear (and
accepted, I encourage not recognizing, or deprecating, unaccepted
proxy designations) responsibility to communicate information coming
from the "center" to the client (or to *not* communicate information
that the proxy deems to be "noise" for the client). Does an
alternate proxy have this responsibility? The core of DP, to me, is
the proxy-client relationship and how it will function.
10. FA/DP is necessarily informal, it creates organizational
structure only in the loosest sense. The relevant Alcoholics
Anonymous tradition (AA is the model FA) says, "AA as such ought
never be organized, but we may create service boards or committees
directly responsible to those they serve." Notice: a "service board"
is a corporation, and it is, in this "exception," not creating a
board -- which will own property as a corporation and thus develop
centralized power -- that is AA or controls any part of AA, and that
is "responsible" to those it serves, which means, in practice, those
who actually form it. They do not obtain the endorsement of AA
itself, they are only quasi-AA, perhaps broadly supported, or
perhaps not. Local Alcoholism Councils are often formed by AA
members to represent the interests of alcoholics and alcoholism
treatment before local governments. These Councils are often created
and dominated by AA members and other interested persons, but they
do not speak for AA, they speak for their members.
11. In the same way, the natural caucuses formed by proxies and
their clients, cooperating, can form, say, Political Action
Committees that exercise real power, collecting donations, or,
because of campaign finance laws, they can easily recommend
individual donations to their members. FA/DP doesn't collect power,
in itself, these PACs are outside the FA itself, it does not endorse
them, though they are free to endorse it. FAs do not take positions
on controversial issues, and this is crucial, if FAs are to be able
to form broad consensus, for to do this, every side on an issue
should ideally be represented. FAs will certainly form with "initial
bias," a bias created solely by the natural bias of its members, and
the power exercised initially through PACs may reflect this bias.
If, as I expect, FA/DP social technology is as efficient as I
expect, these PACs will be successful, and will attract imitation.
If the FA is open to participation by "the other side," it then
becomes a mechanism whereby broader consensus might be found. Each
caucus remains free, if it chooses, to organize its own PACs, but
two PACs opposing each other in the sphere of real power weaken each
other. That's good, when consensus hasn't been found! But
cooperation and consensus are powerful. If agreement can be found,
it will prevail. In order for this to work, joining the FA must not
have any appearance of supporting one side of a controversy. FAs can
report the results of polls of its members, those are just facts.
However, these polls do not establish an "organizational position,"
except in one very narrow sphere, the organization of the FA itself.
Strictly speaking, the non-endorsement of controversial position
practice of FAs is with regard to "outside issues." Within the FA,
there are traditions which maintain the autonomy of all "meetings"
and which prevent coercive power from arising. There is still the
Iron Law of Oligarchy, it is not cancelled by FA/DP, but the leaders
that develop are as they are in AA. "Our leaders are but trusted
servants, they do not govern." Strictly speaking, these "leaders" do
govern, but they only govern their own "meetings," i.e., they govern
their own communication group, their own natural caucus, and because
they cannot compel participation, the collected power is highly
restricted. In AA, the saying is, "The only requirement to start a
meeting is a resentment and a coffee pot." AA turns natural division
and disagreement into fuel for expansion, for the more meetings
there are, the more options there are for members, the more
communication power exists.
12. I imagine FA/DP using all forms of communication technology,
but, it is important to remember, it doesn't depend on any form of
communication in particular, beyond the central idea of two people
communicating directly. DP is established if there is a member list
and a list of assigned proxies (preferably with acceptances, that's
often overlooked by those who are trying to set up a "voting
system.") For voting applications, I recommend, instead of DP, Asset
Voting, with DP as a voluntary and optional method for electors
holding votes to cooperate and coordinate to create representation.
Asset Voting is compatible with existing traditions regarding
deliberative assemblies, creating a default assembly where every
member has the same voting power as every other member. That can be
modified to reduce the voting power of a member whenever an elector
who has transferred votes to that member votes directly, but that is
a tweak, an improvement, I believe, and would establish true and
complete representation in all decisions of an Assembly, which is
important, but it is also possible that such "outside voting" would
be only advisory. Early implementations of Asset Voting should keep
it simple, establishing the principle of full representation by
chosen representative, instead of representatives being chosen in
"contests," for such contests represent not only winners and losers
among candidates, but winners and losers among voters. Are the
losers represented? How can we say that a voter is represented in a
legislature when the voter opposed the choice of a majority, not to
mention the mere plurality that is often the real case? This is the
elephant in the living room of "representative democracy." It isn't.
The people, the "demos" are not represented, jurisdictions are,
districts.
----
Election-Methods mailing list - see http://electorama.com/em for list info