Legal Beagle: Voting in an STV election
115 Responses
First ←Older Page 1 2 3 4 5 Newer→ Last
-
PS I am happy for people to swipe this and post it elsewhere. If you'd like a shorter or edited version, please get in contact. I'm happy to proofread something like this for anyone (even professional publications) without charge.
-
Graeme, does the supercity enabling legislation still block Auckland from adopting an STV electoral system if it so desired?
-
Graeme Edgeler, in reply to
Graeme, does the supercity enabling legislation still block Auckland from adopting an STV electoral system if it so desired?
Not as far as I can see. That was only for the 2010 election.
-
Minor nit: the quota changes downwards during the voting, largely because people don't rank all 40 odd (in Dunedin's case) candidates, the quota is calculated from the number of votes still in play (including fractional bits) and the number of unelected seats still in play.
Of course Hospital Board voting is only silly if the government hasn't taken away your right to vote for your local hospital board
-
Graeme Edgeler, in reply to
Minor nit: the quota changes downwards during the voting, largely because people don't rank all 40 odd (in Dunedin's case) candidates
I did include an acknowledgement of that in an earlier question, but I accept if you were only reading that, then it could be a bit misleading:
If anyone now has a majority of the remaining votes, they’re elected.
-
Graeme Edgeler, in reply to
Hospital Board voting is only silly if the government hasn’t taken away your right to vote for your local hospital board
My view is that we should elect people into positions where they make decisions (especially decisions around funding levels), and not elect people into positions where they are administering decisions made by others.
If DHBs had the power to tax people/businesses/whatever in their area in order to fund healthcare, so that elections were about how much healthcare your DHB would provide etc. I would absolutely support having DHB elections. At present, DHBs appear to be organisations whose main purpose is to provide someone whom the government can blame when the health system doesn't work well.
-
Paul Campbell, in reply to
If anyone now has a majority of the remaining votes, they’re elected.
(still in minor nit territory) true for the last seat, but it's really "if anyone has NUMBER-OF-OUTSTANDING-VOTES/(NUMBER_OF_OUTSTANDING SEATS+1) votes then they're elected" otherwise you discard the lowest polling candidate and redistributing their votes, if they were the last listed candidate on any voter's list NUMBER-OF-OUTSTANDING-VOTES gets smaller
"Distracted Scientist", who sometimes posts here, did a great visualisation of the counting process in a Dunedin election - you can see the slope of the line at the top slowly drooping as the quota drops - also as it drops more (a larger fraction) of the votes for candidates already elected get redistributed to other candidates
https://distractedscientist.wordpress.com/2010/10/14/no-surprises/
-
Graeme Edgeler, in reply to
(still in minor nit territory) true for the last seat, but it’s really “if anyone has NUMBER-OF-OUTSTANDING-VOTES/(NUMBER_OF_OUTSTANDING SEATS+1) votes then they’re elected”
That was because I was using an example where you needed two candidates. Also, it's not number of outstanding seats + 1, the quota is still calculated using number of seats + 1, even after the first person is elected.
-
Interesting. I'd thought giving no ranking would hurt a candidate more than a low ranking.
What happens if you don't number all candidates sequentially? E.g.
Candidate A: rank 1
Candidate B: no ranking
Candidate C: rank 3Invalid?
-
Graeme Edgeler, in reply to
I’d thought giving no ranking would hurt a candidate more than a low ranking.
Lots of people do. That's why I post this every time we have an ranked vote election :-)
What happens if you don’t number all candidates sequentially? E.g.
Candidate A: rank 1
Candidate B: no ranking
Candidate C: rank 3Invalid?
A valid vote for candidate A. In the even candidate A is elected, or eliminated, the vote will not transfer to any other candidate. This is functionally identical to voting Candidate A: rank 1, and leaving the rest blank.
-
Paul Campbell, in reply to
the quota is still calculated using number of seats + 1, even after the first person is elected.
oops you're right - however my main point is that the number of effective votes still in play decreases as people who haven't ranked all 40 (and who can reasonably do that?) drop out. The resulting quota slowly goes down (by ~2000 votes, ~10% in the graph I linked to).
