Sunday, March 16, 2014

State Election Post-Count: Denison

This is the first 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. 

WINNERS: Bacon (ALP), Groom (Lib), O'Connor (Grn), Archer (Lib)
CONTEST: Ogilvie (ALP) vs Amos (ALP) - other ALP candidates getting insufficient preference flow
STATUS: Ogilvie leading and will win count.  If close may be subject to protest but likely to not be close enough.

Sunday 16th:

This thread concerns the post-count and preference cutup in the seat of Denison.  Denison (progressive results here) is the one seat that was absolutely settled at party level on the night.  

On-the-night quota totals are 2.26 Liberal, 2.05 Labor, 1.29 Green, 0.18 PUP, and the remaining 0.22 is split between six independents, the Socialist Alliance and the woefully-performed Tasmanian Nationals.  Scott Bacon and Matthew Groom are re-elected with quota, Elise Archer is re-elected and so is Cassy O'Connor.  

The remaining interest in this seat is in a four-way battle between Labor candidates Madeleine Ogilvie, Julian Amos, Alphonse Mulumba and Sharon Carnes for the seat vacated by Graeme Sturges.  On the night Ogilvie lead Amos by 217, Mulumba by 426 and Carnes by 551.  However these figures are dwarfed by what will be the first major event of the cutup for these candidates - the throwing of Scott Bacon's surplus, currently worth 3336 votes.  Following this will come the exclusion of all the minor candidates including PUP (Barbara Etter at one stage threatened to win a seat in polling, but in the end PUP collapsed to only 3%), at least two Liberals and three Greens.

A situation will arise with Cassy O'Connor elected, Elise Archer elected or nearly so, and with the four Labor contestants, the Greens Bill Harvey and the Liberals' Robert Mallett, left in the count.  Probably the first losing Labor candidate will be excluded before either Harvey or Mallett.  The other three will then have the benefit of that Labor candidate's preferences.  Harvey and Mallett may be excluded with either two or three Labor candidates remaining.  Whenever they are excluded their preferences will be available to the competing Labor candidates.  It is possible those of Mallett's preferences that don't exhaust will favour Amos.  However Carnes' preferences should favour Ogilvie (flow between female candidates is a notable feature in Hare-Clark).  Mulumba is a wild-card in the mix because his migrant background may be attractive to Greens voters - but the Green preferences to be thrown will be those of Harvey, and Harvey's voters should be at the informed end of the Green spectrum and may be aware (if they care) that Mulumba is socially conservative. 

If vote total data by polling booth become available before the count ends, it will be possible to determine how the votes for Labor candidates compare to the Bacon vote by booth, and this may be useful in projecting the outcome.  (For instance if Sharon Carnes is strong in Glenorchy booths and Bacon is also strong there, this may give her a chance despite her low primary.) For the time being my view is that any of the four Labor candidates can still win, but on the throw of Bacon's surplus in a week and a half it should be possible to exclude one or more from contention.

Update Sunday evening:  A potentially disastrous (but hopefully harmless, albeit embarrassing) development with the news that 163 predominantly Liberal votes were destroyed as a result of an accident with a letter-opener.  See the TEC press release. My understanding is that the damage to ballots occurred in the area of the Liberal column, and that this is why Liberal candidates were disproportionately affected.

On that basis it would be nice to be able to say that since there is no contest within the Liberal ticket or between the Liberals and any other ticket in Denison, that therefore the loss of these votes won't affect the outcome compared to if they had been included.  Unfortunately the lost votes do have a potential impact - firstly some #1 votes were for other candidates by the sounds of it, though it is not known to me if the competing ALP candidates were among these.  Secondly assuming the lost votes are mostly 1 Liberal with very few for Bacon means that Bacon's surplus becomes more valuable by maybe 25 or so votes, which could make a difference of a few votes each to the battling ALP candidates.  Thirdly there is some potential for the issue to affect the point at which Robert Mallett is excluded, which might then affect which Labor candidate is elected.  These are example possible impacts; there are others.

The TEC will be hoping that there are no wafer-thin margins between competing Labor candidates in the cutup, so that the issue can't alter the outcome.  However even if it did, it would be surprising if a losing Labor candidate challenged their comrade's election anyway, since a by-election caused by one Labor candidate protesting another would be a terrible look for the party. That said, the Electoral Commission itself might petition as per WA.

