[EM] Election-Day musings. Sainte-Lague is the best PR.

[Abd ul-Rahman Lomax]

At 08:43 PM 12/5/2006, MIKE OSSIPOFF wrote:
>Yes, election day was a long time ago, but I reside in a remote rural area,
>and I don't often get to a computer.

Welcome back, Mr. Ossipoff.

>I already knew this, but it occurred to me again that, if we have to have
>representative government, it would be better if people were voting for a
>party. Voting for a party platform instead of for a personality, a hairdo,
>someone who acts presidential, etc.
>
>In representative govt, that would be PR. I've talked here about how Direct
>Democracy with Delegable Proxy (a term I got from James) is the form of govt
>that I prefer.

I would have said that straight Delegable Proxy *is* DDDP, because a 
proxy is one who exercises a right on behalf of another when the 
client prefers not, or is prevented, and I'd prefer to keep DP as a 
form of direct democracy unless it is qualified.

It's handy that standard proxy already exists and is in wide use in 
corporate governance; generally overlooked is that delegability is a 
property which need not require changes in formal structure, unless 
one wants the formal structure to allow on-the-fly changes in 
delegated proxy assignments. In theory, DP is simply a matter of the 
principals, those who wish to do so, using a DP structure which 
*they* form to create, among other things, recommendations for the 
assignment of formal proxies.

In other words, if you have proxy democracy, those who want delegable 
proxy can have it simply by creating and using it. They do not need 
to get permission from anyone, they do not need to convince a 
majority, they do not need to convince the oligarchs or major 
shareholders, as an example. They just need to recognize the value 
and follow that understanding in actual practice.

>I believe that people would accept it long before they'd
>accept PR. Proxy is reinvented and re-proposed again and again by people new
>to the subject. I've observed that myself at a PR meeting. But the
>proportionalists stubbornly reject it.

It is quite odd to me that proxy democracy is considered some kind of 
new invention. It is quite old. What is new is the idea of applying 
it to political institutions. I suspect that the reason goes back to 
the origins of democracy, to the very concepts of sovereignty. The 
participation of the people in government was not considered a 
property right; it was, rather, a privilege extended to the people by 
agreement with the sovereign. Functioning direct democracies for 
government were rare, so rare that we typically give the only obvious 
contemporary example a specific regional name: New England Town Meeting.

Even with Town Meeting, historically, the state, with its superior 
jurisdiction, whittled away at the independence of towns, 
centralizing power and decision. The Sovereign *grants* towns the 
right to govern themselves as direct democracies. It grants a 
charter. The Sovereign is the source of power. Not the people.

We still see this upheld as desirable: "We do not have a democracy, 
we have a Republic." And, of course, the two major parties in the 
U.S. are named precisely over this distinction. Legally, the U.S. 
form is republican, not democratic. However, the sovereign has 
granted certain participatory rights to the people, and, because the 
people are largely passive, and accept the illusion of democracy thus 
created, this situation continues.

(I am not denying that there are democratic *aspects* to the U.S. 
system. However, the system itself does not consider the people as 
sovereign. "A government of the people, by the people, and for the 
people," is a political slogan, not a factual description, one 
invented in the context of a civil war where one set of people were 
imposing their opinion on another set by force of arms. No, you may 
not leave the Union. That the seceders were themselves oppressive 
with regard to a class of people does not change this. The Sovereign 
will not tolerate succession.)

However, the people *do* have the power, if.

If they are organized. I am, of course, far from the first to 
recognize this. But this realization actually begs the question. How 
can the people be organized?

The road is littered with the wrecks of attempts to organize the 
people, and some of these attempts are still limping along. The 
problem of how to organize the people is exactly the problem of 
government. If you succeed in organizing the people, and the 
structure thus created is not itself democratic, you have merely 
substituted one oligarchical form for another. Because of the 
relatively free nature of our society, many attempts to organize 
"people power" remain at relatively low levels of success. And, 
almost uniformly, they are not democratic in structure.

At least not in *real* structure. Some have structures that allow, 
just like our national government, members to vote for officers. But, 
read the bylaws. Very many of these bylaws specifically prohibit 
proxy voting. *If it were not prohibited, it would be lawful.* Other 
organizations, sometimes, *do* permit proxy voting, but proxies are 
unused. And then someone realizes the power of proxies, uses it, 
gains just enough power to terrify the oligarchs, and proxies are 
then prohibited. There are ancient and powerful forces working 
against democracy, even within organizations that ostensibly promote it.

