[EM] I need an example of Condorcet method being subjected

Abd ul-Rahman Lomax abd at lomaxdesign.com
Fri Jan 22 09:39:32 PST 2010


At 12:03 AM 1/22/2010, Kathy Dopp wrote:

>Terry, just do not imagine that people do not see the "trick" you use
>of redefining words that have had a common meaning for decades.

The biggest is "majority," which has been redefined to mean something 
very different, which is then justified on a bogus analogy with real runoffs.
"majority" has always been shorthand for "containing a vote from a 
majority of ballots cast in an election," but this gets slipped into 
"majority of ballots containing a vote for one of the top two 
remaining after eliminations."

And the dirty little secret is that in most elections where there is 
no majority in first preference, i.e., when the sequential 
elimination retabulation is done, there is no majority -- real 
majority of *voters* -- after retabulation.

If the Robert's Rules method is followed, the elimination continues 
one more step in that case, to find if there is a majority of votes 
who have ranked the candidate instead of refusing to vote for the 
candidate. Education on IRV doesn't say, unfortunately, that ranking 
a candidate is a form of vote for the candidate, and many or most 
voters imagine that ranking a candidate "last" is a vote against the 
candidate. Because it is actually a vote for the candidate against 
all the write-ins or minor candidates who can't be ranked, if the 
ranking is eliminated.

No, if you want to vote against a candidate without taking a stand on 
every possible write-in, don't rank the candidate at all.

That's why this "majority" redefinition is so pernicious; without 
realizing it, voters approving IRV have eliminated a majority 
requirement without being aware of it, having bought the propaganda 
that IRV "guarantees" a majority result. Top two runoff does 
guarantee a majority result (if write-ins are not allowed; if they 
are, it's possible for majority failure to occur, though rare).

The classic case is in San Francisco, where the voter information 
pamphlet panel gave a "neutral" description of the measure that 
claimed something like "winners will be required to gain a majority 
of the votes."

It's very difficult to interpret that in a sane way to make it the 
truth, it's a deceptive statement, and it was, I'm sure, based on 
FairVote propaganda, and if the *accurate* statement had been made, 
the effect would have been very different.

"Winning candidates will have gained more votes than the leading 
opponent after the rest of the candidates and ballots which don't 
rank them have been eliminated. Mathematically, this is a "majority," 
not of "the votes," but of "votes remaining after all but two 
candidates are eliminated and the ballots not ranking those two are 
eliminated as well."

"Will be required" sounds like some standard which must be reached -- 
and thus which could fail -- and, since this redefined "majority" is 
a tautology, a mathematical construct of the method, that's deceptive.

As I've pointed out, use the Robert's Rules method of counting, which 
continues to the last elimination, not terminating when there are 
only two candidates left, but seeking to find a complete majority, 
and we could then make a parallel claim. "Candidate must receive 
unanimous support from votes cast."

In Brown v. Smallwood, the Minnesota court noted, with approval, 
prior judgement in another state that wrote about the issue being a 
"majority of voters" rather than a "majority of votes." Or was it 
"plurality" or language like that, I forget. It's too bad that they 
didn't follow what they apparently did not understand and failed to 
apply. Bucklin seeks to find a majority of *voters* who have approved 
a candidate. Not a majority of "votes." The number of voters is the 
number of valid ballots. If the ballot is valid, votes on it are then 
considered, and if a majority is required, a winning threshold is 
established that then must be found or the election fails. Or the 
election is decided by plurality.

IRV, in every implementation so far in political elections, is being 
decided by plurality, in most of the elections that go to "instant runoff."

But FairVote, even after all of this has been made abundantly clear, 
continues to promote the deceptive arguments, even if modified 
cleverly to make them not-exact-lies. That is, if you know the truth, 
you'd have to say that the FairVote propaganda is true, as to literal 
fact, but only deceptive as to impression created. And, of course, 
individual activists continue to promote the deceptive impression, 
probably not realizing that they are lying, or not caring. We will 
see that FairVote has turned the corner when it becomes willing to be 
"fair." Thus Kathy's Fairytale Vote is quite on point. FairVote is 
selling false hopes, fairytales, based on a collection of 
misinformation and deceptive political argument, designed to play on 
voter ignorance of the complex issues of voting systems. And, long 
ago, Rob Richie indicated his contempt for the "ivory-tower 
theorists" who objected to his deceptions, on the basis that they 
were not "practical" and not "politically realistic." Which means, 
boiled down, that, in his view, to be politically successful you have to lie.

That's not reform, that is the same old shit. With lipstick on it. 
Looks good, if you don't look and smell too closely. Bite into it and 
swallow it and pay attention to what happens.... what happens?

That's why studying actual IRV performance is so important. I didn't 
get half or more of the problems with IRV until I started studying 
actual results and comparing with top two runoff results.

