For Americans who are interested in alternative voting systems, I recommend The Fulcrum: https://thefulcrum.us/ They cover good-government type reforms, including voting reforms.
I'll also plug https://www.electionscience.org, an awesome and underfunded nonprofit that helped get Fargo's approval voting initiative passed. (I've donated to them but have no other affiliation.)
I've been equating RCV with Score Voting, not IRV. I'll review the literature and see where I'm wrong.
I do not support IRV. It's an election integrity nightmare for little or no gain over Approval Voting.
My introduction to election stuff was thru IRV advocacy. Then I learned about tabulation complexity.
Over time, my position changed to PR for assemblies and Approval Voting for executives. I now wonder if Approval Voting could (should) also be used for PR.
That's all to say my strong-opinions-loosely-held about voting systems is motivated by us forging a better democracy, not any particular implementation. I'll realpolitik support any path forward.
But what about the best alternative of all - the single transferable vote? Call me biased as an Australian but it's served us well for giving small groups a chance in the Senate and electing 3rd party candidates such as Adam Bandt.
The reason why our Senate has third parties is because states act as multi-mumber electorates (it's not winner-takes-all as it is for the House of Representatives). This means that a uniform 20% vote for party X in a given state (either first-preference or after runoffs) will result in about 20% of seats being held by party X. The same is not true for winner-takes-all systems -- 150 electorates with 20% of people uniformly voting for party X will yield 0 seats for party X. That's why we only have 4 seats in the House of Representatives being held by third parties, compared with 14 Senate seats (though still not as many as you might like).
To take Adam Bandt as an example, he got 49% of first-preference votes in his electorate[1] -- most third parties aren't going to get that many votes in a single electorate (the centre of Melbourne is probably the only place you could see the Greens getting that many votes but they do have a consistent 10% first preference rate in polls).
Don't get me wrong, I do think our voting system is pretty good and better than most, but it does suffer from some of the issues described we could definitely improve it.
In the US context who would get voted in using the Australian Senate model is an interesting thought.
GOP and Dems would obviously pick up the majority, but I could imagine a fishing/hunting coalition (with possible NRA support) picking up a few seats, a few religious groups similar to Family First doing the same, and a whole lot of others. The scary thought would be a QAnon supported candidate gaining enough votes for a seat.
The point with Bandt and for other 3rd party candidates is that these models give them a chance as it removes the issue of split/spoiler votes and people are free to vote how they want but fall back on who they would prefer from a major party.
> The point with Bandt and for other 3rd party candidates is that these models give them a chance as it removes the issue of split/spoiler votes and people are free to vote how they want but fall back on who they would prefer from a major party.
Multi-member electorates are still preferential voting systems (just like our Senate elections are a preferential voting system) and thus still have that property.
The point I was making is that winner-takes-all elections result in smaller parties being unrepresented because a party needs to get the majority of votes in an electorate in order to get any representation. If you had 10 representatives per electorate and chose them using the Hare-Clarke system (which is used in Tassie and the ACT for the state government elections -- though their districts have 5 members) you would expect at least one representative to be a Green on average (since they poll around 10% nationally). Right now, the Greens have only 0.6% representation (1/150) in the House of Representatives despite having a 10% first-preference rate across the nation.
As for the more general point about spoilers, that is true but a lot of people (about 80%) still vote for the two major parties as their first preference -- possibly because they were never told how preferential voting works (I learned about it in primary school, but I don't know how common of an experience that is -- and it was never mentioned while I was at high school.)
I've been voting 3rd party in the Senate since I first got to vote (originally the Democrats and lately Greens) but probably got that notion from my parents who would have instilled Don Chipp's "Keep the bastards honest" in me. I was hoping Turnbull might have taken up Fraser's manifesto for a new party but that seems to have gone nowhere now.
Interesting to read about the Unity2020 movement who are trying to chart a similar path of taking the centre but are likely doomed to fail unless someone pumps a few billion into their advertising. The lack of preferential voting also doesn't help their cause.
I personally find the calculation of the redistribution and weightings in the Senate to be fascinating - but the math probably goes over the head of a lot of people.
It's worth noting that with the process of eliminating the last candidate, their votes are redistributed by preference and not simply discarded - so in the end your vote goes to one of the last two candidates unless there was a runaway winner and no need for a redistribution. That's not really explained in the OP.
This is an exciting time for ranked-choice voting in the US. In addition to the big victory in Maine, there have been several ballot measures to start using it in other cities and states: https://www.fairvote.org/ranked_choice_voting_wins_in_2020
I'm all about Approval voting, just vote for everybody who you would want as president, and whoever gets the most votes win. Want to vote for everybody? Go ahead!
Besides the spoiler effects listed in the article, voting has been made deliberately difficult in America. There is a deliberate effort to disenfranchise voters to keep things from changing. Anything that makes voting EASIER will create fairer elections, which neither party really wants unfortunately. Approval voting is the hardest for the voter to mess up, and therefore will likely be the fairest. You can screw up Ranked Choice. The Borda / Condorcet methods involve more complex math which can be screwed up. Approval voting is just simpler, which means your vote is more likely to be counted the way you meant for it to be.
This could, of course, be coupled with other measures that make voting easier. Instead of a single day, make it a month so that people can go around their work schedule and lines will be shorter.
In fact, in person voting is easier to scam than mail-in voting, so the fairest option would be to make mail-in voting very very easy. Print off your ballot online or pick one up at the post office or whatever, and send it in. When you vote in person, mistakes/hacking/etc are simpler because you can target a district that seems like it will skew a certain way. Mailed in votes should be scanned and counted and the anonymized scans should be made publicly available so the populace/press can verify the results themselves.
It won't happen because both parties have a vested interest in preventing it from happening. Fairer elections don't produce candidates like Trump and Biden.
For example, the city of Fargo just switched to approval voting -- coverage here: https://thefulcrum.us/voting/approval-voting-2646444809