2022 Hobart Council Count

All figures on this page are unofficial - see TEC for official results when available.

MAYOR (Incumbent: Reynolds):  CALLED (Friday 10:30) Anna Reynolds defeats John Kelly with reduced margin compared to 2018 win.

DEPUTY (Incumbent: Burnet):  Burnet leads Behrakis and will win by very large margin (Called Saturday before final exclusion)

COUNCILLORS (11 recontesting incumbents, 12 vacancies): 

Called: Reynolds, John Kelly (off initial sampling), Zucco, Burnet (off 20% count), Elliot, Behrakis, Bloomfield, Dutta, Harvey (off 50% count)

In The Mix (for three seats): Posselt, Lohberger, Sherlock (these three leading substantially), Fox, Kate Kelly, Briscoe.

Coats, Thomas, Fox, Briscoe currently projected to lose seats.

UTAS MOVE ELECTOR POLL: CALLED No has won.

Introduction

This is my specific page for commentary on the 2022 Hobart council count. Comments on other councils and general themes will be found here. Don't expect to see much on this page on Tuesday; there will be more action on Wednesday.

On Tuesday ballot papers will be removed from envelopes and there may be some computer data entry for some large councils but don't expect official figures.  I may have some indications from scrutineers depending on the pace of counting but the main counting activity for Mayors and Deputies together with the start of data entry is on Wednesday. The TEC expects to have provisional 20% or 50% distributions for all councils at some stage on Wednesday.  These provisional results are what would happen if only the votes entered to that point were used to decide the election.  Usually most of the candidates who are provisionally elected at the 20% stage end up winning, but often some of those elected to the last few positions change.  It is common to see media misreporting that candidates "elected" in these 20% and 50% interim counts have "won" but it is only when the final, 100%, count is done that the winners are official. 

Where the final few seat results in the interim counts are within 100 votes or so (depending on council) then it's possible the winners will change by random sample error alone.  They can also change if a candidate's victory depends on which of two other candidates is excluded at a certain point.  But there is a more important reason why they can change, and it is this:

Beware, Big Word: Stratification!

It's very important to be aware that interim primary vote figures for the various Councils using automated data entry won't necessarily be unbiased samples of the whole count.  Votes do not reach the Electoral Commission in random order, but rather arrive on particular days, and within each day's sample they may be further stratified by post office of origin (for instance).  There will be political differences between the pool of voters who tend to vote early in the count and those who vote late in the count. Also, campaign events that happen during the postal voting period can have more impact during some parts of the count than others.  

The TEC may well make vigorous efforts to reduce this stratification, but nothing will get rid of it completely.  It's also hard to say how compulsory voting might affect it, since the pools of early and late voters will be different.  In 2018 a "green shift" was observed in many (not all) councils, with obviously green/left candidates tending to get a final percentage that was around 1.1 times their vote in the first interim count, and righties tending to drop back accordingly.  The effect might be more pronounced, absent or in the opposite direction this time - who knows?

Hence when the TEC release primary figures for various percentages be aware that these are not true representative samples and that past experience shows that substantial changes may occur.  (For instance in 2014 one candidate for one council who was not getting elected on the 20% sample was elected in third place, though that's an extreme example.)  Any result that is even remotely close will therefore not be called on 20% of primaries. 

Please note that I only "call" a seat when I consider it virtually certain that the candidate in question has won.  I ask that people not describe me as having called any seat where I have not explicitly used the word "called" for it.  A summary of how I think the counts are going appears at the top of the page.

Hobart Lord Mayor

Tues 3:20 In early unofficial sampling Reynolds appears to be leading on primaries but I will post more when sample is meaningful.

Tues 4:50 First scrutineering sample of 258, noting we do not yet know how representative these are: Reynolds 29.8% J. Kelly 20.5 Harvey and Zucco 11.2 Bloomfield and Elliott 10.9 Bai 3.5 Coats 1.8. At this stage it is looking like Reynolds and Kelly will be the top two. Kelly could close the gap in this sample to at least some degree on preferences (three significant "blue" candidates to one Green). Have not sampled prefs yet

Tues 7 pm Preferences are showing a fair rate of exhaust and scatter - a lead of any meaningful size will probably not be caught here. 

