2025 Late Postcount And Expected Recount: Bradfield

And it has come to this ..

BRADFIELD (Lib vs IND est 2.5%) 

Gisele Kapterian (Lib) vs Nicolette Boele (IND) 
Boele led by 40 votes at end of indicative 2CP count.

Kapterian led by 8 votes at the end of distribution of preferences

Automatic recount started Monday 26 May

Boele has taken lead during recount and is expected to be declared winner

Declared winner will be seated, loser will be able to challenge outcome in court.  

AEC challenge appears unlikely

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

Introduction (Monday 19th 4:30 pm)

Welcome to my thread for the end of the count in the extremely close NSW seat of Bradfield.  This has been a roller-coaster postcount where my assessment as of Sunday 4th was that Nicolette Boele was perhaps very slightly better placed but not much.  The news that a prepoll booth had been only part counted appeared to be the end of the contest via a 440-vote boost resulting in me upgrading it to (at different stages) an expected or very likely Liberal win. Several sites and networks including the ABC called the seat for Kapterian on Monday 12th.  At this point I thought Boele's chance was very low (I would have said about 2%) but stopped short of calling it myself.  The next day Boele got a spectacular batch of out of electorate prepolls and today she did the same thing on the very last batch of postals, overturning Kapterian's lead on the final day of the initial count. At the end of the initial count, the self-styled "Shadow Member for Bradfield" leads by 40 votes.  

This thread will follow the count at least until the result is declared including any moves around possible challenges of the declaration.  Bradfield, the 2022 teal seat that missed, was previously covered on the Coalition vs Teals page but following the precedent I set with my Herbert thread in 2016 I now do a new thread for any seat that gets inside the automatic recount margin (<100 votes) at the end of the indicative count.  (Not you too, Goldstein, I'm busy.  Don't even think about it.)

Cases where the indicative count is extremely close cause a lot of confusion.  Herbert 2016 was more confusing than normal because facing an indicative margin of eight votes, the AEC decided to go to a full recount first, followed by the distribution of preferences.  In the case of Herbert it was mathematically possible for a non-major party candidate to win the seat based on final primaries, but it obviously wouldn't happen.  The normal order is the distribution of preferences first then the recount. For Bradfield, with Boele leading Labor's Louise McCallum for second by 6.7% with 14.67% of votes between the Greens and other minor candidates, the AEC does need to establish for sure that Boele does make the final two.  Of course she will, but just imagine going to a full recount then finding that she didn't ...

But haven't preferences been distributed already in the 2CP standings between Boele and Kapterian?  Yes they have but only provisionally, by finding which of the candidates is ranked highest on each ballot.  The distribution of preferences process formally excludes candidates starting from the candidate in last place, and moving each candidate's ballots to the next candidate in line as they are excluded.  During this process all these moved ballots are checked again.  It's common for the candidate who is more dependent on preferences (in this case Boele) to drop back a little in this process, only to make up for it when the full recount is held.  For instance in Fairfax 2013 Clive Palmer led Liberal Deputy Leader Ted O'Brien by 36 at the end of the indicative count, by 7 at the end of the distribution of preferences, and by 53 at the end of the automatic recount.  We don't have comparable figures for Herbert because the recount was done first with the distribution of preferences changing the margin by only two votes. In McEwen 2007 the margin was 6 for Labor at the end of the distribution, 12 to the Liberals' Fran Bailey at the end of the recount and 31 to Bailey after amendment by the Court of Disputed Returns following Labor's failed challenge.

During the distribution of preferences, from time to time the total of votes will change and I will note these as I see them around other commitments above.

An automatic recount rechecks every vote.  This often favours the candidate who received more preferences as the other candidate has more first preferences that have been less rechecked,  

I expect the distribution plus automatic recount could go into early next week but the time frames can depend on scrutineer behaviour among other things!  (Fairfax 2013 was spectacularly slow because of Clive Palmer scrutineers challenging almost every adverse vote.)

What happens then?

