Sunday, March 16, 2014

State Election Post-Count Thread: Bass

This is the fourth of my state election post-counting threads.  There will be one for each electorate and they will be updated frequently.  During the cutup (which starts in the middle of the week after next) I will be on remote fieldwork and will update nightly, but hope to take a day off during the cutup.  

SEAT OUTCOME: 3 Liberal 1 Labor 1 Green.
WINNERS: Gutwein (Lib), Ferguson (Lib), Courtney (Lib), M O'Byrne (ALP), Booth (Grn)

Sunday 16th:

Bass is one of the two economically struggling northern electorates where the Liberal Party has done exceptionally well.  They're currently sitting on 3.44 quotas (57.4%).  Labor is on a distant 1.40 quotas, the Greens have 0.75, PUP have 0.31 and we should not forget the Australian Christians with 0.06.  The two indies don't have much.

Peter Gutwein and Michael Ferguson are elected with both well over quota.  Michelle O'Byrne is far enough ahead of Brian Wightman that I very greatly doubt that he can catch her, especially as Senka Mujcic's preferences should favour O'Byrne.  





So that's three, and as the Liberals have well over three quotas they will have a new MHA who should be either Sarah Courtney or Barry Jarvis, though Leonie McNair isn't totally out of it if she can get a super-strong preference flow from Gutwein and Ferguson.  

On the raw party totals the other seat seems totally cut and dried with Kim Booth (Green) so far ahead of the leftovers for the major parties; it seems like it should be a simple 3-1-1.  

Wightman getting up?  Very unlikely

It's not that simple, and the reason it is not that simple is that Michelle O'Byrne is so far short of quota.  Hare-Clark is about candidate totals not just party totals, and if the 1.4 quotas for Labor splits fairly evenly between two candidates, then they are each not too far behind Booth's 0.75.

This cutup will proceed first of all by the throwing of the surpluses of Gutwein and Ferguson.  They have 0.75 of a quota spare between them and this will mostly go to their ticketmates.  With Michelle O'Byrne still in the count (unlike in 2010) perhaps a shade more than 10% of these surpluses will leak (edit: quite a lot more than 10%, see Monday comments below), dropping the Liberals back to around 3.36.  

From that point on we'll gradually lose the indies, the Christian and the four PUP candidates, the four minor Greens, three minor Labor and probably Leonie McNair (unless she does exceptionally well off the Ferguson/Gutwein surpluses) and be left with a situation in which there are probably still only two elected with Booth (Green), O'Byrne and Wightman (ALP) and Jarvis and Courtney (Lib) all in the count.

Whoever is last here - and it won't be Booth on current figures - loses.  If it is Wightman his preferences elect O'Byrne, creating a surplus that mostly exhausts and probably splits fairly evenly between Booth and the Libs, and Kim Booth wins without problems.

But if it is one of Jarvis or Courtney who is excluded at this point, their preferences elect the other Liberal and then create a surplus.  Most of this surplus exhausts but what is left of it has the choice of Wightman, Michelle O'Byrne and Booth.  Especially if it is Jarvis who is defeated, Wightman is more socially conservative than O'Byrne and could benefit by hundreds of votes vs Booth.

Given that Wightman starts 1138 behind Booth (and Booth should get more from his Green ticketmates than Wightman from his Labor ones) that's not enough, but there are also 4000 preferences that will favour the Liberals and Labor over the Greens to be thrown from the minor parties and independents.  In this electorate PUP only have four candidates, so their votes may be a little less likely to exhaust than in Braddon. 

The case in which Wightman gets up is what chessplayers call a fantasy variation.   Everything must go perfectly for him.  He must get at least a good share of preferences from the other ALP candidates, and have Jarvis and Courtney perform unequally on Liberal preferences (otherwise he can't get ahead of them) and perform strongly on preferences from PUP and others while there are still several candidates in the field and then perform very well on Liberal preferences.  For a one-term incumbent it's really hard to credit all this happening, and the Greens' current 0.75 quotas seems sufficient.  Preference flows are not that strong in Hare-Clark elections.   The most likely situation by far is that Wightman is either eliminated before either Liberal, or else can't get enough from the Liberal surplus to catch up. 

