General Question

LostInParadise's avatar

Is instant runoff voting a good idea?

Asked by LostInParadise (32161points) June 20th, 2010

The problem with having the winner of an election being the person with the most votes is that it discourages people from voting for third party candidates. People feel that voting for a third party candidate is wasting their vote, since the candidate has no chance of winning. Another problem is that a third party candidate can act as a spoiler in a close election, siphoning the votes from one candidate, who would otherwise have the most votes.

Instant runoff voting addresses that problem. The idea is to approximate having a runoff vote among leading candidates in the case that no candidate got a majority of votes cast. See Instant runoff example to see how it works.

Instant runoff is used in Australia and I would be interested in the views of any Aussie voters here. It is also used in San Francisco local elections. It could be implemented in the U.S. in national elections without any federal legislation. Each state is allowed to choose how it votes for senators and congressmen and states can also choose how they distribute their electoral votes in a presidential contest.

I think it is a good idea, but I am not sure. Suppose we used it just for presidential elections and a third party candidate got elected. That would be awful, because the president would have zero support in Congress. If we used it for both congressional and presidential elections then there would be the possibility of having the mess of many candidates from different parties. This may not be so bad, and I am not sure how likely that this would be.

In the U.S. there is also the problem of adding complexity to a vote counting process that already has problems, but I do not see this as an insurmountable obstacle.

Observing members: 0 Composing members: 0

14 Answers

Nullo's avatar

Certainly looks interesting. I daresay that it would resolve concerns over voting fraud.

benjaminlevi's avatar

I think it is a good idea, but since it helps out 3rd party candidates I doubt anyone in the US two-party system would support it.

shpadoinkle_sue's avatar

I think it’d revive the 3rd party system. It seems now that it’s only about taking votes away from the other 2 parties. Why not try it?

Maximillian's avatar

I looked at the example, and I love it. But yes, it does pose the problem of the two major parties saying “NO”. But I love it.

Kraigmo's avatar

There’s no logical reason to not have instant runoff voting.

Zaku's avatar

Yes it’s a good idea.

Dr_Lawrence's avatar

As long as the USA clings to its rigid two-party system, many worthwhile ideas are effectively locked out of the public discussion. Choosing between highly polarized and inflexible points of view has been abandoned in most older and more mature democratic countries.

Runoff voting certainly has its merit but will have minimal application so long as politics is a forced choice between inflexible extreme views that are passed down between generations with little tolerance for children to think for themselves. Crossing traditional voting customs is sometimes enough to get family members marginalized or shunned in America.

mudlock's avatar

No, it is not a good idea.

Firstly, IRV does not eliminate the spoiler problem, it only reduces it. With three candidates, if each one gets at least 25% of the votes, there can still be a spoiler. Any third-party today would be ecstatic over a 25% vote-share, but the moment they reach that plateau, if they spoil an election, voters will still turn on them, and they still won’t win; we’ll end up with two parties, just like we have now.

You mention Australia; look at the composition of their House of Representatives, where they use IRV. There are still only two parties represented (well, one party, and one 60-year old coalition who never challenge the other coalition-member’s incumbents; close enough.) Meanwhile, the Australian Senate, which does not use IRV, is composed 9% of third-parties. (The Senate uses single transferable vote, a form of proportional representation, which IRV is a bastardization of.)

Sure, third-parties have won under IRV. But look at the offices they’ve won: they’re all ones that the same third-party had won previously under the old plurality method of voting! (San Francisco city council, Green party; mayor of Burlington, VT, Vermont Progressive party.) IRV doesn’t help third parties any more than plurality does; and while that’s good enough to win an election in a few places across the country (usually ones where one of the two major parties is terribly out-numbered), it’s clearly not good enough for most of the country.

Furthermore, we have a two-party system not because of some plutocratic conspiracy or voter brainwashing, as others have suggested, but because the voting system we use cannot adequately distinguish among more than two options. The tendency for single-winner plurality elections to lead to two-party systems is known as Duverger’s Law. If we change the system though, we can escape from this trap. But instant runoff voting is certainly not the way out!

There are voting systems that can properly distinguish among three options: approval voting and score voting (also known as range voting). If you’re interested in real electoral reform, look in to them.

