January 3, 2010

IRV=Voter Disenfranchisement

On Tuesday, March 3rd 2009, I voted for Mayor but my vote didn't count and neither did 605 other votes. This is called voter disenfranchisement. So how did this happen? It all started with a voting system called IRV (Instant Runoff Voting) where voters ranked their candidates in order of preference on the ballot.

Unfortunately, under this system, if you chose not to rank your candidates or you chose only one candidate (who happened not to make the final runoff) your vote was not counted.

The final runoff between Kiss and Wright was decided by 8,374 votes, not the total 8,980 votes cast for Mayor. Kiss won by 4,313 votes to Wright's 4,061 votes. Now add their two numbers together and you get 8,374, which leaves 606 ballots in an exhausted pile that's not counted under IRV rules.

I would have liked to have voted in an election between Kiss and Wright as the only two candidates. I did not know or would have guessed that the final runoff would be between those two candidates. This is where IRV fails.

I was disenfranchised from that election because I did not want to rank candidates or strategize over the possible outcome. I just wanted to vote for one candidate and that's it. If it was a traditional election there would have been another separate runoff at a later date where I could have heard more from Kiss and Wright and then had the opportunity to cast a more informed vote, and one that would have counted.

23 comments:

rbj said...

Owen, your argument carries no water. Your vote counted as much as anyone else who ends up in the loser's column. Your candidate lost.

I agree that Andy Montroll (if that was the sole candidate you marked) should have won the election (if we had better rules for tabulation than IRV). But the reason is not what you say. The reason that Andy Montroll should be the mayor is that the majority of Burlington voters chose Andy over any other particular candidate when asked to choose between the two. That is what is simply meant by the Condorcet winner (which is what Andy is). Most often (as it did in 2006), the IRV winner will be the same as the Condorcet winner (because if the Condorcet winner makes it to the IRV final round, he will also win that final round by definition) but sometimes (and 2009 in Burlington is such an example) IRV fails to elect the Condorcet winner.

But Plurality is *worse*. And delayed-runoff (between the top two 1st-choice candidates) is worse. Your vote for Andy (or Dan, if that was your vote) would have "counted" no more effectively with the old 40% Plurality system either.

The correct reform to IRV is Condorcet, *not* Plurality nor Delayed-Runoff.

Owen said...

I respect your opinion and what your saying makes sense but my argument does carry water. My vote wasn't counted. I thought I understood how IRV worked and because I actually didn't, I lost out. I realize now if there is another IRV election in the future I will have to rank my candidates to be sure my vote counts. But I don't want to rank the candidates and strategize over who should be second and third and so on.

Why is that so hard to understand?

I prefer the system we had. It's simple and you vote for one candidate. If I'm stupid because I prefer this then so be it.

rbj said...

If you chose to not register an opinion about a secondary preference (who you would vote for if your first choice wasn't in the running in the final round) then you had no vote in the final round (only) because you expressed no opinion between the two candidates that did make it into the final round. But, even if you marked your ballot in the initial round for only one candidate, you did have a vote. And your candidate lost.

If you voted only for Andy (or Dan or "Homer" Simpson), and expressed no opinion between Bob and Kurt, then you chose to not vote in a possible runoff between Bob and Kurt.

So if your objection is to the ranked-order ballot, I continue to say your argument carries no water. All the ranked-order ballot does is require you to decide what your support is for all the candidates on the initial Election Day. You do not get to wait to see who wins sufficiently for a runoff where (if your candidate doesn't make it to the runoff) you have another opportunity to re-evaluate your political preferences and vote again. You do not get to change your mind who you support after your first choice. With IRV or any ranked-order ballot (and no delayed runoff), you must decide on Election Day. Why is that requiring too much from you?

Really it's the delayed runoff that is "One Person, Two Votes". The Instant Runoff commits you to decide who it is that you are for and who it is that you are against on a single Election Day.

But, if your objection to IRV is specifically how it tabulates the vote and identified Bob Kiss as the winner when more Burlington voters explicitly marked their ballots that they preferred Andy Montroll, then I am with you. And with the Condorcet method of tabulation of the ranked-order ballots, that pathology would not have happened. Another pathology is about the 1513 voters who supported Kurt Wright first and Andy second, their vote for Kurt actually helped elect Bob Kiss. If a few hundred of them stayed home (or insincerely ranked Andy above their favorite, Kurt), then Bob Kiss would not have been re-elected. Condorcet would also have avoided that problem.

