Common Myths About Ranked Choice Voting Debunked
Recently, I wanted to link to a concise summary of the most popular RCV myths, and was surprised to find that there really isn’t a good one. So here’s my attempt at addressing all of the most common false claims in one place.
Wait, what’s “Ranked Choice Voting”?
To be clear, the system I’m talking about here has many different names, such as Ranked Choice Voting (US), Instant-Runoff Voting (US), the Alternative Vote (UK), and Preferential Voting (Australia).
All of these common names are ambiguous, and could just as correctly be used for other voting systems like Contingent Vote, Supplementary Vote, Benham’s method, BTR-IRV, etc. All of these alternative voting systems feature instant-runoff rounds based on ranked-choice / preferential ballots, and the name “Alternative Vote” could apply to literally anything other than our current system. The term “Ranked Choice” has also been used for other voting systems that use ranked ballots without the instant runoff, such as Schulze Method or Borda Count.
Less ambiguous names include Hare RCV and Ware’s Method, though both could also apply to multi-winner Single Transferable Vote (STV). In this article, I’m only talking about the single-winner version that’s being pushed heavily in the US.
The myths
Myth: Ranked Choice Voting fixes the spoiler effect.
Reality: The spoiler effect occurs when a candidate enters a race, and does not win, but their presence in the race changes the outcome (typically helping a candidate win who is less favorable, from their perspective). This happens under our current First Past the Post (FPTP) system when, for example, a third-party candidate enters the race, taking votes away from the mainstream candidate who is more similar to them ideologically, causing the less-similar candidate to win instead.
RCV is good at preventing this problem when the third party candidate is weak (which is why this is the only scenario that RCV advocates ever mention in their marketing). It does not fix the problem in general, however. RCV only counts first-choice votes in each round, just like FPTP, so a strong third-party candidate entering the race can take first-choice votes away from the mainstream candidate more similar to themselves, causing the “lesser of two evils” to be eliminated first. Then the “greater of two evils” beats the third party candidate in the next round and wins. If the third party had stayed out of the election, the “lesser of two evils” would have won, which is a better outcome from the third party’s perspective.
(This is not unique to FPTP and RCV. It’s a fundamental flaw that affects any voting method based on counting only first-choice votes, including FPTP, Supplementary Vote, Contingent Vote, Exhaustive Ballot, Two-Round System, Hare RCV, Top-Two Primaries, Top-Four Primaries, and Final Five Voting.)
A real-world example is the 2022 special election in Alaska. Sarah Palin acted as a spoiler, taking enough votes away from the more moderate Begich that he was eliminated first, despite a 52% majority of voters ranking Begich higher than Peltola, and 61% preferring Begich over Palin. If Palin had strategically dropped out, or if Palin voters had strategically ranked Begich first, then the seat would have gone to a Republican, which would be a better outcome for her supporters. Hare RCV suffered from exactly the same vote-splitting flaws as FPTP, and produced exactly the same outcome.
Myth: Ranked Choice Voting makes it safe to vote honestly for your true favorite, without worrying about wasting your vote.
Reality: Unfortunately, Dennis the Election Koala is wrong. You absolutely can waste your vote, whether you’re using RCV in Australia or America. It is not safe to vote honestly for the Nice Party; and you may be better off voting tactically for the Discreet Scumbags to avoid throwing your vote away. Unfortunately I am not a skilled artist, so I can only debunk this one with text:
As described above, whenever there are three or more strong candidates, the spoiler effect rears its ugly head. If fans of a strong third party candidate vote honestly for their favorite, they can cause their second favorite to be eliminated first, helping a worse candidate get elected and hurting their cause. In these cases, it would be better for them to vote tactically for the “lesser of two evils” and help them win, just like under FPTP.
Under RCV, it’s only safe to vote honestly when your true favorite is one of two strong candidates, or when they don’t stand a chance of winning at all, making it little more than a protest vote.
Better voting methods actually do make it safer to vote honestly for your true favorite, without worrying about throwing your vote away, which increases the viability of third parties and independents.
Myth: Under RCV, if your favorite candidate doesn’t win, your vote will transfer to your second favorite.
Reality: Again, this is only true in certain contrived scenarios used to market RCV, not in the general case. Since RCV only counts first-preference votes in each round, honestly giving your first-choice ranking to your favorite can take away enough votes from your second favorite that they are eliminated first. When your favorite is then eliminated, your vote cannot transfer to your second favorite, as they are no longer in the race. Your preference for them is never counted.
I’ve talked to people who are very confused about this point, and insist that their second choice must be resurrected somehow and their 2nd place votes transferred to them, but that’s not how Hare RCV works.
