Photo finish: McEwen

2007 2004
Booths 37,572
Pre-Poll 3,242
Absent 4,335
Postal 3197
Provisional 70
Total 48,416
ALP (adjusted) 50.0
LIB (adjusted) 50.0

Saturday evening. This post will be used to follow the final stages of the count in McEwen, where at the close of election night counting the ABC computer has Liberal incumbent Fran Bailey 0.2 per cent ahead of Labor candidate Rob Mitchell. On raw figures from the AEC Mitchell is 0.4 per cent ahead, but the Liberals had a relatively stronger performance on absent, pre-poll and postal votes in 2004.

Sunday evening. About 70 per cent of pre-polls have been counted, showing basically the same swing as the booth results. So Fran Bailey’s adjusted lead remains on 0.2 per cent.

Monday 10pm. As in neighbouring La Trobe, counting of postal votes (3,309 out of perhaps 8000) has delivered a huge boost to the Liberals. So far 2063 Liberal votes have been counted compared with 1246 for Labor, a swing to Liberal of 1.62 per cent from 2004. On raw figures this has turned Fran Bailey’s 315 deficit into a 502 vote surplus, and boosted her adjusted lead from 0.2 per cent to a formidable 0.6 per cent.

Tuesday 4pm. No new figures added so far today, but The Australian reports Labor has been “buoyed by the discovery of about 3000 votes wrongly sent to neighbouring Scullin. The ALP believes the votes may be from the working-class urban end of the electorate and may strengthen challenger Rob Mitchell’s position”.

Tuesday 11pm. No sign in the count of any undiscovered votes saving Labor’s bacon here. Instead we have 2688 newly added postal votes breaking 1527-1161 Fran McEwen’s way, giving her an 862 vote lead that might be deemed decisive if it weren’t for the lingering promise of those undiscovered votes.

Friday 3pm. Not sure if this has anything to do with those legendary missing Scullin votes, but Labor has done amazingly well from the counting of 2070 absent votes, slashing McEwen’s lead from 862 votes to 396.

Saturday 6pm. Another coup for Labor with a further 1007 postals going 640-367 their way, slashing the lead to just 111. It seems those absent votes from the previous entry were indeed the Scullin 2000 of legend, having been lodged as absentee votes at a booth near the boundary in Epping and sent to the wrong centre. About 4000 more absent votes remain to be counted, but these will presumably not behave the same way as those from the Epping booth, which come from a Labor-friendly area. It cannot thus be assumed that the strong trend to Labor will be maintained (although comments thread chat suggests the bulk of remaining absent votes were in fact lodged in neighbouring Labor electorates). If anyone’s interested, I’ve corrected a spreadsheet error that was inflating the Liberals’ “adjusted” two-party score.

Saturday 11pm. Another 2823 pre-polls have boosted Bailey’s lead by 39 votes.

Tuesday 11pm. A long-delayed addition of 4820 votes has gone against the much smaller earlier batch to break slightly in Bailey’s favour, increasing her lead by 34 votes. Still more bad news for Labor from booth vote re-checking, which has cost Mitchell 61 votes and Bailey only 32.

Wednesday 8pm. A further 1020 pre-poll votes have split almost evenly, increasing Bailey’s lead from 190 to 194.

Thursday 8pm. The contest has again come to life with 1034 absent votes breaking 569-465 Mitchell’s way, with a further 800 still unaccounted for. With re-checking also costing Bailey 23 and Mitchell breaking even, there are now just 82 votes in it.

Friday 8pm. Mitchell continues to barnstorm home, with a new batch of postals going 113-67 his way and pre-polls going 212-202. Re-checking has turned up eight votes for Bailey and five for Mitchell. Bailey now leads by 32. Still unaccounted for: 423 absent votes, 59 pre-polls and 78 postals. Samantha Maiden of The Australian reports that “an estimated 300 votes” remain to be counted.

Saturday 9pm. Good Lord. The final votes are in, pending a recount, and the last few absent votes have broken 100-93 in favour of Rob Mitchell, postals have gone 37-21 his way, pre-polls have favoured him to the tune of 33-23, and further rechecking of booth votes have cost Bailey 14 and Mitchell eight. The results: Mitchell by seven votes in the closest federal election result in modern memory. This post will now be shutting up shop – commenters might like to take their business to the new post.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

113 comments on “Photo finish: McEwen”

Comments Page 1 of 3
1 2 3
  1. I wonder whether pre-poll and absentee votes will actually be as strong for Bailey this time around. I personally doubt it.

