[EM] Endorsement for STAR voting

[Rob Lanphier]

Hi folks

I was asked to provide a quote supporting STAR voting by the Equal.Vote
Coalition (equal.vote).  The following paragraph is a slightly modified
version of what I wrote to them:

*Our "choose-one" electoral system gets a passing grade in democracy, but
just barely. It gets a "D". We can do better. We have the technology. There
are no PERFECT systems, but there are systems that are better than the
systems used in most single-winner, gerrymandered districts. STAR does much
better than passing; it may even get an A or an A+.*

I wrote the endorsement because I believe that STAR is a much better
ranking/rating system than "RCV" as currently promoted by FairVote.  It's
pretty clear from recent discussion on this mailing list that there are
some folks here that don't believe what the simulations say about approval,
STAR, and other cardinal-flavored voting systems.  They don't believe those
simulations, but haven't provided compelling simulations of their own that
show how STAR, approval, or even score voting would be a disaster over the
long haul.

I don't feel comfortable with the strategy burden imposed on voters by
plain "score voting" (a.k.a. "range voting").  Most of us agree that the
best strategy for "score/range" is a "min/max" strategy....basically, turn
the election into an approval election.  It doesn't seem fair to have a
system that requires voters who want to maximize the utility of their
ballot to know enough about the system to know the "min/max" strategy.
Even worse, all of the "cardinal" systems (except for approval) require a
much more complicated ballot, just so that people who don't have the
"min/max" decoder ring can give partial credit to some candidates.  That's
an awful burden to put on voters for the added complexity of numerical
ballots, when "smart" voters will only vote some variation of "0" or "1".

STAR is not perfect by any stretch of the imagination (c.f. Arrow,
Gibbard).  From a complexity perspective, it's tougher than choose-one
voting, since it presents "0" through "5" stars.  I suspect SOME of the
strategy problems with plain score/range cause it to be beneficial to
voters to min/max their ballots.  But I can't think of any CREDIBLE
electoral scenarios where that would happen.  Given that a 5-star STAR
ballot only allows voters to sort candidates into six tiers (gotta love the
fencepost problem),

And this is where the "political science" behind this goes from being
numeric to human and inexact.  I think approval voting (with "0" or "1")
presents voters with a choice.  Which candidate do I REALLY REALLY want?
Well, I should vote "1" for that candidate.  How about another really good
candidate?  Well....only if their approval polling is above 20% might I
even risk that.  How about a compromise candidate that seems like they have
a good chance of winning?  Well, yes, definitely vote for them, since all
of the candidates that I like have only 20% approval (if that) and almost
certainly wouldn't win anyway.  Are voters going to get everything they
want?  NO.  Voters will need to figure out how to compromise with their
neighbors (and housemates, and spouses).  There may be some "collusion"
that occurs, but let's not talk about what goes on in peoples' bedrooms.
Let's keep this mailing list clean, okay?  :-)

STAR allows for voters to express some nuance, but not too much. "What if I
really really don't like a candidate?", you may ask.  Well, don't give them
any stars!  Only give stars to candidates you "approve" of (i.e. candidates
you would vote for in an approval-voting election).  "Who should I approve
of?"  Well, perhaps everyone that you can barely tolerate, such that you
prevent a bad candidate you most fear from winning.  If you're not afraid
of the bad candidate (because polling suggests they are under 20% approval
and they seem like a crackpot that has 0% chance of winning), then don't
let that candidate guide your voting.

I think both STAR voting and approval voting in single-winner elections
would have similar effects on the electorate in the long term, and I'm
hoping that at least one of them truly catches on.  I think the Condorcet
winner criterion (CWC) matters a lot more than "cardinal" advocates often
suggest, but it's almost impossible for me to imagine a credible election
scenario where the STAR winner and a strictly CWC-compliant method would
differ.  More to the point, with that final pairwise comparison, STAR
virtually guarantees that a majority of voters prefer the winner to the
runner up.  And it's 1000% better than RCV/IRV as promoted by FairVote.

Given that approval is in use in two American cities building a track
record (Fargo and St. Louis), and STAR is close to getting approved for use
in Eugene Oregon, I feel like I'm kinda done waiting around for a better
system.  I think both STAR and approval voting get an "A" or an "A+" (maybe
93 or 96 on a scale of 0 to 100).  I feel like RCV/IRV gets maybe a 61 or a
62 (i.e. barely above choose-one/FPTP, and it may not be better than
choose-one).  Since moving to San Francisco (a city which uses RCV/IRV),
I've become MORE afraid of electoral fraud, not less.  I've seen what a
goat rodeo the vote counting process has been in SFBay jurisdictions that
use RCV/IRV, so the "precinct summability" argument has had more sway with
me now than it once did.

So, the second paragraph of this email is my endorsement of STAR.  I
reserve the right to change my mind, though, so convince me I'm wrong.
Please.

Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20240318/1e806e3f/attachment.htm>