[EM] IRV / RCv advances

Abd ul-Rahman Lomax abd at lomaxdesign.com
Sat Jul 14 11:05:24 PDT 2018

Ignorance about voting systems is widespread. "IRV" was invented in the 
19th century, and is the single-winner version of Single Transferable 
Vote, which is a not-so-bad method for creating, in multiwinner 
elections, something like proportional representation, though a far 
superior variant was invented in the 1880s that would, if impemented, 
create full representation. "No taxation without representation," but 
systems that depend on contested elections leave many and often most 
voters without any chosen representation, and, at best, a weak 
compromise. But because we have never seen anything better, we think 
that we have a democracy, and we do, compared with truly awful systems. 
But the systems we have can fail badly, and we never seem to fix them.

Now, back in the 19th century, single-winner STV was known to be a very 
poor system. Within the context of a strong two-party system, it looks 
good. However, whenever there are three factions, and the third faction 
starts to approach parity, it can fail badly. IRV proponents look at 
what can seem like successes, but ...

IRV was so-named to make it resemble the most widely-used "advanced 
voting system" in actual use, top two runoff. In fact, top two runoff 
showed a very interesting characteristic, in research conducted by 
FairVote. In about a third of the runoffs, the frontrunner in the 
primary lost in the runoff. That is rare in actual IRV elections. Why? 
Well, IRV proponents also claim that Robert's Rules of Order (RRONR) 
recommends IRV. In fact, it recommends a version that requires a true 
majority of the votes for someone to win. If there is no true majority 
(i.e., a majority of all ballots containing a vote for the leader), then 
the election must be repeated. With a top-two runoff? No. That, in fact, 
would cement in the pathology. With a completely new election, new 
nominations, etc. RRONR actually doesn't like the IRV method, but 
suggests it only because it is in actual usage in some organizations, 
and if it is impossible to hold a runoff, IRV can be considered. In 
fact, they know that there are better methods for that contingency, but 
they are not in wide use, and RRONR is descriptive, not prescriptive.; 
(and if one is following the described standard rules, no decision is 
ever made by mere plurality, which IRV commonly does.

And the voters of San Francisco were lied to in the voter information 
pamphlet that claimed a "majority of the votes" would still be required 
to win. The IRV initiative actually removed that requirement from the 
law. But they then redefine "votes." If you didn't vote for a 
frontrunner, one of the top two, our ballot is disqualified. The system 
pretends you didn't vote. I've pointed out that by these rules, all 
elections would complete with unanimous agreement of "all the votes," 
just do the elimination one more time.

I haven't looked at recent elections, but Burlington Vermont was a city 
with three major parties. It elected a candidate from the Progressive 
Party when it was clear from the votes that a majority of voters 
preferred the Democrat over him. Republican voters actually would have 
seen a better result if they had stayed home and not voted. (Because 
they preferred the Republican, the Democrat was disqualified before 
their second-rank votes for hiim were counted, leading the the 
Progressive winning. If one is a Progressive, one might think that a 
good result, but the voters of Burlington proceeded to toss out the 
method. By the way, top-two runoff would have the same problem.

People are interested in other voting systems, because some of them are 
known to work better than IRV. Top two runoff, where write-in votes are 
allowed in the runoff, is better than IRV. Of course, if one thinks that 
runoff elections are Bad, then one might like IRV more. However, there 
are other options that reduce the need for runoffs while preserving 
basic democratic values, as IRV does not.

I'm fascinated how reformers, supposedly interested in improving 
democratic process, will lie to promote their ideas. (But most pushing 
IRV are not lying, they are ignorant, but some who make their living 
lobbying for IRV do lie.)

It is fascinating to me that those with obviously little understanding 
of the issues are ready to dismiss discussions of them by those with far 
higher knowledge and experience as "crazy." It seems to be symptomatic 
of our time, ignorance displayed with high confidence.

A naive appraisal would look at, say, Bush v. Gore in 2000 and think 
that IRV would have fixed it. And it might have. However, the problem is 
that with IRV, third parties would gain an initial toehold and would 
gain much more of the vote. That might be considered good.; But, then, 
the spoiler effect would return with a vengeance, with results like 
those in Burlington, demolishing the benefit. I'm not going into them, 
but there are demonstrated systems with far better and safer 
performance, such as Bucklin Voting, which could be called "Instant 
Runoff Approval." It counts all the votes, and is precinct summable, 
unlike IRV, which is expensive and difficult to canvass, and which ends 
up ignoring many or even most of the votes cast. They aren't even 
counted. Bucklin was very popular for a time about a hundred years ago, 
and was eliminated, not because it didn't work, but because it did, and 
that's obvious from the history. (In some elections under some 
conditions, there was still majority failure, but that was not actually 
a poor result from the system, but a characteristic of some primary 
elections. A version of Bucklin that was actually two-round if needed 
would have addressed this.

And then there are score systems, which allow maximum voting 
flexibility. It is now possible to design voting systems that are vast 
improvements over those in use. The system I mention above as designed 
in the 1880s would transform representative democracy, making party 
affiliation unnecessary, probably removing the corrosive influence of 
money in elections (because it does not waste any votes and you can vote 
for your favorite, period, no assessment of "viability" being necessary) 
... but who actually cares about real democracy?

Very few.

Advanced voting systems can and should be implemented in NGOs, because 
until people have experience with them, they won't have a snowball's 
chance in hell of being implemented in public elections.

On 7/12/2018 9:37 PM, Sennet Williams wrote:
> Sorry I don't get online much, but everyone should know that RCV is 
> getting a LOT of good publicity.
> 1-Maine just had the first statewide IRV election in U.S. history.
> 2-since then, there have been op-ed(s?) in the NYT calling for RCV 
> nationwide
> 3-London Breed has just become the first african american female mayor 
> of SF: thanks to RCV.
> 4-Jesse Arreguin is the first latino mayor of Berkeley, thanks to RCV.
> 5-Jean Quan was the first asian mayor of Oakland, thanks to RCV.
> 6-Libby Schaaf, Oakland's new mayor, was elected thanks to RCV.
> If you want to pay attention, IRV/RCV/ranked pairs are inevitably the 
> future, that is why  I don't understand this craziness discussing 
> outdated election "systems."
> -Thanks for reading
> ----
> Election-Methods mailing list - see http://electorama.com/em for list info

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20180714/5c075336/attachment.html>

More information about the Election-Methods mailing list