The method that IRV uses to count the vote and declare the winner is messed up and the 2009 mayoral race is a perfect example of that. But Plurality Rule (which is sometimes Minority Rule) is even more messed up. And Delayed Runoff decreases voter turnout (on Runoff Day) because it is decidedly less convenient than voting on a single Election Day. Electing someone with reduced voter turnout is decidedly less democratic than with a robust turnout.

Owen said...

I know, I didn't play by the IRV rules and so I lost. I don't like a system that forces people to rank candidates and I certainly didn't realize the winning percentage was going to be taken from a number that's less than the majority of voters that came out to vote.

rbj said...

Thank you for your honesty, Owen. By what you have said, do you realize that essentially, you disenfranchised yourself? (I am assuming you voted for Montroll without expressing an opinion on Kiss or Wright.) The whole idea of IRV or any ranked-order ballot is that there isn't going to be a runoff in 3 weeks, no matter what. So you (as well as every other voter) have to make up your mind on Election Day who you would vote for in a runoff if your candidate doesn't make it. You chose not to do that, but that was your choice, not a disenfranchisement foisted upon you. But no voter should have the need for strategic voting foisted upon them. The most common voting strategy (where a voter does not vote their sincere preferences in order to best accomplish their political interest) is called "compromising". It's what happens in "traditional" elections (what you want us to return to) when people who support a 3rd-party candidate (like Nader) decide to vote for the major party candidate (like Gore) in order to avoid electing their worst choice (Bush).

It happens also with IRV as Republican Prog-haters found out in Burlington 2009.

Owen said...

I think your getting it. IRV creates a system where voters disenfranchise themselves unintentionally. That's part of my concern.

rbj said...

Owen, I got it last March. (Have you seen my paper titled "The Failure of Instant Run-off Voting to accomplish the very purposes for which it was adopted: An object lesson in Burlington Vermont"? email me at rbj@audioimagination.com and i'll send you a copy, if you want.) I even got it before I voted for IRV in 2005. I knew it wasn't Condorcet so it could conceivably elect someone other than the Condorcet winner. But I thought it unlikely because "all" the Condorcet winner need do is make it to the final IRV round and that person will also win the IRV. But Andy didn't make it to the final round so my estimate that it's unlikely for IRV to miss electing the True Majority winner (another term for the Condercet winner) was experimentally disproved.

Owen, I really get it. I've been studying and contemplating the problem of majority rule in multi-candidate elections for 2 decades.

But the anti-IRV crowd does not "get it" if your "better" alternative is returning to the 40% plurality rule. If the old law was in effect in 2009 and Kurt Wright had somehow squeeked by a "victory" with a 40% plurality or (more likely) with a delayed runoff where more of the Republicans bothered to get to the polls on Runoff Day than Dem and Prog liberals, that would have been a worse travesty for democracy than what happened in 2009. Kurt Wright was the third most preferred candidate. Bob Kiss was the second most preferred candidate and Andy Montroll was preferred by Burlington voters over every other candidate. That is precisely what our ballots collectively say.

Owen said...

You can send me a copy at ombreath@gmail.com

rbj said...

it is done.

please tell us what you think after reading it.

Owen said...

OK, give me a few days. Have some other business to attend to. Thank you for the INFO.

Lea Terhune said...

40%, 39%, 51% ? Given the fact that only 27% of registered voters bothered to vote in the last mayoral election, I don't think 39, 40 or 51% is a very big issue. All are somewhere around 12% of the registered voters! The plurality issue is minor. If saving money is important, we should make the plurality 30%. We need to pour our energies into what we can do to increase voter participation. IRV is not the answer and if Deb Markowitz hopes to be Governor, she had better drop IRV from her campaign platform

rbj said...

The plurality issue is minor? If the plurality is 50%, then it isn't an issue, plurality is the same as majority (which is always the case between two candidates, unless they tie). But in the cases where plurality is not the majority, then it is the minority.

Minority rule is definitely a problem, Lee. That's basically saying "my candidate should win even if he is disliked by the majority of the electorate."

the whole point of why Burlington voters adopted IRV in the first place is that the plurality issue is not minor.