    If anything the mood in the polls prior to the election was even more antagonistic to the Libs. Bailey was also caught up in a certain amount of “debate” in local newspapers pointing out that she was making dubious claims about her “contribution” to the local area.

    The Melbourne city booth and two hospital/ nursing home booths are still not counted. Labor won the city booth last time around by a fair percentage. Bailey won the two nursing home booths.

    This is going to end up one of Australia’s most marginal seats, I reckon.



  2. Latest AEC figures (9.33am 25/11) show that after preferences, Mitchell (ALP) is ahead by 570 votes with a swing of 6.8% and ahead on 2PP 50.38% to 49.62%.

    This one is going down to the wire.

  3. Yes, TP.

    The difference between Antony Green and the AEC here is that Antony is assuming similar results from the absentees, nursing homes, postals etc as last time (which did go fairly solidly to Bailey). I’m not so sure that this will be repeated this time. No “Latham factor” to inspire fear in quite so many of the absentees, for example.

    Down to the wire, indeed.



  4. I hope Bailey is beaten then i would not have to see all those absurd airbrushed photos making her look twenty years younger.

  5. Update from AEC:

    1.27pm 25/11 – 73.94% 2PP counted

    Mitchell (ALP) leads by 562 votes with a swing of 6.80% and ahead on 2PP 50.38% to 49.62%.

  6. A comment that applies to all these close races…Liberal apologists/Labor pessimists all expected a narrowing. Now they all expect the Liberals to be saved by voted that were cast well before election day. Spot the contradiction comrades!

  7. The change is from the counting of the votes from the Seymour “Barrabill House Nursing Home” nursing home

    which went 2PP 34-26 to Bailey (a swing against her of 4.11% in that booth).



  8. Another update from the AEC and another narrowing of Mitchell’s lead:

    76.33% 2PP counted

    Mitchell (ALP) now leads by 315 votes. 50.2% to 49.8% on 2PP – a swing of 6.62%.

  9. Mmm. Wonder where that particular set came from? A substantial number of votes in that package , as the total vote has gone up from 73% to 76%. Some absentees from surrounding areas. maybe?

    Still no counting of the Melbourne city booth according to AEC, and it runs Labor’s way usually anyway.



  10. Charlie and Howard Hater, my blue and red figures aren’t the raw numbers – they are treated to compare like results with like. On this basis the swing to Labor is 0.2 per cent short (or at least it was last night). However, I’m not taking into account the fact that different pre-poll figures come from different areas, so that might be skewing things a bit.

  11. Update from AEC and Bailey (LIB) is now ahead:

    26/11 4.41pm

    Bailey (LIB) ahead by 506 votes – 2PP 50.31% to 49.69% – swing is back to 6.11% to the ALP.

  12. Bailey now up by 506 votes as at 16:41 on Monday – huge catch up for the Libs in one day, suggesting momentum is with Fran and will be v. difficult for Labor to get in front again. Back to helping Mr Helper, Robert.

  13. Of course the postals will go frans way (always do)… and so will the ones on Wednesday… Thursdays count is where its at!!!

  14. just some news on mc ewen

    Labor’s hopes of ousting former tourism minister Fran Bailey in McEwen have been buoyed by the discovery of about 3000 votes wrongly sent to neighbouring Scullin.

    The ALP believes the votes may be from the working-class urban end of the electorate and may strengthen challenger Rob Mitchell’s position although he slipped backwards after the counting of postal votes yesterday.

  15. Looks like a late one for the AEC as the McEwan update didn’t come through until 10:10pm. The count has cut the Bailey (Liberal) lead to 396 votes with 84.46% of the vote counted, which now includes 2070 absentees.

    It has been mentioned elsewhere on this site that last night’s count was meant to include the “missing” 3000 votes. The gain to Mitchell (ALP) in last night’s figures, however, seems to have come entirely from the absentees which are running 61% in Labor’s favor.