I've also been asked about the chances of the Liberals getting four by the same sort of method.  This seems even more remote.  The Liberals will gain on Christian and PUP preferences but they would likely need to each get at least as much of the Labor surplus (Wightman's preferences) as the lone Green in Booth and I cannot see that happening, even if the Liberals are ideally evenly split.  In such a contest Booth would be the incumbent.  (Edit: because of higher Liberal leakage (see below) we can rule this out entirely; the Liberals won't get four.)

Courtney vs Jarvis

So it's most likely Bass is 3-1-1 and that the real issue is which of Courtney and Jarvis gets the third Liberal seat.  That could be clearcut once the two Liberal surpluses are thrown.  Courtney has some possible advantages here - she is likely to have more profile across the electorate while Jarvis's support is more regional, she is ahead on primaries, and the remaining Liberal candidate to be presumably excluded is female.

(Further reading on why 0.75 quotas plays 3.44 and 1.40 is even potentially an issue for those who really need to know: Ginniniderra Effect.  Ultra-wonky article detailing a case in the ACT in which a Green with 0.60 quotas actually managed to lose to a Labor ticket with 2.41.  But on current figures the gap is nearly twice as large and despite unfriendly preferences Booth seems OK.)

Sunday 16th evening update: Today's extra figures made no real difference, which is good news for Booth.

Monday 17th clarification: In today's Mercury a preference flow from PUP (45% Liberal, 40% Labor, 15% Green) was given with me as source.  That flow was actually from the federal election not the state election.  The state flow will be different, particularly because a fair percentage of the PUP votes will exhaust.  Also, Kim Booth is higher profile than the federal Green candidates.  Nonetheless I expect PUP preferences to be unhelpful to the Greens in general.

Today's updated figures have strengthened Booth's position slightly, but indeed any day that goes by without him appreciably dropping back is a good one.

Monday night update: leakage higher than expected? I am getting some interesting scrutineer reports of very high leaks in certain booths from the Liberal ticket to the Labor ticket in this electorate.  There's a reason for them that I have realised - leakage from the Liberal surpluses should be way higher than in 2010 not only because O'Byrne is still in the count but also because both Gutwein and Ferguson are elected.  In 2010 Ferguson's surplus leaked 8.9% directly to other parties, but 69.8% of it went to Gutwein.  When that portion going to Gutwein was thrown as a surplus, 10% of it leaked in turn.  This means that the total leak from Ferguson's surplus (had Gutwein already been elected) would have been 16%.  With O'Byrne still in the count and with Wightman also an incumbent (meaning Labor has two incumbents compared to zero in 2010) it could be even higher this time.

Why does this matter?  Apart from it putting paid to any chance of four Liberals winning, the main recipients of this leakage will be Labor rather than the Greens, so there could be a 400 vote or so leg-up for Wightman out of it.  Not likely to be enough for him to win since he was projecting as a thousand or so behind otherwise, but enough that I am still not calling this seat.

(Update Monday 24th: These reports proved unrepresentative when preferences were distributed.  It may be that the extent to which Labor were on the nose reduced leakage from the Liberal ticket sharply compared to what it would have been.)

 Extra modelling  To try to look more closely at Wightman's chances, I've set up a model in which I assume that Wightman and O'Byrne split all votes reaching (or thrown within) Labor equally.

I've also assumed:

* Exhaust rates of 60% for Liberal votes (could be higher), 40% for PUP votes, no exhaust assumed for Christians and ungrouped (though I suspect there will be some)

* That Wightman beats the fourth Liberal

* That because there are less votes to be thrown within the Greens ticket that can leak than the Labor ticket, the Greens have a 250 vote leakage advantage

This model required Labor to get 89% of available non-exhausting preferences (including Liberal excess, PUP, Aus Christians and ungrouped) to Booth's 11%, which is not realistic, though they may get close. 

I conclude from this that not only do Labor need to do super-strongly on preferences from other parties compared to the Greens, but they also need Brian Wightman to somewhat outperform Michelle O'Byrne through the cutup in general.

Given she's outpolled him on primaries by a ratio of nearly 1.3 to 1, I don't think that is likely.