Tues 8:30 I now have 550 votes sampled and Reynolds has a fight on her hands! (Not saying she won't win the fight, just that it's a good one). I now get Reynolds 29.6 Kelly 22.4 Zucco 12.4 Harvey 11.1 Elliot 10.7 Bloomfield 8.7 Bai 3.4 Coats 1.6.  Preferences from Harvey are favouring Reynolds but those of all others except Bloomfield are favouring Kelly. After applying my sample of preferences it's 50-50 (in fact Kelly is a very meaningless whisker ahead, given the margin of error in this sample is about 5%.)

Tues 9:35 Added another 93 votes to sample; Kelly won those 41-39. Done for the night.

Wed 11: Today's votes coming through in processing are considerably to the left of yesterday's. In another sample of 200 I had Reynolds ahead 35-15 on primaries, another scrutineer also finding Kelly not far clear in second off 100.  Across all my sampling I now have Reynolds ahead about 52-48 after preferences off a sample of approaching 850 primaries.

Wed 12: Lord Mayor votes are now being thrown into large trays and it's hard to sample anything while this occurs but I'm tracking the ratio of Reynolds and Kelly votes and that was 1.44 in a sample of 526 going to one or the other, cf 1.48 in my previous sample - still projecting to about 52-48 after preferences but a very long way to go and my sample of votes for others with preferences is only a few hundred.

Wed 4: On rough tallies of totals being added up on tables now I have Reynolds 28.0 Kelly 20.2 Zucco 12.1 Harvey 11.6 Elliot 11.2 Bloomfield 9.4 Coats 3.7 Bai 3.9 and this projects off my relatively small preference sample to 52.7-47.3 to Reynolds. I am a little mistrustful of the Bloomfield component of that sample (her preferences are favouring Reynolds) so that may yet prove generous. These rough totals are from about three-quarters of the primary votes. Final primaries should appear tonight.

6 pm: Corrected a minor error in my numbers above but in any case Reynolds improved in votes yet to be added (a very noticeable left shift in the last few thousand) putting her lead at around 28.4-19.4 which projects to about 53-47. Again I am not calling this, the preference sample I have isn't big enough.

7:40 Final primaries are up and Reynolds' lead is 28.8-17.4.  Should be enough, as that is projecting out to 54+.  Now we just wait for the preference distribution.  

9:15 "Update, 8:30pm - Some errors have been found and a recheck of first preferences is in progress with an update shortly.".  The errors probably won't be major but let's see.  There seems to be quite a substantial error here which is that several hundred votes for John Kelly are being erroneously credited to Coats, probably just a simple adding up error or a confusion of bundles during rechecking.  Coats and Bai have been in lockstep in the race for last position since counting began so there is no way Coats should be 700 up on Bai.  The numbers also disagree with rough totals I compiled together with another scrutineer as concerns Kelly.  

On the assumption that the error is what I think it is (edit: confirmed and Reynolds leads 28.8-19.9), I have run some random error modelling on my preference estimates and I still have Kelly with a significantly non-zero chance of winning this (say, 10%), so I will not be calling this tonight.  

Thursday 9:00 Bai has been excluded and exclusion of Coats is about to begin. Kelly made a slight gain (169-141) off Bai, and now trails 29.38-20.69.

Thursday 9:30 am The throw of Coats looks like closing the gap by maybe 150 (from about 2750) which may sound like very little but there are also votes flowing from Coats to Elliot and Zucco that will help Kelly later. There will also be minor adjustments to primary totals after a mis-sort spotted by yours truly (16 Elliot votes in the Coats pile); this will be patched pretty fast I think. Bloomfield probably next to go which will be a very interesting exclusion.

Thursday 10 Count suspended. Further mis-sorts have been identified and the number of votes involved is now a few dozen - unlikely to affect the eventual winner but with potential to affect the order candidates are excluded in. I am expecting a full recheck and a delay of a few hours.

3 pm: count resuming soon.

3:50 count has resumed with a re-throw of Bai. The exhaust rate off Bai is about 30%; I think that fully optional preferencing for mayors and deputies should be reviewed as I am seeing too much exhaust for my liking; perhaps semi-OPV as used in the Legislative Council would be better.