Once the distribution of preferences and the automatic recount (assuming it is correct) are required, the result will be declared, unless it's a tie (see below).   The result will then be open to challenge in the Court of Disputed Returns.  The last CDR challenge to a close Reps finish was in McEwen 2007.  Common grounds for challenge include excluded votes (voters whose votes were unable to be counted through no fault of their own) and complaints about formality interpretations on challenged votes.  The winner is seated while the challenge is heard.  Often losing parties decide not to challenge - see Antony Green's piece for some reasons why.  

The Court of Disputed Returns after examining all evidence can determine that the original winner still wins, that the other candidate wins or that the election for the seat is voided (which means the seat becomes vacant and a by-election is held).  

The AEC can potentially challenge the result itself if it believes the outcome is unsafe.  The main basis for which this might occur is multiple voting.  In the case of Herbert (final margin 37 votes) the AEC conducted urgent investigation of possible multiple votes and was satisfied that there were not enough of them to change the margin.  Many steps taken in recent years, including increased use of automatic markoff and the establishment of a special roll for suspected multiple voters, have reduced the scope for multiple voting to cause a seat to be voided, but the potential is still there.  In general cases of multiple voting are unintentional - voters with mental health or memory issues or confusion about the voting system.

In the event of a tie, the AEC will conduct a further recount. If the result is still a tie then the AEC must return a finding that the election for the seat has failed.  This could then be challenged in court, where either side could try to establish that they had in fact won based on differing interpretations of contested votes - but it is likely some basis would be found for voiding the original election for the seat and holding a by-election anyway.

Past close margins

I am often asked questions about past close margins at this time.  Unfortunately lots of people are asking AI software about past close margins, and the AI sites are frequently getting it wrong or even obfuscating their way out of answering the question.  The closest margin that has stood in raw vote terms is 7 votes, Werriwa, 1914.  The closest in 2CP terms was the 1939 Griffith by-election, 8 votes (50.007 2CP).  The closest in 2CP terms at a general election was Hawker 1990 (14 votes) and there was a 12-voter in Stirling 1974.  Closer margins of 1 vote in Ballaarat (as then spelled) 1919 and 5 votes in Riverina 1903 were voided and rerun because of electoral irregularities.  The famous tied election that Australia has seen was in Nunawading 1985, a Victorian upper house state seat where a winner was drawn out of the proverbial hat.  This too was voided for irregularities.  

How much difference can recounts make?

I've mentioned above the cases of McEwen 2007 (18 votes difference) and Fairfax 2013 (46 votes difference) in terms of the changes recounts can make.  A report from 2008 gives the history of previous recounts that were allowed (p 24).  The 2007 election pre-dated the current 100 vote rule and a recount request for Bowman (64 votes) in that year was denied.  The average difference for the recounts included in that report is 22 votes with one case (Bass 1998) of a 64 vote difference.  

Distribution Of Preferences updates

To be added here when distribution is underway and anything happens.  The AEC states that the distribution is expected to take the rest of the week and that progressive updates won't be published.  In past elections there have been minor changes to the Virtual Tally Room totals during this process so I will note if any of these are seen.

It is also common around this time to see claims concerning votes not admitted to the count; I will note any here that I see, or any other incidents.

Monday 19th 10 pm:  Boele is down two votes and Kapterian one.  Boele leads by 39.  There are changes shown in two booths, Chatswood (Bradfield) and Willoughby PPVC.  Individual booths can have multiple changes during the distribution process.  

Tuesday 20th 5 pm: The margin briefly dipped to 38 but is now back to 40.  In total both candidates are down four to informal since the distribution started.  Eight booths have had changes so far today.  

Tuesday 5:40 pm: Kapterian has dropped another 5 to informal, trails by 45.

Wednesday 21st 11 am: Only one booth is showing corrections so far today - Kapterian is down another vote to informal and trails by 46.  Small sample size but it is interesting that Kapterian, the primary vote leader, is so far losing a lot more preferences to informal than Boele.  

1:40 pm: Lead is 43.