  16. I’ve been doing some further numbers…

    * Ordinary (polling booth) results are about 3500 below what might be expected based on the 2004 results, adjusted for the 9% increase in enrollment in McEwan from 2004. Consequently, the “3000 missing votes” may yet turn up.

    * Based on 2004 returns, there are likely to be approximately 4000 absentee, 1500 pre-poll and 1000 postals yet to be counted, or a little over.

    * If remaining absentee, pre-poll and postal returns remain consistent with trends to date this results in around 3115 extra votes to the ALP and 2385 to the Liberal Party. This would be enough for the ALP to win the seat by a bit over 200 votes.


    * The current percentage of absentees going to the ALP (61%) is dramatically higher than 2004 (44%) yet postals and prepoll are more consistent with 2004 (2004 – 39% pre-poll, 39% postal to ALP; 2007 – 45% prepoll, 40% postal to ALP). This suggests that there might be some weirdness in the early absentee counting. If the 61% absentee ALP vote doesn’t hold up then the Liberal party looks pretty safe.

    * We’ve been lead to believe that if these 3000 mystery votes ever appear they’re likely to favor the ALP. Although it’s possible that they’re still out there I personally won’t be holding my breath.

    Conclusion: the ALP remains a real chance here, but given the above uncertainties my money is on the Liberal Party retaining the seat by a nose.

  17. Rob Mitchell has cut Bailey’s lead from 800 odd votes to 396!!!
    This from the AEC at 5pm Friday:

    Turnout: 84.46%
    BAILEY, Fran Liberal 42,905 50.23
    MITCHELL, Rob Australian Labor Party 42,509 49.77

  18. I understand that the ‘missing’ votes from Scullin were counted yesterday afternoon – hence the halving of the margin overnight.

    They weren’t missing per se, but were from a large Epping booth which is located in Scullin – but where many McEwen voters from north of Epping and Mill Park lakes voted. These people were voting as absentee voters.

    On that basis, the Libs’ chances in McEwen are now much better than 50/50.

  19. potato_masher @ 20 –

    Hmmm… I’d like to see the numbers behind your odds… 400 votes seems to be a huge amount to pick up given that in 2004 only 44% of McEwan’s absentee votes went to Labor. (Not to mention remaining postals and prepolls, 55-60% of which are still flowing to the LP.)

    Even if the remaining absentees follow the seat’s overall swing of 6% to the LP this simply reinforces the status quo at 50/50. The ALP’s only hope in this seat is if the absentees stay on 61/39%, which seems unlikely… unless there’s something I’m missing?

  20. Actually I have changed my mind re Mcewen…but only slightly…

    There are still an amazing 7000+ absent votes to count in McEwen as well as 5000 pre-polls

    Expect that the Libs will gain a few hundred on pre-polls

    But the absents will mostly be concentrated on the southern metro boundaries of McEwen…predominantly strong Labor outer northern suburbs such as Epping, Mill Park as well as less strong areas near Diamond Creek and Greensborough…these are absent votes from people who live in McEwen but have voted down the road (in the same or neighbouring suburb) which is in Scullin or Jaga Jaga.

    With 7000 votes at say 55%, Labor could gain as much as 700 votes…and I would put 55% as a moderate level. The Libs will have to depend on keeping up their leading trend in the remaining pre-polls and postals or will be overtaken

  21. I’ve been up in Queensland for the last few days and I’d just about given up on McEwen from the numbers I saw earlier in the week. Got quite a shock at Rockhampton airport on the way home when I saw this afternoon’s figures.

    Bailey must be seriously sweating , I reckon, and not just because of the hot weather that I gather we’ve had here while I’ve been away!

    I’d be very surprised if Labor don’t pick up this seat next time around if they don’t make it this time. Bailey is getting old. If she stands again in 2010 then I suspect that the national incumbency will boot her out. If she doesn’t stand then loss of the local member factor and probably some demographic changes will make it a Labor seat on Green’s prefs.

    If Labor doesn’t win McEwen in 2007 then it is going to be a mighty enticing prospect for a good candidate next time around! The demographics suggest it is likely to simply keep on getting better for Labor barring a major redistribution. But it looks to me as if there is still a 40% chance that Mitchell will take it for Labor this election



  22. What will probably save her skin is the forthcoming Victorian redistribution due in 2009. The current enrolment of 104,509 is already on the high end of the range.

