Western Australian election count endgame

As Labor locks in 41 seats in the lower house, the past form of below-the-line votes suggests some interesting new possibilities in the upper.

Saturday 4.50pm

And now the North Metropolitan preference distribution. The danger spot for the third Liberal, Tjorn Sibma, was the point at which the third Labor candidate, Kelly Shea, was excluded. At this point, Sibma had 33,298 votes to Shea’s 31,561. Had this been reversed, Sibma would have dropped out and his preferences would have elected One Nation to the fifth seat; after which Shea would have dropped out, and Xamon would have been elected to the sixth seat. As it transpired, Xamon was elected to the fifth seat after Shea’s exclusion with a large surplus, which then flowed to Sibma ahead of One Nation.

Saturday 4.10pm

Here’s a pic of the South Metro distribution (HT: Grahame Bowland). With a quota of 49,570, the Liberal #2, Simon O’Brien, emerged with 51,332 to 47,591 for Lynn MacLaren of the Greens. The most vulnerable point for the Liberal Democrats was the exclusion of the Daylight Saving Party, who had been given second preference by everyone in the Glenn Druery network. The Liberal Democrats had 14,105 at that point compared with 13,112 for Daylight Saving, and then received the latter’s preferences as the last party standing in the Druery network. That pushed them ahead of Australian Christians and One Nation, and the preferences of the latter put them well over a quota with 59,150. Their surplus of 9580 votes then went mostly to O’Brien ahead of MacLaren, deciding the result.

Saturday 3.55pm

North Metropolitan: Liberal 3, Labor 2, Greens 1. Final result: Labor 14, Liberal 9, Nationals 4, Greens 4, One Nation 3, Shooters 1, Liberal Democrats 1.

Saturday 3.15pm

South Metropolitan: Labor 3, Liberal 2, Liberal Democrats 1. So Lynn MacLaren defeated, and Labor and the Greens only to total 18 in the chamber of 36. Simon O’Brien re-elected for the Liberals, and the government’s efforts to persuade him to become President set to begin in earnest.

Saturday morning

Results below show the four regions that have been called, plus my readings of the situation in the two that haven’t been.

For those of you who have just joined us, yesterday delivered the shock result in East Metropolitan of One Nation’s Charles Smith squeezing out Liberal incumbent Alyssa Hayden for a result of Labor three, Liberal one, Greens one, One Nation one. Two counts remain to be conducted, and are scheduled to take place this afternoon: South Metropolitan and North Metropolitan. South Metropolitan is particularly interesting, as a Labor-Greens majority depends on Lynn MacLaren of the Greens overturning a lead for the second Liberal candidate, Simon O’Brien, based on the above-the-line vote. It also remains to be established who out of the Liberal Democrats and the Daylight Saving Party will become the election’s sole beneficiary of Glenn Druery’s preference network. I was earlier floating the possibility of One Nation edging out a Liberal in North Metropolitan, while not doing so in East Metropolitan. However, the situation in North Metropolitan is different in that One Nation is relying on the Liberals to drop out before Labor, and so far it’s been Labor that has been losing more ground on below-the-lines.

Friday evening

The preference distribution sheet for East Metropolitan can be viewed here, and One Nation’s shock win can now be understood. It was not, as most of us initially assumed, due primarily to unusual behaviour by below-the-line votes, the addition of which took 0.5% from the Liberals’ vote share and 1.0% from Labor, while adding 0.7% to the Greens and making little difference to One Nation. Rather, it was due to an overlooked technical peculiarity of the counting system used in Western Australia, which produced a different result from what would have occurred in the Senate.

When a distribution of a candidate’s preferences occurs in the Senate, either due to their exclusion or their election (applying only to their surplus votes in the latter case), all the relevant votes are usually redistributed before the count moves on to the next step. The exception is when some of the votes being transferred have been reduced in value because they have already been partly used to elect somebody. In this case, distinct counts are conducted for each bundle of votes with a particular transfer value, in order from highest to lowest, each of which can end with a candidate reaching a quota and being elected.

