Making STV work tactically

September 23rd, 2013

Wellingtonians are remarkably confused by their STV voting system. I’ve heard few who are confident they know how to maximise the chances of getting what they want with their vote. Perhaps some are quietly confident – they know how, and are happy that others will mess up their voting and have less impact.

The working rule is really simple, but before telling you – a bit of history.

Nigel Roberts on Stuff over the weekend sang the praises of STV, calling it “fair”. That premise about what we need most from a voting system also opens the the Royal Commission report that gave us MMP. It is dopey. The test should be what is most likely to deliver decisive government which can nevertheless be easily tossed out. That is democracy's vital work, tossing out the passengers after we've seen they're useless, and those we're sick of. Its only clear advantage is getting rid of leaders who should go. "Fairness’ systems fudge and negate that advantage.

As always I’m interested in how systems work in practice, not in how they look to those who use self-referential slogans like “fairness”.

I sat on the Select Committee that gave local government the choice of voting systems. I’m sure that few of us understood how it all worked even after multiple explanations. There was particular mystery for those who found ‘maths’ hard at school. The Sainte Lague vote counting method was eventually approved by the majority on trust, with open admission that that they did not have a clue how it worked. Rod Donald was the major influence. I was intrigued by his conviction that Green voters would be best at working voting tactically. He said they would make sure they educated their supporters, the way Aussie parties do. I suspect that he was actually counting on the ‘sounds nicer’ bias among young voters, and expecting to get more dopey second, third and fourth preferences.

He may have been right. Neil Harrap in a letter in today’s DomPost summed up STV as “ending up with whoever the fewest voters dislike”. When most voters don’t understand that any vote rank, however low, is treated as some degree of liking for the candidate, ninnies who say nice things are more likely to succeed than those who say what they mean, even if a majority agree with them. The passenger types will get more second and third preferences.

Now – the simple rule.

Vote only for people you would be happy to see winning. Stop there. Do not rank candidates just to show who you least prefer. Do not rank candidates to help make sure that at least a dog beats the genuine idiot.

Remember – all vote rankings are positive. It is a no brainer. Your vote will be transferred if your favoured candidate does not win. Only vote as far down the sequence as you actually prefer. Stop so that you do not give any votes to candidates who should not be elected at all, even if they’re better than the worst.

Having been an Alliance Party candidate disqualifies Jack Yan, and somewhat surprisingly I have too little information on John Morrison, to vote for him first, after he supported the damaging ‘living wage’. Though the incumbent is a decent person, she has dithered too much to deserve the leadership.

So I’ll vote first for Nicola Young, then Morrison, then no more.

In ward voting the same approach will maximise the impact of your vote. For example, do not list Helene Ritchie at all, even last. Her graceless stupidity got a fresh outing against one of my colleagues supporting personal responsibility for offensive drunkenness last week, but even without that reminder I’d have thought of her as one of the main beneficiaries of ignorant voting on name recognition.

Filed under Constitutional, Politics

Comments

[…] Stephen Franks has good advice for people confused by STV: […]

Thanks for the reminder – But I thought the reason why Kerry Prendergast lost to Celia Wade-Brown at the last election was because many Prendergast supporters ticked Prendergast only & no other names, while many Wade-Brown supporters ticked a range of candidates whose votes were then transferred to Wade-Brown. Is that not correct? Kathryn

Yes excellent Stephen, all votes are positive. Edgler has a point about putting people you hate last. This has an emotional satisfaction. For instance the Australian Labour party request to voters , Vote liberal not Pauline Hanson, use hate.
I say that the preferential votes of the second highest poller be counted before the low ‘wacko’ votes.
I would hate to see this sytem in general elections though. Imagine labour second preference Green, Green second preference Labour, we would need a brutal campaign and arm twisting Nat second preference NZ First. On second thoughts this is crazy,
can we have first past post again please.

Not quite, Kathryn, but I think I know what you’re getting at.

Kerry Prendergast’s supporter’s voted for her, by giving her their first preference. Celia Wade-Brown’s supporters voted for her, by giving her their first preference. We don’t know what other (subsequent) preferences both Prendergast’s and Wade-Brown’s supporters gave to other candidates, because the votes given for both of them were never looked beyond the first preferences. There was no need to look further, because Wade-Brown was never excluded from the count, and, while Prendergast was ultimately excluded, the votes given for her (including those transferred to her during the count) were not on-transferred (because, by then, the winner had been found).

Now, as the other candidates were progressively excluded from the count, through lack of support, those who had voted for them favoured Wade-Brown over Prendergast, by a margin of about 2 to 1. In other words, two out of every three of those voters ranked Wade-Brown higher than Prendergast on their voting papers – Celia 2, Kerry 4; Celia 4, Kerry 5; etc.

The numbers say it all. Prendergast received 21,809 first preference votes. Throughout the count, a total of 2,896 votes were transferred to her. Wade-Brown received 18,560 first preference votes. Throughout the count, 6,321 votes were transferred to her, 3,425 than to Kerry, and a ratio in her favour of 2.18:1.

At the end of the count, of the 49,586 voters whose votes then pointed either to Wade-Brown or to Prendergast, 24,881 supported Celia, and 24,705 supported Kerry, a margin of 176.

At the time, Prendergast complained that, of the candidates standing, she was the only one on the Right (of politics), and that that was why she lost. In other words, there were no other Right-leaning candidates contesting the election whose votes would largely have transferred to her as they were successively excluded from the count.

While that was probably true, it is by no means the whole story. As we see above, it was not the number of candidates who opposed her that mattered, but the number of voters.

In this regard, it could be said that one of the reasons why Prendergast received 3,249 more first preference votes than did Wade-Brown, was because there were indeed no other candidates of the Right for Right-leaning voters to vote for.

So, what she gained at the start of the count, 3,249 (by having no competition on the Right), she lost as the count proceeded. What Wade-Brown lost at the start (by having competition on the Left), she gained (6,321
– 2,896 = 3,425) as the count proceeded. Celia’s gain during the count (3,425), less Kerry’s advantage at the start of the count (3,249) = 176.

Either way, when all was said and done, a majority of voters supported Celia over Kerry, and that’s all that mattered.

Nigel Roberts has long been a doyen of the left in NZ Political Science. You can guarantee any advice he gives will be designed to help secure the relection of that hard core socialist do gooder green, Celia Wade Brown.
Without analysing the issue, you could be sure that once the drumbeat supporting the STV voting system started a decade ago, you could be sure that such systems are nor more fair, what they are, in the case of the STV system is a voting system designed to load the dice even furthur loaded in favour of the left that MMP. In essence the extra work the voter has to do to analyse which way to vote, combined with the inevitable vastly greater number of left wing candidates who appear electable and plausible means your average right centre voter and ordinary professional or worker is somewhat confused and miscalculated how to vote to get Kerry Prentergast reelected. Its hardly worth mentioning that all the usual suspects are working overtime to secure the reelection of their beloved interfering wowser killjoy girl Celia. Some of the worst slanders occurred a week half a ago on the hard left 9 to 12, Laidlaw commentary and interview programme in which the Media Review segment, enabled a barely disguised hard core slander of the main Celia opponent, Mystery Morrison, slandering Morrison with accusations of various crudities and effectively suggesting Morrison was propositioning,female public servants for sex sessions in the shower.