4:15 I was watching a throw of Bai but it has turned out Coats is last and is being thrown first. (The Bai throw will be a quick one.) On the recheck Kelly gained 44 to 5 for Reynolds but this was offset by Harvey gaining 114 so it won't have any real net impact on the leaders.

5:00 Bai throw on, then Bloomfield then Harvey. Tracking the gain rate Kelly needs to win it started at .175 votes per preference and these two exclusions won't change that much. In my sampling Kelly was getting approaching .4 off Zucco and Elliot but going backwards at about the same off Harvey. So all else being equal he needs to make a serious gain off Bloomfield.

5:35 In sampling Kelly does seem to be making a gain off Bloomfield but it's a very slight gain if so. The candidate making the biggest gain is Elliot who I expect to overtake Zucco and finish third. However Elliot is too far behind to catch Kelly.  Bloomfield preferences to Elliot will help Kelly down the track too (by about .25 votes/preference in my sample).  The difference to my sample takes my projection down to about 52-48, and Kelly is still (for all I know) in it.

6:00 Exclusion of Bloomfield took just under an hour; they have three to go and will try to finish tonight but there is a constraint with having to move to another centre, meaning they can't count beyond about 9. Reynolds actually gained 11 votes on Kelly at this exclusion but there are still more unfavourable votes pooling with Zucco and Elliot. I now have a projection of just 51.5 to Reynolds.

7:05 But now Reynolds is doing far better off Harvey than in my sampling, getting maybe 60% right off the bat with Kelly getting very little. That reverses what happened with Bloomfield and Reynolds comes up to about 52.7 in my projection again.

8:10 Reynolds only ended up getting 50% not 60% of that transfer and Kelly is still close enough if he can do it. He needs a .395 votes per preference gain rate off Zucco and Elliot, which he was very close to getting in primary vote sampling, but the actual flows might be a bit weaker. Still significant life in this one.

8:50 This can still get very close. On sampling Zucco's preferences I have Kelly on track to roughly hold the gain rate required going into the final exclusion. It will be harder then though as Zucco's preferences are mostly his own whereas Elliot's include a lot of votes from Bloomfield and even Harvey.

9:40 The Zucco throw is done and it all comes down to this. Reynolds 43.01% Kelly 32.36% Elliot 24.63% Kelly needs to gain at .432 votes per preference. I think that's too much and that even getting .4 won't be easy. I am pretty sure Reynolds has won now but this close to the end may as well wait for the final throw. Play resumes at 9 am at the Baptist Church Taroona.

Friday 10 am: All Elliot's votes have been thrown and they are being checked and tallied. I was only able to sample the non 1 Elliot votes but they were breaking about 60-25 which isn't enough. The size of the piles is also strongly suggesting Kelly's not getting enough and that Reynolds wins about 52-48

Friday 10:40 One small table of non 1 Elliot votes had a gain rate below .1. Votes are being tallied now.

11:30 Reynolds wins, unofficially 53.4-46.6, official figures soon.

4:00 In the end the gain rate for Kelly off Elliot was rather weak, only 0.17 votes/preference.  The obvious "blue" candidates did not recommend preferencing Kelly on their how-to-vote recommendations on Facebook - interestingly, I've had it confirmed that they did not have his permission to name him.  That said they were under no compulsion to say voters should leave other boxes blank (which they did).

Sunday: A comment on exhaust rates in the Lord Mayor contest.  4035 votes exhausted, an exhaust rate of 12.9% of the total but 25.0% of votes that were distributed.  In 2018 exhaust was 17.6% of all votes but 27.7% of votes that were distributed.  However, in 2018 there were 11 candidates not 8, which would increase exhaust rates - any impact of compulsory voting on exhaust rates will need to be checked across multiple councils (and yes I'll do this sometime!).  The number of exhausted votes from Zucco and Elliot alone was greater than the margin, but many of these voters would have exhausted their votes anyway, and many of Elliot's votes especially were votes for other candidates.  Thus, the preference recommendations issue discussed above did not decide the contest.  My estimate is that even had preferences been compulsory, Reynolds would have still won, but it would have been closer (maybe 52-48.)