But in Western Australia, each bundle of votes received, regardless of transfer value, constitutes a distinct count. This has become consequential now that we have micro-parties, in this case Fluoride Free, cobbling together large shares of the vote from a bewildering array of sources, in this case sufficient to turn their base vote of 1262 to 29,860 by the time they were excluded. Under the Senate system, all the votes accumulated by Fluoride Free would have been transferred in a single count, since all had a transfer value of one. The result would have been the election of Greens candidate Tim Clifford with a surplus of nearly 6500, mostly in form of the party’s own votes and preferences from Labor, which would then have flowed to the Liberals.

What happened instead was that Fluoride Free’s preferences were dealt out incrementally, up until the point where Clifford crept over the line. If I understand the situation correctly, the Fluoride Free votes were passed on in the order in which they had received them — first their own votes, then preferences from five small-fry independents (most or all of whom were fronts for the Glenn Druery bloc), the Julie Matheson group, Flux the System and the Daylight Saving Party, of which all but Matheson and one of the independents had the Greens ahead of One Nation. The votes accumulated by Clifford to this point where then used up in electing him, and no surplus was transferred.

However, many more of Fluoride Free’s preferences remained to be distributed, and proceeded to do so at their full value between the only remaining candidates: One Nation’s Charles Smith, on 35,759 votes, and the second Liberal, Alyssa Hayden, on 36,718. These were the larger transfers Fluoride Free had received later in the count, and they included the 4552 votes of Animal Justice, which would have mostly been spent electing Tim Clifford under the Senate method, but instead went to Smith ahead of Hayden. Also favouring Smith were the 8292 votes Fluoride Free had received from Australian Christians, 6491 from Shooters and 2269 from the Micro Business Party. Only the Liberal Democrats, with 3355, and Family First, with 3201, had favoured Hayden.

The vote recorded at the final count is 48,178 votes for Smith and 43,327 for Hayden, but the more telling point is that Smith’s score is only 347 votes clear of a quota. Still undistributed at the end of the count was the 1609 vote Greens surplus, which would have gone to Hayden, and 2388 remaining to Fluoride Free, which I’m unclear about.

Friday afternoon

5.45pm. I’m informed that that’s it for the day. South Metropolitan and North Metropolitan are now set to be finalised at 3pm tomorrow.

4.30pm. A shock in East Metropolitan — One Nation wins a seat at the expense of the second Liberal, for a result of Labor three and one apiece for Liberal, Greens and One Nation.

3.30pm. The WAEC is now going through South West, East Metropolitan and, most interestingly, South Metropolitan, with North Metropolitan to be conducted tomorrow. The South West result is in, and as expected, it’s two for Labor and one apiece for the Nationals, Liberals, Greens and One Nation.

Thursday evening

It seems the WAEC’s indication the four outstanding regions would be resolved today was over-optimistic, and we will now have to wait until tomorrow. The WAEC still hasn’t published the preference distribution files, but you can access them here for Agricultural and here for Mining and Pastoral. To summarise what happened:

Mining and Pastoral: Out of 49,311 formal votes, Robin Chapple’s win for the Greens in Mining and Pastoral came down to margins of 189 votes over Labor at the penultimate count, and 736 over Shooters Fishers and Farmers at the end. The quota was 7045, of which Labor had two off the bat, the Nationals and Liberals one apiece. One Nation got there when the exclusion of Australian Christians unlocked preferences from the Liberals, whose group voting ticket had the Christians second and One Nation third. The final seat came down to Shooters (3507 votes), Greens (2873), Labor #3 (2821), Nationals #2 (2380) and Flux the System (2478), who gathered the micro-party preference snowball. Nationals preferences then flowed to Shooters, and the Flux pinata scattered every which way, leaving the score at Shooters 6563, Greens 3854 and Labor 3665. Labor preferences then flowed to the Greens, who finished with 7389 to the Shooters’ 6653. A different Liberal preference strategy might have elected Shooters ahead of One Nation, or Labor #3 ahead of the Greens (though not both).