But as long as proxy rights are respected, the delegable form can be 
instituted, in effect, without central or collective approval. I 
think this is a major realization, and it is the basic idea behind 
FA/DP: Free Association with Delegable Proxy.

FAs are, by definition, informal; that is, they may have formal 
structure, to some degree, but that structure does not bind or 
control. Rather, it communicates, it deliberates, it suggests. The 
members retain actual power, and where that power is to be exercised, 
they may directly name a proxy. What the informal DP structure adds, 
in theory, is the ability to do this on a large scale, with massive 
collective effect, or, at least, effect proportional to the numbers 
of members who utilize it.

Is FA/DP practical for government? Probably not. FA/DP is thoroughly 
libertarian. But if the people, for government, want delegable proxy, 
or proportional representation, or, indeed, whatever they choose from 
what they can collectively control, they can have it. They can even 
do this if the present government is far less democratic than what we 
already have, as long as they have freedom of association.

Now, the last and most crucial vestige of centralized sovereign power 
is the power of history. We are bound by what our forefathers 
decided. We have rule of law. Commonly overlooked, however, is that a 
majority may change law, a majority, distributed, can even change the 
U.S. Constitution. (The requirement of a 3/4 vote is a vote of the 
states, and, as we have often ruefully noted, this can involve less 
than the support of a simple majority of the people, due to non-PR 
representative forms of electoral decision.)

That the majority may change the structure itself is actually 
embedded in Robert's Rules. Usually one would say that it takes a 2/3 
vote to change bylaws, under RR. But, the details.... an absolute 
majority of members may change the bylaws at any time, and notice is 
not even required, as I recall.

Absolute majorities are very hard to assemble in organizations, unless.

Unless there is proxy representation in wide use.

Proxy representation empowers the majority, and this is its danger. 
And, of course, its promise as well.

Without proxy representation, organizations are run by the active, an 
informal oligarchy which exists even within otherwise quite 
democratic organizations. The active are those who have the resources 
(the time or money) to actually participate, and, in addition, they 
have the motivation. Thus there is a tendency, without the 
*restraint* of proxy representation, for the active to be more 
extreme in view than the general membership. The two-thirds rule 
provides some level of restraint: generally, absent an absolute 
majority, a two-thirds majority of a quorum, after notice, may amend bylaws.

Quorum requirements become almost unnecessary or irrelevant if proxy 
representation exists and is in use.

Now, to the original purpose I set out to expand upon Mr. Ossipoff's 
comments: there is a form of PR that, in theory, is almost *perfect* 
in proportionality, and it can also incorporate some of the features 
of DP. This is Asset Voting, which was named by Warren Smith, whom I 
generally credit with its invention, though the idea was actually so 
obvious that I used to think this is what STV, Single Transferable 
Vote, was. From the name, I imagined that you could vote for someone, 
and they could transfer your vote.

Mr. Smith calls these transferable votes "Assets." His system allowed 
voters to assign fractional votes to as many candidates as the voters 
cared to empower. My own variation on this is Fractional Approval 
Asset Voting, which is much simpler to implement: the ballot is a 
standard ballot, same as is used for Plurality or Approval. One may 
vote for one candidate, in which case that candidate receives one 
vote as an asset. And one may vote for N candidates, in which case 
each candidate receives 1/N votes.

Single-winner, FAAV would, I think, be quite interesting. It would 
convert an election method into a deliberative one. Asset Voting, in 
general, is a variation on delegable proxy, used for a single 
election. (It is delegable proxy if candidates receiving votes may 
then assign these votes to another, who may then reassign them, etc. 
That is, if the votes are like assets, transferable at the sole 
discretion of the possessor.)

However, for PR, FAAV gets *very* interesting. Consider this scheme:

We are attempting to create an assembly with N members, each of whom 
will exercise equal voting power. (DP generally does not set this 
limitation when used directly for governance, but there is a certain 
appeal to peer assemblies.)

V Voters assign their votes to candidates as in FAAV. The quota is V/N.

