Thanks Kathy for your comments >Is it also "reasonable" when IRV elects a winner who the majority of >voters oppose in an "instant"?
>Don, your conception of "equality" is surely different than my own. The relative question is "Does IRV violate the Equal Protection Clause of the Constitution"? If it does not them the state or city has a right to adopt IRV. It is not the responsibility of the U.S. Supreme Court to say "the voters in a city cannot adopt IRV, because there are better methods available or that it fails some criteria". Plurality voting (first passed the poll) does not violate the Equal Protection Clause. Equal Protection does not require an election to yield a majority. The Supreme Court has said that, in effect, voters in a city (or State) are free to adopt (by a vote of the people) any voting method they what (even some really bad ones) as long as it provides Equal Protection to all (and does not violate other parts of the Constitution). See Washington State Grange. v. Washington State Republican Party (2007). In the opinion, the Court said "Finally, facial challenges threaten to short circuit the democratic process by preventing laws embodying the will of the people from being implemented in a manner consistent with the Constitution." This case relates to the "top two primary". The primary selects two candidates by plurality (the first round) to be on the general election ballot (the second round). The primary itself has only one round, and all candidates who have less votes than the top two are eliminated. This type of primary system has some major problems including "cloning", and monotonicity. The "Top-two primary" as well as any Run-off election are conceptually the same problems as IRV (although not instant). It would seem that these election types have already passed the "Equal Protection Clause" standard (see the courts' opinion). INSTANT: Let us assume that a run-off election is held a month later. Well with addition time and information a voter may decide to vote for a different candidate. Let us now assume that a run-off election is held a day later. With very little new information the odds are you would vote for the same candidates as the original vote. With some addition thought some voters, but perhaps very few, may decide to vote for a different candidate. I don't know how may would change their vote a day later. But I will speculate it could be less than 0.5 percent (only a wild guess). Let us now assume that a run-off election is held 1/1000th (or less) of a second later. It is theoretically impossible for a voter to change their mind in an "instant". The "instant" you drop your ballot in the box you are voting (or had the right to vote) in every single round of the election. That is why they call it "instant" run-off voting. I believe this is the definition of "Instant run-off voting" REASONALBLE: When I used the term, I as only referring to a what the U.S. Supreme Court would determine as the "reasonable intent of the voters". If a voter votes A>B>C or A>C>B I believe it is reasonable to assume that the intent of the voter is, if C is eliminated, then the voter would vote for A over B in a second round. The U.S. Supreme court has used the "Reasonability" standard many times. My opinion is not important, what is important is what would the Court assume is the intent of the voter (under the reasonable standard). >Is it also "reasonable" to assume that all voters whose candidates are >eliminated before the final counting round would chose not to vote in >the final round in an "instant"? Assume that A and B are the only candidates in the final round, and that C is eliminated. If I voted C>B>A, then if you look closely at the final vote tally (under IRV) you will see my vote for B. All voters whose candidates are eliminated before the final counting round votes (C>B>A and C>A>B) DID vote in the final round. >And is it also "reasonable" to assume that all voters whose first >choice loses and whose second choice is never considered are OK with >never having their second choice counted in an "instant"? This sounds the same as your statement above. Maybe I don't understand your point. The only thing I would say is that all votes are counted in the second round. Except for a C>blank>blank vote. The question is, what is the intent of this voter? The voters' intent seems quite clear, the voter wants to vote for C in the first round and if C is eliminated then the voter did not wish to vote in the second round. Assume that the voter votes blank>blank>blank. The voters' intent also seems clear, the voter did not wish to vote in the first or the second round. If a voter does not vote in the first round (blank>blank>blank) their vote is not counted, if they don't vote in the second round (C>blank>blank) their vote is not counted in the second round. >And is it also "reasonable" to assume that all voters whose 2nd and >3rd choices are only considered after it is too late to help their 2nd >and 3rd choices avoid elimination are OK with never having their 2nd >or 3rd choices considered in a timely fashion? Sorry Kathy I don't understand this comment. Are you saying that the Condorcet method is better than IRV. I believe I answered that in the beginning above. Don Hoffard ---- Election-Methods mailing list - see http://electorama.com/em for list info