[EM] Why I think IRV isn't a serious alternative

[Abd ul-Rahman Lomax]

At 05:06 PM 11/25/2008, Kevin Venzke wrote:
>If you want multiple parties in
>order to represent more interests, best go to PR in the first place.
>I want it to be possible to have multiple viable "parties" in order
>to make it more likely that the median voter can get what he actually
>wants.
>
>For the latter, I don't think it's clear that if Condorcet can't succeed,
>nothing can.

What's remarkable to me is that a solution was suggested over 120 
years ago, and was simply neglected and forgotten, even though the 
proponent was quite famous, Lewis Carroll.

The basic voting method was STV, optional ranking. And Carroll 
recognized that requiring voters to deeply rank many candidates was 
asking them to do something they probably were not good at.

So he suggested what Warren Smith called Asset Voting and Mike 
Ossipoff, before him, called Candidate Proxy.

If a ballot is exhausted, let it continue to be exercised by the 
candidate in first rank on it, "as if it were his property."

Now, my thinking: we do not have representation if our "choice" of 
representative is coerced, if there is contention involved in it. If 
my neighborhood elects a representative to the Amherst Town Meeting 
(I don't live in Amherst, I live across the river), and I didn't vote 
for that representative, I'm only accidentally represented, I have 
not chosen the representative. The idea of contested elections for 
representative only makes sense when your system grows out of one 
where representatives were chosen by the sovereign. The public, 
voting in an election, is the sovereign. These is not true democratic 
representation; but, because it approximates it to some degree 
(rather badly sometimes, but often better than that), it has been tolerated.

Real representation is chosen representation. And Asset Voting allows 
that, without thereby creating an assembly that is far too large for 
practical conduct of business. The Amherst Town Meeting is 
ridiculous, way too large to be truly functional; it hangs on because 
of long tradition; in two recent elections, it avoided being replaced 
with a Mayor-Council government by a handful of votes. It's called 
Town Meeting, but that term is usually reserved for direct democratic 
assemblies of town voters, that what most small towns here have; but 
when the town gets large, Town Meeting becomes cumbersome and 
impractical, it takes up too much time when there is a contentious 
issue. (When there is no contentious issue, many fewer voters show up 
and it can still be functional.)

Asset should work with simple vote-for-one voting. It should allow 
total and complete sincerity: vote for your favorite, the candidate 
you most trust, out of what could be a large universe of candidates. 
It then uses deliberative processes (negotiation) together with 
post-election amalgamation (the asset holders recast their votes) to 
put together an assembly that could approach or even meet full 
representation, where every vote is represented by the holder of a 
seat whom they voted for, or who was chosen by someone they voted 
for. You would know who was elected, specifically, with your vote, 
together with the other votes needed to meet a quota, if asset 
holders, I'll call them electors, transfer them in precinct blocks, 
so that all the votes the elector received from a precinct (or nearly 
all) went to a single seat. That's not a rigid requirement, but it 
would normally be possible to do that, and would probably increase 
the sense of connection between voters and those who represent them 
in the Assembly.

But, damn! It makes all this fancy election stuff largely moot! 
Offices can be elected in the Assembly, and, hae you ever noticed 
that no assembly uses, in its proceedings, instant-runoff voting? Or, 
indeed, any advanced method? You don't do that when you have an 
assembly that can hold multiple votes on an issue, as many as needed. 
Instead, the simple process of motion, second, discussion, amendments 
and votes on amendments, close of discussion, and vote on a 
single-issue question (Yes/No), have we noticed?, is 
Condorcet-compliant, but more than that. It is intelligent, 
interactive, and considers preference strength. It's 
Condorcet-compliant because if there is a candidate who pairwise 
beats the current proposal in the motion, a motion to amend should 
succeed. If members care! But preferences shift during the process, 
it's not so neat as the set of preferences we use to analyze election 
methods. It's messy and unpredictable. Like other living things.

So what do we need as an election method? If we must have a single 
ballot, and a single winner, period, Range Voting is actually a 
trick: it is the only relatively objective method of assessing the 
expected voter satisfaction with an outcome, turned into an election 
method. It's ideal because it's designed that way. (The only fly in 
the ointment is the charges about strategic voting, but I've been 
arguing that this is based on a total misconception of what we are 
doing when we vote.)

But why in the world would we limit ourselves to that? Bad idea. 
Single-winner elections are focused toward creating office-holders 
who are very difficult to fire, they have fixed terms. That's not the 
way you would hire servants, or officers, for a business. You hire 
them "at will." The continuity argument is bogus. Sure, continuity is 
valuable, and that is one of the things that one considers when 
hiring and firing.

So: Parliamentary form of government, more advanced than the hybrid 
royal/elected representative system used in the U.S. (The President 
is a King for a fixed term and with very substantial restrictions, 
but King nevertheless, not like the President of a corporation, who 
is a hired servant, who can be fired by a simple majority vote of the Board.)

Then the problem reduces to how to manage the parliament, or 
Assembly, as I call it. And that is actually, again, simple, so 
simple that one wonders why, really, it's never been done. There were 
some efforts in the U.S., in the early part of the 20th century, to 
set up proportional voting systems, where holders of seats on a city 
council voted, in the council, the number of votes they received in 
the election. Such a system would allow very broad representation, 
though I think Asset, creating a peer assembly, on the face of it, 
would do better. It didn't succeed anywhere, as far as we've found.

The trick of Asset would allow something much better than fixed-term 
seats. The electors, all of them, could form a virtual uber-Assembly 
that I call a penumbra around the Assembly, the electors would have 
the right to *vote* in the Assembly. When they vote, their vote would 
be subtracted from the votes cast by the candidate(s) they elected to 
the assembly. (I expect that such direct voting would be exceptional, 
that normally it would not be enough to shift results, but that it 
*could* shift results is important.) And the electors could fill 
vacancies that appear, or remove seats and replace them. Thus the 
Assembly could become continuously representative of the electors, 
who were purely chosen by the voters without contention. (Indeed, as 
I'd do it, any voter can become an elector. Just register, minimal 
fee to be published in a directory, smaller fee without publication, 
but you get a number that's used on the ballot. And vote for 
yourself. Elector holding one vote. Down side, perhaps? That vote is 
then exercised publicly, it's public record. Care about that? Don't 
register as an elector, instead, choose the one you most trust. 
Elections might be held once a year. In my opinion, the Assembly 
would be *almost* the same as a pure direct democracy, as close as is 
practical when the scale is large.

I do think that something like this is where we will end up. And I do 
know how to get from here to there. One step at a time. It starts 
with communication, and with establishing networks. It's happening.