Ranked Choice has a Monotonicity problem. i.e. it’s possible for a candidate to lose if a more people rank that candidate higher on their own ballot without changing any other ballots.
This has happened in recent RCV elections, and resulted in the candidate’s ideological opposition winning.
There’s a group called FairVote that’s been pushing RCV since the early 90s despite the many flaws of the system. Flaws that have been known since the system was first designed in 1788.
Seriously, Instant Runoff Voting was invented by the Marquis de Condorcet in 1788 as an example of a broken election system that can eliminate candidates preferred by a majority of voters.
It was later reinvented in the late 1850s by an Englishman who presumably never learned French.
Anyway a modern voting system for consideration is STAR, it was developed in 2014 by people who have read Condorcet, the the works of Kenneth Arrow from the 1950s. (Arrow’s Impossibility Theorium)
STAR and Ranked Robin both are a form of RCV. Most people are familiar with RCV so I feel it’s easier to sell the whole concept under the RCV label and then just directing people to the Equal Vote Coalition, like you did, which does a great job of explaining the nuances.
Seriously though, STAR or Ranked Robin are the best choices if we want more progressive wins overall.
Regular RCV is weighted more towards center with a higher chance of the least liked candidate winning compared to STAR or Ranked Robin although less than our current system.
Even RCV is still leagues better than what we currently have and it still shows the actual support numbers for progressive politicians and policies. However, if we’re making a final choice, I’d be happy with STAR = Ranked Robin > Score Voting > Ranked Choice Voting >>>>>>>> First Past the Post (The current system in most of the US).
RCV, or IRV as it’s known elsewhere in the world, is an Ordinal voting system. That means spoiler effect and enforced two party dominance.
STAR and Ranked Robin are on the Cardinal side of things. No spoiler effect, because votes you can give equal support to multiple candidates.
If I really like two different candidates and don’t particularly care which one wins, I can say that on a STAR or Ranked Robin ballot.
In RCV I have to make a strategic choice, and if I get that choice wrong, it’s not just my guy who loses, but possibly my entire side.
So no, they are not the same. And it’s worse because of RCV’s lack of Monotonicity.
Non-negative responsiveness or monotonicity is a property of a social choice rule, which says that increasing a candidate’s rank on some ballots should not cause them to lose (or vice versa, that decreasing a candidate’s rank should not cause them to win).[1] This means rankings can be interpreted as ordering candidates from best to worst, with higher ranks corresponding to more support. Voting systems that violate non-negative responsiveness can be said to exhibit negative response,[2][3] perversity,[4] or an additional support paradox.[5]
And then explicitly;
Runoff-based voting systems such as ranked choice voting (RCV) are typically vulnerable to negative response. A notable example is the 2009 Burlington mayoral election, the United States’ second instant-runoff election in the modern era, where Bob Kiss won the election as a result of 750 ballots ranking him in last place.[16] Another example is given by the 2022 Alaska at-large special election.
So no, RCV is actually somehow worse than First Past the Post.
Also, that rule about needing more than 50% of the vote? Yeah, that too is a lie. Ballot exhaustion means that it’s 50% of the ballots that are left in the final round. If you didn’t guess correctly who made it to the final round, your ballot is just thrown away. It’s called ballot exhaustion.
Another popular lie is that your vote transfers in order on your ballot. It transfers in order to the candidates who are left. That’s why Bob Kiss won from being ranked last on various ballots. The people before him were eliminated before the first name on the ballot, so the vote skipped a bunch of names and transferred from the first name to the last.
All this because when you hold an RCV election, what you’re really doing is just holding a series of First Past the Post elections on the same ballot. You can’t fix the problems of First Past the Post by repeating First Past the Post a bunch of times.
I agree completely, they’re much better for all of the reasons you stated. I’m saying they’re Ranked Choice in a dumbed down sense, because it’s a bit too complicated for many people to really grasp the specifics. You are still functionally giving your choices a rank, and that’s the easy to explain part for most people to get about STAR and Ranked Robin. So I’m making an appeal to something most people are familiar with already to really sell the two other systems.
I’d say it’s important to use the word “rating” when talking about STAR. This is actually by design. Introduce it to people by saying “Give the politician an honest rating 0-5 stars”. And if they’re all rated as ones and zeroes, I don’t see a problem.
STAR can also be in a ranked format as well, with Ranked STAR being mathematically equivalent to regular STAR, but I agree that regular STAR leans more towards a rating structure.
STAR uses ratings, not Rankings. There is Ranked Robin, but that’s not STAR.
Ranked Robin has more complexity in the count, but the good thing is that it always elects the Condorcet winner, because he sort of invented the system. Well, an early version. Ranked Robin is the formalized modern version, and you could argue that since equal ranks are allowed in Ranked Robin, it’s closer to ratings.
