Sunday, March 29, 2015

New South Wales Postcount Thread

Legislative Assembly seats won: Coalition 54 Labor 34 Green 3 Ind 2
As of 10 April counting in all LA seats has concluded

This seat follows post-counts for undecided or otherwise interesting seats in the NSW state election, including the Legislative Council.  Each seat has an overview with updates scrolling to the top. I've decided that Lismore, while not quite as thrilling as Prahran, is juicy enough to merit a thread of its own, so that's here: Lismore Postcount.

My review of the result can be found at New South Wales: Decisive Win For Coalition.

I am going to start by posting the Legislative Council and then edit the remaining unclear seats into the post progressively later today.


Legislative Council

Result Coalition 9 Labor 7 Green 2 Shooters+Fishers 1 Christian Democrat 1 Animal Justice 1
Animal Justice defeated No Land Tax and Coalition for final seat
As of 17 April counting in LC has concluded

Friday 5:30 More on this: No Land Tax were flogged senseless on the minor party preferences that were distributed, even from potentially friendly parties in some cases.  Bearing in mind that most preferences exhaust, those voters for the named parties that were excluded very heavily favoured Animal Justice over No Land Tax:

Greens favoured Animal Justice 5631-744
Coalition favoured Animal Justice 1700-1406
Outdoor Recreatuon favoured Animal Justice 2056-1490
Motorists favoured Animal Justice 610-417
Voluntary Euthanasia favoured Animal Justice 1799-711
Socialist Alliance favoured Animal Justice 167-36
Fishing favoured Animal Justice 235-150
Australian Democrats favoured Animal Justice 232-116
Cyclists favoured Animal Justice 300-114
No Parking Meters very narrowly favoured No Land Tax 1411-1393
Building Australia favoured No Land Tax 236-179

Also among the unnamed group tickets the most popular one favoured Animal Justice - Group D 269-185; the rest polled very few votes.

Friday 4:55 Distribution of preferences published.  The AJP made significant inroads on No Land Tax at earlier stages of the preference distribution and were about 2000 behind them at the exclusion of the final Green candidate.  At this point they got a massive (by NSW standards) 11.5% preference flow.  At this stage 4.2% would have been sufficient assuming the remainder of the 11.5% they actually did get did not go to their rivals.  Calculations that the AJP would need 10% of Greens preferences assumed they would not gain on other parties' preferences, but they did.  As it turned out, it didn't matter as they got the 10% anyway.

(Earlier I had a figure of 28.5% up but that was just the AJP share of all Greens preferences that went somewhere other than exhaust.)

Friday 3:30 The AJP have actually won quite comfortably. The margin was first reported as 3583 votes but I think it is actually 3177.

Friday 2:00  Beyond some comments that Green preferences were critical, there is no news about the numbers behind the preference distribution yet.  I am unsure it will be published before next month.

Friday 11:45: In a slight turn-up the Animal Justice Party has won the final spot! The preference distribution will be interesting and I'll have news on it when I have seen it.

Friday 11:10 The Button has been pressed and a result is expected within half an hour.

Friday 17th a bit past midnight: The day of judgement is here.  Antony has responded to my comments about 2007 as a precedent for strong Green preference distribution by pointing out that these were actually BTL preferences with odd histories as a result of the highly unusual situation in which the second Green candidate that year had a surplus from the exclusion of the third. (Those with further preferences were also prioritised as a result of NSW provisions that try to limit the passing on of exhausting votes in a surplus.) So 2007 isn't relevant.  (Nor alas is 2003, in which, as with 2011, the final Green candidate wasn't excluded.) Antony has also mentioned the relevant point that AJP are a long way from the Greens on the ballot paper, which doesn't help them.  So I'll concede on this one and agree that NLT are probably favourites.

Thursday 16th: Final primaries are up and the Coalition appear to have done their dash, falling to third.  The question now is whether Animal Justice have enough below-the-line (and above-the-line non-exhausting) action from the Greens and other minor preference sources to catch No Land Tax,  Antony doesn't think this is happening because preference flows above 10% in the Legislative Council are rare.  

However the Greens' preferences weren't thrown in 2011.  They were in 2007, when of 21 candidates remaining in the count, 46/342 (13.5%) preferences went to the Australian Democrat, Arthur Chesterfield-Evans.  

This year at the time the Greens are excluded the only remaining candidate choices are likely to be: No Land Tax, Animal Justice, Liberal/National, Labor, Voluntary Euthanasia, Shooters and Fishers, Christian Democrats.  So only seven other parties to give preferences to.  Such Green votes as followed the How To Vote recommendation will go to Animal Justice, as will some others.  Most votes will exhaust but it's quite possible the AJP will get enough. 