6:00 pm: Five booths have seen corrections today and the lead is 44.  As noted in comments it may be that Boele is gaining on right-wing exclusions that would tend to favour Kapterian and therefore as the distribution continues Kapterian may gain.

Thursday 22nd 10 am:  A further five booths had corrections later last night and one so far today with Boele's lead down to 39.  Commenter Michael notes that Labor has been excluded so the preferences being corrected now will be those from the Labor exclusion, which is most of the distribution, and this may favour Kapterian (we will see).  

11:30: A further six booths have reported changes but the margin is only down by one to 38.

12:00: Margin now down to 33 (I've seen a comment that suggests it got down to 30) without any increase in the number of booths reported, so this may have been formality rulings on earlier votes.  

12:35 And back to where we started, 40.  

2:15: 35

6:53: 32.  I expect the distribution to run through tomorrow.

8:00: 28 ...

Friday 22nd 11:40  Well well well if it isn't our friend St Ives PPVC again.  A surprisingly large correction this morning with 11 votes switching from Boele to Kapterian has Boele's lead down to seven.

1:00 Lead now four - again, it is possible that even if Kapterian leads after the distribution Boele could get the lead back on the recount.  But the chance is rising of a micro-close outcome where the risk of the seat being voided increases.  

2:52 Two!

4:35 LIVE COUNT TIE

Tired: "Shadow Member For Bradfield"

Wired: "Schrodinger's Cat Member For Bradfield"

4:53 Kapterian in front for the first time in the distribution, by one vote.  

5:10 Kapterian by six!  I saw an incorrect suggestion that this has something to do with Boele's open how to vote card (1 Boele votes aren't being checked at the moment) but this did make me think that Boele could have issues on the recount if she had more subtly informal votes as a result.  (The obvious ones where a voter just votes 1 for her and stops were already caught).  (5:20: Back to 4)  I expect the distribution to finish very soon (not sure how we will know that it is finished).

6:40: Back to six.  No updates for half an hour, not sure if that is all for the distribution yet or not. 

7:15 Kapterian by 8.  10:00: The AEC has confirmed details for the recount, which could take up to two weeks.

Recount Updates

Monday 26th 2:45: And they're off!  So far Kapterian has lost two to informal and Boele has lost one so Kapterian leads by 7.  Three booths have changes showing.

5 pm: Boele has lost another vote so the margin is back to 8.  

10:30: Not a good start for Boele, she's lost one vote net to Kapterian who now leads by 10.  Eight booths have been updated.  

Tuesday 27th 7:00 Today has not gone well for Boele either, fifteen booths are updated and she has a net loss of 8 to informal while Kapterian has a net loss of 2. 

Wednesday 28th 11:30 Margin is back from 14 to 10.

3:45 And now margin is 5, so for the first time Boele has a net gain on the recount; will it continue?

5:30 Well it will to the tune of one vote anyway .. margin is 4 with a gain of 10 for Boele from today.  And only five booths with change today so still scope for someone to get a meaningful break.  Meanwhile some sideshow colour with Sky reporting Paul Karp was signed in as a Greens scrutineer!

7:00 The margin is 2.  Another crossover point soon?  Ben Raue's article about the recount process is excellent reading.

Thursday 29th 4:30 Very slight movements today so far, at one point down to 1, now up to 3.  Only three booths showing changes.  

6:30 Seven booths with changes now and the lead is 4.  So no continuation of the trend in yesterday's counting as a micro-close result is getting harder to avoid.

Friday 30th 1:33  Oh hai we are here again.  Boele has made a net gain of eight on the recount margin so far in the recount which still has a long way to run with many booths still unrevised, so it could be things are trending her way as expected, but if it continues will she be able to get out of the margin of litigation?   

2:00 Or not - Kapterian back to three in front.  

5:50 Lead change - Boele ahead by one!  The booth that put Boele ahead was St Ives PPVC making its presence felt in the postcount for the third time, but for a large booth that was friendly on primaries to Kapterian Boele could have done with a bigger gain here.  

6:41 That didn't last long, Kapterian back ahead by one.  

