IRV's Achilles Heel

Instant Runoff Voting, or IRV, is a good election method. By allowing voters the chance to rank their preferences it can do what a primary or runoff does, all in one election. I support IRV because it supports majority rule, and democracy will be stronger for it.

However I must make it clear that IRV takes a small step further, one which is not clearly justified by majority rule, and in my mind may become the undoing of this progressive movement which I support.

IRV uses an elimination method which can be called "bottom up" which means every round one candidate with the least votes is eliminated. So if there's 10 candidates, there can be up to 9 rounds of elimination and recounting.

In the end of this LITERALLY convoluted process, TWO candidates will survive and will EARN the chance to compete head-to-head against each other in the final round.

What IRV supporters don't tell voters is that IRV, like all runoffs, is a "cut throat" game of power. It is a game of KING OF THE HILL where a large set of candidates have to FIGHT each other "on the bottom" to earn the right to compete in the final round. Voters have TWO possibly contradictory goals – they want someone to make the final round AND they want someone who can WIN the final round. Unfortunately IRV and all runoffs don’t treat candidates equally. They only treat the “final two” equally, even if a different candidate might beat either of them head-to-head.

I consider bottom-up elimination as "mostly harmless". It is "democratic" in the sense of "maximizing the chance for the maximum number of candidates" to reach the final round. However this "maximization" doesn't necessarily change the power of voters to influence the result.

In short, I will claim that "majority rule" as a principle does NOT support such an extended runoff process. If a majority of the voters are supporting the top two candidates, IRV will still fiddle along meekly with weaker candidates, possibly "against the will" of a majority of voters for stronger candidates.

Right now Minneapolis has a "top two primary". All the candidates run as equals and the strongest two candidates move forward to the general election for a majority winner. Simple, clear. If you're a candidate, make the top-two or perish.

Now I wonder what will people say, what will candidates say, of IRV if it DISALLOWS a second place candidate a legitimate claimed right to compete head-to-head against the top (plurality) candidate?

The reason this issue is usually hidden, is that it requires four candidates and 3 strong candidates to make a actual difference, but this is not necessarily unusual.

If I'm allowed to project a party example for clarity, consider this imagined IRV election:
Round 1: Rep=34%, Ind=30%, DFL=28%, Green=8%
Eliminate Green
Round 2: Rep=34%, DFL=34%, Ind=32%
Eliminate Ind
Round 3: Rep=56%, DFL=42%
Winner republican with 56%

A nice IRV election, a clear majority winner, right? Well at least we KNOW the republican beat the democrat.

However the Ind. candidate, being in second place in round 1, never got a chance to compete head-to-head against EITHER competitor. SURE, she was THIRD in the second round, but in a top-two runoff she would have made the threshold for viability. Why are we discounting this previously sufficient measure of success?

Would it surprise you that a top-two election might provide the Independence party candidate as the majority winner over the republican?
Round 1: Rep=34%, Ind=30%, DFL=28%, Green=8%
Eliminate Green and DFL
Round 2: Ind=54%, Rep=44%
Winner Independence with 54%

It wouldn't surprise me, and these numbers come from real ballots, even if a different election. I took a rank ballot polling election I held for "Favorite season" with 50 voters. I merely remapped the four seasons onto four parties for this example. ( I'll let you guess which seasons they each represent, and only hint that greens are most valued in the winter! )

Are there spoilers here? It depends on WHO you’re asking. There’s TWO different majority winners depending on who makes the final round!

Maybe in a happy cooperative progressive city like Minneapolis, bottom-up IRV is agreeable, that a second place candidate will contentedly allow herself to be eliminated after a top-two showing WITHOUT a "fair" shot at a head-to-head competition. Maybe.

I mean, SURE, I'm a nice guy. I know politics is a tough place, and nothing is fair to everyone. Someone has to give, and I'll play Al Gore and back down honorably when my time comes, sure I will!

However, even if that is true in progressive Minneapolis I expect that if IRV is implemented in as a bottom-up runoff in a partisan election, SOMEDAY a nice 3-way race like this WILL happen, and courts will be involved by the injured party, and IRV will be rejected, all because it "took that extra step" that was not called for by majority rule.

And sadly we'll be back with plurality again, AS IF, it was the best we could do.

In conclusion, I support IRV, but I believe that as long as we disallow THREE candidates the right to a head-to-head competition against each other (all runoff methods disallow any more that this), a top-two runoff is the best we ought to strive for, instant or not.

It's "closer to what we have", and a more conservative step for change. I takes sides in favor of the candidates supported by the most voters in the first round. It tells candidates "Make top-two or Perish". It says clearly to voters "Vote for someone who can make the top-two OR vote for someone who can WIN the final round."

I don't mind if people CHOOSE "bottom-up IRV". I merely want to make sure people KNOW WHAT WE WANT, and don't take this version of IRV as-is simply because we didn't realize there was a choice and a simpler two round option equally available for consideration.

It's hard to easily explain this difference, and it is "small", perhaps only affected a few percent of elections, but ought to be considered BEFORE we face them.