Tuesday 18th Figures added today have not affected my assessment above.  Booth's notional lead continues to grow in raw vote numbers but in proportional terms is about stable.

Wednesday 19th: This morning's figures have changed little (I suspect they are mostly primary vote rechecking) but for what it is worth Booth's position has improved very slightly.

Wednesday evening: Some scrutineering sampling with a good sample size (see comments) and suitable source suggests that Courtney will win the third Liberal seat and that the split in the Liberal vote is unequal enough to get Wightman over the 4th Liberal hurdle and into whatever showdown he might be able to manage with Kim Booth. 

Thursday 20th:  Minimal changes today but Booth's notional lead before preferences has increased slightly.

Friday 21st: Further improvement in Booth's position.  The 89% target for Labor in my model above is now up to 90.2%; of course they won't get that but any other scenario in which they might supposedly win becomes more difficult.

Monday 24th: The TEC has provisionally delivered the Gutwein and Ferguson surpluses.   Liberal leakage hasn't ended up being all that high; they are running at about 3.37 quotas which is more consistent with my original 10% estimate than some of the figures in circulation.  (Indeed it was a mere 8.6%!).  It also doesn't seem like Labor has that big a leg up from Liberal leakage from the surpluses, and indeed the improvement for Wightman so far is negligible.  (The target figure in my model has "improved" to only 90.1% of distributed preferences.)  Courtney's lead over Jarvis is enough (over 800) that her chances are very good.  Whether Wightman stays ahead of Jarvis remains touch and go.

Looking at the Liberal leakage figures further it turns out that of the leaks, Michelle O'Byrne got 144, Wightman 135, Booth 103.  Booth thus got 27% of those going to one of these three candidates (cf the mere 10% needed in my model).  The only reason the model does not show his position improving is that leaks do not exhaust.  We may see weaker preference flows for the Greens later (since some voters vote across parties for all the incumbents) but I am not seeing realistic chances for Wightman here.

Tuesday 25th: We've now seen the exclusion of several minor candidates.  Labor has now gained 554 votes compared to 190 for the Greens, but this comparison flatters Labor since more Green minor candidate votes have been distributed while Labor have a lot of leakage to come.  So far O'Byrne has also received more preferences than Wightman which is not helping the latter.  Wightman is still in danger of getting excluded behind Jarvis. Indeed assuming the O'Byrne/Wightman split on Labor preferences is about even and the Courtney/Jarvis split likewise, then Wightman needs hundreds of votes more than Jarvis out of PUP preferences and Green leakage.  So he's in quite some danger of not even clearing that hurdle.  Everything points towards a Booth victory at this stage.

9 pm:  Exclusion of Labor candidates continue to favour O'Byrne over Wightman which is more bad news for any chances some may think Wightman might have.  At the moment Wightman could need around 400 PUP preferences more than Jarvis to pass him.  That's no guarantee.

10:50 pm: Senka Mujcic is now being excluded which will leave us with a clear gap between Booth and Wightman as candidates without any votes to come from their own party - probably about 1400-1500 votes.  After that come the Palmer United preferences and then the votes of Leonie McNair, at which point we will see if it is Wightman or Jarvis to go next.

Wednesday 26th 12:45:  I'm typing from the electorate of Bass - the middle of Wingaroo Nature Reserve, Flinders Island, to be exact!  The PUP preferences are all distributed and Wightman has not gained on Booth on them at all; indeed Booth is not that far short of quota now and is almost ahead of O'Byrne.  Jarvis now needs over 1500 more than Wightman, from Leonie McNair's 3800+, to get over Wightman, but my view is whether he does so or not is academic.

5:30 pm: Wightman has outlasted Jarvis by 339 votes.  The Liberals have 5457 votes that will eventually be distributed to O'Byrne, Wightman, Booth or exhaust.  If O'Byrne gets 1700 she will pass a quota. However more likely so many will exhaust that none of the remaining candidates will make quota and Wightman will simply be left last.  Indeed the effective gap between Booth and Wightman has changed little all through the cutup,and Labor has made limited inroads in total vote terms.

The Greens ran an ad saying that the only thing worse than having Kim Booth in Parliament is not having Kim Booth in Parliament.  There will be a probably very grumpy Kim Booth in Parliament this term.