This is an out of character event from an electoral commission that is generally a world leader in quality service and speed of information.  In all the TEC counts I have witnessed over 26 years I have never seen anything like it.

Counting update Sunday night: Ogilvie leads Amos by 188 (narrowed) but has gained over Mulumba (496) and Carnes (606).  It would not surprise me if Bacon's preferences helped Carnes; she should not be written off.

Tuesday 18th: On a part-rechecked update Ogilvie leads Amos by 191, Mulumba by 524, Carnes by 636.  Bacon's surplus is now worth 3840 so proportionally the gap to Carnes is much the same as on the night, while that to Mulumba has grown and that to Amos has shrunk.  There was some small good news in the destroyed ballots saga with eight Humpty Dumpties readmitted to the count, reducing the loss to 155.  The Mercury carries the news that 72 of the destroyed ballots have no number in the ALP column.  However as noted above, the loss of these 72 votes could still affect the outcome by altering the quota total, which affects the value of Scott Bacon's surplus, the value of the Liberal leftovers and potentially the exclusion point of Robert Mallett.  On the other hand, even if there is a margin of, say, fifty votes between the last two Labor candidates, it is possible in theory that incomplete data from the damaged ballots can be used to determine that they could not have affected the outcome.

Scrutineering Update:  I have received some scrutineering figures concerning the Bacon surplus that show that across the electorate Julian Amos is performing very well on Scott Bacon's surplus.  If these figures are representative (the sample size is decent) then Amos will pass Ogilvie on Bacon's surplus and could lead by 400+ votes, while Carnes and Mulumba will be out of it.  Remaining issues will then include whether the Carnes preferences favour Ogilvie, how voters for the minor Green candidates respond when they choose between an economic conservative by Labor standards (Amos) and a social conservative by Labor standards (Ogilvie), and the known unknowns of Denison's 6.7% of PUP/micro/indie riffraff.   There's a bit of a case for installing Amos as the slight favourite at this point - especially given his higher profile - but without any real level of confidence.

Interestingly Amos has been quite vocal about the TEC envelope bungle while the other Labor aspirants have not yet spoken on it.

Tuesday Evening: Ogilvie's primary lead is now up to 202.

Wednesday 19th: Ogilvie's primary lead is now at 206.  A possibly more important development is that through the count the Liberals have been gaining (now up to 2.30 quotas despite having some of their votes snippered) and the Greens slipping back (now down to 1.26).  If this is confirmed and not just an artefact of some parties being more checked than others then it is a positive sign for Amos.  As mentioned above we don't really know yet what lurks under the Green preference rock, so the potential leader of the inter-Labor race would want whatever it is to be as small as possible.

Thursday 20th:  Not a lot to add here with Ogilvie's primary lead now 213.

Friday 21st: Ogilvie's lead now 216.  As votes are added it's to be expected her lead will grow slightly.

Monday 24th: We're still waiting for the last of the postal votes but the TEC have sped things up by provisionally distributing Scott Bacon's surplus.  The provisional distribution sees Julian Amos indeed in the lead as foreshadowed based on the scrutineering samples mentioned above, and that it is a race in two between Amos and Ogilvie.  However the gap is less than projected at only 190 votes which is not a comfortable advantage for Amos.

It's also not yet a comfortable advantage for the TEC given the saga of the paper-eating letter-opener.

Leakage from the Bacon surplus has been fairly low.

Tuesday 25th: We've now seen the Groom surplus and the preference distribution for the nine candidates who embarrassed themselves by being beaten by the Socialist Alliance.  Amos might have been considered a chance to profit from Liberal leakage but in fact he has made no progress today with Ogilvie now 188 behind.

9pm: A few more exclusions (including one Independent and the Socialist Alliance) and Ogilvie has made a little more progress with the lead down to 180 votes.

Wednesday 26th 1 pm: Amos has stretched his lead to 237 votes thanks to preferences from Marti Zucco.

6 pm: The margin continues to move back and forth by small amounts.  Ogilvie gained 12 over Amos on leakage from Penelope Ann (Green) and another two on leakage from Deborah De Williams (Liberal) but then Amos gained seven on the exclusion of Barbara Etter (PUP).  We now have one of the big ones: Alphonse Mulumba, the fourth ALP candidate, who has finished 194 behind Sharon Carnes.  Carnes should be next after that and we may have a much clearer picture of this seat at this point.  Incidentally some sampling reported to me showed high leakage rates in Mulumba's vote.

