tip off

South Australian election late counting

A progressively updated post following the counting of over a quarter-of-a-million outstanding votes in South Australia’s cliffhanger election.


5.3pm. Further counting in Elder reversed the trend just noted, breaking 1485-1254 Labor’s way and putting the lead at 757, which is very likely more than the number of votes still outstanding. Counting also favoured Labor today in Ashford (347-287, increasing the lead to 820) and Newland (671-652, putting it at 656).

Noon. Declaration votes are flowing fairly strongly the Liberals’ way in Elder, a second batch breaking 528-429. That reduces the Labor lead to 526, and could send it below 300 if the trend continues.


8.30pm. The final score for the day from Colton shows Labor gained 1196-1078, putting Paul Caica’s lead at an unassailable 570 with perhaps 1000 votes remaining to be counted. The tide of late counting continued to flood the Liberals’ way in Hartley, today’s batch favouring Liberal candidate Vincent Tarzia over Grace Portolesi by 1685-1204, pushing his lead out to 1131.

5pm. A big addition of votes in Newland breaks 1426-1332, putting Labor’s Tom Kenyon 637 ahead and confirming his victory. In Mitchell, Labor clawed back 39 votes out of 3203 added, but it’s too little too late. Pretty much impossible now not to see a result of Labor 23, Liberal 22, independents two.

4.30pm. Labor has pulled a further 107 ahead with the addition of 1415 votes in Ashford, where Steph Key’s lead is now at 760 and unlikely to change much with perhaps 1000 votes still outstanding. Sykesie reports Labor now 588 ahead in Colton, with postals favouring the Liberals by an insufficient 52-48.

1.30pm. Well-informed commenter Sykesie relates that the morning’s counting in Colton has broken Labor’s way 349-262, putting Labor 539 votes in front and making life all but impossible for the Liberals. Their only hope of making it a twenty-third seat is an unlikely late reversal in Newland, where Labor leads by 543 with declaration vote counting still at an early stage, with about 4000 votes still to count.


11pm. It appears a move from postal to pre-poll counting also staunched the flow in Ashford, where Labor’s Steph Key now looks home and hosed after today’s batch broke only 728-720 to Liberal. This leaves Key 653 in front and projected to win by about 500. Nothing today from Elder or Newland.

6pm. Labor nerves will have steadied considerably with the addition of 970 votes in Colton, which I understand to be pre-polls. These have broken 491-479 in their favour and held their lead at 452. Projecting the existing declaration vote shares over an assumption of about 3000 outstanding votes, Labor emerges over 250 votes in the clear.

1pm. Mitchell continues to trend the Liberals’ way, 980 newly added votes breaking 563-417 and pushing the margin out from 373 to 519.


6pm. It appears the votes counted today were mostly if not entirely postal votes, and they are playing according to the script of favouring the Liberals by virtue of not reflecting the move back to Labor in the final week. On top of what was mentioned previously, today’s counting favoured the Liberals 808-634 in Ashford and 888-767 in Elder, and while that’s likely to be too little too late in Elder, the projected Labor win in Ashford comes down to double figures if the final declaration vote total is presumed to be 6000.

5pm. Encouraging first set of declaration vote numbers for the Liberals in Colton, breaking their way 556-425. If that trend were to play out over a total of 5000 declaration votes – 4000 having been the norm last time, but many more pre-polls apparently having been cast this time – the Liberals would finish about 100 in front. However, it may be that these are absent votes cast over the boundary in a Liberal-leaning part of the electorate, or representative of a particularly strong result for the Liberals on either postals, absent or pre-poll votes that won’t be replicated among the other vote types. UPDATE: I’m told on Twitter that these are postal votes. ECSA doesn’t do breakdowns of declaration votes, but in the corresponding federal seat of Hindmarsh, the Liberal two-party vote in September was 55.4% compared with 53.9% for pre-polls. Absent votes favoured broke 52-48 to Labor, but that’s unlikely to be instructive with respect to Colton.

4pm. The first 1424 added in the only seat that might get Labor to a majority, Mitchell, have broken 782-642 the Liberals’ way, increasing their lead from 233 to 373. If that keeps up, their winning margin will be around 800.

