Saturday, October 15, 2016

ACT Election Live and Post-Count Comments

Labor minority government returned - ALP 12 Liberal 11 Green 2.  
This was also the interim outcome

Postcount Comments

(NB Through the week I will be posting comments here only in the evening, because of work commitments)

What happens now is that votes are gradually entered into the computer and a series of interim preference distributions are released, leading up to the final "button press" which is expected to happen next Saturday.  As we get closer to the full count these should become more accurate as an indication of what will happen, but for the time being it is useful to keep an eye on differences between the current primary count (which is based on more votes) and the primary count for the interim distributions, and also to bear in mind that the counting of postal votes may favour the Liberals.

Saturday: It's over!  In Brindabella Nicole Lawder (Lib) has won the final seat from Labor's Angie Drake by 553 votes.  Drake outlasted Steven Bailey by 307 votes at the second-last exclusion, but had Bailey survived that exclusion he would have lost anyway, probably by more than Drake did.  In Ginninderra the Ginninderra Effect has indeed struck again - Labor with 2.74 quotas at the final exclusion have won three seats while the Greens with 0.78 at the same point have missed out.  This is because two of the Labor candidates had a very evenly split share of their party's votes, and indeed it wasn't all that close between them and the Greens (701 votes for Tara Cheyne over Esguerra with Gordon Ramsay actually ahead of Cheyne at that point).  In Murrumbidgee the Greens are up by 800.  The other two electorates were decided by thousands.



All this confirms an undignified feat for the Canberra Liberals - an Opposition with a reduced share of the seats.  In Australian electoral history at state and federal level this isn't as rare as one might suspect, happening about 20% of the time federally for example.  Federal drag (the Liberals in charge federally this time whereas Labor held office federally last time) clearly makes this more likely to occur but all the same the result can't be considered good.

Friday: Further distributions are further locking in the expected result of 12-11-2.  In Brindabella, Bailey is now over 200 behind Drake though he will have no hope of catching Lawder if he does get ahead.

Thursday: Could be the final nail for the Greens in Ginninderra ... Gordon Ramsey has passed Chris Bourke in the fight for the third Labor seat.  Ramsey does quite well out of Bourke's preferences and is 454 votes ahead of Esguerra in the current distribution.

Wednesday: The new distribution for Brindabella has Bailey four votes short of overtaking Labor, off primaries of 2.45 quotas for the Liberals, 2.03 for Labor, 0.48 for Sex Party.  The actual current primaries have the Liberals on 2.51 and Labor on 2.01, so for the time being Bailey is probably making the final pair.  However, if he does make the final two his position is actually hopeless, because so much of Angie Drake's vote will be soaked up putting her ticketmates over quota, leaving Bailey needing nearly all the surplus to win.  It seems highly likely now that Lawder (Lib) will win the seat.

For Murrumbidgee the current distribution has Le Couteur winning by 0.135 quotas off primaries of .678 Q for the Greens and 2.505 for the Liberals.  The current primaries are .635 and 2.570 respectively, but the gain for the Liberals is split between three candidates.  So Le Couteur's real lead is probably about 0.07 quotas.

It's still Groundhog Day for the Greens in Ginninderra.  They're only losing by 84 votes (.022 quotas) in the distribution but the real margin is probably about three times that.

Tuesday: There is a new distribution but the progress in data entry isn't massive and at a quick look it doesn't tell us anything much.

Monday night: There is no new distribution but there are primary vote changes.  In Brindabella, the Liberals are back up again to 2.52 quotas with the Sex Party back to 0.48 quotas.  On these figures I would expect the Liberals to win.  In Ginninderra, the Greens are down slightly to 0.59 quotas and Labor is down slightly to 2.50.  When both these parties fall, Esquerra loses whatever the Greens lose, but Labor losses are spread across three candidates, so this has actually helped Labor.  No real news in Murrumbidgee, save that no news is good news if you're in front.

Sunday 1:40: In Murrumbidgee, Caroline Le Couteur is .163 quotas clear of the Liberals in the latest interim distribution, since when she has gone up slightly on primaries and the Liberals have gone down slightly.   That's equivalent to about 2.7% of the vote, so massive changes would be needed for the Liberals to beat her there.

Sunday 12:50:  The second interim distribution for Ginninderra has the Greens' Indra Esguerra still missing out, except now she finishes sixth.  In that interim distribution Labor is on 2.524 quotas, the Greens on 0.618 and the Liberals on 1.864.  At the critical point Labor would have 2.723 quotas and the Greens 0.7955, but each of Labor's individual candidates would have more than Esguerra, so the Ginninderra Effect would strike the Greens again in the same electorate!