Myth: Ranked Choice Voting guarantees that the winning candidate has majority support
Reality: This “majority” is fabricated by eliminating candidates, and isn’t meaningful. For example, let’s say there are 5 candidates, with approval ratings like so:
- Alice: 98%
- Bob: 95%
- Carol: 90%
- Dave: 20%
- Egbert: 15%
Because Hare RCV suffers from vote-splitting when there are multiple good candidates, it’s possible for the instant-runoff rounds to first eliminate Alice, then Bob, then Carol, until only Dave and Egbert are left.
In this final round, one candidate is guaranteed to have a majority of support over the other, since there are only two candidates left. Does that mean that Dave is actually the favorite of the voters? No. The voters would likely have preferred any of Alice, Bob, or Carol over Dave. Under RCV, candidate Y can win the election even when a supermajority of voters preferred candidate X over Y.
Eliminating candidates to manufacture a majority is about as meaningful as an evil dictator poisoning all their opponents before the election and then claiming they were voted into office with unanimous support, because they are the only candidate left on the ballot.
This “majority support” claim is false in another way, too:
If voters don’t fill in all of their preferences on the ballots, and all the candidates they do rank are eliminated, then their ballot is now “exhausted” and has no effect on the remaining instant-runoff rounds. So even if candidate X gets a majority of first-choice votes over candidates Y and Z in the final round of an election, those votes may actually only represent a minority of voters.
Myth: RCV is better than FPTP because it gathers more information about voter preferences.
Reality: Ranked ballots allow you to express your preferences between every candidate. This is better than FPTP, which only allows you to say which candidate is your favorite, and express nothing about the rest, right?
Unfortunately, although Hare RCV does gather more information, it then throws much of it in the garbage; many of the preferences voters express on their ballots have no effect on the outcome of the election.
For instance, let’s say you and a friend prefer candidate A over B, and dislike candidate C the most, while two other voters have the opposite preference (C > B > A), and a fifth voter fills out their ballot with B > C > A. These are the five ballots cast:
- A > B > C
- A > B > C
- C > B > A
- C > B > A
- B > C > A
In this scenario, B gets the fewest first-place votes and is eliminated first. That voter’s 2nd choice C then takes priority, and C, your least favorite candidate, wins the election, with 3 out of 5 remaining votes.
But B was preferred over C by a 3/5ths majority of voters, so why didn’t B win? Voter 3 and 4’s preference for C > B was counted while eliminating B, but your opposite preference for B > C was never counted. You could have swapped that preference on your ballot (voting A > C > B) and it would make no difference to the outcome of the election.
Under Hare RCV, some voters’ expressed preferences affect the outcome of the election, while other voters’ preferences do not. This is inherently undemocratic and unfair.
Myth: Ranked Choice Voting is more likely to elect moderate candidates with broader appeal
Reality: Ranked Choice Voting suffers from the center-squeeze effect, which means it is biased against moderates, and in favor of polarizing candidates.
The center-squeeze effect was first noted by Samuel Merrill in 1984. It occurs whenever there are three or more candidates spaced across some kind of left-right political spectrum or multi-dimensional ideological space, while the voting system counts only first-choice preferences in each round (as in FPTP, Hare RCV, Contingent Vote, etc.)
The candidate near the center of the voter ideology is preferred by majorities of voters over every other candidate and could easily beat any of them head-to-head, has the highest overall approval rating, is the best representative of the ideology of the average voter, and by all accounts should win. Yet, because there are competing candidates to either side, the first-choice votes of the outer wings of voters go to those partisan candidates instead, while the moderate only gets first-choice votes from a small sliver of voters in the middle, causing the most-representative candidate to be eliminated first. So these systems are actually biased against moderate candidates, and more likely to elect more polarizing candidates to either side.
Other ranked voting systems count all preferences simultaneously, rather than only counting first-choice preferences, or measure voter approval directly, which eliminates this bias against moderate candidates.
Beyond Ranked Choice Voting
As mentioned above, there are many different voting systems (dozens, if not hundreds) and these problems are not universal to all of them. No voting system is perfect, but some are much better than others. (I prefer STAR Voting, Condorcet RCV, Approval+Runoff, and others.) Please research the alternatives and don’t just go with RCV because it has the biggest marketing push behind it.
Here are some other listings of RCV myths that I’ve found:
- Rebutting false and misleading testimony from RCV advocates
- Responding to the FairVote presentation on May 18th from Diane Silver
- Claims made by T.G.Bouricius and FairVote
- Refutation of False Statements about Instant Runoff Voting (IRV)
- The Limits of Ranked-Choice Voting
[I intend to update and expand this over time, but I was being too perfectionist about it and decided to use RCV Day to motivate myself to publish what I have so far. The clickbait title and subtitle were written by ChatGPT; don’t blame me.]