2pm. 1218 votes have been added in Newland, breaking 632-586 to the Liberals and reducing the Labor lead from 589 to 543. If that trend continues, the Liberals will only be able to wear away about 200 votes. However, trends in late counting can be variable, particularly in relation to pre-poll and absent votes which might be cast in particular parts of the electorate or neighbouring electorates. Unfortunately, ECSA doesn’t distinguish between different types of declaration vote in its published results.

Monday night

This post will follow the crucial late counting for the South Australian election, which has so far only dealt with re-checking of the polling booth votes counted on election night. Counting of an estimated 260,000 pre-poll and postal votes begins today, with the Liberals needing multiple miracles to boost them from their likely total of 22 to a majority of 24, and Labor hoping they might yet get there through what presently seems an unlikely win in Mitchell. Labor’s narrowest leads are of 571 votes in Colton (1.6%) and 589 votes in Newland (1.8%), while the Liberal lead in Mitchell is 233 (0.7%).

  • 101
    Posted Tuesday, March 18, 2014 at 8:17 pm | Permalink

    Actually if your name is Hitler, A you probably deserve everything you get.

  • 102
    Posted Tuesday, March 18, 2014 at 8:22 pm | Permalink

    I’m not sure who does the maths in the electoral office. This morning the Commissioner was saying there were 280,000 then 320,000 votes to count. If 770,000 is 69% then 1,090,000 is 100%. It has never occurred to get anywhere near 100%. About 92 or 93% is about average so likely about 240,000 votes all up counting todays c32,000.

  • 103
    Posted Tuesday, March 18, 2014 at 8:32 pm | Permalink


    Perhaps they were talking about potential votes, as they couldn’t know exactly how many postal votes will come in over the next week (I think they get two weeks to arrive).

  • 104
    Kevin Bonham
    Posted Tuesday, March 18, 2014 at 8:48 pm | Permalink



    I did a quick and dirty analysis of remaining vote as well. I assumed the final total will be the final total from 2010 (I couldn’t find the total number of declaration votes per seat listed anywhere).

    Are you assuming the same vote total as 2010 or the same %age turnout?

  • 105
    Posted Tuesday, March 18, 2014 at 8:51 pm | Permalink

    The 2010 total of valid TPP votes (as per the ABC elections page- down the bottom of each electorate page).

  • 106
    Kevin Bonham
    Posted Tuesday, March 18, 2014 at 9:03 pm | Permalink


    The 2010 total of valid TPP votes (as per the ABC elections page- down the bottom of each electorate page).

    Isn’t that a risky assumption then, if there has been enrolment growth arising from past population growth and/or net inwards migration?

    To take Colton as an example, enrolment is 25512. But in 2010 Colton’s turnout was 22144 which was 94.2% of enrolment, implying c. 23507 enrolment. So the final formal vote total should be larger.

  • 107
    Posted Tuesday, March 18, 2014 at 9:27 pm | Permalink

    106 KB – quite right. And projecting on initial trend is also risky as there are different groups of votes involved.

  • 108
    Posted Tuesday, March 18, 2014 at 9:41 pm | Permalink

    Well thats true but for some reason the ECSA doesn’t report the number of declaration votes sent or received like the AEC does so I just had to use the previous total valid votes.

    Another 1000 voters in Colton (assuming unchanged declaration vote patterns) would tip it into the Lib column though- I just tried it- so it is particularly important for that electorate.

  • 109
    Posted Tuesday, March 18, 2014 at 9:53 pm | Permalink

    GhostWhoVotes @GhostWhoVotes
    #Galaxy Poll SA Seat of Fisher Which party should Bob Such IND support for Govt: ALP 27 LIB 66 #savotes

    GhostWhoVotes @GhostWhoVotes
    #Galaxy Poll SA Seat of Frome Which party should Geoff Brock IND support for Govt: ALP 35 LIB 53 #savotes

  • 110
    Posted Tuesday, March 18, 2014 at 9:56 pm | Permalink

    At the present rate of counting a result won’t be known until the end of the month.

  • 111
    Kevin Bonham
    Posted Tuesday, March 18, 2014 at 9:58 pm | Permalink


    106 KB – quite right. And projecting on initial trend is also risky as there are different groups of votes involved.

    Yes. And we saw in the federal election that even different groups of the same kind of dec vote behaved very variably. Not just absents (which can come from locations with a political slant) but also postals.