Saturday 31at 2:30:  Kapterian now ahead by three, gain of two at Roseville.  Whoever wins this even by one vote, assuming someone does win it, gets seated and gets to be an MP for several months even if the seat does end up being voided and rerun.  So there is that. 

4:00: Kapterian lead out to five, gains another two at Roseville West.  Now starting to look a lot less like a trend to Boele and a lot more like a random drunken walk that is likely to end up by crashing into the doors of the Court of Disputed Returns.  But still time for someone to avoid that.  The best hope for Boele could be the Turramurra PPVC with over 6000 votes, the other big booths are all gone and I believe most of the postals are too.  There will be a fresh distribution of preferences at the end but that's unlikely to move even a few votes. 

5:00 And back to three. 

7:00 BIG SHIFT! Well in the context of this nailbiter anyway. Boele's last big chance was the Turramurra PPVC but it has delivered for her with Kapterian losing 16 (in general primaries to informal).  Boele now leads by 12 and her chances of being declared the winner whatever happens next are much better now.  

10:40 Someone asked so for what it's worth, I calculate there are 23777 booth votes awaiting checking.

Monday June 2 11:30: An interesting sideshow in Bradfield is the 2PP realignment as there has been some speculation Labor could win the 2PP in this seat.  Not looking at all likely at this early stage as Labor are trailing Boele's 2CP on dec prepolls by a few points and postals by 12 points, though they are probably early postals which will be worse than later ones.  

1:30: Three booths processed today so far and Boele's lead is now 15.  

2:15: 18 now,up another three in one booth (St Ives North).  Three of the booths checked today have had relatively high Kapterian primary votes, for what that's worth.    

2:25 20!  Boele must now have a serious chance of winning by enough for multiple votes to not void the seat - in which case the onus would be on the Liberals to decide whether to challenge and by how much.

4:40 21.  No new booths added, so I suspect that was a formality referral.

5:00 Even now 25, what a great day for Boele so far.  14 booths to go before the distribution starts but five of them are tiny,  It's expected the recount will take the rest of the week to finish.

Ben Raue has another great podcast episode talking about the recount with scrutineers. It sounds like the distributions of preferences are going on booth by booth as votes are checked so there may not be a long distribution phase once the recounts are done.  

Tuesday June 3 11 am:  Boele is now ahead by 34 with 11 booths remaining, three of them tiny.  She did especially well in the St Ives and Turramurra area booths.  It is very likely that Boele will now go on to be seated and there is now a good chance that the margin will be outside the margin of a potential AEC challenge over multiple voting.  This means Kapterian would have to challenge and would need to have grounds to do so, none of which have yet been suggested.  

12:50: Per the AEC briefing today I have the following details:

* the recount is likely to end late tomorrow or possibly Thursday, though the level of scrutineer challenges is always unpredictable in time terms

* the Turramurra PPVC swing was not a data issue but simply a large number of formality changes, similar to what we have seen in smaller numbers in many of the day booths in Turramurra and St Ives

* the number of multiple votes is low, possibly single digits - suggesting to me strongly that at the current margin the AEC would not be lodging a challenge 

1:05 Teh Narrowing! Kapterian has recovered one vote, nine booths to go.

2:10 The gap is now 30, it came down another three at Warrawee, however that was likely on paper to be Boele's worst booth in terms of its size and her having more primaries than Kapterian.  

3:30 The Willougby booths are now gone with Boele leading by 27 25 27, there are only three Wahroonga booths to go (two are small and none seem likely to help Kapterian) plus two tiny booths.  Unsure if there are any non-ordinary counts remaining.  (There was a late change from 27 to 25, possibly on referrals).  

5:30 The margin is still 27 and only three booths remain, two of them tiny.  

Wednesday 11:00 Only one tiny booth remaining now - not sure if there is anything else to come after that or if we could be getting a statement that the recount has finished soon.

11:20 I've seen that the final announcement on the seat is likely this afternoon after the NSW Electoral Officer rules on remaining referred challenged votes.