Agricultural: There were 88,175 votes in the count and a quota of 12,597. The Nationals started with two quotas, Labor and Liberal with one apiece. The micro-party snowball didn’t amount to much, with the Liberal Democrats going out before the Greens. Labor was then elected with preferences not just from the Greens, but also the Daylight Saving Party, Flux the System and the enigmatic Family First. Presumably this reflected a partial deal with the Glenn Druery bloc, in which the Liberal Democrats and Fluoride Free refused to participate. That left One Nation on 10,777, Shooters on 8,752 and Liberal #2 on 4,241, after which preferences from elected Labor and excluded Liberal went to Shooters, who defeated One Nation at the final count by 13,299 and 10,820. In other words, Liberal preferences could have elected One Nation if they had gone that way, but they actually had them third behind Shooters.

From these results, the effect of adding below-the-lines to the above-the-line totals is to add 0.3% to the Greens vote, and take 0.3% from the Liberals and 0.5% from Labor. If that was borne out in the other four regions, the One Nation possibility in North Metropolitan would disappear, as it is contingent on the Liberals falling behind Labor; and the Greens would fail to get ahead of Liberal #2 in South Metropolitan, in part due to a reduced Labor surplus, but would win their fight against Labor #3 in South West.

Thursday noon

The button has been pressed in Agricultural, delivering the long anticipated result of two Labor, two Nationals, one Liberal and one Shooters. The other four regions were, as of yesterday, scheduled to happen today.

Wednesday 2pm

We have the first finalised upper house result in, so I’m bumping this post. The Greens are relating that Robin Chapple has been re-elected in Mining and Pastoral, in what can only have been a close race against Labor’s third candidate, which presumably means a result of two for Labor and one apiece for the Liberals, Nationals, One Nation and Greens. More on that shortly. The other five regions, last I heard, are set to be determined tomorrow.

Monday morning

Preference counts have been completed and published for 21 of the 59 seats. The key point of interest relates to Kalgoorlie, where until now we could not be sure that the Liberals had won the seats over the Nationals, since there remained the possibility that the Liberal would fall out at the second last count. In the event, Liberal candidate Kyran O’Donnell had 8533 (34.7%) to 6656 (33.6%) for Labor’s Peter Forster, with Nationals candidate Tony Crook bowing out with 4812 (31.7%). With the distribution of Crook’s preferences, O’Donnell prevailed over Forster by 8533 (56.2%) to 6656 (43.8%).

The tables below compare the behaviour of preferences from the 21 seats counted so far with those from all 59 seats at the 2013 election. This is done by adding together preferences as allocated from the relevant party’s point of exclusion, at which point there may or may not be candidates other than the two major parties to pass on preferences to. “Pollution” refers to the combined share of the vote passed on which the relevant party had acquired indirectly as preferences. One Nation preferences did appear to flow quite solidly to the Liberals, certainly in comparison with the federal election, where the split was all but 50-50 (11 of the 21 seats counted had One Nation candidates). Conversely, strong evidence emerges that Greens preferences flowed more heavily to Labor than in 2013. The only seats out of the 21 with Nationals candidates were Geraldton, Kalgoorlie and Murray-Wellington, which may not be representative.


Formal distribution of preferences starts today, which presumably means we’re pretty much at the end of the primary vote count, so I haven’t bothered with another update of the results. The picture is clear, with all the close seats (Pilbara, Jandakot, Joondalup, Murray-Wellington and Kingsley) having broken to Labor. Still very much of interest though is the upper house, where we have the rather peculiar situation of what is presumably an all-but-complete count of above-the-line votes, but no information whatsoever on below-the-line votes. The latter are simply being put through the data entry process, without being used to update the first preference count. So we’re pretty much in a state of suspense until the button gets pressed, which is expected on Wednesday or Thursday.