LostInParadise's avatar

@mudlock , Welcome to Fluther! You have apparently looked into this. I will definitely check out approval voting and score voting.

You will have to explain to me how a third party can still be a spoiler. Suppose three candidates get 40%, 33% and 27% of the vote. The candidate with 27% is eliminated after the first round and the votes are distributed to the other two, one of whom will now have a majority. I do not see how anyone would complain about that.

Kraigmo's avatar

@mudlock , Wouldn’t the combination of Instant Runoff Voting and Proportional Representation address everything you mentioned?

By proportional representation, I mean for example if Greens and Libertarians get 5% each, that means they get 5% each of the seats in the legislature, etc.

I guess you could argue that proportional representation would make instant runoff voting moot, but I think the combination of both would have the strongest positive effect.

mudlock's avatar

@LostInParadise:

You have to look at a voter’s entire ballot, not just their first-choice preferences, to know what they will consider a spoiler. It also helps to formally define what a spoiler is: A spoiler is a candidate whose presence changes the outcome of the election, but not so that they win the election; in other words, them being on the ballot ruins the election for someone else. Using your example, and expanding on it:

40% want candidate A, and then candidate C
33% want candidate B, and then candidate C
27% want candidate C, and their second choice is evenly split between A and B.

Under IRV, A wins this election. But if B wasn’t in the election, C would have won instead. So B is a spoiler for C. (Meanwhile, A is not a spoiler: if they weren’t running, C would win, but when A is present, A wins; that’s not spoiling, that’s winning! Neither is C a spoiler: without C, A wins; with C, A still wins.)

This particular failure of IRV is known as “center squeeze”, where a moderate, centrist candidate, who could defeat either of the other candidates in a one-on-one contest, is eliminated first, and the final round of the election is between the two extremes.

@Kraigmo

Interestingly, IRV was developed from a proportional representation (PR) method, called single transferable vote (STV). STV uses the same ranked-order ballots, and the same elimination rules (with different thresholds), but elects multiple winners from each election. IRV is simply STV that elects only one winner.

And STV isn’t that bad; it’s what’s used for the Australian Senate, about 9% of which is members of third parties. It still tends to amplify small majorities; so the two major parties control a number of seats in excess of what their true popular support is; a problem which becomes more severe the fewer seats are elected from each election, until, at “one” (i.e., at “IRV”), all proportionality is lost.

But I mentioned score voting (also known as range voting) earlier: there is a PR method based on score voting, called re-weighted range voting (RRV). There hasn’t been a lot of research done on it, but considering the effectiveness of score voting compared to IRV for single-winner elections, it’s possible that RRV could be a better PR system than STV.

For further information, let me refer you to some websites:

The Center for Range Voting a group advocating these methods
The Least of All Evils my blog on this issue

LostInParadise's avatar

Interesting, but I wonder how realistic these scenarios are. You say that B is a centrist candidate, where presumably A and C represent extremes. If that is the case then it does not seem plausible that the 40% who favor A would have C come second.

I know these types of analyses, even with just 3 candidates, can get rather involved. One of the problems is with cases where “strategic voting” is involved, meaning that people will vote for someone that they do not prefer in order to prevent someone else from getting elected. This happens all the time in the usual voting system. I don’t know how often it turns up in practice with other voting systems.

mudlock's avatar

@LostInParadise

No, C is the moderate. Sorry, I should have specified that. Does this make more sense now?

As for “how often does this happen,” yes, those answers get kind of complicated. And I can field all sorts of arguments about voting system criteria and expected incidence rates of certain failures under various voter models… But there’s a deeper question behind your question; allow me to answer it instead. What electoral system results in the greatest societal benefit?

According to computer simulations, under all conceivable conditions, it’s score voting; and by a surprisingly impressive margin.

Rather than repeat myself, I refer you to this entry I wrote just last month, What do you mean by ‘Best’?

LostInParadise's avatar

Thanks for the clarification. You have given me quite a bit to look at and think about.

For something as important as voting you would think that there would be some international commission to investigate all the various voting procedures and to make a selection based on a set of criteria.

Answer this question

Login

or

Join

to answer.

This question is in the General Section. Responses must be helpful and on-topic.

Your answer will be saved while you login or join.

Have a question? Ask Fluther!

What do you know more about?
or
Knowledge Networking @ Fluther