    Probably room for some kind of multi-source analysis that weights the dec vote trend in a seat so far, the dec vote trend from last time, subjective estimates about how much swing back to the government there was (etc).

    Under normal circumstances leads >1% 2PP wouldn’t get pulled back very often.

  • 112
    Posted Tuesday, March 18, 2014 at 10:10 pm | Permalink

    OK…OK….I decided to play with Kevin’s method. So we are assuming the same turnout as 2010, and using the total number of electors on the roll in 2014, and assuming all declaration votes behave the way they did on day 1 of declaration vote counting…..

    Ashford: ALP by 290 votes (Lib TPP 49.3%)….becomes 49.6%
    Elder: ALP by 394 votes (Lib TPP 49.0%)…….becomes 49.4%
    Newland: ALP by 438 votes (Lib TPP 49.0%)..becomes 49.2%
    Light: ALP by 711 votes (Lib TPP 48.3%)……becomes 48.4%
    Colton: ALP by 128 votes (Lib TPP 49.7%)…becomes 50.5% and flips
    Florey: ALP by 591 votes (Lib TPP 48.6%)…becomes 48.9%

    So with the new analysis, it would become 22 ALP – 23 Lib – 2 IND
    (Colton flips to Libs with no other changes. I haven’t bothered looking at Adelaide or Mitchell as they won’t flip even if there are more declarations).

    Indeed, what Kevin said actually makes a big difference as it would mean the Libs would be well favoured to form a minority government.

  • 113
    Posted Tuesday, March 18, 2014 at 10:10 pm | Permalink

    i.e. same turnout percentage

  • 114
    Posted Tuesday, March 18, 2014 at 10:18 pm | Permalink

    FWIW I’m told that the counting in Colton today was all postals, which have also previously favoured the Libs. Pre-polls and absentees in 2010 were apparently much closer to the vote on the day, and those are what remains to be counted.

    The X-factor is the apparently late swing, so will pre-poll votes be weighted more to the Libs due to being cast before opinions changed? There was a poll a week out showing Colton as 50-50, so if that was accurate at the time, Caica is probably safe. I guess we’ll see in the next couple of days.

  • 115
    Greensborough Growler
    Posted Tuesday, March 18, 2014 at 10:25 pm | Permalink

    The big question is “Does the late swing actually exist”?

  • 116
    Posted Tuesday, March 18, 2014 at 10:32 pm | Permalink

    The big question is “Does the late swing actually exist”?

    I’m pretty sure it was there but it was only about 2% at most.

    The worst polls had the Libs at about 55% and it looks like ending up 53%.

  • 117
    Posted Tuesday, March 18, 2014 at 10:34 pm | Permalink

    There was a poll a week out showing Colton as 50-50,

    The polls were actually pretty bloody accurate.

    The betting market was a shocker.

  • 118
    Posted Tuesday, March 18, 2014 at 10:34 pm | Permalink

    I don’t know that much good evidence exists that shows late swings in campaigns. It seems mostly chatter about campaigns from people who get funded to campaign or jounos paid to cover campaigns.

    A fair few people make up their mind at the time of voting although often this will be non-consequential decisions eg to Vote Family First then Liberal as compared with Liberal then Family First.

    For people doing pre-poll and postal voting the decision is brought forward. The polls were fairly consistent with about a 53/47 2PP preferred to Libs across the State but that failed to account for where the swing of about 1.5% cf 2010 was occurring.

  • 119
    Posted Tuesday, March 18, 2014 at 10:48 pm | Permalink

    Even if Colton flips, I still cannot see any way the Libs can get to a majority. Am I kissing something? If not, they must win the hearts of the independents, and Such seems unlikely to vote for them. I am having trouble seeing how Marshall can be premier.

  • 120
    Posted Tuesday, March 18, 2014 at 10:49 pm | Permalink

    Doh! Missing something…. Stupid fingers :)

  • 121
    Posted Tuesday, March 18, 2014 at 10:52 pm | Permalink

    The difference if Colton flips is Libs only need 1 indie to support them. Or both Indies to say – let Libs form a minority govt and we will vote for legislation on its merits and then Libs just need 1 indie each time to support them.