TTR has its defects, for sure, though not as many as FairVote has 
alleged, their campaign has, almost entirely, been against top-two 
runoff, and, as a result, they have reduced election fairness in the 
actual jurisdictions they have targeted, based on deceptive arguments 
about cost, and plastering over the "majority" issue. Those 
jurisdictions, all of them, thought finding a majority was important, 
so the cost of runoffs was worth it. Obviously! But tell them, 
deceptively, that you can find a majority without the expense and 
inconvenience of a runoff, why, sure, what a great idea!

But it was a lie, originally, and it remains as deceptive propaganda 
that is turned into lies by those who don't understand it.

Terry, as long as you support this, you are properly tarred with the 
same brush. If you don't like the tar, stop supporting the spewing of 
deceptive propaganda by FairVote. Take a stand on it, both within 
FairVote and in public fora.

There are preferential voting methods that clearly will improve 
voting system performance, or, at worst, do no harm. Take top two 
runoff, and use an advanced preferential voting method for the 
primary, seeking a true majority (or possibly some lower criterion 
that solidly predicts that a majority would be found in a runoff, not 
the naive "40%" which does no such thing), and then use the method 
also, possibly with reduced ranks, for the runoff. Use an intelligent 
method for finding the best two candidates for ballot position, and 
allow write-in votes in the runoff, using a spoiler-free method in 
the runoff (so that write-ins can't spoil the election through 
vote-splitting, unless the voters really can't stomach voting for the 
lesser evil as well, in which case a majority failure indicates a 
real failure. As you know, Robert's Rules would continue seeking a 
majority, it is an unconditional requirement unless bylaws permit an 
election by plurality, which parliamentarians strongly discourage.)

Stand up for the best. You can still argue, at this point, that IRV 
is a better method than plurality, but only with partisan elections. 
Kiss was a better result than Wright, and Wright *might* have won 
under Plurality, and Wright was the worst result (though not by a 
large margin compared to Kiss). But in nonpartisan elections, there 
is no longer any reasonable argument that IRV is better for the 
jurisdiction adopting it.

Stand up to the FairVote ideologues and demand that FairVote start to 
work for real reform. Insist that FairVote cooperate with the voting 
system expert community. Insist that FairVote support and aid in 
reasonable implementation of other alternatives than IRV, and you 
will find that the expert community will be far more moderate about 
STV used for proportional representation, which was the original goal 
of the predecessor organizations to FairVote. STV is reasonable for 
PR. I'd argue that there are better methods, but I'd be thrilled to 
see an STV-PR proposal on the ballot or in legislation, and I'd 
support it, even though I know it is not ideal.

But IRV on a local ballot, in nonpartisan elections, is actually 
worse than plurality at much higher cost, and I have come to the 
position of strong opposition (from an initial support), based on 
study of actual elections. In partisan elections? I'd be more 
divided, but, given the Burlington results, the claim that the 
pathological behavior of IRV would be rare -- often it was claimed 
that there were no examples, but the reality was that there wasn't 
the data, Australia did not disclose the necessary ballot data to do 
real analysis -- can be seen to be false. And there is no reason not 
to use, instead, the much better performing American Preferential 
Voting, as it was called in the political science literature of the 
time, instead, plus if jurisdictions want a majority, they can use 
APV (Bucklin) in a runoff voting context, it will avoid many or most 
runoffs, reducing cost and inconvenience. It works better at finding 
majorities than IRV, because it does count all the votes. If IRV is 
going to be used, it should be in a runoff voting system, and top-two 
is lousy with IRV, if it means that lower-preference votes can't be 
counted. IRV with all votes counted would be far better for this purpose.

Later-no-harm makes sense with multiwinner elections, to a degree. 
Single-winner, it is a pernicious criterion, one much better off 
violated than respected. Allowing equal ranking with IRV would allow 
the voter to decide which they prefer: protection of the favorite, or 
casting an effective vote for a major candidate. But once we are 
going to do that, why not go all the way, and use instant runoff 
approval, which has a very solid theoretical basis in repeated 
election theory, a sliding down of approval cutoff to find a 
compromise? Instant runoff approval is a closer simulation of 
repeated balloting without eliminations, the basic Robert's Rules 
method, than is sequential elimination. IRA? Also known as Bucklin. 
That is exactly what it is.

Support experimental use of other methods, particularly ones which 
are low-cost, as any approval method on the table is, without question.

If you don't do this, Terry, fine, it's your right. But since you 
have entered this arena and have publicly supported IRV, and even 
incorporated some of the deceptive propaganda into the Vermont 
legislation you introduced, we, as well, have the right and possibly 
the obligation to expose your behavior and that of FairVote.

Instead, Terry, join us and keep us honest. By all means, don't allow 
us to make false and misleading statements, either. But don't attach 
yourself to lies. If you do, the flimsy lipstick will break away and 
you will indeed be smeared with shit. 




More information about the Election-Methods mailing list