Any candidate receiving this number of votes or more is elected, or, 
alternatively, may directly appoint members of the assembly, using 
V/N of his or her assets to do so. (Proposals have been made that 
candidates would publish assignment lists in advance; however, while 
candidates should be free to do this, they similarly should be free 
*not* to do it. Voters can then decide whether such promises are 
important. Personally, I'd rather pick *one* person whom I trust, or 
perhaps a small group, and leave it to them. I can then focus on 
finding the single candidate whom I most trust. This candidate then 
becomes *my* personal representative, and functions as such, directly 
or indirectly through those the candidate has chosen.)

Now, this gets interesting: when transferring votes, candidates 
transfer them in precinct blocks. For this to work, a certain small 
wastage of votes is required; however, in my view, the benefits would 
far outweigh this small loss. I'm not going to go into the math of 
wastage; but I would assume that only complete blocks would end up 
being assigned to each representative. The point, of course, is that 
voters will know exactly whom they elected. Because of secret ballot, 
the representatives won't know, unless the voters tell them, but the 
voters will, of course, know what precinct they voted at. This, also, 
is public record.

This is, in effect, gerrymandering on the fly. Candidates assembling 
votes will be assembly districts. Districts need not be contiguous, 
they might, for example, represent other kinds of constituencies: 
auto workers, for example, in Michigan, to give a stereotypical example.

And they might well be contiguous districts. A candidate receiving a 
large number of votes might decide (and might even promise) to assign 
these votes to create local representation.

It could get quite complex, but the complexity is not in the system 
itself, rather it is in how it might be used. In this case, 
complexity is a synonym for intelligent.

This would create an assembly which is truly proportional. It is not 
based on political parties, but I assume that parties would continue 
to be important, and, if the voters think party affiliation is 
important, they can create party blocks in the assembly; what is new, 
however, is that *independents* also get proportionally represented.

What happens with the dregs? Well, I said we were attempting to 
create an assembly with N members. There are two major possibilities 
as to how to handle the dregs. The first one is that we don't 
necessarily end up with N members. The assignment of votes in 
precinct blocks will necessarily end up in a certain unevenness of 
the number of voters represented by each winner. I'd prefer, I think, 
that it take a *minimum* of V/N votes to create a winner, so, then, N 
members becomes next to impossible. This could be handled by setting 
N higher than the expected number of winners; if somehow it works out 
that all votes get assigned to create the full number, it is not 
really a problem that there is an extra member or two in the assembly!

Or, alternatively, the dregs, excess votes assigned to create 
winners, could be collected and used to create another winner or 
perhaps a few. These winners would *not* be precinct-assigned, they 
would be raw votes.

And what could be *very* interesting, is that excess votes could be 
used to create members with reduced status. The problem with large 
assemblies (and, indeed, the problem with direct democracy on a large 
scale) is that they become procedurally impossible. However, this 
problem is not created by voting rights, it is only created by the 
right to participate in deliberation. Reduced-status members might 
still have the right to vote, but not to address the assembly or 
introduce motions without permission. Reduced-status members would 
not, however, hold a full vote. They would hold fractional votes, 
based on the number of votes assigned to them. This could be made 
standard, i.e., Associate Members might have, say, a tenth of a vote, 
but this is actually more complex than simply allowing associates to 
have any fraction at all. I would state the voting power of 
Associates in actual votes held, which would then be divided by V/N 
whenever vote totals were compiled. What gets interesting is that an 
associate could be created with a single vote. If, however, there 
were very large numbers of associates, attendance of a large number 
of them might become impossible; so ... proxy voting.... It's already 
used in some state assemblies when working in committee....

(And how would they ask for permission? Through a member who *does* 
have full rights. "Mr. Speaker, I rise to request permission for 
Associate Member Smith to address the assembly." The Speaker replies, 
"Without objection ... the Associate may rise and speak." And, of 
course, any full member may object... If there is an objection, it 
would be put to a vote. Can Associate members vote on this? Yes, of 
course. Voting rights should be inalienable, the problem of scale 
properly only applies to actions which take the time of the assembly. 
But Associate members would only have fractional votes)

But the complication (and power) of having Associate members is not a 
core feature of the idea. It is merely one way of dealing with stray 
votes. The other way is simply to waste them. The wastage would be 
quite small and would be distributed among all the voters of the 
candidate who was reassigning the votes to create winner. The smaller 
the precinct size, of course, the smaller the wastage. It would be, 
essentially, invisible to the individual voter, who would have an 
assigned representative, one he or she either voted for directly, or 
who was chosen, directly or indirectly, by that directly chosen rep.

Accurate PR with your own personally elected representative. Don't 
you think this is interesting?