8:30 pm: A major development with Ogilvie gaining on the preferences of ticketmate Alphonse Mulumba to close to within 78 votes of Amos.  Although Ogilvie and the candidate now excluded, Sharon Carnes, have very different factional backgrounds such distinctions are frequently lost on voters and it would not be surprising to see more support for Ogilvie here.  Wherever we are after this, there will be only the preferences of Bill Harvey and Robert Mallett to go.  Leakage from minor Greens candidates so far has favoured Ogilvie over Amos while Liberal leakage has been a mixed bag.  At the moment I prefer Ogilvie's chances.

9:10 pm: We are most of the way through the exclusion of Carnes, and Madeleine Ogilvie is back in the lead for the first time since the primary count.  She leads Amos by 33 votes, with 927 votes received by Carnes from other sources, and an 11-vote smidgin from Carnes via Elise Archer, to go before Carnes is finished. If the current margin was final it may not be the end of the matter, so whoever wins would also want to win comfortably enough to eliminate any doubt caused by the damaged ballots saga.

March 27 11am: The Carnes exclusion is finished and Ogilvie now leads by 74.  It was not clear that the remaining Carnes preferences would favour Ogilvie given that many came from Bacon (whose initial preferences favoured Amos) so this is more good news for Ogilvie.  We now have 4001 Bill Harvey preferences and 4408 from Robert Mallett to come to decide whether we have a clear winner today or whether the answer could be "see you in court".

12:42 pm: The Greens' preferences have favoured Ogilvie.  She now leads by 507 with only the 4645 Liberal votes to come.  Most of those will exhaust so the flow to Amos over Ogilvie will have to be quite pronounced to even get him inside the proverbial "margin of litigation" arising from the damaged ballot papers.  (Incidentally, at this stage it is not clear exactly what that margin is; whether, for instance, a 50-vote margin would survive challenge is something that would require very careful analysis.)

2:13 pm: Amos has made some inroads on the primary-vote share of Mallett's votes but it's nowhere near enough.  The gap is 366 with only 999 left to throw.  Those remaining are votes that came to Mallett from Matthew Groom and Scott Bacon.  Many of them will exhaust making closing down the remaining margin highly unlikely.  After finishing second-last in this seat on primaries in 2010, it looks like Ogilvie has won this time around.

3:30 pm: It's over; Ogilvie has won by 331 votes.


  1. Thanks as always for your analysis, Kevin.

    Is each ballot individually entered into a computer at some point (as I think it is with the Senate)? If so, is there anything stopping 'the button' from being pushed earlier? If at every point of exclusion the margin required to change the result is greater than the total number of possible remaining postal ballots for that electorate, could the results be declared?

    Even if the margins are smaller than the number of possible remaining ballots, could the button still be pressed and re-pressed as ballots arrive? People such as yourself could still provide the chances for each candidate based on that - many would still be in the dark, but some candidates might be revealed as winners regardless of how the count progresses.

  2. Ballots are not entered into a computer and there is no "button". All distributions are conducted by hand. The vote totals are entered into a computer which then instructs which part of the preference distribution process is to be conducted next.

  3. OK thanks. That was my original understanding, but I thought things may have changed. Does the Senate, perhaps, have a software "button", or did I dream it?

    1. Yes, the Senate has computerised entry of below the line votes and a "button press". This is much easier in the current (and hopefully soon to be ex-) Senate system because so many votes are 1 above the line and don't need to be individually entered manually.

  4. Interesting to hear Ogilvie is a bit of a social conservative. From what ive read, shes very progressive on asylum seekers.

    Pro life?

    1. Primarily anti-euthanasia. Has also supported some changes to recent abortion law amendments (at least in the area of what doctors who might have moral objections or non-mainstream views about health risks are allowed/required to tell patients). And has supported (albeit in a questionnaire only to my knowledge) the so-called "Nordic model" of attempting to control prostitution by criminalising the purchase of sex. (Support for this model does not necessarily entail a conservative viewpoint overall as some "radical" feminists also support it.)

  5. Having spoke to Madeline several times, I believe the social conservative tag doesn't quite hit the mark. She may not be quite in Greens supporter territory, but she is definitely in the centre-left on social issues.

    1. I don't think anyone here has used the "social conservative" tag without some degree of qualification - eg "by Labor standards", "a bit of" etc. On those terms I can't see why it isn't valid. In terms of the general left/right assessment of a range of social issues her positions on specific issues would be very variable.


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.