Anecdotally I've seen claims that Greens voters giving preferences often didn't follow the card.  If the AJP is beaten it could be a case of what comes around goes around - Greens were furious when the AJP preferenced the Liberals in the ACT Senate in 2013 (though as it turned out even with AJP preferences the Greens would not have won).  

Wednesday 15th: The Button will be pressed at 11 am on Friday.  That's not necessarily the end of it as we will need to see the preference distribution and the margins to see whether AJP has any kind of a case for a Dispute of Return, assuming that they do not win.

Monday 13th Ben's latest assessment is a bit less bullish for No Land Tax than Antony's, finding it "entirely possible" that either the Liberals or the AJP are going to catch No Land Tax on preferences.  I haven't had time to attempt any independent modelling of the situation yet and am not sure there is really any reliable way of modelling the ability of these parties to catch up to NLT.  It seems unlikely NLT will get any preferences to speak of, but there may not be the same degree of below-the-line preferencing against them as there was against Pauline Hanson last time.

Friday 11:00 pm Note that if the AJP lose very narrowly not to the Coalition but to No Land Tax, then Antony's scenario of the seat being conceded to prevent a full by-election goes out the window.  

Friday 11:00am Ben Raue's model has a similar outcome to Antony's revised model yesterday: the Liberals fall behind No Land Tax and it is doubtful that Animal Justice get close enough to win (though they might get close enough to challenge).  It appears that No Land Tax are in fact in the box seat.  

Thursday 9:40 I was starting to wonder about Antony's projections when I saw how much the gap was coming down without the Others pile being touched. It has turned out there are more below-the-line votes than Antony previously thought.  Antony's revised projection has the Coalition falling back to between No Land Tax and Animal Justice on primaries, with the final seat remaining a mystery until the press of The Button.  I hope that as we get closer to the final count we can do better than that.

Wednesday 8:00 The gap is still closing but Antony Green has some projections up including details of below the line votes.  His suggestion is that the Coalition is likely to hold on to their 10th seat.

Wednesday 11:20 The gap continues to close as absent votes are added.  The Coalition is down to 9.54 quotas from 9.58 over the weekend, now with 86.5% counted.  While No Land Tax are notionally next in line, I think the Animal Justice Party are the more serious threat because they are likely to attract preferences.  In the event of a very close AJP loss and challenge, Antony Green raises the prospect that the Coalition may be wise to simply concede the seat in order to avoid a 21-seat by-election.

Tuesday 7th 11:20 I haven't added much for a while since there's not been much to add, but with the count now at 85.6%, the Coalition's lead over No Land Tax has dropped back from .20-ish of a quota to .18-ish.  The real action on this front should come as the Others votes are sorted, including below-the-lines.

Thursday 12:00 The Tally Room has a projection of where the primary count might finish up - because no Absent votes have yet been counted, the likelihood is that the Coalition will fall back towards the clutches of No Land Tax and Animal Justice.  As Animal Justice are likely to benefit from left-wing preferences (including from the Greens) they would be in with a serious chance on Ben's numbers, while the uncertainty levels in all such estimates mean that No Land Tax can't be ruled out either.  

Wednesday 12:30 Consistent with its general postcount performance the Coalition has now pulled further ahead with 9.53 quotas and the Greens are back to 2.11.  I doubt the Coalition would have much trouble winning ten if the count stopped now.

Tuesday 3:00 The Coalition has lost some of the gains vs No Land Tax while the Greens continue to drop back.  

Tuesday 12:00 Up to 73% counted and according to the ABC the Coalition has improved a little with 9.49 quotas, No Land Tax have dropped back slightly to 0.39 and the Greens have dropped back to 2.17. 

Monday 2:00 An excellent piece at Tally Room on this count.  Ben Raue notes that the booths included in the count so far have a slight pro-Coalition skew and also (not surprisingly) that the Greens will do better on below the line votes once included.  So there is some prospect that the Greens can get up into the mix for a third seat though they are well behind at present.  And also, there's the elephant in the room: the AJP were potentially disadvantaged by an iVote bungle that could result in them challenging should they narrowly miss out.

Monday 12:45 ABC figures currently show Coalition 9.475 quotas, Labor 6.80, Green 2.24, Shooters and Fishers 0.83, CDP 0.58, No Land Tax 0.407, Animal Justice 0.385.  Both No Land Tax and AJP have improved substantially, though the Coalition has also improved slightly.  If this continues it will be worthwhile having a close look at the possible flows to the AJP especially.  The Greens how to vote card preferenced them third, after Australian Cyclists who will be eliminated.

