A likely FairVote talking point against St. Louis's Approval + Top Two ballot measure

While I think St. Louis could start a revolution for Approval Voting, I’d like to point out that FairVote will have a field day pointing out that Approval + Top Two is, in the long run, actually more expensive than RCV. The cost of holding all those runoffs will greatly swamp the millions of dollars it costs to upgrade voting machines once and for all to RCV. At the very least, it may have been pertinent to skip the runoff if someone in the primary earned 50%+ of the vote, even though this may have appeared worse in quality to St. Louis Approves advocates.
A much cheaper way of finding the majority support STL Approves sought would’ve been to just go with STAR Voting. However, I am still confident that the Approval Voting ballot measure will pass and have a huge impact on the voting reform movement, so I don’t think it’s that serious either.

Cost is unimportant compared to social utility efficiency and constitutionality

Then there is the argument that most likely the runoff will be two people of similar ideology, although that would then give advantage to whomever was slightly closer to center…

It is a very relevant factor, if you want your system adopted and to spread in the first place.

Not necessarily. If the runoff picks the top two utilitarian candidates, the more majoritarian of the two wins. This actually means the runoff could hurt election results overall, though this is where I favor taking a primary candidate with >50% of the vote as the winner, as a runoff where nobody has a majority of the vote should serve to expand, and not minimize the support each candidate seeks.
Also, having a runoff in the first place makes it a bit easier to be strategic, “hedge your bets”, and get that majoritarian-utilitarian candidate into the runoff by any means to win instead of the more utilitarian candidate. I suppose VSE might even regard this as a good thing, based on how it judges STAR!
Edit: However, I’m unaware of whether honest Approval Voting is significantly likely to elect a utilitarian candidate rather than the Condorcet winner. It seems that with or without a runoff, as long as you have polling Approval will tend towards the median voter. So your point stands :slight_smile:

Is FairVote campaigning against Approval in St. Louis? I thought someone else posted that IRV isn’t an option there since the machines can’t run it, so the local IRV organizations have shifted their support to the Approval campaign.

Not that I’m aware of. I’m just pointing out the logical contradiction FairVote can nitpick on if they’re asked about St. Louis Approves, which allows them to continue making Approval look inferior: STL Approves wanted to avoid paying for new machines by paying way more for runoffs.
They can even say that this proves Approval isn’t good enough on its own, because its advocates weren’t confident enough in it to skip the runoff.

(May be a double post – internet is weird)

STAR fails in this scenario:
51% A5 B4 C0
49% C5 B3 A0

IRV also fails this scenario (but STAR does not):
39% A5 > B3 > C0
10% B5 > A1 > C0
10% B5 > C1 > A0
41% C5 > B4 > A0

You might have meant to put this in the other post.
I think the interesting thing about the first example is that the minority has an incentive to run a second utilitarian candidate to crowd out the runoff. Yet this also destroys their own chances of winning. Also, if the field expands to 4+ candidates, then at least 2 of them (presumably the candidates who are the secondary preferences of their own side) would likely have an incentive to transform into consensus candidates anyway.
The second scenario highlights a strategic conundrum that’s even available with IRV: voters must choose to compromise and put B as their 1st choice, or take a risk and put their true favorite first. Essentially, the same risk that cardinal systems offer, yet here B’s true support would be obscured by polling and final results, making people potentially abandon that candidate and not even think of strategically putting them 1st (or they might flock towards B, if they are aware that strategy works in IRV.) In a cardinal system, people would likely be honest at least the first time a poll included a given candidate or party, meaning that everyone would have a clear sense of the true support that candidate had before making any strategic moves and tanking or increasing their support in the polls. That might make it a lot easier for candidates like B to appeal to everyone, since people would see that the candidate legitimately has enough support to govern, not just be the lowest common denominator compromise. Maybe that’s where ranking goes astray and people begin to oppose Condorcet sometimes.

I believe Approval Voting would be used in St. Louis’s current primary elections, with the runoff taking place during the general election. So there is no additional separate election, and thus no extra cost. So this is actually way cheaper than RCV.

The St. Louis initial approval election, then (assertably single-selection) top-two election method – would far better be described as an initial approval election followed by a (corporeal, not virtual) top-two runoff election. Calling the initial election a ‘primary’ is flirting with terminological madness.

Conventionally, political parties hold ‘primary elections’, and governments hold ‘elections’. Although they may hold certain privileges such as having their candidates names published on official ballots, parties are strictly private organizations. The problem here is that governments are not conventionally said to hold ‘primaries’.

The terminology of election methods analysis is already a horrible no-good rotten dreadful mess.

It’s called a primary because of the date it’s held, not because they are choosing party nominees.

So, we are gifted yet another rusted-out water heater in the lexicological landfill.