For this reason, I’m now offering two scenarios for my simulations – one based on the raw numbers currently in the system, and the other with rough adjustments made according to how much difference the addition of below-the-lines made to first preference totals in 2013. What this amounts to is an improvement for the Greens (who I’ve bumped 0.3% in Agricultural and Mining and Pastoral, and 0.6% everywhere else), and cuts for the Liberals (down 0.4% in Mining and Pastoral and 0.8% everywhere else) and the Nationals (down 0.4%).

This suggests the Greens are not out of the hunt in South Metropolitan, contrary to what the above-the-line projection has been suggesting, and could deprive the Liberals of a second seat in South Metropolitan — which is to say that it’s not out of the question that there could be a Labor-Greens majority. However, the politics of this are complicated by the fact that the second Liberal is Simon O’Brien, and reports yesterday suggested Labor hopes to tempt him with an offer of the President’s chair. Either way, Labor and the Greens between them would have 18 seats out of 35 on the floor, with the President only exercising a vote in the case of a tie (which would not arise when all members were present).

Another result to emerge from estimating the numbers inclusive of below-the-lines, which I was not expecting to see, is the fairly dramatic possibility of One Nation stealing the Liberals’ third seat in North Metropolitan. The ABC projection shows the danger point is after the exclusion of Family First at Count 19, leaving very little to separate the four remaining candidates: Greens (11.78%), Liberal #3 (10.48%), One Nation (10.45%) and Labor #3 (10.15%). So the Liberals only need to fall back by 0.33% at this point to be excluded, in which case their preferences would elect One Nation. Below-the-line votes should ensure the Greens keep their head above water here, so they look sure to take the other seat that’s outstanding at this point. My cursory examination of past form suggests One Nation at least holds its ground when below-the-lines are added, so this could get interesting.

In another development, the above-the-line projection now shows Labor in danger of losing its third seat in Mining and Pastoral to Flux the System and the micro-party preference snowball, but that won’t happen if below-the-lines behave as they normally do. I’ve also got the Greens improving in the race for the final seat in South West at the expense of Labor #3.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

113 comments on “Western Australian election count endgame”

Comments Page 1 of 3
1 2 3
  1. So, the big question is: if Labor needs Green votes in the upper house, what will the Greens try to extract? Will the Greens allowing big budget cuts without McGowan slugging the miners? We shall see.

  2. If they do get the chance, the first thing Labor should do is enshrine a one vote one value law, exactly as happened in Qld and SA after the gerrymanders were finally broken. If it takes giving the Senate presidency to a Liberal or even the Flux member, so be it.

  3. In the Lower House, eight seats including Joondalup and Kingsley are now shown as “awaiting declaration” but I cannot see full preference cutups for them.

  4. Hard to see a result other than Libs and LDP winning last 2 South Metro seats but always worth hoping. Only about 16,000 BTL votes and Greens are behind about 5500 on ATL. Lib vote not likely to drop as much as usual because quite a few Libs would vote BTL to avoid preferences going to PHON. Plus Lib voters BTL much less likely to be fooled by LDP position. And a some % of Labor BTL will not preference Greens ahead of Libs.

    Also hard to see DSP heading LDP. ATL gap is about 800. LDP vote will no doubt drop to maybe 1% BTL as most potentially fooled Lib voters avoid them. But that would still keep them level on first votes with DSP for BTL votes. Everything else pretty being much as expected or random so hard to see gap closing much.

  5. Antonbruckner11

    Hoping tax the miners. The federal government has failed to tax them fairly or fully. And managed to redistribute 70% of WA’s GST, despite decreasing revenues.

    Well I hope so.