Opening comments for Legislative Council (29 March):
The NSW Legislative Council is elected on a very different system to the Australian Senate.  21 seats at a time - half of the Council - are elected statewide with a quota of about 4.5%.  Voters can vote above the line, but if they just vote 1 then their vote exhausts when the party they have voted for leaves the count.  Voters can also distribute preferences to multiple parties above the line, or between candidates below the line (which requires voting for at least 15).

Many of the seats are filled by the Coalition, Labor and the Greens based on whole quotas.  Then, while preferences have some influence in close contests, the remaining seats tend to be filled on a first-past-the-post like basis to the parties with the largest remainders.  

There was confusion on election night between the totals provided by the ABC and the totals provided by Elections NSW, which had the Coalition and Labor with noticeably lower percentages of the vote.  However the explanation for this is that the percentages shown by Elections NSW are of all votes including informal votes and unallocated below-the-lines, while those shown by the ABC do not include informals.  

With this taken into account Elections NSW currently have figures of Coalition 43.03%, Labor 31.58, Greens 9.96, Shooters and Fishers 3.91, Christian Democrats (Fred Nile Group) 2.77, No Land Tax 1.67, Animal Justice 1.59 etc.  The ABC has slightly higher vote totals and insignificantly different figures from these.

On current figures the Liberals win nine seats off whole quotas, Labor six and the Greens two.  This leaves four seats for remainders and the highest remainders are currently Labor (.94 quotas), S+F (.86), CDP (.61) and Liberal/National (.467).  

The Shooters and Fishers and Christian Democrats votes shouldn't move around a lot compared to the votes of the Coalition and Labor, so I see no reason to doubt that those parties will each win a seat.  The real issue here is that if the Coalition vote falls from the current 43.0% to around 42.6% then the tenth Coalition candidate drops into a battle for the final seat with No Land Tax (.368 Q) and the Animal Justice Party (.349 Q).  That could get complex, especially since the Greens currently have a leftover .19 of a quota, and some Greens voters will preference the AJP.  The Green vote might also rise as the trivial number of below-the-line votes are entered in, though it is hard to see them getting a third seat at this stage.

More likely the Coalition vote will stay the same or rise slightly in late counting, but with such a high percentage of votes in the post-count we do not know that yet, and there is still a long way to go.  The current figures have the Coalition on a surprisingly good primary vote, only a few points below what they got in the lower house.  

It matters little to the new Upper House - the Christian Democrats and Shooters and Fishers will each be able to pass government legislation by themselves whether the Coalition win the tenth seat or not.  But it is interesting to see if a micro-party can get in from one and a half percent of the vote.



Lismore has its own postcount thread.

Gosford (Lib 11.9%) Chris Holstein (Liberal) vs Kathy Smith (Labor)

Wednesday 8:30 Labor's claim to have officially won by 76 was obviously premature - the count wasn't finished!  The good news for them though is that counting changes were in their favour; they have won by 203.

Tuesday 11:50 The Daily Telegraph reports there is still some checking going on.

Tuesday 7:00 Some apparently faux drama today with the official count going backwards and forwards before settling at the familiar sounding margin of 76 votes.

Sunday 11:00 Two days after Labor claimed to have officially won the seat (which might go to a recount anyway) there has still been no official confirmation of the win published.  Either someone has got ahead of themselves or someone else is slow updating their website.

Friday 2:00: The official site has updated and yes, what happened is that Labor's performance on the remaining absent votes turned around significantly, such that they went from trailing 52:48 in the first several hundred to a virtual 50:50 split, which is slightly better than booth-vs-absent projections off the 2011 results.

Friday 1:15: There are unverified reports (including from the NSW Labor twitter feed) that Labor has won the seat by 76 votes! As NSW legislation only allows four days for postal votes to be received after the election it is possible the count is indeed finished.  The official site shows a Labor lead of 18 votes but the automated entry process may well have picked up errors.  If Smith has indeed won it will be interesting to see what has happened with the late additions (most likely it would be remaining absents breaking to Labor).  Awaiting confirmation or otherwise.

Thursday midday: Smith back in front by 33.  This is off new enrolment votes breaking 188 to 124 to Labor (compared with 102 to 73 last time).  Absents added today continued trending slightly to Holstein.  I still favour the Liberals' chances but as we saw with the Fisher (SA) by-election, when it comes down to the last dribbles of small parcels and projected margins in the dozens you never know what is going to happen.