    In the next redistribution McEwan will almost certainly move northwards, cutting out a few of the Labor voting booths in the southern end of the electorate.

    The only thing that might cost her the seat is an early DD election.

  23. That was a huge surge for Mitchell in the last batch of counting given that the only change from yesterday is the addition of a little over 1000 postals.

    Mitchell, therefore, has picked up almost 60% of this latest batch. Quite a dramatic change from the split in earlier postals.

    This makes some sense, since to date the swing to labor in postals has been <1%, compared to the swing in the seat overall of around 6%. If Mitchell can pull the postals up to the average swing in the final batch then he is indeed in with a fighting chance.

    Although I’d still call it 50/50, the token $10 I put on Mitchell to win might be safe after all… the kicker will be whether the swing on absentees is a little above or a little below the seat average.

    If remaining absentees reflect the overall -6.35% swing to date Mitchell likely gets over the line. If they lag behind the seat average then Bailey survives, bearing in mind that in 2004 only 44% of absentees went to the ALP.

  24. Mitchell is murdering poor little Franny on absentees. 111 is the margin and now and there are still many thousand to be counted. Is this the revenge of the shift workers in the northern suburbs?

  25. Anyone know where the remaining absentee votes are from?

    There will no doubt be some from booths in Scullin, Jaga Jaga and Calwell, but no doubt there will be some significant numbers from less Labor friendly areas such as Indi, Murray, and maybe for some other rural voters, Bendigo.

    I wonder which have been counted and which are still to go.

  26. The only additional votes showing this afternoon so far (on the 5.54PM update) are actually pre-polls, which have pulled Bailey back to a lead of 286 votes BUT the overall percentage of pre-Polls she got from this lot again was substantially less than the earlier ones.

    Still no sign of additional absentees on the AEC website.



  27. Sorry, don’t know how I imagined that figure – should have said “have pulled her back to a lead of 150” (not 286 – that is merely the overall difference on pre-polls – sorry about the brain fade)



  28. In response to

    # Rob on 01 Dec 2007 at 9:48 am

    “What will probably save her skin is the forthcoming Victorian redistribution due in 2009. The current enrolment of 104,509 is already on the high end of the range.

    In the next redistribution McEwan will almost certainly move northwards, cutting out a few of the Labor voting booths in the southern end of the electorate.”

    I have to disagree with your assertion here.

    McEwen is currently the largest electorate in Victoria by number of electors, but if you look at the surrounding electorates – it is likely to move South.

    To its North – Indi, to its South-East – McMillan, to its West – Bendigo, to its South – Ballarat, Lalor, Calwell, Scullin, Jagajaga, Menzies, Casey and LaTrobe.

    If you look at the sizes of these electorates.

    VIctorian Average – 93,030.

    McEwen – 104,509 (Victoria’s Largest)
    Indi – 90,871
    McMillan – 87,092

    Bendigo – 97,197
    Ballarat – 93,624
    Lalor – 103,761
    Calwell – 95,622
    Scullin – 88,924
    Jagajaga – 93,765
    Menzies – 89,968
    Casey – 89,313
    LaTrobe – 92,029.

    Looking at that, surely the most likely outcome is that parts of Northern McEwen will be chopped off and given to Indi (Eildon, Alexandra, Yea perhaps) – which can only expand southward, and perhaps parts will be given to McMillan (In the Yarra Ranges). The rest could stay relatively the same, but the relative weight of the southern areas of McEwen will be increased.

    I do agree with the contention of # Rod on 01 Dec 2007 at 1:05 am, that it is likely to fall to Labor the next time it goes to the polls in the near future. Either at a by-election or the 2009/2010/2011 election – whenever it is held.

  29. I’d been thinking along similar lines, myself, McEwen Man, but another distinct possibility may be that they will simply shuffle off the rapidly developing Mill Park / Mernda growth corridor area along the northern extension of Yan Yean and Plenty Roads into Scullin. This, in the short term, could be used to restore basic numerical balance (though Scullin itself would in the medium term then be likely “outgrow” its neighbours, too, I guess.)

    Still no sign of any more of those absentee votes!