  6. And Kay
    Ah Kalgoorlie.
    Everyone votes for their own reasons, not necessarily along classical ideological lines. Remember that some of the Green redistributed vote started in Flux and FFS, sorry I mean SSF.
    Some of those voters may have just been voting for the littlest parties first and leaving the majors (for Kal) to last. Maybe the candidates were both their friends? Maybe they were anti the sale of public assets and after PHON their vote went to ALP 😀

  7. In regards to preferences in Kalgoorlie, I think the far more interesting factor is that more preferences from One Nation voters went to the National rather than the Liberal (with plenty of other going to Labor as well). 814 to the Nats, 792 to Libs, 515 to ALP – so not much more than a third to the Libs. Assuming the One Nation how to vote card recommended giving next preference to the Libs, even when one takes into account that 15% or so of those One Nation preferences have been ‘polluted’ by coming from other parties first, it’s not a great endorsement of the Liberal-One Nation preference deal.

  8. Andrew – someone from WA can confirm, but I thought the Lib Phon preference deal only applied in LC. But I’ve no idea what appeared on Phon HTV for lower house.

  9. Wfield, the main point of the preference deal was that ON would prefer the Libs in Assembly districts in return for the Libs giving ON a high place on their GVTs for the Council regions. Was always going to work better for ON than for the Libs – all that ON needed was for Liberal-voting sheep to put a 1 in the Lib box and the Lib surpluses would flow to ON automatically. Whereas for the Assembly the Libs needed ON people to be on the ground handing out HTVs (or the Libs to supply people in ON T-shirts to hand them out), and they needed ON voters (rebellious, fed-up sort of people) to then sheepishly follow the party’s recommendation. As Andrew has shown, that didn’t work, at least in “Kagoorlie”.

  10. The Mining & Pastoral result is a bit of a surprise! Will be great to see the actual figures, but one would think that the Greens would be feeling confident in South West and hopeful in South Metro (although you’d want to see if there is a highish rate of Liberal BTL as well in view of the One Nation deal).

  11. We’ll all know for sure soon enough, but Robin Chapple’s win gives stronger hope that the Greens may also get over the line in South Metro (although there are factors unique to Mining & Pastoral which don’t apply in South Metro, so it’s no certainty the same thing will happen)

    Apart from this giving the party 5 seats – equalling their best ever outcome – when at one stage in the counting it looked like they may only get 2, it would also mean a clear ALP-Greens majority even with a Labor person as President.

  12. Yes Andrew it’s looking very like 14 ALP, 5 Green. So if the Greens can win that many on silly boundaries, will they be prepared to vote for 1 vote 1 value boundaries (or at least something much closer to it than the present fudge)?

  13. Considering the Greens have had a swing against them and picked up more seats, I’d think that would indicate to them that just because they did well under this system in this *particular election*, it mightn’t treat them well in the future. It’d still be in their (and our, being the general public) interest to change it.

  14. @ Airlines, the ‘swing against them but they did better’ is simply because the minor parties were more active this time, going from 6.78% to 19.3%. Generally, minor party votes in the senate flow to the Greens before the Liberals or Labor. The Greens going from 8.21% to 7.96% (and likely higher as BTL votes seem to be coming in strong for the Greens) is a very small swing against them considering how many new parties there were (particularly, Animal Justice Party on >1%, compared to 0 votes last election)

  15. As Labor learned during the Gallop government’s first term, they need either a) 19 votes on the floor to eliminate malapportion, or b) for the Greens to change the law to give the President a deliberative vote if there are only 18.

  16. I think changing the role of president is likely more difficult given precedent in the Westminster tradition, the only argument that could be made is to the Federal Senate where it’s held that you’re disadvantaging the represented area by denying the President a vote but that argument is founded on principles diametrically opposed to fixing the malapportionment. You’d still need 19 votes for that anyway wouldn’t you ? Barring abstention that would result in a tie allowing the speaker to vote.

  17. I don’t understand why the principle involved in arguing for WA to adopt the same approach as the Senate would run counter to the principle involved in fixing malapportionment? They are both based on encouraging more equal representation for the voter.