4:24: More votes added and Holstein's lead is out to 20.

3:50: Another similarly sized bundle of absents have just been added reducing Holstein's lead by ... one vote!

Wednesday 11:30: And verily it did come to pass that the Liberal Party tooketh the lead in Gosford.  Chris Holstein is a massive five votes ahead.  This includes 591 absents which broke to Holstein 267 to 225, but we don't know where they are from or whether they are representative (the usual problem with absents - projecting from small proportions of them to the final total is not recommended).

Tuesday 5:35: Holstein took a big step forward with a gain of over 400 on the postals included so far, but nearly as large a step back with a big gain to Labor in the Woy Woy prepoll.  As a result Smith leads by 37.  Holstein is likely to gain further on remaining postals.  The picture with absents is less clear since in 2011 they were actually slightly poor for Labor on a 2PP basis compared to the rest of the electorate.  Assumptions that such behaviour will hold up from election to election aren't all that reliable so Labor are not out of this.  By standard methods it still projects to a narrow Liberal win but when things get down to the tail end of the postcount like this, past patterns don't always hold up.

Tuesday 3 pm: Finally the 2PPs for various prepoll centres have been added, cutting Smith's lead to 85 votes.  That's not the end of it though as the 2PPs for the postals received to date have not yet been added.  On these Holstein outpolled Smith by 338, about 300 more than expected from votes overall, and the Greens also performed poorly on these votes, so I would expect Holstein to take a lead of at least 250 votes when they are added.  There are also still the Woy Woy prepolls to come.

Monday 11:45 Along similar lines to Ifonly's comments, Tally Room projects the Liberals taking the lead by as much as half a point.  I agree that the Liberals are now well placed to defend this seat.

Monday 5:20 As Ifonly notes in comments, prepolls added to the primary vote (but not yet the 2PP vote) are ugly for Labor in this seat and we should expect the Liberals to take the lead soon.  Absents may then turn things around for Labor but Labor's position is shaky.  I'll try to do some more serious number-crunching of Labor's chances here later tonight.

Monday 12:40: The ABC site is running slightly ahead of the official site on preferences but not primaries and has Smith's lead down to 272 votes (50.45%).  I believe that this means the preferences for the iVotes have been added.

Sunday midnight: This is a straightforward two-party contest.  First-term incumbent Liberal Chris Holstein leads on primaries but is presently trailing after preferences.  At present Smith's lead is 328 votes, giving her 50.58% 2PP. 1738 postals and 2155 iVotes have been counted for primaries but are not yet in the 2PP count.  Holstein's postal vote is running 14 points above his booth vote so when those postals are added the gap will narrow.

The Entrance (Lib 11.5%) Michael Sharpe (Liberal) vs David Mehan (ALP)

Wednesday 8:30  A blowout in Labor's margin, presumably in checking before the distribution of preferences; they have won by 338.

Tuesday 11:50 There have been some very slight changes with the gap at 145 votes but it is generally agreed Mehan has won the seat.

Sunday 11:00 See comment above re Gosford.

Friday 1:20 The NSW Labor Twitter account is reporting that they have won this seat.  Awaiting official confirmation.  (Unlike in Gosford there is not a specific claim of a win with a given margin, so it may be that they have called the seat for their candidate with very few votes left to count.)

Thursday 1:55 Labor lead out to 161 as a result of a strong performance on new enrolment votes.  As with Gosford there were more of these than last election.

4:22 Some more votes (not  absents) have knocked Labor's lead down to 74.

3:50: A small and possibly unrepresentative batch of absents has broken 171-153 to Sharpe with 36 exhausting.  Some other votes, probably postals, have been added, and Labor's lead is down to 94.

Wednesday 11:30: Mehan has improved slightly to lead by 178.  No absents have yet been included in either the primary or the 2PP total.

Tuesday 8:00: Ben Raue has saved me the time for that close look by running a projection which currently projects the Liberals getting over the line by three votes.  In other words, it's a tossup.

Tuesday 6:00: The addition of Liberal-friendly postals to the 2PP count has narrowed Labor's lead to 157.  Labor seemed to be travelling OK on prepolls when I last checked (their 2PPs are mostly not yet added) but I'll have a close look later to see if Labor are at serious risk here.

Monday 5:20: A bunch of prepolls have been added to the primary count but not yet the 2PP.  Labor is roughly holding station in terms of gap to the Coalition on these compared to the booth votes.  We'll have to see the preference spread but this is looking better for Labor than Gosford at present.

