Adam Tarr explains why IRV leads to 2-party domination

(Executive Summary)   (oversimplified argument)   (simpler examples #1 and #2)   (Return to main page)

#voters Their Vote
45 Worst>Middling>Best
10 Middling>Worst>Best
10 Middling>Best>Worst
35 Best>Middling>Worst

In this 100-voter example election, under IRV, "Worst" wins. ("Middling," with the fewest top-rank votes, is eliminated in the first round then Worst wins 55 to 45 over "Best".) This is a fairly realistic scenario of the sort that arises with voters and candidates positioned along a one-dimensional line.

Now suppose eight voters in the bottom faction cleverly (but dishonestly) switch their top-two preference order. Result:

#voters Their (new) Vote
45 Worst>Middling>Best
10 Middling>Worst>Best
18 Middling>Best>Worst
27 Best>Middling>Worst

Then the IRV election proceeds differently: "Best" is eliminated in the first round, and then in the second round "Middling" wins over "Worst" 73-to-27. From those 8 voters' point of view, this was an improvement.

The moral is: Dishonest voting ("exaggeration") pays. These 8 voters of course regarded the candidates to be ordered Best>Middling>Worst (the candidate-names have been chosen to reflect their point of view). But they realized that honesty in their vote wasn't going to work: "Worst" was going to win. They saw that "Best" had no chance of winning. So then they "betrayed" Best by intentionally ranking him second and dishonestly ranking "Middling" top. This tactical move allowed the "lesser evil" Middling to win instead of Worst. (Incidentally, they also could have changed their vote to the even-more-dishonest "Middling>Worst>Best"; that also would have been a tactical success.)

The phenomenon in this example goes a long way toward explaining why it is that all three IRV countries are (and historically have been) 2-party dominated. Voters figure the third party has no chance and they are best off exaggerating their view of the top two parties so as not to "waste their vote". As you can see both from this and the other example, in IRV that strategy often works. But if enough voters act this way, then third party candidates can never win an IRV election since they'll be eliminated in early rounds!

(Incidentally, there is another example which works for both IRV and the "BTR-IRV" system. So please do not expect BTR-IRV is going to save you from all this. Another (simpler) example. And another. And another.)

This example directly contradicts the false claim commonly made by proponents of IRV voting that IRV "eliminates the problem of wasted votes." Some third parties such as the USA's Green Party have actually endorsed IRV, not realizing that was a suicidal move which would simply preserve the present USA state of 2-party domination. We hope they instead choose a course aiming toward survival, such as adopting Range Voting.

Still not convinced?

Maybe you object that this example sounds rare and contrived and unlikely to occur in practice. (If so, you are wrong. And in fact a betrayal-pays scenario like this recently happened in Peru.) But that does not matter much. What matters in the eyes of these 8 voters is: Which is more likely?

  1. That the third-party candidate "Best" is going to be elected, but if we betray him by making this strategic dishonest voting move, we'll foolishly prevent that.
  2. That our making this move will successfully cause "Middling" to win.
See, our 8 voters don't know what the other voters are going to do. They cannot see the whole table. They can only guess roughly. So they cannot be sure which vote is the best – the honest one, or the altered one. So they have to assess the chances. In the USA, third party candidates win less than 1% of the time, historically. As I write this in 2005, there is exactly one member of the 435-strong US House of Representatives who is non-Democrat and non-Republican. (Bernie Sanders.) In the 150-member Australian house (elected by IRV!) during 2004 & 2007, the total number of House seats owned by third parties was zero out of 300. So it is reasonable for our 8 voters to believe the chances of the third-party candidate "Best" winning are very small. But the chances of "Middling" winning are quite large, historically about 50%. Now factor in the probability that a situation like the one in the table (where exaggeration pays) is going to arise. Reasonably probable – situation illustrated, as we said, is pretty realistic for 1-dimensional "left" to "right" approximate political situation. Conclusion: Even if situations like the one we illustrated are so rare they only happen 5% of the time, then it still is strategically the right move always, when chances are considered, to exaggerate in your IRV vote.

So the rarity objection won't cut it. As they say down South, "that dog won't hunt."

Logic isn't enough for you? Try facts.

Maybe you think I'm a dangerous mathematician type who can wave his magic arguments to convince you of anything, but you aren't going to buy it. Nope, logical arguments aren't going to be enough for you.

Fine. Then forget logical arguments and take a look at IRV reality. No theory – just the facts ma'am.

The facts are there are three IRV countries: Ireland (mandated in their 1937 constitution), Australia (adopted STV in the early 1900s, but in 1949 added "reweighting" to STV in their multi-winner elections, a change which does not matter for us since we are only considering single-winner elections – Australia and Ireland have both kinds of elections), and Malta. (Later note: a recent addition is Fiji, but it unfortunately then got subtracted due to a 2006 coup.)

All three became 2-party dominated in their IRV seats. And this is despite the fact that in addition to IRV single-winner elections, they all also have multi-winner STV "proportional representation" (PR) elections, and they are parliamentary rather than presidential. Both of these two factors mitigate toward having more than 2 parties (the parliamentary countries with PR essentially all have many more than 2 vibrant political parties). But despite those multiparty-genic factors, the effect of IRV in these countries has been enough to drag them back down to 2-party domination status! So given that, you can bet your bottom dollar that the USA, were it to adopt IRV but still to remain presidential and without multiwinner PR elections (i.e. wholy with single-winner elections), will definitely stay 2-party dominated.

Third parties, are you listening? Hello?

We sure hope so. Because it is your survival at stake. Get it right. Do not advocate IRV like a suicidal idiot.

Return to main page