Hobart Deputy

Tues 3:20 In early unofficial sampling Burnet is getting the most primaries so far followed by Behrakis but I will say more when I have a meaningful sample.

Tues 5:00 Sample of 250 Burnet 30.8 Behrakis 18.8 Dutta 12.4 Briscoe 10.4 Posselt 8.8 Thomas and Christie 6 Spender 4.4 Jackson 1.6 Davies 0.8. The votes outside the top two don't have a strong right/left split so on those numbers Burnet is in a very good position.

Tues 7 pm: high degree of exhaust, if Burnet's lead holds in further sampling she will retain.

Tues 8:45 With a sample of 504 I now have Burnet 29.6 Behrakis 16.9 Dutta 13.5 Briscoe 11.9 Posselt 9.5 Christie 6.9 Thomas 5.8 Spender 3.6 Davies 1.6 Jackson 0.8. Burnet easily beats Behrakis if they are the top two as she is beating him or level on everyone's preferences. Maybe someone else gets over Behrakis into the final two but I think they're much too far behind if they do especially with c. 30% of preferences exhausting.

Wednesday: No official counting yet. 

Thursday 10 am: Official counting is starting following the delay in the mayoral count.

3 pm: Final primaries soon and that will be it for the day.

3:30 Unofficially (edit: confirmed) it will be something like Burnet 26.4 Behrakis 16.9 Dutta 14.7 Briscoe 14.6 Posselt 9.0 Thomas 7.4 Christie 4.7 Spender 3.4 Jackson 1.5 Davies 1.4. Not as large a lead for Burnet as in my sample; she easily beats Behrakis. I did not sample contests with other candidates but she should have enough if Dutta or Briscoe do get into second. (Briscoe appears to be a ghost ship in this count having probably lost his council seat, which also suggests his flow here won't be great.)

Friday 3: Plowing through the minor exclusions and Behrakis made some gain off Spender but is still not that clear of Briscoe and Dutta. The coming exclusions are Christie, Thomas and Posselt (Posselt will be interesting.)

Friday 6: Thomas excluded (helping Briscoe and Behrakis) and the standings are: Burnet 30.00 Behrakis 21.51 Briscoe 19.18 Dutta 17.54 Posselt 11.78.  This is interesting because Posselt has run a leftish campaign and his votes could put Dutta over Briscoe (over Behrakis as well perhaps a bit hard?)  But if Dutta doesn't get over Briscoe then he could put Briscoe over Behrakis.  Briscoe preferences won't necessarily help Behrakis against Dutta and Burnet.  Not clear here who makes the final two. 

Friday night:  no more tonight but counting resumes Saturday so result tomorrow I expect. 

Saturday 11:15: Dutta gained on Briscoe off Posselt but not enough with Burnet gaining more, so Dutta is out.  Briscoe needs 2.53% off Dutta's 20.53% to overtake Behrakis.  With votes also splitting to Burnet (who will get a lot of these) and exhaust I suggest that is very unlikely and so Burnet vs Behrakis will most likely be the final two with Burnet winning very easily.  Burnet on 34.13% is way down on the 48.6% she had at the four-candidate stage in 2018 but in that case none of her opponents were incumbents while in this case they all are.  

12:50: Briscoe got surprisingly close to Behrakis but didn't get there.  Burnet will win again by heaps; I'm not even sure she needs any preferences here.  

2:00 Burnet wins 61.13-38.87.  A swing of maybe 3.5% to Behrakis from 2018 when there was no final margin because Burnet crossed 50% against two opponents.  (Behrakis was an off-council candidate then.)

UTAS Move Elector Poll

Tues 3:20 my first unofficial sample of 100 formal votes NO leads 73-27 and it is notable NO is attracting support from voters for candidates who are not clearly anti-move. On this basis looks very likely NO will win, not calling yet .

Tues 5:00 No on 74.5% with a sample size of 235. CALLED No has won 

Saturday:  No wins with 74.38%.  This was going to be delayed until after the Hobart Mayor/Deputy counts but the latter two dragged on for so long I suspect staff counting other councils switched to it.