  18. WB

    for the Greens to change the law to give the President a deliberative vote if there are only 18.

    But even if there was only 18, and the president has a deliberative vote, wouldn’t they need one more for a majority? 18-18 isn’t going to pass bills.

  19. The Senate principle is that states deserve equal representation (irrelevant of their vast difference in population) and that denying the President a vote would violate that.

    Fixing malapportionment in the WA LA regions relies on the argument that districts do not deserve equal representation but rather people do.

  20. 18-18 isn’t going to pass bills.

    But how would 18-18 happen in the first place? You’re starting out with 36 members. One sits as president and doesn’t get a vote (except if there’s a tie). Assuming no absences or abstentions, a tie will never happen and legislation either passes or fails on 18-17 (or wider margins if more people support/oppose). With one absence or abstention, you can get a tie on 17-17. Then the president can cast the deciding vote, for 18-17 again.

    How/when is 18-18 possible?

  21. The difference between Green + Labor controlling the LC will probably come down to a few hundred votes in South Metropolitan Region, either way. Interesting, and we might know today!

  22. To change the relevant act, you need to have the votes of an absolute majority of all members, i.e. 19. But the President only gets a vote if it’s a tie – and an 18-17 vote in your favour isn’t a tie. So you either need to win the vote 19-16 on the floor, or win it 19-17 by giving the President a deliberative vote.

  23. AR

    How/when is 18-18 possible?

    Never, unless the president has a deliberative vote (and if the president is silly enough to go ahead with a vote without checking the numbers). As it is, assuming every MLC is present, you would get a government bill defeated 17 to 18 if a bill is that contentiously close.

  24. @Elaugaufein I see your point. But I think the argument could equally be made that each electorate/region deserves equal representation – with or (ideally) without malapportionment.
    @William – I am assuming a decision to give the President a deliberative vote only needs an 18-17 vote, rather than 19-16?

  25. @Andrew B

    I always thought that Victoria has a fair model for the LC. You get one vote one value, but unique representation for each rural region despite the smaller numbers. However, it might not translate well to WA as perhaps Victoria has a slightly better balance of metropolitan to rural population (correct me if I’m wrong).

  26. Where, Wakefield? Can’t see anything on the general ABC News site, and Antony is still showing only M&P and Ag results. And yes I hit refresh.

  27. Raaraa: Melb. has a population of 4.1m, in a State of 5.9m. Perth has a population of 2.1m, out of a State of 2.6m. Significant difference, indeed – from ~68% to ~80% in the capital. Also, Melbourne accounts for 5 out of the 8 VLC regions (62.5%) – Perth accounts for only 3 out of the 6 WALC regions (50.0%).
    Combine these, and you get around an 8-9% malapportionment for the VLC, but a 30% malapportionment for the WALC. It’s a bloody disgrace!

  28. Indeed Matt. Thanks for providing the numbers.

    And I think I’m with Wakefield on this one. I don’t think the results has been finalised yet in any but the 2 regions.

  29. Thats a pretty bad end result for Labor in practical terms by my read unfortunately. No “neutral” party got up, which means they can’t give an Independent the Presidency and there’s no reliable path to 18 votes (let alone 19) absent the president’s vote.

    The Nats, LDP and Shooters might each be amenable on certain things but only the LDP is even going to theoretically back malapportionment reform (which fails to make 19 unless the President gets a deliberative vote).

    Though denying the speaker a vote while requiring an absolute majority in an even chamber is the kind of idiocy that it’s hard to believe anyone overlooked .

  30. Yes Elau, but that’s not yet the result. I think Wakefield is misreading Antony’s page – easy enough to do because it’s presented rather confusingly, with a disclaimer that needs very careful reading.

Comments Page 1 of 3
1 2 3

Leave a Reply

Your email address will not be published. Required fields are marked *