-
BTW I'm an STV fan, but I should point out a couple of down sides:
- scrutineers are practically impossible, at least here in Dunedin, 'counting' (typing it in to the computer) is done 300km (in Chch) away over a period of 3 weeks - a scrutineer would have to relocate for close to a month to do their job - it's so hard that in practice no one does it
- a recount is either impossibly expensive (requiring all the votes to be typed in again), or farcical (running the STV program again on the same data)
In the example I linked to above the last candidate who lost lost by 43 votes but could not afford a real recount
-
Moz,
STV ... It is generally considered a proportional voting system.
My understanding is that with single-member electorates it is not proportional at all, and with multi-member it is somewhat locally proportional.
In other words, if across the country each electorate has one winning candidate, a party with a solid 55% two party preferred wins 100%, and the other parties get 0%. In practice it's more complex, but the outcome is the same - in Australia 10% of the vote gets The Greens 1/150 of the seats, 42% of the vote gets the Coalition 50.6% of the seats (76/150). In a PR system, including NZ's compromised version of MMP, that 10% would translate to at least 10% of the seats. But 4.9% of the vote... zero seats thanks to the anti-democratic threshold, and those excluded votes are normally left out of the electoral statistics (you don't see party vote vs MP totals add up to less than 100% and a "discarded votes: 6%" entry).
With fewer, multi-member electorates there's somewhat less disproportion and it tends to happen lower down the list. With, say, 3 members per electorate, the quota is 1/(3+1) = 25%. A party that consistently gets 20% of the votes after preferences will equally consistently get 0% of the seats.
-
I understand STV as the best method when (unfortunately) electing a single candidate to a position.
What I don't understand are the ramifications of multi-member STV contests. In Wellington, we have two or three seats in each ward, and (some) parties choose to endorse candidates - but they only ever endorse one candidate in any ward.
Is that because putting up more candidates would reduce the chances of any one of them winning?
And it appears to limit the ability to change the alignment of councils - if there's a Green, Labour and National candidate with name recognition and good campaigns, they'll all get elected even though the Green might be on 40% and the Nat on 25%.
(The same applies in FPP, only more so)
-
If you live in the Greater Wellington Area all of your elections will be STV
Hold on. I live in Lower Hutt, and get to vote for members of the Greater Wellington Regional Council, but the Hutt City Council elections won't be STV, will they?
-
Moz, in reply to
Is that because putting up more candidates would reduce the chances of any one of them winning?
In general, yes. Australia has settled on "how to vote cards" given to voters as they enter the polling booth, and widespread publishing of party lists (including on the ballot paper). New Zealand doesn't allow either AFAIK, making it more critical that parties don't split their vote between "people who remember candidate A" and "people who remember candidate B". In theory STV means most of those people will rank either A,B or B,A, but it's that other group who exhaust that matters. Only putting A when B is winning means your vote is wasted (where A,B means you vote for the winning B).
In Tasmania recently there was an upset - the ALP machine put one candidate well down the list, she campaigned for her supporters to vote "below the line" (rank each candidate), and beat out the person above her on the list. In some ways it would be nice if NZ did that, but it would please the 1% who are political geeks at the great expense of the 99% who don't care.
(In single-member electorates Oz does one candidate per party).
-
Paul Campbell, in reply to
What I don’t understand are the ramifications of multi-member STV contests. In Wellington, we have two or three seats in each ward, and (some) parties choose to endorse candidates – but they only ever endorse one candidate in any ward.
I think it's more an issue of branding ....
Is that because putting up more candidates would reduce the chances of any one of them winning?
No, provided that the faithful rank their candidates as #1 and #2 (and I don't think it matters which way around, but I need to think about that more carefully, I can't find an example where it makes a difference)
-
Rich of Observationz, in reply to
*All* methods of directly electing a single candidate to an office such as mayor / president are disproportionate and undesirable -> a candidate can be elected on 51% of the votes and the 49% that didn't elect them will have no say.
(and in the case of a council, the mayor won't have a majority behind them on many issues).
There is a good reason why most countries (of those that bother with the voting thing at all) have an indirectly elected central government.
-
izogi, in reply to
'counting' (typing it in to the computer)
Is this done manually? Somehow I'd assumed there was a scanning process to read voting papers, maybe with real people looking more closely at the rejects.