  30. I still can’t understand the “declaration Vote Scrutiny process” page for McEwen.

    (see )

    The number of absentee votes so far received seems very low and the tallies simply don’t seem to match, with more votes being counted than “envelopes issued” in some categories.

    Is it likely that some of the “absentee” votes have incorrectly been added to the “pre-poll” and “postal” categories, perhaps, or am I simply misunderstanding the meaning of the figures?

  31. McEwen Man – Your numbers are good.

    My call was a gut reaction based on my experience with particular outer-suburban seats that have in the past tended to move progressively away from Melbourne following each redistribution. I’ve now had a look at the actual numbers and it does indeed seem that the next push will be towards the city.

  32. Rod on 02 Dec 2007 at 9:14 am –

    I would guess that the numbers ISSUED are those issued *by* the McEwan AEC office or booths, but not necessarily *for* McEwan. So, for example, “absentees issued” will all be for ballots issued to people enrolled in other electorates who are voting at McEwan booths. (While McEwan voters voting elsewhere won’t be included in this total.) The ‘issued’ numbers therefore may not be a good guide of what to expect.

    A better guide is the 2004 returns, to which you might want to add 9% for seat growth. For the record:

    2004 totals:
    ABSENT 6,188
    PRE_POLL 3,451
    POSTAL 6,798

    By this account, pre-poll looks done and postals look close to done, with only a couple left to trickle in from Tierra del Fuego (or wherever) over the next week.

    We know that we’re waiting on around 1091 provisionals, of which about 600 are likely to be valid. We’re also waiting on around 4000 absentees.

    On the other hand, if you look at turnout, we’re currently on 92,230 but you would reasonably expect to receive around 100,000 at 96% turnout. We therefore know that we have around 7500 ballots left to count overall – most likely the extra 2400 will split mostly between absentees and a few more postals. (Also, if you’re only looking at the 2PP totals you need to subtract around 3.5% from the above figures to allow for informals.)

    Therefore, the seat will be decided on absentees and Labor needs to hold at least a 6% swing on absentees to be in with a chance. My guess it that it’s looking good for Mitchell, although there’s no doubt it will go down to the wire.

  33. Rob wrote:

    I would guess that the numbers ISSUED are those issued *by* the McEwan AEC office or booths, but not necessarily *for* McEwan.

    Perhaps, Rob, but if you have a look at the very “neat” set of figures from neighbouring Macmillan at or in the Queensland seat of Dickson, for example, you will see that the McEwen numbers look exceeding strange!

    My suspicion is that the number of ” issued” number is, indeed, meant to be for this electorate, not from this electorate. I think either there is some category misattribution here, or there are data entry errors for some categories, or some of the figures for votes received have not been properly updated on the web site.

    Regardless, it is pretty clear that there are still enough absentee votes waiting to come in to give the seat to Rob Mitchell if current percentages hold, as you suggest.

  34. I wrote My suspicion is that the number of ” issued” number is, indeed, meant to be for this electorate, not from this electorate.

    This piece of nonsense was, of course, meant to say:

    “…….the “number of issued” number is, indeed, meant to be for this electorate, not for those voting ‘absentee” for other electorates at McEwen booths .”

  35. Rod and McEwen Man: I think it’s too hard to work out what would happen to McEwen in the next redistribution.

    While McMillan is short on numbers, you have a mountain range between it and McEwen, and I doubt the AEC would move the Yarra Valley to McMillan.

    I see two options – either little bits of urban sprawl get hived off to seats like Scullin, Casey and Menzies (this would probably help the Libs in McEwen), or somewhere like Seymour, and/or Alexandra, Eildon and Yea get moved to Indi (which would definitely help Labor).

    One one side of the electorate, the Diamond Creek area is growing rapidly, while houses are going up everywhere in the new suburbs north of Scullin. The population projections would suggest quite a lot of voters will need to be taken out of McEwen to cope with future growth.

    I don’t think the AEC will fiddle much with Bendigo, and if Ballarat is to absorb more voters, they’d probably come from Lalor.

    My rough guess is that seats like Scullin will be moved a bit north, because their population projections wouldn’t be as healthy. And that would help Bailey (presuming she survives this election – it’s going to be close).

Comments are closed.

Comments Page 1 of 3
1 2 3