Lea Terhune said...

Ha! When 27% of registered voters show up, the difference between 39% and 51% is insignificant, and all are a minority of registered voters. The significant issue is that we need to inspire people to vote. Any ideas? -- and puhleese do not tell me the answer is IRV or Condorcet. None of these voting methods matter if people don't care to vote. I'm hoping that if we get issues on the ballot that people care about, they will vote!

rbj said...

"Any ideas? -- and puhleese do not tell me the answer is IRV or Condorcet. None of these voting methods matter if people don't care to vote."


Well, the City of Burlington could pass election rules that say, when it's time to decide who'll the next mayor is, to send a single ballot to my address on North Cove Road and I'll fill it out and send it back and that would determine who the mayor is.

it's not IRV or Condorcet.

Anonymous said...

"I voted for Mayor but my vote didn't count and neither did 605 other votes."

This is sheer & utter nonsense!

"Unfortunately, under this system, if you chose not to rank your candidates or you chose only one candidate (who happened not to make the final runoff) your vote was not counted."

Baloney...this is simply not true, period end of story! "Bullet voting" is a purely reasonable way to vote under an IRV system.

"I would have liked to have voted in an election between Kiss and Wright as the only two candidates."

Then you should have ranked either one of those candidates in the first place, moron!

"If it was a traditional election there would have been another separate runoff at a later date"

...where waaaaay less people would have shown up, period.
---------------------------

"The reason that Andy Montroll should be the mayor is that the majority of Burlington voters chose Andy over any other particular candidate when asked to choose between the two."

This is, again, sheer nonsense. Voters were NOT asked to rank merely between two candidate, they were asked what their preferences were between FIVE candidates.
------------------------------

"My vote wasn't counted."

Yes, it was, moron!

------------------------
"Another pathology is about the 1513 voters who supported Kurt Wright first and Andy second, their vote for Kurt actually helped elect Bob Kiss. If a few hundred of them stayed home (or insincerely ranked Andy above their favorite, Kurt), then Bob Kiss would not have been re-elected."

This kind of "analysis" fails to even try to understand the will of voters on election. These voters from above merely wanted Wright to be mayor more than they wanted Montroll to be mayor, period. Asking these voters to "stay home" or vote against their conscience is completely & totally undemocratic.

-------------------

"IRV creates a system where voters disenfranchise themselves unintentionally"

...if they are as stupid as you are Mulligan...ugh...

-------------------

"We need to pour our energies into what we can do to increase voter participation"

...and since IRV in Burlington, voter turnout is UP!

rbj said...

well Anonymous, i don't think you even know who's on your side or not.

i'm voting against the repeal question. but i do not ignore or deny the pathologies that happened in the mayoral election of 2009 (that did not happen in 2006).

the only reason one can claim that Bob Kiss should be mayor over Kurt Wright is because 249 more voters marked their ballots that they preferred Bob over Kurt (4313 to 4064). that's a good reason: why should "Candidate C" ("Curtis") be elected when a majority of voters agree that "Candidate B" ("Bob") is a better choice?

but, by the same token, why should "Candidate B" ("Bob") be elected when a majority of voters agree that "Candidate A" (guess who the "A" stands for?) is a better choice? that's a pathology. there is no rhyme or reason you can point to (other than "them's be the IRV rules") that supports the choice of Kiss over Wright that does not also support the choice of Andy over Bob.

there are two other pathologies:

Kurt Wright was, in fact, a spoiler. if Wright was not in the race and everybody came to the polls and ranked the remaining candidates the same, then Montroll would have ended up in the final IRV round and we know that Montroll would beat any other candidate in the final round, had he been in the final round. so there was a spoiler, not a good thing.

the last pathology (that you've written off) is that more Wright supporters marked Montroll as their 2nd choice (a 3 to 1 margin to the other way around). call them what you may, but i think "GOP Prog-haters" might not be a bad label for these folks. what happened for these GOP Prog-haters is that their primary support for their favorite candidate (Wright) actually elected their last choice, Bob Kiss. that's an undeniable fact which gives these GOP Prog-haters a legitimate gripe about how the IRV worked.

none of the pathologies would have happened in 2009 if the same ballots were used, but the election was resolved using Condorcet rules rather than IRV.

facts speak louder than ad hominems.

