by Clay Shentrup and Warren D. Smith

Many people believe that Instant Runoff Voting allows voters to "vote their hopes, not their fears", because it takes away the incentive to vote strategically instead of sincerely. It is not hard to understand why this notion is popular, considering statements like this one:

"A voter's best strategy [with IRV] is to sincerely rank the candidates."
From an article by FairVote founder Rob Richie, with Terrill Bouricius & Philip J. Macklin in Science Magazine 294 (May 2001) 303-306

This statement is simply false. This is one of the most basic and well known facts about IRV, and Richie et al are simply telling a flat-out lie. Not only in IRV, but in fact in every ranked-ballot voting system (in which unanimously top-ranked candidates win and in which there is not always a "dictator" voter who can singlehandedly decide everything), in at least some elections, it is not a voter's best strategy to sincerely rank the candidates. This is called the Gibbard-Satterthwaite theorem and is one of the most famous and well known (if not the most famous and well known) theorems in all of voting theory. (So FairVote's ignorance of this theorem is rather instructive about their level of expertise.) 

Why "lie"?

With Instant Runoff Voting, supporters of anyone other than the (typically two) apparent front-runners have to decide between sincerely top-ranking their true favorites, or insincerely top-ranking their "lesser evil" candidate. To determine which is the better option, those voters have to think which is more likely:

1. That sincerely top-ranking their favorite candidate (who is not a front-runner) will help him win.
2. That sincerely top-ranking their favorite candidate (who is not a front-runner) will lead to a worse result than if they had insincerely top-ranked their favorite front-runner.

Mathematical analysis shows that the latter situation is more probable, and therefore it is (effectively) always a wise IRV strategy to simply betray one's sincere favorite for the preferred front-runner. Here are the two hypothetical outcomes as simple examples of why this happens, using the two major U.S. parties plus a Green. These are the outcomes that minor party supporters must consider in determining the best strategy (we can ignore the other possibilities, where the result is the same whether the Greens vote Green or Democrat).

Scenario #1 - Top-ranking G pays for G's supporters.

% of voters - their vote
28% G > D > R
23% D > G > R
  3% D > R > G <-- Voters that would give D a better chance against R, than G would have against R
46% R > D > G

Scenario #2 - Top-ranking G costs G's supporters, and leaves many of them regretting the decision - because they could have gotten their second choice instead of their last, by strategically top-ranking D.

% of voters - their vote
28% G > D > R <-- Voters who could have gotten a better result by insincerely top-ranking their second favorite
21% D > G > R
  5% D > R > G <-- Voters that would give D a better chance against R, than G would have against R
46% R > D > G

If #1 is more likely then it is strategically best for G's supporters to honestly top-rank G.  But if #2 is more likely, it is strategically best for most of G's supporters to top-rank D (because that is their second choice). So how do we determine which is more likely?  Well, it basically comes down to asking, which will there be more of?

D > R > G voters

or

G > R > D voters

If there are more of the former, then strategically top-ranking D becomes their best bet, in any case where there is even the least bit of preference for D over R.

Don't like logic?  Try facts.

This is reason enough (and there may be others) that IRV has always led to two-party domination where it has seen long-term use.  In Australia they even have "How to Vote" cards, that tell voters exactly how to game the system.  Not surprisingly, third parties in the Australian House of Representatives (where they've used IRV since 1918) garner a tiny fraction of first-place votes.  As I write this (in 2007) their IRV seats are 0.18% minor party.  Those in the U.S. are 0.09% minor party - with plurality voting, and in spite of the fact that we don't have proportional representation, like they do in their Senate.  So it appears that even after decades of use, IRV is so hard on third parties, that they can't win a substantial number of elections with IRV, even when they are healthy in P.R. branches of government, and have that momentum going for them.  This has been the same trend as in Ireland, Malta, and Fiji. It is thus no wonder that Australian political analysts at AustralianPolitics.com say that IRV "promotes a two-party system to the detriment of minor parties and independents". The Libertarian Reform Caucus has called IRV a "bullet in the foot".