[EM] A rant about IRV and it's recent history in Alaska.

[robert bristow-johnson]

> On 12/05/2024 7:55 AM EST John T Whelan <john.whelan at astro.rit.edu> wrote:
> 
>  
> That, like RB-J's original statement, is something of an oversimplification.

What, specifically, in my post is an oversimplification?

>  Yes, Peltola was the Condorcet winner in the 2022 general (when she had the advantage of incumbency)

Consider the election was less than 3 months after the special election in August.  Same three major candidates, same campaigns, same election cycle.  But in the "do over" voters have information from the first election and, for 99.9% of the voters that the information is incomplete (leading to false conclusions), they use that information to inform their do over vote in November 2022.

Now what are most naive voters gonna think after August?  They will mistakenly think (from reported tallies of the August election result) that Begich was a weak candidate and they will jump ship.  Indeed, Palin went on the campaign trail telling Alaskans that Begich should drop out "get out of [her] way" so that she could beat Peltola in November.  When exactly the opposite was correct; Palin couldn't beat Peltola head-to-head.  But Begich could.  Palin needed to drop out and get out of Begich's way.

But what really happened was that other Begich voters jumped ship (betrayed their favorite) and marked Peltola #1 in the November election (because they were petrified of Palin just as decent Republicans are, along with me, petrified of Trump).  That net transfer from Begich to Peltola changes the number of voters ranking between these two and that put Peltola ahead of Begich, head-to-head, in November 2022.

> and Begich was the Condorcet winner in 2024 (when the electorate in general shifted to the right compared to 2020 and 2022).

Begich's margin over Peltola was slightly less in 2024 and the percentage margin was less than half in 2024 than it was in 2022.

>  Ultimately, the fundamental fact remains that FairVote claims IRV eliminates the spoiler effect, when it's been demonstrated twice now that this is only true until the "spoiler" has enough support to eliminate the Condorcet winner before the final round.
> 

That's right and that can happen whenever there is a close 3-way race.  IRV does fine if the would be spoiler is not in the final round.  But when it fails, it's always the loser in the final round that is the spoiler.

And voters voting for the loser in the final round *never* get their second-choice votes counted.  These Palin voters had a reason for marking Begich (or a very few, Peltola) as #2.  They were assured that if they couldn't get their first-choice candidate elected, that they're vote would count for their second-choice, if they marked a second-choice.  But that's a false promise in *any* IRV election having 3 or more candidates and more than 1 round.

But if we didn't have 3 or more candidates in an election, we would have no need for RCV and FPTP would do just fine.  So in promoting this reform, it's necessary to consider what the risk of failure is, not with the entire set (700+ U.S. RCV elections), but with those elections where the reform *actually* does something different than what existed before.

More than half of all single-winner RCV elections in the U.S. had 2 or fewer candidates.  In the earlier days it was 300 out of 500.  It's only when there are 3 or more candidates that RCV can *possibly* exhibit the fruits of its reform.  Of those election, it's only in the "come-from-behind" elections that demonstrate that RCV actually *did* something different.  There used to be about 15 of those elections.  I think I read somewhere (FairVote, perhaps) that this number is now about 25.  The Burlington 2009 election is one example, but the Alaska 2022 election is not.

Those are the only trials where the proponents of reform can point to them and say that their promoted reform actually did its reform.  Then the portion of failure is hardly "vanishingly small".  Because when RCV actually does any good, that portion of success is also vanishingly small.

> This assessment is independent of the fact that I personally would rather see Peltola in that seat than Begich.

Me too, in spades.

It's just that elections, even election in Vermont where I live, do not exist to elect to office the candidate Robert Bristow-Johnson would rather see elected.

So, particularly in the context of voting reform, I object just as much if it's the candidate that I like who is the beneficiary of a spoiled election.  But for sure, most other people on the Left will not have the introspection to see that when it's their own who benefitted from an RCV election, whether it was spoiled or not.  That's where the intellectual dishonesty really resides.

Just imagine if the numbers with the three top candidates were mirror imaged and it was the candidate on the Left who was the spoiler and the candidate on the Right who was the beneficiary of the spoiled election.  Would these "intellectually honest" IRV proponents be singing the same tune, when it's their side that was screwed?


--

r b-j . _ . _ . _ . _ rbj at audioimagination.com

"Imagination is more important than knowledge."

.
.
.