but, you and i are voting the same way in a month.

rbj said...

a number i forgot to include, while the IRV final round shows that Bob beat Kurt by 249 votes, applying the same counting another pair shows that Andy beat Bob by 587 votes. Andy's "defeat" of Bob is more than twice as strong as Bob's defeating of Kurt. but IRV didn't consider that other defeat.

a point i wanted to make about the GOP Prog-haters' votes actually causing Bob Kiss to be elected is that, *if* IRV survives to 2012, these voters will have to consider (strategize) their vote and, if they want to be savvy, have to consider ranking their 2nd choice (the Dem) above their 1st choice (the GOP) just to avoid electing their last choice (the Prog). that means that the 2009 election encourages strategic voting where the type of strategy employed is called "compromising".

Anonymous said...

"but, by the same token, why should 'Candidate B' ('Bob') be elected when a majority of voters agree that 'Candidate A' (guess who the 'A' stands for?) is a better choice? that's a pathology."

No, it's merely a trick that's trying to be pulled on people that don't understand theoretical math. The "woulda, coulda, shoulda" elections didn't happen, period.

"Kurt Wright was, in fact, a spoiler"

...if, by "spoiler", you try & twist the meaning of a word that most people know the definition of in the first place that is. Please, to argue that voters should "stay home" & not vote for a person that they really, really want to see as their mayor is completely & totally undemocratic...just like advocating that one candidate should really, really stay home & not run for office, period end of story.

Anyone that knows Kurt Wright knows that he's wanted to be mayor for many, many years, which is why he ran in the first place!

"the last pathology (that you've written off) is that more Wright supporters marked Montroll as their 2nd choice (a 3 to 1 margin to the other way around)."

GOP Prog-haters?? Who's said that before?? Not I my misguided friend.

Once again, you simply try to ignore the will of the voters, which was, in this case, that they really, really wanted to see Wright elected over Montroll, and Kiss as a last resort.

Gee whiz...Dems really didn't want Wright (the most "conservative" candidate in the race), and GOPers really didn't like Kiss (the most liberal candidate in the race)...someone alert the BFP! LOL...

Look, I like Andy Montroll...he's a really nice guy...a smart guy...and his election results reflect that. Voters really liked Andy, but not enough to be their mayor. It's as simple as that...live with it...

"but, you and i are voting the same way in a month"

...so, why don't you STFU & not give false "facts" to those that aren't on your side in the first place??

Anonymous said...

"but IRV didn't consider that other defeat"

...because there was no "defeat". Once again, for those that have short memories apparently, the election was between FIVE different candidates, NOT two candidates.

rbj said...

you're telling me to STFU about false facts? all of my facts are supported by... the facts. the ballots are public record. anyone can get them. anyone that knows how to can write a simple computer program to run whatever count they want to, and the rabid IRV opponents (like Tony Gierzynski) have done that.

by ignoring and denying the facts, all you do is toss red meat to the other side. then they take those facts, add to it your ignorant and self-serving denials, and come to the false conclusion that returning to Plurality would fix the problems.

"theoretical math". duh.

whatever is the "theoretical math" that is done in the IRV final round is exactly the same math that is done to determine who the true majority winner is. but we apply it to all candidates, not a single pair of candidates that survive the IRV elimination mechanism that has its own goofy math.

the "theoretical math" is pretty simple: if a ballot is marked with Candidate A ranked higher than Candidate B, it's counted as a vote for A. if it's marked with Candidate B marked higher than Candidate A, it's counted as a vote for B. that's it. count the votes, do this for every pair of candidates and see who wins in each little runoff. it's simpler than IRV, fairer than IRV, and for sure fairer than the old Plurality method.

"if, by 'spoiler', you try & twist the meaning of a word that most people know the definition of in the first place that is." duh.

the meaning of "spoiler" that i am using is a loser whose presence in the race actually changes who the winner is. "no spoiler" means that if this loser is removed from the race and the ballots, the winner would remain unchanged. if Wright were removed from the election, Andy would have beaten Bob in the final IRV round. that's a simple fact supported by the public record.

it doesn't really matter if Andy is a really nice guy or not. he could be the devil from hell, but that doesn't change the fact that a majority of Burlington voters would rather see Andy elected mayor than any other candidate, including Bob Kiss. it is only because of convoluted and arbitrary IRV rules that this explicit voter preference did not get considered in the election count last March.

you're doing your cause far more damage than you claim i am doing. you should be the person who should STFU. you're just being stupid and polemic. you'ld rather just rant your slogans than solve a problem (because you deny the problem). there are anti-IRVers that do the same thing. you're no better.

