SoylentNews is people
SoylentNews
Common Dreams reports
Election reform advocates on [April 18] praised a decision by Maine's Supreme Court, upholding the use of ranked-choice voting for the state's upcoming primary elections, saying the ruling demonstrated that the court heeded the demands of Maine voters.
[...]Unlike in traditional voting, in which the candidate with the largest share of votes wins--even if he or she is far from capturing a majority of the support--in ranked choice voting, voters rank each candidate in order of preference. If no candidate has a majority after the first count, the least-popular contender is eliminated, voters' ballots are added to the totals of their second-ranked candidates, and the ballots are recounted. The eliminations and recounts continue until one candidate has a majority.
Supporters of the system say it increases voter turnout and proportional representation.
Maine's June 12 multi-party primary elections, in which voters will choose candidates for governor and congressional districts, will now make history as the first state election to use ranked-choice voting.
Fifty-two percent of Maine voters supported the system in a November 2016 ballot initiative, but lawmakers passed a bill last year delaying its implementation until December 2021 and argued that the state could not use a new voting system without direction from the legislature. The state Senate also threatened to repeal ranked-choice voting altogether if it could not pass a constitutional amendment by then.
More than 77,000 Maine residents signed a petition saying any repeal of the system by the legislature should be voided.
"The Maine legislature has changed or repealed all four of the initiatives passed by Maine voters in 2016", said Kyle Bailey of the Committee for Ranked Choice Voting in a statement on Tuesday. "Today's decision by the Maine Supreme Court confirms that the Maine people are sovereign and have the final say."
The Portland Press Herald, Maine's largest circulation daily newspaper, has extensive background details in their April 17th story: Ranked-choice voting will be used for June primaries, Maine supreme court rules.
(Score: 2, Informative) by Anonymous Coward on Thursday April 19 2018, @12:18PM (22 children)
Isn't that the one where if you have three candidates, where half the population prefers candidate H and hates candidate T, and the other half prefers candidate T but hates candidate H, and the only one everyone can accept is candidate Q, Q will be eliminated first, and one of the two most hated candidates (let's say T) will win?
(Score: 2) by acid andy on Thursday April 19 2018, @12:44PM (4 children)
You're missing the other side of the coin, where half loves T and the other half loves H. Or, at least they prefer them to Q.
I suppose the outcome you wanted could happen if each of the choices were weighted using some multiplier and then added up at the start. So, people ranked T or H 1st, so those votes are multiplied by 1 (kept unchanged), but everyone ranked Q 2nd and those votes are multiplied by 0.5 or 0.7 or something. The hard bit would be deciding what would be fair multipliers for each rank.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @01:34PM (3 children)
See my reply below [soylentnews.org].
As I see it, the answer isn't forcing voters to state simple preferences ("A>B>C") with no information on the strength of those preferences, then applying arbitrary multipliers to assume strengths.
Instead, just let the voters state their preferences in a form that includes strengths ("A:99 B:90 C:00" or "A:99 B:10 C:00"), and use the information they give you.
(Score: 2) by acid andy on Thursday April 19 2018, @05:58PM (2 children)
I agree with you, completely that letting the voter specify the weighting they wish to give each candidate is the only truly fair system. I suspect the trouble is that many of the proles just won't understand a system of that complexity and either won't use it at all or their vote will be rejected as invalid or they'll vote in a way that doesn't represent their true feelings due to misunderstanding the process. You could end up with 20-30% of the ballots being rejected. Hey, on second thoughts, that's probably still better than the current system!
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 2) by bob_super on Thursday April 19 2018, @06:36PM
> on second thoughts, that's probably still better than the current system!
Choosing a bus driver is a complex process with many criteria, and an assessment by knowledgeable people in the field. We don't want to make a mistake and have people get hurt.
Choosing the people in control of the US ? Meh! That's a popularity contest based on whatever lies sound the best. Who could possibly get hurt?
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @09:27PM
This is a fairly big problem with ranked (ordinal) ballot systems, because it typically looks like:
Candidate A [ ]
Candidate B [ ]
Candidate C [ ]
Candidate D [ ]
and has instructions that each box must be left empty or filled with consecutive numbers from 1 to n, where n≤(number of candidates). So
Candidate A [1]
Candidate B [3]
Candidate C [2]
Candidate D [ ]
is a valid ballot, but
Candidate A [2]
Candidate B [4]
Candidate C [3]
Candidate D [ ]
or
Candidate A [1]
Candidate B [2]
Candidate C [2]
Candidate D [ ]
or
Candidate A [1]
Candidate B [4]
Candidate C [3]
Candidate D [ ]
are all invalid.
With score (cardinal) ballots, the ballots are (conceptually, at least) identical, but the rules and semantics are different; each box may be left blank or filled with any number from 0 to 9, 0 to 99, or some such range. It's incredibly simple, and the only way to spoil a ballot is to write illegibly, to write something other than a number, or to write a number outside the range.