I used the word formalized because STAR and Ranked Robin both have specific written procedures for the ballot appearance, the counting, and basically everything else you need to run an election using the system.
It makes adoption easier when you don’t have to design anything, just follow directions for a fairly representative election.
Ranked Choice has a Monotonicity problem. i.e. it’s possible for a candidate to lose if a more people rank that candidate higher on their own ballot without changing any other ballots.
This has happened in recent RCV elections, and resulted in the candidate’s ideological opposition winning.
There’s a group called FairVote that’s been pushing RCV since the early 90s despite the many flaws of the system. Flaws that have been known since the system was first designed in 1788.
Seriously, Instant Runoff Voting was invented by the Marquis de Condorcet in 1788 as an example of a broken election system that can eliminate candidates preferred by a majority of voters.
It was later reinvented in the late 1850s by an Englishman who presumably never learned French.
Anyway a modern voting system for consideration is STAR, it was developed in 2014 by people who have read Condorcet, the the works of Kenneth Arrow from the 1950s. (Arrow’s Impossibility Theorium)
Find more info at www.equal.vote
STAR and Ranked Robin both are a form of RCV. Most people are familiar with RCV so I feel it’s easier to sell the whole concept under the RCV label and then just directing people to the Equal Vote Coalition, like you did, which does a great job of explaining the nuances.
Seriously though, STAR or Ranked Robin are the best choices if we want more progressive wins overall.
Regular RCV is weighted more towards center with a higher chance of the least liked candidate winning compared to STAR or Ranked Robin although less than our current system.
Even RCV is still leagues better than what we currently have and it still shows the actual support numbers for progressive politicians and policies. However, if we’re making a final choice, I’d be happy with STAR = Ranked Robin > Score Voting > Ranked Choice Voting >>>>>>>> First Past the Post (The current system in most of the US).
STAR and Ranked Robin are definitively not RCV.
RCV, or IRV as it’s known elsewhere in the world, is an Ordinal voting system. That means spoiler effect and enforced two party dominance.
STAR and Ranked Robin are on the Cardinal side of things. No spoiler effect, because votes you can give equal support to multiple candidates.
If I really like two different candidates and don’t particularly care which one wins, I can say that on a STAR or Ranked Robin ballot.
In RCV I have to make a strategic choice, and if I get that choice wrong, it’s not just my guy who loses, but possibly my entire side.
So no, they are not the same. And it’s worse because of RCV’s lack of Monotonicity.
And then explicitly;
So no, RCV is actually somehow worse than First Past the Post.
Also, that rule about needing more than 50% of the vote? Yeah, that too is a lie. Ballot exhaustion means that it’s 50% of the ballots that are left in the final round. If you didn’t guess correctly who made it to the final round, your ballot is just thrown away. It’s called ballot exhaustion.
Another popular lie is that your vote transfers in order on your ballot. It transfers in order to the candidates who are left. That’s why Bob Kiss won from being ranked last on various ballots. The people before him were eliminated before the first name on the ballot, so the vote skipped a bunch of names and transferred from the first name to the last.
All this because when you hold an RCV election, what you’re really doing is just holding a series of First Past the Post elections on the same ballot. You can’t fix the problems of First Past the Post by repeating First Past the Post a bunch of times.
I agree completely, they’re much better for all of the reasons you stated. I’m saying they’re Ranked Choice in a dumbed down sense, because it’s a bit too complicated for many people to really grasp the specifics. You are still functionally giving your choices a rank, and that’s the easy to explain part for most people to get about STAR and Ranked Robin. So I’m making an appeal to something most people are familiar with already to really sell the two other systems.
I’d say it’s important to use the word “rating” when talking about STAR. This is actually by design. Introduce it to people by saying “Give the politician an honest rating 0-5 stars”. And if they’re all rated as ones and zeroes, I don’t see a problem.
STAR can also be in a ranked format as well, with Ranked STAR being mathematically equivalent to regular STAR, but I agree that regular STAR leans more towards a rating structure.
STAR uses ratings, not Rankings. There is Ranked Robin, but that’s not STAR.
Ranked Robin has more complexity in the count, but the good thing is that it always elects the Condorcet winner, because he sort of invented the system. Well, an early version. Ranked Robin is the formalized modern version, and you could argue that since equal ranks are allowed in Ranked Robin, it’s closer to ratings.
I used the word formalized because STAR and Ranked Robin both have specific written procedures for the ballot appearance, the counting, and basically everything else you need to run an election using the system.
It makes adoption easier when you don’t have to design anything, just follow directions for a fairly representative election.