Political scientist here. Please don't pelt me with rotten fruit.

We generally use three categories to differentiate electoral systems. The
number of categories depends on who's writing, but everyone pretty much
agrees on three: district magnitude (1 in your case), ballot type
(categorical in your case), and then allocation rule (plurality in your
case).

Shugart, Latner, and I argued here
https://protectdemocracy.org/work/toward-a-different-kind-of-party-government/
that 'FPFP' did not really exist in the US due to the widespread use of
primaries, some of which have been replaced with nonpartisan winnowing
rounds (AK, CA, etc). FWIW, Burnett and Kogan (2015) noted
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2519723 this elision
in their conclusion nearly a decade ago.

Other examples of plurality allocation with categorical ballots and
multi-seat districts:

• multiple non-transferable vote (incl. as limited voting)
• single non-transferable vote (incl. as limited voting)
• cumulative voting
• etc...

I generally stay quiet, but this issue is fundamental enough, I think, to
merit the above contribution. FPTP often comes across as an imaginary
target.

A purist might insist on calling IRV 'plurality' as well, so long as it
does not require the voter to rank all choices.

Jack

On Mon, May 6, 2024 at 3:10 PM Closed Limelike Curves <
closed.limelike.curves@gmail.com> wrote:

Jack, I think that your argument FPTP does not exist in the US because
of primaries is a sleight of hand, so to speak.

FPTP is still FPTP. If it walks like a duck and quacks like a duck....
You could equally argue that FPTP does not exist in the UK because, as I
have known locally in the past and no doubt present, a 20-person local
party committee or caucus steers the selection process. That is a
"primary" of sorts.

The contention has interest as an illustration of how far academics will
go to apologise for FPTP. Why extenuate such an indefensible system?
FPTP is not complicated, it is simply incomplete. FPTP is not an
election, it is only the first stage of an election. It would be nice to
hear that loud and clear from academics, and, conceivably, political
beneficiaries.

Regards,

Richard Lung,

On 06/05/2024 20:39, Jack Santucci wrote:

This is great, thank you so much! No objections from me to political
scientists, especially not any on this list. Definitely not any who are
interested in social choice. :)

This left me a bit confused, though—

1. Would you think of the terms "First-past-the-post" and "Plurality
   voting" as interchangeable in a paper, when used without qualification?
2. What do you mean by "Categorical ballots"? I haven't heard that term.
   Most social choice theorists think of FPTP as a ranked ballot rule (one
   that just happens to ignore everything after the first rank).

With regards to the paper, I don't think ballot exhaustion is a problem for
majoritarianism; exhausted ballots still produce simple/relative majorities
in the last step. No method can guarantee an absolute majority (>50%) for
one candidate over another, except by forcing voters to express opinions
they don't really hold (e.g. rank candidates they know nothing about). The
same problem is true of two-round systems as well—the apparent absolute
majority is created by ignoring the voters who don't turn out in the second
round.

The main reason IRV and two-round aren't majoritarian is they can override
the will of a majority of voters, even when there's no need to do so (i.e.
no cycle). People who hear John won because he had "majority support"
naturally take that to mean most voters preferred John to his opponent
Jack; most would be surprised if the opposite were true (as can be the case
in non-Condorcet systems). Social choice theorists use the term
"majority-rule" the same way (it means if most voters support A over B,
then A wins, unless someone else beats A).

In other words, simply having a majority of voters in the final round isn't
enough: if most voters preferred someone else, you're not really the
majority choice*. *(Was Chirac really the majority choice because he had a
majority against Le Pen? That seems like a low bar, given a soggy baguette
could've beat Le Pen as well!)

This is what most of us on this list mean when we say IRV is
plurality-like: the eliminated candidate is determined by the loser of a
plurality vote at each step, so majority support isn't enough to win. You
need to have several pluralities as well (one for each round).

On Mon, May 6, 2024 at 12:39 PM Jack Santucci jms346@georgetown.edu wrote:

While in accord with the sentiments of this post, John Stuart Mill MP
when in Parliament pointed out that "majoritarianism" isn't
majoritarian, a majoriity of a majority in Parliament may be a minority,
a half of a half is a quarter. He was advocating "Mr Hare's system" of
"Personal Representation" so grotesquely misrepresented, here, as IRV.

On 07/05/2024 18:27, Closed Limelike Curves wrote:

OK, I think I understand now; are you taking the term "plurality" to mean
the social choice function still returns a winner, even if that winner
doesn't have a score higher than 50%, similarly to how Nagel uses it here
https://www.cambridge.org/core/journals/american-political-science-review/article/abs/effect-of-approval-balloting-on-strategic-voting-under-alternative-decision-rules/35809BC96CFA0DC2B50E1550A190D2BF?
I suppose that makes sense, although it feels a bit surprising. Is that a
mathematically meaningful definition? Couldn't I define the social welfare
function for FPTP as (1 + fptp_vote_share) / 2 to guarantee a winner always
has a "majority" without changing anything meaningful about the method?

On Mon, May 6, 2024 at 12:39 PM Jack Santucci jms346@georgetown.edu wrote:

Perhaps? I don't know that article very well, but I am broadly familiar
with how the ideas evolved into the 2007 piece on the 'Burr dilemma.'

For me, "plurality" means "the most votes" -- or, in a district of M seats
it means "the M highest vote totals."

On Mon, May 13, 2024 at 4:22 PM Closed Limelike Curves <
closed.limelike.curves@gmail.com> wrote:

Ahh, got it, that makes much more sense. I was confused since it seemed
like using the term "plurality" seemed like it would describe all voting
rules—after all, it's not like we're going to elect the candidate with the
second-best score. But you might if you have a multi-winner election!

On Mon, May 13, 2024 at 2:21 PM Jack Santucci jms346@georgetown.edu wrote: