There is, however, an argument that the longer it takes to produce a result from an election, the less confidence people have in the result, either because they lose the cause-and-effect feeling between going to the polling station and learning the result, or because bad actors deliberately try to undermine the process by filling social media with false narratives (and encouraging intimidation of poll workers) while the count goes on.

Don't get me wrong, I'm not saying the result should be announced immediately, using electronic counting (and definitely not online voting), but 13 counts and then 10 days of recounts for a single district (Dublin South-Central[0] in 1992) is hard to justify, and I think that the voting system makes the process slower/more expensive and more error-prone. For comparison, apparently the most recounts for a UK general election seat was seven.[1]

Also, I believe that IRV/STV/RCV doesn't support the summability criterion[2], which leads to more centralisation of the vote counting process.

[0] https://www.thejournal.ie/closest-irish-election-results-in-...

[1] https://en.wikipedia.org/wiki/United_Kingdom_general_electio...

[2] https://en.wikipedia.org/wiki/Summability_criterion

Even after the 1992 incident happened, the system continued to be used. I didn't live here back then, but, according to my local friends, after that, counting was automated for a couple elections and, then, it was agreed that while automation offered faster results, the manual process was more trusted by the people.

> bad actors deliberately try to undermine the process by filling social media with false narratives

I'm glad the voting system also weeds out extremists. The problem is that, in the US, said extremists are already in power, because current voting systems don't punish extremism enough.

The article seems to confuse RCV (voting by ranking a list of candidates) and IRV (a specific way of choosing the winner from such ballots). I'm not sure to which the law applies. Wikipedia has a disambiguation page for RCV that lists these 2 concepts.

There are ranked voting schemes that obey the summability criterion, although they aren't as commonly used as IRV and STV.

It's also quite practical to use a non-summable method with a limited set of candidates since any non-contrived method can use a summary array of length n! by using a count of each possible ranking. E.g. a large 10 candidate election could use only a 15MB summary array.

Delayed counts, recounts, and challenged results with our FPTP elections are now the norm.

How could RCV be any worse?

Congress has an approval rating of 20%

There is no faith in the system as is

I don't know that it's universally positive. In San Francisco it backfired and resulted in electing an extremist District attorney who refuses to prosecute criminals.

From the perspective of a major Boudin critic, I don't blame RCV/instant-runoff voting at all. Boudin ran on a platform that had broad appeal, albeit more appealing to some than others--thus his third-round win. The problem is that the appeal of the platform came mostly from how voters understood the issues in a national frame of context, not a local frame of context. SF criminal justice policy already had baked in all the reforms which were part of the national discussion. So ill informed people thought they were voting for one set of policies (policies they already had, in fact) but were getting something far more extreme.

It does speak to the argument that simple majority voting provides a clear mandate--clear, at least, in terms of the American tradition of elections. But that whole mandate argument isn't very persuasive. If there was a traditional runoff, Boudin might still have won, yet we'd almost certainly still be in the same position, including a recall. But let's say Boudin wouldn't have won--SF criminal justice "reform" might have continued drifting toward ever greater extremes as the Overton Window slowly shifted. (The platform had broad appeal, so it still represented a center gravity.) There's something to be said for giving people what they think they want and letting things play out unencumbered, and RCV/instant-runoff theoretically does that better than first-past-the-post or traditional runoffs on average. If preferred policies are truly poor ideas and things fall apart, it then becomes much more difficult for advocates and voters alike to dissemble later. Boudin's policies have forced every SF politician and plenty of voters to reconcile to some extent their politics with reality, and effectively forced the centrist politicians to firmly declare that reforms have become too radical and counter-productive. Previously everybody had been able to avoid that.

I find the whole mandate argument disingenuous. In theory it might make sense in some rare situations, but in practice it's merely part of the rhetoric of a national political party who couldn't care less about mandates--they'll claim a mandate whenever it suits them, and reject a mandate whenever it suits them. And, frankly, voters will often do the same. Boudin's platform had broad appeal, but that won't stop some people from claiming that RCV/IR resulted in voters being cheated; they'd say the same thing even if he won a traditional runoff.