In practice, to avoid issues with handwriting, and to enable machine counting, a related form is generally suggested:
___________ (0) (1) (2) (3) (4) (5) (6) (7) (8) (9)
Candidate A [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
Candidate B [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
Candidate C [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
Candidate D [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
In this case, at most one box/oval/hole is to be checked/filled/punched per row. Again, incredibly simple, and the only way to spoil a ballot (punching two holes in the same row) is very similar to spoiling a plurality ballot by punching holes for two candidates. People accustomed to the existing system should be able to get that.
That's a valid concern with either ranked or score ballots -- confusion between 1 for 1st place, 2 for 2nd, etc. vs. highest score wins is entirely possible in either direction. But honestly, while I don't want to disenfranchise anyone, even those who can't read and follow directions, or set up any sort of capability test for voting, I'm prepared to tolerate some confusion the first few cycles with a new voting system -- it will decrease over time as people become familiar with the new system.
(Score: 5, Informative) by MrGuy on Thursday April 19 2018, @12:52PM (4 children)
No system is perfect, and this is certainly a risk.
However, it's much more often the case (you see this fairly commonly when there's a single viable independent candidate in a US election) where the candidate pulls much more strongly from one of the major candidate's bases than the others. In tight races, this tends to produce pressures that distort the electorate and/or frustrate people from expressing their true preference.
Consider a hypothetical. For sake of argument, let's call then Don, Hill, and Bern. Don and Hill are the major party candidates. Bern is an independent. Most of Bern's supporters prefer Hill over Don, and would vote for Hill if Bern wasn't in the race (Don is their least preferred candidate). Hill has significantly more voters than Bern (being the major party candidate), but Bern has a significant number of voters. Let's say Don has 45%, Hill has 40%, and Bern has 15%. The combined Hill and Bern voters slightly outnumber the Don voters, but the Don voters are a plurality (there are more of them than either the Hill or Bern voters individually).
In the current single-candidate system, there's tremendous pressure on Bern to drop out of the race. "He's taking votes away from Hill!" "He's throwing the election to Don!" If Bern stays in the race, he risks his (and his voters) least-preferred candidate winning. And even if he drops out, the result won't simply be all his voters going to Hill - some might decide not to vote at all because they'll be discouraged, so even if he drops out the election will be very close.) And if Bern stays in the race, there will be huge pressure on his voters to vote "strategically" - rather than vote for the candidate they prefer, they should vote for the "more electable" candidate, or they risk "throwing their vote away" in a tight race. If Bern and/or his voters do not bend to pressure, their least-desired result (Don is elected) will happen, frustrating the fact that the majority of the electorate does not want Don elected.
The new Maine system handles this case better (with "better" defined as "comes closest to getting everyone to vote and having a result most in line with what the voters want"). In this system, there's no need for there to be pressure on Bern to drop out, and no voters get discouraged. Everyone goes to the polls. Bern supporters vote Bern-Hill-Don. Hill supporters likely vote Hill-Bern-Don. Don voters lead with Don, and split their second and third preferences. Bern, as the low polling candidate, is eliminated, and his votes are redistributed to his voters second preference (which is largely Hill), leading to Hill being elected.
(Score: 4, Interesting) by Grishnakh on Thursday April 19 2018, @05:37PM
In this system, there's no need for there to be pressure on Bern to drop out, and no voters get discouraged. Everyone goes to the polls. Bern supporters vote Bern-Hill-Don. Hill supporters likely vote Hill-Bern-Don. Don voters lead with Don, and split their second and third preferences. Bern, as the low polling candidate, is eliminated, and his votes are redistributed to his voters second preference (which is largely Hill), leading to Hill being elected.
Or, it turns out the pollsters were wrong, and Bern was more popular than they thought, and Bern edges out Hill, but not enough to win over Don under our current system. But under the ranked-choice system, Hill is dropped, and her votes go to their 2nd choices, which are almost all for Bern, and Bern wins. Here, it turns out the voters choosing the "unelectable" candidate were right after all, despite what the media and pollsters were trying to convince them of, and the candidate most preferred by the electorate wins.
This might not happen, but it might; in 2016 we found out in a big way just how wrong the pollsters can be. They insisted up and down that there was absolutely no chance that we'd get the outcome we got, and here we are.
(Score: 3, Informative) by tangomargarine on Thursday April 19 2018, @06:47PM
No, The Donald lost the popular vote in this "hypothetical" scenario.
"Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @07:25PM
Just what I always wanted, giving voice to communists.