Thursday 27th 9 am:  The distribution of the Jarvis preferences is complete leaving only the Courtney surplus to be done.  That's 1632 votes and Booth's lead over Wightman is 1214.  Even if none of the remaining votes exhausted Wightman would have no chance given that Michelle O'Byrne is also still in the count, but in fact the number staying in the count at all will be less than the Booth-Wightman margin.  Wightman has made some progress through the count but nowhere near enough.

11:10 am: And it is done.  In the closest inter-party result thus far, Kim Booth has retained his seat defeating Brian Wightman by 1034 votes.  As expected, the candidate margin is closer than the party margin (the Greens as a party are 2196 votes closer to their next quota than Labor are) but it doesn't change the outcome.






9 comments:

  1. Jarvis is somewhat advantaged over Courtney because he will end up with a triple donkey, compared to Courtney with only a double.

    ReplyDelete
  2. It appears there are forty different sheets (in Bass) because of the different numbers in the various parties. Some of those may be replicates.

    The booths I saw today suggested Courtney may be better preferenced than I'd anticipated in my counts.

    ReplyDelete
  3. Some stats from today:
    - The cut up of the Lib votes from Ferguson and Gutwein runs about 5:3 in favour of Courtney (n=193).
    - The consequential leakage (i.e leakage that ends up somewhere other than Courtney or Jarvis at the exclusion of Wightman) from the Liberal vote is around 6.3%. (n=300)

    ReplyDelete
    Replies
    1. Ta. If that is representative then Courtney will very likely win seat 3. Do you know if those Lib votes are from specific booths and if so which ones?

      Re leakage, do we know Wightman is getting excluded before (presumably) Jarvis? If Jarvis is getting only 3/8 of the Liberal vote and some of it is leaking then that puts him on about 6K, not counting ALP leakage and fourth-party preferences. Wightman really should just over that - by a few hundred - unless O'Byrne really hogs the minor Labor votes.

      I very much appreciate the scrutineering comments I'm getting from people on the site + via email + Twitter especially as I have not had time to do any myself because of upcoming fieldwork. However I will keep updating every night while away, technology permitting.

      Delete
  4. Having now modelled those figures, I get Jarvis falling behind Wightman prior to his exclusion, which would lead to the interesting situation of Courtney being elected (without requiring too many votes), then O'Byrne (to whom Liberal leakage heavily favours), leaving whichever of Booth and Wightman come out on top afterwards; Booth needing maybe 1500 to the 2500 needed by Wightman.

    ReplyDelete
    Replies
    1. As you can see above I get something similar - Wightman should beat Jarvis. I'm assuming the 1500/2500 needed are figures to go to quota. If that is so then the problem for Wightman is that most of the Lib votes will exhaust on Courtney's surplus (Lib exhaust rate is usually around 60% in this situation) and he'll need a massive flow of non-exhausting votes.

      Still if he is within 1000 at this stage and the Libs have 3500+ left, he might have a show. Comparable case is the 2006 cutup with Booth vs Reissig (ALP). Reissig was low profile but the votes that had gone Fry-Gutwein-Napier and then run out of Liberals split 70:30 to Labor. Admittedly Fry was an ultra hardliner with a big anti-gay reputation, hardly someone whose supporters would preference the Greens. Votes with the history Jarvis-Courtney (presumably often going through Gutwein and Ferguson too) might split 75% to Wightman over Booth if he is lucky, on account of Wightman's higher profile.

      Delete
  5. Ballots filled past 12 are running from 15-40% depending on the booth; higher in the lower socioeconomic areas. Most of those not done so are filled to 5.

    ReplyDelete
  6. Hi Kevin,

    Could you please put dates next to the days you list. Just easier to keep track of everything, especially for newcomers to your site. Also thank you for keeping the updates coming, no Antony Green to rely on.

    Cheers,

    ReplyDelete

The comment system is unreliable. If you cannot submit comments you can email me a comment (via email link in profile) - email must be entitled: Comment for publication, followed by the name of the article you wish to comment on. Comments are accepted in full or not at all. Comments will be published under the name the email is sent from unless an alias is clearly requested and stated. If you submit a comment which is not accepted within a few days you can also email me and I will check if it has been received.