  • 122
    Posted Wednesday, March 19, 2014 at 7:07 am | Permalink

    So how likely is Brock to vote for the Liberals?

  • 123
    Swing Required
    Posted Wednesday, March 19, 2014 at 10:41 am | Permalink

    The Galaxy polls are one thing, but without checking the figures, I’d say the ALP vote in Such’s electorate was extraordinarily low, as Labor voters went for Such, knowing Labor wouldn’t win the seat.

    Brock also depends totally on Labor preferences, I think.

    I’d say both independents future survival would be more assured electorally by supporting Labor, but that might not be a consideration to them due to age or health.

    Haven’t had time to check that, but perhaps others have?

  • 124
    edward o
    Posted Wednesday, March 19, 2014 at 10:52 am | Permalink

    If Such wanted another term, and he went in with the Liberals, what would all the Labor voters do? Vote Liberal ahead of him in 2018? I doubt it, they’d want to keep him, rightly. Independents keep “talent” from the other side out of parliament and make that other party have to spend resources in otherwise safe seats. Also, they might change their mind at the next election, rather than be someone who will definitely vote for the other side. Such is damaging to the Liberals by occupying their seat, ALP voters would be mad to retaliate even if he comes down against them this time.

  • 125
    Patrick Bateman
    Posted Wednesday, March 19, 2014 at 11:39 am | Permalink

    I haven’t had time to keep up with things due to work –

    Can anyone give me a 2 line summary of where things appear to be heading?

    Or is it genuinely unclear what is going to happen?

  • 126
    Posted Wednesday, March 19, 2014 at 11:43 am | Permalink

    If Such goes Liberal and that makes him unpopular with Labor then they might run a strong campaign. If the Liberals also run a strong campaign then there would be some chance of Such falling to third and then his preferences electing one or the other side.

    Last election the 2PP was only 52.9 – 47.1 to the Liberals.
    This time that will blow out as the Liberals ran very strongly and Labor put very little effort into the seat.

    One fact that Such has continually mentioned is that Fisher has the highest amount of public servants. If Marshall reduces the public service then that might hurt the Liberals a lot in Fisher.

  • 127
    Independently Thinking
    Posted Wednesday, March 19, 2014 at 12:09 pm | Permalink

    Sorry haven’t responded as my computer had home had been lent out to my wife’s friend for some unfathomable reason and the kids had taken their laptops out with them until late…then still wouldn’t let me use them! Thanks…

    Anyway, regarding the late counting and Everything’s and KB’s comments; I have info to indicate that prepolls in Colton will favour the Libs comfortably despite the late swing back to the ALP – and that the postals may fluctuate a bit but they did not come from a particular area or anything – it was random. With the absentees expected to break close enough to 50-50 I think Colton will flip, and the Libs are quietly confident of it too.

    Now what that means is that the seat total become Lib 23 ALP 22 Ind 2 – that set of numbers combined with the Liberal 2PP of around 53%, the poll showing most people in both Independents’ seats want a Liberal Government and Such saying he would look at where his voters put their 2nd preference – it all adds up to the Liberals. However, that is not necessarily how the Independents may choose. They could split.

    Fat lady ain’t singing yet.

  • 128
    Posted Wednesday, March 19, 2014 at 12:14 pm | Permalink

    My attempt at a summary:

    Whilst the outcome of Colton, and to a lesser extent Ashford may be still somewhat uncertain (but certainly both more likely to stay Labor), I think the outcome will be either one of three scenarios, in increasing order of likelihood:

    1)Libs win both Colton and Ashford and have a majority. Seems incredibly unlikely.

    2) Libs win Colton or Ashford. Not especially likely. This would mean LIB 23, ALP 22, IND 22. Liberal minority government.

    3) ALP hold Ashford and Colton. ALP 23, LIB 22, IND 2. Labor minority government. This option is probably twice as likely as option 2 and several log orders of magnitude more likely than option 1.

    Quite simply, in my opinion, the party with more seats will form government.

  • 129
    Posted Wednesday, March 19, 2014 at 12:18 pm | Permalink

    IT – the ALP scrutineers are still cautiously optimistic in Colton. One suggestion was between 75-150 votes in favour of Caica based on their modelling. It will obviously be close though.