Looking at the current primaries, Labor is on 2.525, the Greens on 0.609 and the Liberals on 1.831.  So unless this turns about it will be Groundhog Day for the Greens in Ginninderra.

Sunday 12:45: A second interim distribution was released late last night.  In Brindabella, off primaries of Liberal 2.507 quotas, Labor 2.031 quotas, Sex Party 0.426, it had the Sex Party .074 quotas short of getting over the third Labor candidate.  The actual primaries are 2.442, 2.033 and 0.507, so on current numbers the Sex Party are probably just making the final cut (but whether this remains the case will depend on their performance in remaining counting.)  If we are lucky the next interim distribution will have the Sex Party making the final contest, and that will provide a basis for modelling both Liberal vs Sex Party and Liberal vs ALP for the final seat as more votes come in.

--------------------------------------------------------------------------------------------------------------------------

Election Night Comments

Wrap (10:30) That's just about all from me for this evening I think.  So what we have is a result with very little primary vote change, with a modest swing against the Liberals and a gain by a disorganised gaggle of micros and independents. For the most part the fourth party challenges have not amounted to much, but a sensational result for the Sex Party in Brindabella puts them in with a shot at a share of the balance of power.  It remains to be seen whether Labor has just retained its previous position, or strengthened it to an outright win or an alternative path to passing legislation.

In the absence of any polling, the election was a shocker for those pundits who expected a Liberal win, a close call or a heaving crossbench.  As noted just after polls closed, they ignored the impact of federal factors, and that the ACT is not an easy place to win in.  When Katy Gallagher led Labor to victory in 2012 she did so through the force of her own popularity with her party on the nose at federal level.  It is easier for Andrew Barr with the Coalition in power federally, and this has mitigated any "it's time" factor.

10:00 At the moment Brindabella is the only seat I see major doubt about, but I'll keep an eye on the interim distributions for the others.  For Brindabella it still looks like the Liberals should just hold, but that is no sure thing.

9:30 The Sex Party continue to improve in Brindabella, to the point that on current numbers they would more than likely get over the third ALP candidate (but this may change after postals).  The Liberals are down .12 quotas and the Sex Party up by the same amount since the interim distribution, which had the Liberals .22 quotas ahead of the Sex Party at the point where the latter lost.  That might sound like the Sex Party are winning, but the Liberal losses are distributed across multiple candidates, so they're probably not quite there at present.  However if they do get over Labor's third candidate, Labor's third candidate then puts Labor's first two in with surplus, and that surplus might put the Sex Party over the third Liberal.  The difficulty for the Sex Party could be postals, where I still think the Liberals should improve.

9:20 I've had a look at Kurrajong where the Liberals have fallen well below two quotas but they seem to still be fine for two seats by some margin there.   

8:55 The Brindabella black swan scenario is getting really interesting.  The Liberals have now dropped 0.10 quotas off the interim distribution, with most of it now going to the Sex Party.  That's putting Labor ever closer to majority government, except that it's now lineball whether the Sex Party outlasts their third candidate.  The Sex Party is now on a strong 8.1% and is in contention for the seat on current numbers, but probably still short of getting over the third Liberal.  It looks like a Liberal hold still (especially with postals) but the next interim distribution will be interesting.

8:32 Le Couteur's posiiton in Murrumbidgee on primaries has improved a lot since the interim distribution with the Liberals down about 0.07 quotas and the Greens up 0.02.

We should also keep an eye on the shifts in Brindabella.  The Liberals have dropped 0.07 quotas there too since the interim distribution, but it has gone to the Greens and Sex Party, not to Labor. At present, although the Liberals lead Labor by 0.45 quotas, the seat is actually pretty close.  The notional distribution sees a party total gap of 0.52 quotas close to 0.13 - the causes being preferences (0.20 quotas) and the even spread of votes between the Labor candidates.  Assuming half the change since comes off each Liberal, their notional lead now for the final seat might be down around 0.09 quotas.  One question is whether the Labor team can stay that even.  Postals would also be likely to help the Liberals.

8:22 Antony has a theory that as the evening goes on the spreads between votes within parties are increasing (electronic voting encourages "linear voting" is the idea.)  That could help the Greens in Ginninderra and the Sex Party in Brindabella, though both are currently a long way short.

8:05 Interim distribution for Yerrabi is 3-2 to Labor by a very large margin.

