- Date published:
12:15 pm, September 26th, 2019 - 21 comments
Categories: local body elections - Tags: stv, voting
It’s that local democracy time of year again, and thanks to initial efforts at local body voting reforms, even more councils1 are adopting Single Transferrable Vote for wards, have it for District Health Boards, and use Instant Runoff Vote (which we weirdly also call STV in New Zealand- it’s basically STV where only one person can win) for Mayor.
Without going into full voting system nerd mode, (I can do this if requested, and tell you all about further local government voting reforms that would be better) STV is a good improvement on FPP, and a quasi-proportional system where you can still elect individual candidates. It’s not ideal, but if we used it for General Elections, it’d be about half as much of an improvement over First Past the Post as a reasonably proportional system like our implementation of MMP would be. While I normally talk about electoral issues in terms of making them more proportional and accessible to the public, I want to actually do a brief PSA in the voting period on how to vote effectively in STV.
In New Zealand, we rightly recognize that sometimes voting in these local elections are exhausting and you don’t want to do the research to rank all candidates. It’s permissible to partially order an STV vote for this reason, as it makes it harder to cast an invalid vote.
That said, it’s not optimal to do so. If you have time and energy to figure out your relative preferences for all candidates, ranking them all, (or you know, leaving out the last one, because that’s functionally the same thing) can actually swing the election your way a bit more. Yes, I mean you should even rank candidates that you know are “lesser evils,” the only time you should not rank a candidate is if you can’t make up your mind based on the information you can find easily. If you can’t do that, that’s fine, do your best. But if you want more from your vote, I think it’s important people have information on how to vote effectively.
Why should we rank “lesser evil candidates?”
Well, it’s in the name- your vote transfers under STV, but contrary to some people’s understanding, it doesn’t actually transfer in whole all the time if you’ve picked any popular candidates, and in fact in any STV election with more than 2 candidates, it’s rare (and increasingly so as we add more candidates) for a voter’s entire vote to count for one candidate, and mostly only happens if you happen to have the winner of the last round as your first preference.
Looking at a particular ward for example, voters who picked the first winner in Wellington’s Lambton Ward from 2016, Iona Pannett, as their first preference actually only used 62.4% of their vote (also known as a candidate’s “keep value”) to get her elected. Because she was more than a thousand votes over the necessary quota to be elected, the remaining 37.6% of their vote went on to count towards their next preference, if they had a next preference listed. If those voters “bullet-voted” with just a first preference, though, they lose the remaining 37.6% of their vote and it simply doesn’t transfer, lowering the quota instead and leaving subsequent decisions to other voters.
If you’re an informed voter, you should want that residual vote left over from each candidate’s “keep value,” no matter who it will apply to, until it’s all used up. To guarantee you get it all, list all the candidates you reasonably can in order of best to worst. If you don’t have the energy to figure it all out for everyone, that’s okay, vote with an incomplete list. But don’t deliberately leave compromise candidates or lesser evils off the lower part of your list.
Dunedin City Council
Kaipara District Council
Kapiti Coast District Council
Marlborough District Council
New Plymouth District Council (1st time)
Porirua City Council
Ruapehu District Council (1st time)
Tauranga City Council (1st time)
Wellington City Council
Greater Wellington Regional Council
Palmerston North City Council
I don't see the point TBH.
If electing a single candidate, unless there is a massive swing STV is just FPP with warm fuzzies.
You can feel happy voting for your preferred candidate but that vote shifts to the other guy when your preferred candidate didn't win on first round.
It makes the small candidates feel better because they get more first preferences than they would have got under FPP but they still don't get to win.
The main effect is it boosts the main candidates' votes because people will put them in for 2nd (or 3rd) preference instead of those minor candidate votes being entirely 'wasted' in FPP.
But it doesn't make sense for multiple candidates either.
For my Local Board I have FPP for 7 members.
I tick the 7 City Vision members in the list, ta-da.
For my District Health Board I have STV to elect 7 members.
I rank the 7 City Vision Health members in the list 1-7 which takes several seconds longer but has the same effect.
If the 7 City Vision Health members all get in my 7 ranks all had effect.
If less than 7 of them get in then less than my 7 ranks had effect.
Either way there is no point ranking more than 7 candidates so they might as well have just been 7 FPP ticks like the Local Board.
You don't see the point in what, STV? That's not exactly what this post is about. It's telling voters in areas that already have STV how to vote effectively.
But if you want to talk election systems briefly…
"If electing a single candidate" That's IRV- it gets a different name because it falls into a different class of election systems, but for some reason we've incorrectly labelled it as STV in New Zealand, too. It's a dumb idea for voter reform and only worth doing to align single-winner elections with STV voting methods for multiple winners, and you are correct in saying it isn't much of an improvement. It works fine when there are no more than two leading candidates- if you have a close three-way contest the lack of monotonicity comes into play and you can end up needing to vote your most preferred candidate second sometimes. It's rather confusing. STV actually functions a lot better than IRV comparatively, but it's still the rough equivalent of FPP in the class of "proportional multi-winner candidate systems."