  • 130
    Edwina StJohn
    Posted Wednesday, March 19, 2014 at 12:29 pm | Permalink

    Paradoxically it seems like a good one for labor to lose. Better to go out with 22 seats and maybe come back in four years (Victorian style) then cling on and get thumped for a minimum of 2 terms ( NSW style) in four years or less.

    Besides look at how much of labors team are has beens whose best years are behind them.

  • 131
    Patrick Bateman
    Posted Wednesday, March 19, 2014 at 12:35 pm | Permalink

    Thanks sysksie.

  • 132
    Kevin Bonham
    Posted Wednesday, March 19, 2014 at 12:45 pm | Permalink

    I’ve seen people getting this wrong so just to be clear about it: in, say, Colton, there are not 6000 votes to throw. Yes there is 76% showing counted with 19.5K but the 76% is of total enrolment not of total actual votes. c. 6% won’t have voted meaning that the remaining vote number is probably c. 4500 of which 100 or so will be informal.

  • 133
    Posted Wednesday, March 19, 2014 at 12:56 pm | Permalink

    “The betting market was a shocker”

    not for the bookies :)

    backed in from $1.30 to $1.01 means some will punters will be sweating up quite badly this month

  • 134
    Posted Wednesday, March 19, 2014 at 1:12 pm | Permalink

    Kevin is right, there is no way there is more than 5000 votes to be counted in Colton, and it highly unlikely to be much more than 4500.

    With Caica leading Barry by 440 votes, Barry needs to get 54.9% of the remaining votes if there 4500 votes left to count.
    He is doing much better than that but if the absentee votes break at 50-50 then it will depend on how many of them there are.

    IT do you have any information on how many absentee votes there are in Colton?

  • 135
    Posted Wednesday, March 19, 2014 at 1:12 pm | Permalink

    Can anyone explain the huge 20%+ swings to Liberal in Port Augusta and Stirling North? eg. the Port Augusta booth had a 55% ALP 2PP in 2010, but in 2014 it was 68% Liberal 2PP!?!

    This is what has caused the state-leading 13% swing to Liberal in Stuart, the swings in the non-Port Augusta booths were much smaller on average.

  • 136
    Posted Wednesday, March 19, 2014 at 1:34 pm | Permalink

    in, say, Colton, there are not 6000 votes to throw.

    We don’t say “throw” here on the North Island. We say “distribute.”

  • 137
    Independently Thinking
    Posted Wednesday, March 19, 2014 at 1:39 pm | Permalink

    Sorry EM I do not – nor does my contact. I have been told to expect about 1,500 postals for Colton, about 2,000 prepolls but no idea about absentees.

    I agree with KB that the total number of votes will add up around the 94% mark and there will be some informals in there too as part of that 94%.

    I am told to expect that Barry will get 57% or a little higher of the postals and prepolls but closer to 50% of the absentees. It wouldn’t surprise me if the absentees went slightly ALP but that is my gut feel.

    Adding in all those figs assuming Libs stick to 57% on the postals and prepolls and 50/50 on the absentees then the ALP will win by about 40-50 votes, which is what Sykesie seems to suggest the ALP number crunchers feel.

    My Liberal contact swears the votes counted yesterday in Colton were postals.

  • 138
    Posted Wednesday, March 19, 2014 at 1:47 pm | Permalink

    New batch of votes added for Mitchell, where it’s pretty clear the Liberals are home.

  • 139
    Posted Wednesday, March 19, 2014 at 1:53 pm | Permalink

    Kevin Bonham and Edi_Mahin. According to figures from 2010, there was a turnout of 22,000 in Colton with an informal vote of about 1000. Redistribution has added 1000 potential voters from West Torrens. Using 2010 turnout as a guide, this year’s turnout will be about 22,500 which means there would not be more than 3,000 votes to count. Given that the ALP’s lead is a little more than 400 votes and we have only seen a slight advantage to the Libs with pre-polling and postals, I think it is a safe bet that the ALP will retain the seat. The remaining votes will need to be over 57-43 to the Libs.