To be clear I am not calling Bernie a communist, he clearly sold out to capitalist
billionairesmillionaires.(Score: 0) by Anonymous Coward on Friday April 20 2018, @04:11AM
Keep in mind that the ranked choice voting here only applies to Primary elections to nominate party candidates, not to the General elections which actually elect people for office. The original initiative was intended to do so, however this was unanimously ruled unconstitutional by the Maine supreme court as the Maine state constitution explicitly defines the voting system to be based on a plurality, and the initiative was not for a constitutional amendment.
https://assets.documentcloud.org/documents/3728801/Maine-Supreme-Judicial-Court-ruling-on-Ranked.txt [documentcloud.org]
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @12:52PM (1 child)
That's the current system, but nice try.
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @04:49PM
No, the current system is the one where a small group of people gets to decide that no matter how much you hate Trump you'll just have to vote for him anyway, or your punishment will be Hillary.
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @01:27PM (4 children)
Yes, it's a deceptive name for IRV, which behaves exactly as you describe. It's deceptive because "ranked-choice" literally describes a type of ballot, and there are several ways to determine a winner from a stack of ranked ballots. IRV is just one of them (and one of the worst), but its proponents try to remove the others from discussion by equating ranked ballots with IRV.
Ranked ballots in general have the fundamental weakness of pretending all preferences have the same strength, rather than allowing voters to specify relative strength of preferences. But Condorcet methods at least listen to all the preferences expressed. IRV, on the other hand, ignores most of the information on the ballots in each round
Range/score ballots are not only more expressive, but can be counted in one round, and summed by precincts, exactly as we do with plurality votes.
Your described situation is a real problem for ranked ballots, because there's no way to tell whether H>Q>T means "love H, like Q, hate T", or "love H, hate Q, hate T just a tiny bit more" -- is Q universally disliked, or universally tolerated? Ranked ballots simply don't let the voters tell you (most ranked systems won't even let you state an explicit non-preference -- there's no way to vote H=Q>T or H>Q=T); you need range ballots to make this distinction.
But in a situation with more than three candidates, it can become pretty obvious that the unfairly eliminated candidate was in fact weakly preferred rather than weakly despised; consider the four-candidate race where the vote is more-or-less evenly split among these three ballots:
A>D>x>x
B>D>x>x
C>D>x>x
Here D, despite being the universal second choice, will be eliminated in the first round for not being anyone's first choice.
(Score: 2) by tangomargarine on Thursday April 19 2018, @06:45PM
Why would you even rank a candidate you hated? Unless you have to rank everyone, just don't write anything next to the people you hate; that way there's no possible way you can accidentally vote for them.
"Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
(Score: 2) by HiThere on Thursday April 19 2018, @08:45PM
I prefer Condorcet voting, but IRV has the advantage that its easy to describe. And it's so much better than plurality wins that I would hesitate to criticize it.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @09:26PM
Read half of your post. But with that situation "love A, hate B, hate C just a tiny bit more", what if you actually hate it and leave it blank? Then if A gets eliminated as an option because it was the least popular that vote will be counted as blank as there is no option B. Nonetheless for voters who did not hate B or C that will be counted as they will fill in the option with they second favourite candidate
(Score: 0) by Anonymous Coward on Friday April 20 2018, @07:27AM
That's probably the example I was thinking of, I just didn't remember the details. The problem is much more obvious with four choices rather than three.
(Score: 3, Informative) by Thexalon on Thursday April 19 2018, @04:26PM (1 child)
First, read up on Arrow's Impossibility Theorems [wikipedia.org]. That was proof that it's mathematically impossible to correctly reflect the position of voters with 3 or more candidates running.
The arguments about which voting system is least bad are about which of the flaws in the system are expressed. For example, in a plurality system, third parties are usually given short shrift, but in approval voting can get outsized power.
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 2) by bob_super on Thursday April 19 2018, @06:52PM
> proof that it's mathematically impossible to correctly reflect the position of voters with 3 or more candidates running.
The answer is proportional voting, but that get screwed up by party bickerings that cause the extremists to wield outsized power (Israel these days).
So the real answer is proportional voting with no parties allowed, where you have 3 or 4 people per district, each wielding a percentage of one vote matching the local results, and are not allowed to talk to each other outside of official bill writing sessions. Which is highly idealistic and utterly impractical.
(Score: 0) by Anonymous Coward on Thursday April 19 2018, @06:39PM
Well, that's got to be the first time a thread got Arrow's-theorem'd
(Score: 2) by HiThere on Thursday April 19 2018, @08:42PM
There is no perfect voting system possible. This has been rigorously proven. In the process they also evaluated the number of edge cases where the result was inferior. Plurality wins (the normal US election method) was the second worst voting method. The worse one involved the government putting a bias on how it counts the votes (this includes "only allow one candidate to be offered).
The admitted reason plurality wins became dominant in the US was because it was the simplest choice. (Also, many of the alternative methods had not at that point been proposed.)
I suspect that a fair lottery would still be a better method, but keeping it fair might be trickier. OTOH, be prepared for a proliferation of parties.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 2) by FatPhil on Thursday April 19 2018, @08:45PM
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves