Comment Re:Well when you put it that way... (Score 1) 1312
My numbers aren't THAT contrived, just because they are specific. When you imagine there being three viable parties, roughly how much of the vote do you imagine each one having? There's actually quite a bit of room where the effect I showed happens.
IRV chokes when there are two parties with roughly equal sizes, and one of the parties (G) 'feeds' heavily into another party (D), but not so much the other way around. In this easily imagined setup a small point swing decides whether G or D is knocked out first, and that causes either a cascade of votes into D, or all the votes fleeing both G and D because G cannot get enough votes from (D) to stay alive. The G people think their votes will feed harmlessly into D, but this can't happen if G knocks out D first, and in fact the G people are in for a rude shock: they just screwed themselves out of their second choice.
You say: In both cases, the candidate that the majority does not like will not take office.
This isn't really true because no one had a majority among the first choices. If you're just saying that the majority of LAST choices didn't win in my examples (meaning G), that's not saying much, and not always true in IRV, and how do you know what the last choices are in standard voting?
Sure, it looks ugly when someone wins without a majority of the vote, but I think my example of the 2% shift from G D R to D R G producing a 36-point swing from R to D shows that the final tally in IRV is bullshit anyway.
In your supposed counter-example, your 2% swing produces a 2% swing. I don't think you understand. I'm not complaining that it's possible for 2% to swing the election; I'm pointing out that in a standard vote it's easy to know what you need to do (assuming the public knows generally who is more popular and there is time to campaign), whereas IRV is LESS representative of your preferences because it tricks you into thinking it will represent them accurately.
People who voted G D R got their least favorite candidate because they ranked their favorite first; the problem isn't that R wins in a close count, the problem is that a glitch in IRV is chosing R over D, when the relative ranking of D and R didn't change in those 2% anyway.
Also consider this: if that 2% had changed from G D R to R G D (having some epiphany that R was the one true leader) then the vote looks like this:
31%: G, D, R ("We lost 2%? Pshaw!")
32%: D, R, G ("We've been united for clarity!")
2% : R, G, D ("We're with you, R!")
35%: R, D, G ("No! Get away get away!")
Now instead of D being eliminated and R winning as in the first example, G is eliminated and D wins.
If that 2% REALLY wanted R to win, they shouldn't have placed him first like that. What fools! Don't they realize that sometimes in IRV that putting someone as your first choice can make them lose?
The straight up vote system has the virtue that most people can understand that you don't back the loser, and the benefit of voting strategically is apparent to everyone and so very many people DO vote strategically. Casting a 3rd party vote in the current system is effectively saying "yes I know my guy can't win, but I actually don't have a preference between the other two."
(You can even argue that it helps third parties by giving them more leverage and power against a more mainstream party because the votes really are being taken away; in an IRV where G is only 11% of the vote and D knows that it can get almost all of G's second votes -- it pays no attention to G whatsoever.)
Try this thought on for size: IRV makes primaries obsolete
I seriously doubt it. The party that concentrates their efforts on one candidate has a huge campaigning advantage and is likely to win. Anyway it could easily still boil down to a 3-way race at some point between two R-ish people and a D, say, and you're back in wacky land.
The current voting system is like a car without any doors. Yes, it's a piece of crap car and you have to be careful not to lean too far over when you drive fast, but when you hear passengers say "I'm going to act like this car has doors no matter what" it's hard to sympathize with them when they fall out.
IRV by contrast is like a car with doors that appear sturdy so you think you can lean on them but in fact sometimes they swing open at random times and dump you out on the street, so maybe you shouldn't be leaning against them, but in fact it's even possible you just get sucked out of them for just sitting there (i.e. voting for your favorite).
Which one would you rather have? Maybe IRV is okay for relatively unimportant elections--i.e. driving slow, when it's no big deal if you fall out.
IRV chokes when there are two parties with roughly equal sizes, and one of the parties (G) 'feeds' heavily into another party (D), but not so much the other way around. In this easily imagined setup a small point swing decides whether G or D is knocked out first, and that causes either a cascade of votes into D, or all the votes fleeing both G and D because G cannot get enough votes from (D) to stay alive. The G people think their votes will feed harmlessly into D, but this can't happen if G knocks out D first, and in fact the G people are in for a rude shock: they just screwed themselves out of their second choice.
You say: In both cases, the candidate that the majority does not like will not take office.
This isn't really true because no one had a majority among the first choices. If you're just saying that the majority of LAST choices didn't win in my examples (meaning G), that's not saying much, and not always true in IRV, and how do you know what the last choices are in standard voting?
Sure, it looks ugly when someone wins without a majority of the vote, but I think my example of the 2% shift from G D R to D R G producing a 36-point swing from R to D shows that the final tally in IRV is bullshit anyway.
In your supposed counter-example, your 2% swing produces a 2% swing. I don't think you understand. I'm not complaining that it's possible for 2% to swing the election; I'm pointing out that in a standard vote it's easy to know what you need to do (assuming the public knows generally who is more popular and there is time to campaign), whereas IRV is LESS representative of your preferences because it tricks you into thinking it will represent them accurately.
People who voted G D R got their least favorite candidate because they ranked their favorite first; the problem isn't that R wins in a close count, the problem is that a glitch in IRV is chosing R over D, when the relative ranking of D and R didn't change in those 2% anyway.
Also consider this: if that 2% had changed from G D R to R G D (having some epiphany that R was the one true leader) then the vote looks like this:
31%: G, D, R ("We lost 2%? Pshaw!")
32%: D, R, G ("We've been united for clarity!")
2% : R, G, D ("We're with you, R!")
35%: R, D, G ("No! Get away get away!")
Now instead of D being eliminated and R winning as in the first example, G is eliminated and D wins.
If that 2% REALLY wanted R to win, they shouldn't have placed him first like that. What fools! Don't they realize that sometimes in IRV that putting someone as your first choice can make them lose?
The straight up vote system has the virtue that most people can understand that you don't back the loser, and the benefit of voting strategically is apparent to everyone and so very many people DO vote strategically. Casting a 3rd party vote in the current system is effectively saying "yes I know my guy can't win, but I actually don't have a preference between the other two."
(You can even argue that it helps third parties by giving them more leverage and power against a more mainstream party because the votes really are being taken away; in an IRV where G is only 11% of the vote and D knows that it can get almost all of G's second votes -- it pays no attention to G whatsoever.)
Try this thought on for size: IRV makes primaries obsolete
I seriously doubt it. The party that concentrates their efforts on one candidate has a huge campaigning advantage and is likely to win. Anyway it could easily still boil down to a 3-way race at some point between two R-ish people and a D, say, and you're back in wacky land.
The current voting system is like a car without any doors. Yes, it's a piece of crap car and you have to be careful not to lean too far over when you drive fast, but when you hear passengers say "I'm going to act like this car has doors no matter what" it's hard to sympathize with them when they fall out.
IRV by contrast is like a car with doors that appear sturdy so you think you can lean on them but in fact sometimes they swing open at random times and dump you out on the street, so maybe you shouldn't be leaning against them, but in fact it's even possible you just get sucked out of them for just sitting there (i.e. voting for your favorite).
Which one would you rather have? Maybe IRV is okay for relatively unimportant elections--i.e. driving slow, when it's no big deal if you fall out.
