[EM] Yet another IRV problem

[Bart Ingles]

Anthony Simmons wrote:
> 
> From: Bart Ingles <bartman at netgate.net>
> Subject: Re: [EM] Yet another IRV problem
> 
> >> Anthony Simmons wrote:
> >> >
> >> > >> From: Forest Simmons <fsimmons at pcc.edu>
> >> >
> >> > >> Why is IRV considered better than plurality when it fails
> >> > >> this consistency test and also fails monotonicity?
> >> >
> >> > >> Is it only that it allows more ballot expressivity and
> >> > >> more or less eliminates spoilage by tiny parties?
> >> >
> >> > I think the main reason is that it prevents what happened in
> >> > Florida last year.  This is what people see, so this is what
> >> > they want a cure for.
> 
> >> What exactly does it prevent?  Do you mean the situation
> >> where Nader "spoiled" a close election, which would
> >> otherwise have gone more decisively to Gore?  Don't
> 
> It would have eliminated the spoiler effect caused by
> candidates who get a few percent of the vote. Like Nader.
> That's what everyone saw, so that's their idea of what reform
> should deal with.  At least that certainly is the impression
> I've gotten.
> 
> >> forget, it could easily have gone the other way, in a
> >> different election (or the same election, in a different
> >> state):
> 
> All kinds of things could have happened.  But what people
> saw, and therefore what they're reacting to, is the spoiler
> effect, specifically in which the spoiler gets a small
> percentage.
> 
> >> 50%   Bush
> >> 45%   Gore
> >>  5%   Nader, Gore
> 
> >> In other words, a decisive plurality election could have
> >> been turned into a tie runoff.
> 
> That I don't quite follow.  You couldn't mean that Bush and
> Gore could have tied, could you?


There were actually at least two issues:

(1) The spoiler effect, which was mostly of concern to Gore voters this
time around (actually of concern to others as well, but it didn't cost
anyone else the election).  Although it's not clear how many votes
Buchanan and Browne drew away from Bush.  Nader himself claims that for
every 100 votes he received, if he had dropped out 38 would have gone to
Gore, 25 to Bush, and the remaining voters would have stayed home.

(2) The closeness of the election, which may have actually been of
concern to more people than the spoiler effect.  Unfortunately, rather
than giving a decisive win to Bush, the spoiler effect this time caused
a near tie, which IRV (or any other method) could have avoided.  My
point is that across many elections, the reverse is every bit as likely
-- IRV could turn a decisive FPP election into an IRV draw (I wonder
what the recount controversy would be like then?).

Since most of the press coverage focused on (2), that's what I assumed
you were talking about.

Actually, I think (1) received less coverage than either the recount
controversy OR the electoral college paradox.

Bart