  • 140
    Frodo Baggins
    Posted Wednesday, March 19, 2014 at 1:55 pm | Permalink

    I can’t see the Indies going with Labor, not after 12 Years of a Stale Labor Government and especially not after what happened to Oakeshott and Windsor(onto the Dole Queue)

  • 141
    Toorak Toff
    Posted Wednesday, March 19, 2014 at 1:56 pm | Permalink

    Does anyone know at what stage the ‘informals’ which have only one box marked are included in the count?

    These, plus the donkey vote, will help Labor in the vital seats.

  • 142
    Posted Wednesday, March 19, 2014 at 1:58 pm | Permalink


    If the figures of 1500 postals and 2000 prepolls include what has already been counted and the total number of declaration votes is going to be about 5500, a figure Kevin and I both agree with, then there would be about 2000 absentee votes to be counted.

    That would imply that Barry would need to make up 440 votes out 2500 and that leaves Caica with just under 100 vote lead if those votes break 57% to Barry.

    Still very much in doubt as those figures do not have to be far off for Barry to win.

  • 143
    Posted Wednesday, March 19, 2014 at 2:03 pm | Permalink

    Yes Mitchell is definitely over although many have said it has been over since quite early in the counting in the seat on Saturday night.

  • 144
    Posted Wednesday, March 19, 2014 at 2:07 pm | Permalink

    Mitchell was a goner on the night for the ALP

    Just down to Colton. My gut says Caica will hold on even though it’s tight

  • 145
    Independently Thinking
    Posted Wednesday, March 19, 2014 at 2:09 pm | Permalink


    Put simply, yes.

    But it would also mean Barry needs to gain about 52.5% of the absentee split which I can’t see him doing. Even the Libs are not saying that is going to happen but it is possible that when things are this tight a rogue batch can throw the result one way or another.

    I hope they don’t lose any ballot papers like in WA.

  • 146
    Posted Wednesday, March 19, 2014 at 2:27 pm | Permalink

    Toorak, the ALP scrutineers tell me that those “informals” with only one box filled in were added in on Sunday.

  • 147
    Posted Wednesday, March 19, 2014 at 2:28 pm | Permalink

    All we need now is some more results from Colton but I have to go to work extremely shortly.

  • 148
    Kevin Bonham
    Posted Wednesday, March 19, 2014 at 2:28 pm | Permalink

    Re #139 we actually know the enrolment in Colton; it is 25512, up from 23517; the enrolment increase is more like 2000 than 1000. In Colton in 2010 there were 21118 formal ballots so if that proportion holds true there should be 22909 this time. There are so far 18860 formal so on that basis 4049 to go.

    However, there is a small catch here which is that for whatever reason the informal rate thus far is 2.8% compared to 3.7% in 2010. If we assume the new informal rate holds for the rest of the count then projected formal voting increases to 23201 leaving 4341 formal votes to be received.

    So assuming the end %age turnout is the same as 2010 and the lower informal rate holds then the Liberal target is marginally over 55%.

  • 149
    Posted Wednesday, March 19, 2014 at 2:47 pm | Permalink

    This is how I see it. In 2010 there were 23,517 voters enrolled in Colton. 1,373 didn’t vote. This election there were 25,512 enrolled. If we deduct say 1,400 non-voters this election, the total votes cast would be approx 24,112, say 24,000. 19,409 have been counted, leaving 4,703 to be counted. Some of these but probably not many would be informal. So say 4,700 in round numbers. If these on average break 55-45 Barry’s way he would pick up 470 votes thereby winning by 20 votes. If they break 54-46 in his favour he would only pick up 376 (not enough to win).

  • 150
    Kevin Bonham
    Posted Wednesday, March 19, 2014 at 3:06 pm | Permalink

    On informal dec vote vs booth vote rates in Colton, in 2010 booth votes were 3.9% informal and dec votes 3% informal. So it could be that we’ll see an informal rate more like say 2% in the remaining dec votes, but this only increases the number of votes very slightly (I get 4377). There is no way there will be only three informal dec votes. :)

    I’ve just had a look at Ashford and the thing to note here is that final turnout in the seat in 2010 was very low at 91.8%. If that repeats then the Liberals need over 60:40 which is surely not happening. If turnout goes up to say 94% they might have a shot.