In each round of the Ranked Choice Voting (RCV) tally, you’re only able to support a single candidate at a time, just like in Choose One Voting. This means that RCV suffers from a similar vote splitting and spoiler effect that Choose One Voting does, as we saw in the Alaska 2022 Special General Election. Voting methods with vote splitting only work well with two viable candidates. RCV upholds two-party domination for the same mechanical reasons as Choose One Voting.

If you’d like a more in-depth explanation, read on.

RCV does mitigate the spoiler effect, but not in the way you think. The spoiler effect under Choose One Voting causes two ramifications relevant to third parties:

• Third parties have a difficult time winning elections.
• Third parties can leverage the spoiler effect to force major parties to shift on certain issues by threatening to spoil their elections.

RCV strongly mitigates the second ramification without fixing the first.

Third parties spoil close elections when they’re able to convince a handful of voters to vote for their party when those voters otherwise would have voted for a major party. Under RCV, those voters would simply rank the major party candidate second and their votes would transfer after the third-party candidate is eliminated. This removes the leverage of the third party and ensures the “correct” duopoly party candidate is elected. This is why RCV has higher accuracy than plain Choose One Voting but has accuracy comparable to Top Two Runoff (TTR). TTR, in practice, does effectively the same thing as the RCV tally: it eliminates all the candidates who are not viable and then gives the voters a chance to pick directly between the (two) viable candidates without interference from spoiler candidates.

(In the above graph, RCV is called "IRV" and Choose One Voting is called "Plurality". Smith/Minimax is similar to Ranked Robin.)

Some argue this is exactly what they want from a voting method. They like the two-party system and just want to ensure the “correct” major party wins. However, 63% of Americans want a viable third party. So how does RCV fare for that?

Not much different from the current system, as demonstrated by Australia’s lower house. It's still largely two-party dominated after a century of using RCV despite having compulsory voting, which boosts the turnout of moderate voters relative to the rest of the electorate.

When people first learn about RCV, their intuition often tells them that it creates a path for a third-party candidate to win. To understand this, we’ll walk through this intuition and then demonstrate why it doesn’t work.

Many voters vote for a major party candidate not because a duopoly candidate is their favorite, but because they want their vote to help determine who wins. To achieve this, they dishonestly give their vote to a viable candidate instead of their genuine favorite.

RCV would (supposedly) allow these voters to support their honest favorite as their first choice and have their preferred major party candidate as their backup in case their first choice gets eliminated. Either way, they would be helping to prevent the greatest evil from winning. If RCV is implemented, then perhaps enough voters would now feel safe enough to honestly vote for their favorite candidate – a third-party candidate. If enough voters do that, then a third-party candidate can actually win! All that was needed was for voters to feel safe enough to vote their conscience.

Unfortunately, this is not how it plays out.

There are several problems, but the simplest one is that voters who don’t prefer a major party are not a monolith. Even though 43% of Americans now identify as independent while the major parties have fallen to 27% each, it’s important to recognize that those independents are even more diverse than the two major parties are from each other. While the biggest chunk is in the center, the most politically active are on the fringes. The Democrats and Republicans are genuinely the two largest political factions in the US, and independents will never coalesce around a single candidate. Independents on the far left will often rank the Democrat second and the same happens from the far right and the Republican. Those fringe candidates won’t make it to a competitive final round of RCV.

Even a “Great Mediator” wouldn’t do the trick. While they’d be the first choice of a decent chunk of voters in the middle, they’d be the second choice of the two largest chunks of voters: Democrats and Republicans. Since RCV ignores most second choices through the first competitive round, the Great Mediator will be eliminated before either of the major party candidates, despite being preferred over either head-to-head. Those second choices from major party voters are then never counted in favor of the Great Mediator. There have already been multiple American RCV elections where we know this happened: Pierce County, WA, 2008, County Executive; Burlington, VT, 2009, Mayor; Alaska, August 2022, US House. There have likely been others that we don’t have the ballot data for and we also have reason to believe voter and candidate behavior causes this to happen invisibly before elections.

Fortunately, there’s a solution to this problem: count all of the ballot data. If those second choices are always counted, then the third-party candidates actually have a shot at winning. STAR Voting counts all of the ballot data, and if you want to stick with a ranked ballot, Ranked Robin solves this, too.