-
Tim McKenzie, in reply to
I understand STV as the best method when (unfortunately) electing a single candidate to a position.
That's dubious. STV doesn't satisfy the Condorcet criterion; that is, it can fail to elect someone who would beat every other candidate in a two-way race. There are systems that do satisfy the Condorcet criterion.
-
izogi, in reply to
that is, it can fail to elect someone who would beat every other candidate in a two-way race. There are systems that do satisfy the Condorcet criterion.
I haven't yet thought this through to the point of my brain sizzling, but wouldn't there be theoretical scenarios where there simply wouldn't be a candidate who could beat every other candidate in a two way race?
ie.
Voters prefer A over B because of x.
Voters prefer B over C because of y.
Voters prefer C over A because of z. -
Moz, in reply to
Yes, but that's not the question being asked. "A candidate who would beat each alternative in a 2-way match" by definition isn't your scenario. Your scenario can, and possibly does, occur, but the condorcet systems are designed to hide that possibility.
STV is designed to make is possible for a party to build support within a single-member system, while normally single-member electorates are used to achieve the opposite.
The more general issue is people's reluctance to vote for parties or candidates that cannot win (or are seen as such). A lot of effort goes into creating that perception in Australia, both Labor and the constellation Coalition parties consistently talk about "wasted votes" (ie, those not going to them) and disparage the possibility of fringe parties winning anything. The Senate voting rules were recently changed in an effort to handicap minor parties, but then the Prime Munster called a double dissolution election which undid that (halving the quota for a seat meant that even though it was nominally harder to get elected each candidate had twice the chance of election. Kind of "10% harder because of the rule change, but 50% easier because of the double dissolution = 45% easier overall")
-
Graeme Edgeler, in reply to
Hold on. I live in Lower Hutt, and get to vote for members of the Greater Wellington Regional Council, but the Hutt City Council elections won't be STV, will they?
You're right. That was misleading. I listed all the councils that will use STV and Hutt City Council elections won't be run with them, so you'll have to used both :-)
-
Graeme Edgeler, in reply to
I haven’t yet thought this through to the point of my brain sizzling, but wouldn’t there be theoretical scenarios where there simply wouldn’t be a candidate who could beat every other candidate in a two way race?
Yes. It's possible to have a race with no condorcet winner. The complaint about STV is that *if* there is a condorcet winner, STV does not guarantee that that person will win.
eg imagine there are four candidates for a single position. 3 of them get around a third of the first preference votes each, the fourth candidate gets no first preference votes, but gets ever voter's second preference. In a match-up between them and any one of the other candidates, they'd get around 2/3 of the vote, and are the clear consensus choice. STV (or Alternative Vote, which is STV in single member elections) does not gaurantee they will win.
When there are more than two candidates for an election, there is no perfect voting system that will meet all the criteria you might want in a voting system.
I quite like STV because it meets the later no harm criterion (where you cannot harm the election prospects of a candidate by ranking additional candidates lower than them). But it's a judgment call as to what particular criteria you want a voting system to meet. It cannot meet them all.
-
Tim McKenzie, in reply to
I quite like STV because it meets the later no harm criterion (where you cannot harm the election prospects of a candidate by ranking additional candidates lower than them).
As opposed to ranking additional candidates above them? I'm being a bit cheeky there; I assume that what you meant is that your preference among your less-preferred candidates (including whether or not you even have and express such a preference) doesn't affect the chances of your more-preferred candidates. But STV places unnecessarily heavy weight on your higher preferences.
Suppose your first preference is A, your second is B, then C, etc., all the way to Z. STV seems to presume that you care more about A beating B than you care about B beating Z, which is highly dubious if it's a multiple-winner election, and you want both A and B to win, and you really don't want Z to win, even if it means you have to settle for B and C instead of having A as one of the winners.
(If we're still talking only about single-winner elections, then I'm pretty sure that any situation in which a Condorcet-satisfying system fails to satisfy the later-no-harm criterion will be a situation in which there is no Condorcet winner, in which case the tie among the Smith set can be broken by STV, which would presumably satisfy both of us.)
But yes, it is impossible to satisfy all reasonable desiderata for any election system in which there are three or more candidates.
Post your response…
This topic is closed.