What you're describing for your local board is called Block Voting, and is the usual reason for critiques of voting “at large,” because it achieves its simplicity by ignoring proportionality. It's even worse than STV. It has no good use case at all, its only advantage is that it's simple. If you wanted something that simple for you to use, you could use RAV, (Re-weighted Approval Voting) and you could tick any number of preferred candidates instead of the specific number to be elected. It's basically a very simplified form of the RRV (Re-weighted Range Voting) I alluded to, and keeps all the complexity on the counting end of the equation.
Suffice to say, there have been numerous Monte Carlo method studies done on candidate voting methods, (usually referred to as "Bayesian regret studies," after the measure used to quantify voting system performance) and while some of them might be more complex to use, they often come with far more proportional and expressive results that better express the preferences of electorates. FPP is really only good to use when there are precisely two options and one winner is required- although I don't like IRV as a single-winner method, I accept it is indeed an improvement over FPP at least, my beef with it is that it absorbs a lot of energy that could be used for better electoral reforms.
STV has to be the sytem that only a few understand well, another group misunderstand it , all all the rest couldnt care.
The Australian type preferential voting seems to best understood of the preferential number voting systems.
1) You need 50% +1 to be elected
2) minor choices do matter and the order you rank matters [#2 is good and #15 isnt !]
3) the least popular choice is eliminated first.
The whole idea of 'excess votes' just makes peoples head hurt and contrary to idea of voting to eliminate the least likely candidates first if there isnt majority.
…Australian preferential voting is STV, Duke, if you’re referring to their Senate.
If you’re referring to their House, it’s IRV, which is just STV with only one winner, like our votes for Mayor in Wellington, or other cities that have gone full-STV. It’s actually a much worse system, as it has all the disadvantages STV has for being a multi-winner system without the benefits of proportionality.
“Excess votes” is actually not contrary to the idea of STV at all. You get exactly one vote, but it can be split into fractions if more people than necessary vote for a candidate. All multi-winner proportional systems work in similar ways.
You mean the system known as STV? That is exactly what the aussies use for their senate and most of the state elections. I think that their lower house uses something similar
Yeah the lower house uses a easier to calculate form – Two Candidate Preferred (TCP)
The crucial part of that is this
In other words as far as the voter is concerned they do exactly the same as a STV – fill out all preferences. 27 of them in the case of the Auckland
City CouncilDistrict Health Board.
Australian Senate is 'preferential voting' [their words] with bells and whistles ( and defies normal description)
eg short version
:If you vote above the line, you need to number at least six boxes from 1 to 6.
If you vote below the line, you need to number at least 12 boxes from 1 to 12.
Once a candidate for Senate is past the quota only then are their excess votes distributed ( as described for here for STV)
House is preferential voting as described [https://www.aec.gov.au/Voting/counting/index.htm] when the lowest ranked candidate has their second choice distributed and so on .
Excess votes of the leading candidate arent used as described for STV in NZ
Not all the states require every box to be numbered preferentially.
This is how AEC describes the vote counting
House of Representatives count on election night
Immediately after the polling place doors close, polling officials open and empty the House of Representatives ballot boxes. The green ballot papers are unfolded and all the number '1' votes (first preferences) are put into separate piles for each candidate and counted. Informal ballot papers are counted separately (ballot papers that are not completed correctly are referred to as informal ballot papers)…
Following the first preference count, polling officials conduct an indicative two-candidate-preferred (TCP) count – a distribution of ballot papers to two selected candidates. This result is then phoned through to the DRO.
The two selected candidates are those expected to receive the most first preference votes. The TCP count is conducted to give an early indication of who is most likely to win each seat, as this is not always clear from first preferences.
Senate count on election night
Following the House of Representatives count, polling officials open and empty the Senate ballot boxes. The ballot papers are sorted into first preferences for each group above-the-line (ATL) and below-the-line (BTL) and first preferences for each ungrouped candidate as well as those which are obviously informal.
The senate voting is in fact just straight STV, there are only two differences between our use of it in local elections and theirs in federal ones:
1) In New Zealand you can partially list the candidates to whatever degree you so choose, so long as you express a first preference.
2) Australia has the option for parties to recommend a preference order for you, in New Zealand we restrict candidates from telling you how to vote for other candidates in any way, regardless of their party.
I can GUARANTEE you that votes above the quota are redistributed as I described. In fact, I'll let you be lazy: look here, under "transferring the surplus." The Aussie Senate uses STV, and votes above the quota are ALWAYS redistributed in STV, not doing so breaks the proportionality of the system. The only way you differ the transfer is that there are a few different quotas you can use. I believe we both use the Droop quota.