Sunday midnight: This is the seat left vacant by Chris Spence, who retired facing ICAC investigation over electoral funding issues.  It's another straightforward two-party contest where the Liberals have a slim (2.2%) lead but are currently being overhauled on preferences.  At present, Mehan's lead is 402 votes, giving him 50.65% 2PP. His gain rate on (mostly Green) preferences is .234 votes per minor candidate vote cast.  2187 iVotes and 1417 postals have been counted for primaries but are not yet in the 2PP count.  Sharpe's postal primary is running 7% above his booth primary.

East Hills (Lib 0.2%) Glenn Brookes (Liberal) vs Cameron Murphy (ALP)

Wednesday 8th: Final result is a win to Brookes by 372 votes.  Murphy will have 40 days to challenge if he wants to file a dispute of return.

Wednesday 11:30: With official confirmation of the size of Brookes' lead it would take an astonishing performance from Murphy on absents to win this seat now.  In 2011 absents were actually slightly worse for Labor than booth votes so there's nothing more to see here. I've added it to the list of apparent wins.

5:55 pm: Brookes' lead has blown out to 767 presumably with the addition of postals.  Could be all over bar the court case.

Tuesday 3 pm: The ABC site shows Brookes' lead cut to 278.  I suspect the iVotes, which were not too bad for Labor, have now been included in the 2PP count.

Monday 11:45 pm: Tally Room projects the margin to close from here but not by enough.

Monday 12:35 pm: The ABC site is way ahead of Elections NSW and has Brookes leading by 622 (50.76% 2PP).  However it is not clear what kinds of votes have been added.

Sunday midnight: Murphy is not quite done for yet, currently trailing by 420 votes, with Brookes on 50.75% 2PP. Murphy is not getting a very strong preference flow, but a large Christian Democrats vote isn't helping there.  As with the above seats the Liberals will benefit when the 2PPs are added for postals, in which Brookes is currently running 10 points above his booth primary vote.  It is common for conservative incumbents to strengthen in post-counts but in this case the greatly increased proportion of prepoll voting, and the introduction of iVote, mean we have to wait to see if those patterns hold.

The hope for Murphy is that because the smear campaign came late in the piece that prepoll votes would be better for him and not reflect it.  No evidence of this so far in the postals.

Background: It's quite surprising that the most marginal Liberal seat appears on track to be retained, but on polling day it was revealed that Murphy, the son of former High Court judge and Attorney-General Lionel Murphy, was the target of a smear campaign in which his "corflutes were plastered with stickers" calling him a "paedophile lover".  I am unaware of any basis for the campaign other than Murphy jnr's former presidency of the NSW Council of Civil Liberties.  In that role, he, for instance, opposed laws allowing for the "relocation" of convicted sex offenders living in public housing.  Another such example issue involved "outing" of child sex offenders.

Murphy's Liberal opponent has denied any knowledge of the smear prior to seeing media coverage, and also noted that some of his own posters had been vandalised. Being unauthorised, the stickers are clearly illegal, but I doubt Murphy has any effective redress if he loses, at least unless some link between the stickers and the Liberal Party was shown.


Ballina has its own postcount thread.


  1. In regard to the LC election ballot paper - At my local polling place I watched how the papers were handed to voters. The paper was gently folded (but not creased) several times. and handed over with the first groups of candidates visible at the top.

    When people got to the stall it was very hard for them to manage to spread the paper out - if they bothered to try.

    Do you think this might be conducive to them voting for candidates in the A to D groups or leaving the paper entirely blank?

    1. Quite likely such issues would have slightly advantaged candidates with good ballot draws or perhaps encouraged informal voting.

      I have a range of concerns about the sheer size of the NSW ballot paper. There is a need for the NSW constitution to be amended to adapt to the use of above-the-line boxes and reduce the number of candidates required to stand for a party box.

  2. looking at Gosford, it would appear that the vote on the night was 31,000 41% to 39% with 28,000 flowing to TPP 49.42% and 50.58%. The postal and prepoll show an additional 7,400 votes with the percentage split 47% 35%. Giving a new primary split of 43% 38%. The absentees are still to come but it looks to me that Gosford is going back to the coalition.

  3. Update on Gosford, I'm going to call Gosford Liberal. The TPP has about 480 votes but this total doesn't include postal which in first preferences has about 500 more for the liberal. The absentee I think will give a more to Labor +Green but be a net gain to Liberals because of exhaustion (no one is giving out how to votes for absentees)