Hobart Councillors (12 to be elected)

In a sample of around 250 votes Reynolds and John Kelly are way over quota and Bloomfield, Elliott, Burnet, Harvey and Zucco seem to be polling significantly while Thomas especially is struggling so far among the incumbents. Very early days yet. I will add more if I get to 500.

Tues 6 pm: In a sample based model of 523 I have Reynolds on about 2.2 quotas, John Kelly about 1.8. Zucco is just on quota in this sample and Bloomfield is on about 0.8 with good flows from the leaders. Burnet and Harvey get flows from Reynolds putting them up to around three-quarters of a quota each (that might rise); Elliot is in a similar position. Lohberger (!) and Behrakis are doing reasonably well in my sample. (Behrakis doing better in preferences on another scrutineer's sample than mine, by the way ) Most obviously in the mix behind them I have Dutta, Fox, Posselt, Coats, Kate Kelly, Sherlock (a noted preference sponger so my model may underrepresent how she would go) and Corr. Thomas and Briscoe are both struggling in this sample though by no means out of it yet.

Note that this is a small sample and a lot can change but I expect most or all of the first seven listed should get up, the rest is just a hint. 

Wed 11: I will have a full councillor sample when I get past 1000 votes but the left shift mentioned for mayor is on here too, especially favouring Dutta in my sampling while Bloomfield is not as high as yesterday (but still clearly at least competitive.)

Wed 12:30 20% count is up and provisionally winning are Reynolds, J Kelly, Zucco, Burnet, Elliot, Behrakis, Bloomfield,  Harvey,Dutta, Fox, Posselt, Sherlock with Lohberger and Kate Kelly 13th and 14th. Detailed comments soon.

Have looked at the provisional distribution and in it Reynolds and John Kelly win on primaries and Zucco on surpluses.  The next six are well clear but there is then a five-way fight for three places involving Fox, Posselt, Sherlock, Lohberger and Kate Kelly. In the current distribution Kate Kelly is overtaken by Sherlock and eliminated by a margin of 29 votes.  Kate Kelly's surplus then puts Sherlock over Lohberger by 14, with Posselt 21 ahead of Lohberger and Fox 25 ahead.  So any of Sherlock, Posselt and Fox could conceivably miss out to Lohberger.

We may be heading for a 7-5 left-right council here.  None of the other "blues" are in the top 14 and even if one of Thomas, Briscoe or Coats gets into that mix they don't have anywhere to get preferences from, so all these three are in difficulty.

Wed PS: Jeff Briscoe has graciously conceded defeat on Facebook after 28 years on Council, the second longest in Hobart history behind Zucco who will be re-elected.

Thurs 3 pm: 50% count coming soon.

Thurs 7:30 In the 50% count Lohberger now replaces Fox on the provisional winner list and the margin is quite comfortable (indeed he is 11th ahead of Sherlock). Kate Kelly is 76 behind Sherlock in their within-ticket battle.

Thursday late night: I have looked at the 50% count compared with the 20% count and in general in the 50% count it is Kelly and "blue" candidates who have come up and left candidates who have dropped, particularly Fox and Posselt.  Nobody has changed by more than 1% and most have changed by 0.3% or less.  Lohberger wins after using Briscoe's preferences to overtake Sherlock and Fox; Lohberger is 86 votes (0.5%) ahead of elimination behind Briscoe, so it's not completely clear he gets past that point.  I don't know if there's a chance Lohberger's votes will save Briscoe if Lohberger is excluded (Lohberger is well left of Briscoe but there is common ground on the Utas issue.)  I'm now satisfied that the top nine have all clearly won and have called them.  Incidentally if Briscoe does overtake Lohberger, this may instead cause Fox to win.

Friday 6:15 Expect it will be a few days before all the ballots are entered here.   Thomas has also now graciously conceded.  

Tuesday 10:00 RESULT CONFIRMED.  The final winners are the same as the 50% winners but in a different order.  John Kelly, Elliot, Bloomfield, Posselt and Lohberger are in and Thomas, Briscoe, Coats and Fox are out.  The order of election is Reynolds, John Kelly, Zucco, Burnet, Elliot, Behrakis, Dutta, Bloomfield, Harvey, Posselt, Sherlock, Lohberger.  None of the critical margins were close.