I said the Senate uses STV ' as you described' for the counting. However they call it preferential voting.
You might be interested in this from Andrew Geddis:
(especially if you're in Wellington, faced with a Regional Council that destroyed a perfectly good bus network – with help from the local WCC; or some of the candidates putting themselves up for a ride on the Ticket Clipper's Express to Fame and Ego. Pale, stale and thick as pigshit beneficiaries should show their concession cards on boarding. There’ll be a 15 minute stop at Fame only, and there’s been a decision to put extra First Class carriages on for the comfort of the “well-deserving” candidates that can afford it)
Thanks, I hadn't seen that. Geddis is saying the same thing I am here, coincidentally, but figures he'd be up on how STV actually works.
And yes, I agree with you on GWRC. The only incumbent still running that I'm rating to any degree is Daran Ponter. I'm giving my top rankings to Nash, Victoria Rhodes-Carlin, and Lee. The problem is of course that there will be some right-wing winners to GWRC, and the other cities in the area are not as likely to care about GWRC screwing up our bus system. I wouldn't have minded electing Sue K again, but she didn't want to run again for GWRC.
"I tend to agree with you Matthew". :p I've just done the same thing re rankings. I once worked with Ponter in the ps and will give him a high rating as well – I'm just hoping his cajones are still growing
Oh, btw, hopefully you'll rate Sue K on the Health Board – plus anyone that's actually been a medical practitioner or been involved in health delivery and in contact with actual human patients
"faced with a Regional Council that destroyed a perfectly good bus network –"
You do know that it wasnt the current council , but the previous one that voted to give the contract to a new operator ? ( under government rules that said lowest must win)
Yes Duke, and I'm well aware of Steven Joyce's hand in it all. The current council(s) are by NO means blameless in as much as implementation fuckups of a bad design (Where to begin!), and silo thinking between local and regional councils.
The policy I saw had Gerry Brownlee's signature at the bottom. It required a tender process that was required to allow bids for parts of the network to encourage multiple operators, and all contracts to be selected solely on price. Design was not helped by finding that Snapper data could not be used, or by the need for new depots for the different operators in different places. The underlying aim may have been not to reduce costs, but to make the system disfunctional . . .
Yes, this is the current PTOM. Labour have been slow on repealing it, although I believe the Greens have been pushing them to get it done, and a recent development* actually renews my hope that Labour are considering this but there are technical issues or higher priorities making it wait.
*I may do a post on this soon, but basically Labour has just opened up the possibility for GWRC and other regional councils to delegate public transport functions to city councils. You can imagine how enthusiastic I am to get Wellington's bus system into WCC's hands, as WCC has been much better and more responsive in the wake of the bus scandal. I'm imagining how much hell Iona and Sarah Free would raise with Tranzit et al right now and it is glorious.
Good point. The 'system design' part is best handled by City councils for their urban areas.
From memory Auckland has slowly grown its circular routes around CBD while keeping the old routes that run from isthmus suburbs to the city.
Further out they built the hubs around train stations and terminated local routes at the station instead all running to CBD, but still had a single regular service that covered that route.
For the North Shore they had the busway.
Wellington seems to have invested nothing in bus infrastructure and wanted change soley from route redesign , which was really about saving money
Well, suffice to say, the Wellington system has been an exercise in how NOT to do things.
The really funny thing is that now, a good many of those standing for council (both local and regional) are making promises to fix it – an most of them, but a couple, still with lots of cloth in their ears.
For the Auckland DHB (our STV vote) there were 27 candidates.
First I picked who I absolutely wanted on it and who I absolutely did not.
So Peter Davis at 1 (he has been involved in preventative health policy for years at a preventative stats level, best place for health policy, and I know his ideas) and Doug Armstrong at 27 (obstructionist who adds a negative influence only).
I then processed to do everyone I knew like that because I usually either want them or don't.
Then I put remaining City Vision at the top and the remaining C&R at the bottom.
Frankly after decades upon decades of C&R I can't see that they have added anything except wasted time to my city. Now with the growth of the city the obstructionism and wasted time is directly visible in a creaking infrastructure.
Some of the CV people are a pain, but they do tend to work together for the greater good.
Then I looked for people involved in treating health as a service rather than health as a business either to the community. Health as a business went to the bottom – basically it isn't when you are looking at basing it around a preventative health model. Service to the top with an emphasis on maori (appalling health track record) and non-surgeons (bottom of the cliff usually).
That left me with about 6 in the middle. That was pretty random
Sounds like a great approach. 🙂
Years ago when I stood for ADHB they divided the Isthmus into about 3 wards so the 'tickets' ( which I wasnt on) only had 2 ? candidates each. Now of course everything is supersized up