Anonymous said...

"all of my facts are supported by... the facts"

...in your wildest, theoretical dreams that is my friend.

"by ignoring and denying the facts, all you do is toss red meat to the other side."

WRONG...by confronting the intentional misinformation by some (and I personally don't include you in that group...yet, since you appear to be able to cut through the baloney & realize that IRV is worth saving) I (and many others I would note) are better informing voters of what the stakes *really* are.

What you are merely doing, as I've already pointed out (and I won't do again my friend!), is giving a bunch of false talking points to the opponents of IRV, period end of story.

"'theoretical math'. duh."

No, not "duh"...that's what ALL of these "pathologies" are based on...elections that could have happened (or that some think should or would have happened if this, that, or the other theoretical thing had happened) but DIDN'T happen! I've already successfully debated these bogus issues before, and I'm a real live scientist my friend. So, don't even bother...

"whatever is the 'theoretical math' that is done in the IRV final round is exactly the same math that is done to determine who the true majority winner is."

This is sheer & utter nonsense, and you should know better than to claim that my misguided friend. Counting ballots based on one's known choices is NOT the same as assuming this or that untruth & then trying to re-count those ballots.

"not a single pair of candidates that survive the IRV elimination mechanism that has its own goofy math."

LOL..."goofy math"...yea, maybe for those that don't understand simple addition & subtraction I guess...please...

"the 'theoretical math' is pretty simple: if a ballot is marked with Candidate A ranked higher than Candidate B, it's counted as a vote for A. if it's marked with Candidate B marked higher than Candidate A, it's counted as a vote for B. that's it."

Once again, it's apparently "too bad" and "irrelevant" that the election was actually between FIVE different candidates and NOT just two! LOL...and 'round & 'round we go...

"duh."

No, again, not "duh"...you're trying to twist the word spoiler into meaning something that it simply doesn't mean, period end of story!

"the meaning of 'spoiler' that i am using is a loser whose presence in the race actually changes who the winner is"

...which, in this case, is the person in the election that came in SECOND...LMAO! You really are too much my friend...give it up...all of this has been tried before...

"that's a simple fact supported by the public record."

No, it's a simple & completely illogical LIE, that's what!

"it doesn't really matter if Andy is a really nice guy or not."

Yes, it really does, since, as I've explained to you before, it explains why he got so many second & third place votes. People liked him, but not enough (when compared to everyone else that was on the ballot at the time) to be their mayor in the end.

"it is only because of convoluted and arbitrary IRV rules that this explicit voter preference did not get considered in the election count last March."

There is simply nothing "convoluted" or "arbitrary" about IRV my friend. IRV merely allows voters to express their personal preferences for who should win an election. People express their personal preferences about many, many things every, single day...it's really not rocket science yanno.

"you'ld rather just rant your slogans than solve a problem"

I haven't ranted any "slogans" my friend...YOU'RE the one doing that...and they are ALL the slogans of the anti-IRV crowd...now wake up!!

Joyce McCloy said...

A Feb 5, 2010 headline in a San Francisco newspaper reports about a lawsuit based on this very issue:

"Group sues to stop instant runoff elections in SF"

Bay City News February 5, 2010

SAN FRANCISCO — Six San Francisco residents sued the city in federal court Thursday to challenge its instant runoff voting system.

The lawsuit claims that the way the city implements the system is unconstitutional because some voters are denied the ability to have their vote counted in later rounds of balloting.
http://www.sfexaminer.com/local/Group-sues-to-stop-instant-runoff-elections-in-SF-83607307.html#ixzz0edR8t4WT

Anonymous said...

This lawsuit will be thrown out of court as the courts have already ruled on the issues that are in play here. IRV is totally Constitutional, period.

Also, as per the comments on that article:

"the lead plaintiff is losing candidate Ron Dudum. In spite of the previous comments, the suit has nothing to do with ranked choice voting per se, but rather questions the use of voting machines that limit the voter to fewer rankings than there are candidates."

This is just another case of sour grapes.