8:03 Interim distribution for Murrumbidgee has Caroline Le Couteur (Green) beating third Liberal Peter Hosking by 0.059 of a quota.  This one is likely to be too close to call for a long time.

8:01 Interim distribution for Kurrajong is 2-2-1. Nothing to see here.

7:58 Interim distribution for Ginninderra has the Greens eliminated in eighth.  They are extremely close to getting over at least one candidate from each major party, but need to get over two from at least one party (which is tricky because the preferences from the excluded candidate would boost the other candidates for their party).  So at the moment the Greens are struggling there.  Current outcome 3-2 to Labor.  Final Liberal-Labor seat margin can't be calculated because Labor currently has four still in when Liberal 3 drops out.

7:53 Interim distribution for Brindabella has the Sex Party being eliminated in seventh because of the even spread of the three Labor candidates. At the moment they are 0.1 of a quota short of making it into sixth (which would not necessarily help them).  Current outcome 3-2 to Liberal. Final seat margin is 0.13 quotas.

7:51 Interim distributions are up and I am checking them now.

7:32 36% counted in Yerrabi now, which is looking a great deal like 3-2 to Labor.  Major turnarounds in the Liberals' favour are now needed for them to win three seats in each of three electorates, and the most serious fourth-party threat is against the Liberals in Brindabella, so it is already looking difficult for the Liberals to form government.

7:22 The Sex Party is now on 7.1% in Brindabella, however that's usually not enough for a seat in a 5-seat Hare-Clark electorate, and it probably isn't enough on current figures yet.

7:15 Antony mentioned having seen an interim 3-2 result for Brindabella a little while back, though it hasn't appeared on the ACT site yet.

7:00 My broad impression on the electronic votes counted so far is that all of Liberal, Labor and Green are down a little bit compared to the overall result so far, with others picking up the slack.  I expect others to continue to gain a collective swing, because there are more of them this time than last time and based on the general anti-big-three vibe seen in the federal election.  On the votes cast so far if the button was pressed now I would expect something like 12-12-1 or 12-11-2 (Labor first).  The question is whether either major party will pick up on remaining counting and threaten to win a majority.  On current figures quite substantial gains compared to the early votes are needed for either to do so.

6:52 Ginninderra has about a quarter counted and there is a big vote for ungrouped candidates, but it splatters between many of them (only Kim Huynh getting anything significant).  On current figures it looks like three for the left and two for the Liberals with Labor and the Greens fighting for the last seat on the preferences from the tiddlers.

6:45 An interesting result so far is that the Sex Party are ahead of the Greens in Brindabella based on prepolls.  Antony is saying this is likely to be an artefact of electronic voting - ie that electronic voting is attractive to young voters who are more likely to vote for the Sex Party.  But that argument also applies in spades to the Greens. While the Sex Party would currently be a chance for the final seat if preferences from the Greens helped them, in reality on the current votes they would struggle to match the totals of individual candidates, so on the current votes Brindabella would probably be 3-2-0 to the Liberals.  But should the Liberal vote go down and Labor up from these figures, things will get very messy there indeed!

6:35 Elections ACT website performing poorly so far - sometimes working, mostly not.

6:20 OK, it seems like nothing has happened yet with no votes posted on the Elections ACT website.  The ACT is interesting in that interim distributions of preferences (showing who would win based on the votes counted to that point) are posted, based on data entry and electronic early voting.  However because certain parties perform better or worse on late counting, these interim distributions tend not to be the final story.  We saw this in 2012 with the Ginninderra Effect (I am contractually required to link to that extremely wonky link before votes are counted) being not present in the early distributions, but emerging as the primary count unfolded.

I've had no time to pay even subjective attention to the campaign, so anyone else's guess is as good as mine as to what will unfold.  What I will say is that two big-picture patterns almost totally ignored in the pundit predictions you can see around the place are the ACT's general leftwards lean and federal drag (the tendency of voters to vote against a federal government at state level).  The Liberals last won government in the ACT in the dying days of the Keating government and were only able to last two terms.  If they win from opposition while holding office federally it will be a very impressive achievement.

The performance of fourth parties will be very interesting tonight.  They should get a big share of the vote between them, but will any get enough to threaten for seats or will the vote just spray between several.

6:15 I'll be starting some comments soon, just need a bit of time to get my head around what's happening so far.  (Very busy weekend for me between weeks of fieldwork, so please excuse the slow start!)

1 comment:

  1. Thanks for your work Kevin. Always a pleasure to see Hare Clark voting system at work and your expertise on display.

    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.