Victorian election: photo finishes

A closer look at the yet-to-be-decided seats for the Victorian election.

A full display of the election results, with complete booth figures, swings and probability estimates, can be found here. This will be updated with the latest figures at irregular intervals.

Tuesday afternoon

4pm. Ali Cupper has reportedly emerged the winner in Mildura after distribution of preferences by 254 votes.

2pm. No official figures available, but the preference count in Prahran has established that Greens incumbent Sam Hibbins has prevailed over Labor’s Neil Pharaoh, apparently by around 200 votes, with the latter conceding defeat on social media. So the Greens have maintained their lower house status quo of three seats, losing Northcote but gaining Brunswick, although they stand to be gutted in the upper house, where they went in with five and will come out with one or two.

Tuesday morning

Preference distributions will apparently start being conducted today, and it doesn’t seem there are more than tiny handfuls of votes remaining to be cleared up in the primary and two-party counts. So unless the preference distribution process turns up a misplaced bundle, it would seem Labor has won Bayswater and Hawthorn, the Liberals have held on in Caulfield and Ripon, and independent Ali Cupper has scraped home in Mildura. The only significant action in the close seat counts yesterday was in Caulfield, where postals continued to save the day for David Southwick, the latest batch favouring him 145-61 and extending his lead from 338 to 410.

Monday morning

The most interesting development over the weekend from my perch was that 525 postals were added in Mildura, which cut Ali Cupper’s lead from 252 to 189. With the dealine for postals to arrive being tomorrow at 5pm, there will presumably be only one more, smaller batch to go, and very little chance that it will overturn Cupper’s lead. Tiny additions to the vote in Hawthorn and Ripon made no appreciable difference to the situation there, as related in the previous entry. Kevin Bonham has been doing good work following the count in Prahran, where only scrutineers’ reports offer any guidance as to the flow of preferences between the Greens and Labor, which stands to decide the result. Suffice to say that it’s going to be very close. The other potential wild cards when preferences are distributed are Melton and Benambra, which Labor and Liberal respectively have at least some chance of losing to independents. Then there’s the upper house …

Friday evening

Just as my interest in the count was winding down, along comes Caulfield — Liberal member David Southwick led by 1101 yesterday, and trails by 118 today. Southwick is one of three Liberals on very narrow deficits, hoping they might be overturned on the final batch of postals. The damage to Southwick was done on absents — not so much due to the swing, which was actually modest (3.6% to Labor, compared with 7.0% of ordinary votes), but the surprisingly high number cast (5692 as compared with 3130 in 2014). Labor scored over 60%, for a split of 3439-2253. Postal votes have been overwhelmingly favouring Southwick — 2682 to 1030, or 72.3-27.7 — so it will only take a small number of late arrivals behaving according to form to get his nose back in front. But there will very little in it either way.

Provisionals and a handful of pre-polls have made next to no difference in Hawthorn, where the Labor lead goes from 156 to 163. A big day of counting in Ripon (or maybe two — I don’t think I looked closely at the numbers yesterday) has failed to settle the matter — Labor has moved to a razor-thin 73 vote lead due to the latest pre-polls, which they won 2211-2059. There was nothing in it on absents (1296 to Labor and 1289 to Liberal), and Labor made their usual small gain on provisionals (147 to 124).

In Mildura, independent Ali Cupper got a handy 195-144 break on provisionals, cancelling out a 41-22 loss on pre-polls and 20-10 on absents in a race where every vote counts. She now leads by 303 votes, which will presumably be enough.

Friday morning

As the count dries up, the in doubt seats are increasingly looking less so. Labor’s lead in Hawthorn grew from 47 to 156 yesterday, as they gained the edge on absents (281-207), postals (109-86) and pre-polls (27-15). They should gain a bit more when provisionals are added, leaving John Pesutto needing something pretty extraordinary on late postals. Labor’s lead nudged from 236 to 266 in Bayswater, after provisionals broke 120-88 their way and postals went 69-67 to the Liberals. The Greens’ lead in Brunswick grew from 353 to 414 with small additions of absents, postals and pre-polls, at which point you would be pretty much calling it. It no longer seems necessary to continue following Nepean, where Labor leads by 794, or Sandringham, where the Liberals lead by 451. No progress today in Mildura, where independent Ali Cupper leads the Nationals by 281.

Thursday morning

Another good day for Labor overall, who seem to be doing better from votes cast out-of-district, whether as absent votes or pre-polls, than the in-district pre-poll votes that were counted on election night. However, I’m not clear if absent votes are all being added in one hit per electorate, or if further additions can yet be anticipated where results have already appeared. I’m tending to think the latter — since absents are usually the best part of late counting for Labor, a fair bit hinges on this.

In Hawthorn, absent votes turned yesterday’s 235-vote Liberal lead into a Labor lead of 47. Labor got 56.8% out of 2498 absents, above the 53.0% I was projecting. I was also projecting there would be 3792 in total, so I am guessing there are another 1000 or so still out there. These will be decisive if so, but it can’t be said how they might behave — batches of absent votes can behave very differently depending where they were sourced from. The Liberals also got only 50.7% out of 4242 new pre-polls (2150 to 2092) added yesterday, compared with their 56.3% of the first 7148 counted.

Things continue to go Labor’s way in Bayswater, where their lead grew yesterday from 165 to 236. Labor got 55.2% of the absents added yesterday — exactly as I had anticipated, but they were 2705 in total rather than my projected 2054. It was also a good day for Labor in Nepean, where they won a batch of 3673 new pre-polls 1903-1770 — 51.8% compared with their 46.4% from the first 14,903. Labor now leads by 492, and most of the outstanding votes are likely to be absents, none of which have been added, so the balance would seem to be tipping their Labor. My lineball projection as of yesterday is now for a Labor winning margin of 0.5%.

One late counting bright spot for the Liberals is Sandringham, where 4464 new pre-polls behaved very much like the first 9424 in breaking 2489-1975 their way. Furthermore, absent votes were added and while they went 1139-884 to Labor, there were less of them than I was anticipating (2023 rather than a projected 2566). However, I’m not sure if this is all of them or not. In any case, the Liberal lead is now 497, and with only a few scraps still outstanding, this will be hard for Labor to rein in.

The Greens’ lead in Brunswick grew from 218 to 353, but they underperformed my projection out of 2653 absent votes counted in Brunswick, scoring 1475 to Labor’s 1178 — 55.6% compared with my projected 61.8%. However, that’s also about 1000 less than I was projecting, so there are presumably more of these to come. There would seem to be another 2000 of these as well. Labor will need about 56% of what’s to come.

The Nationals have come storming home in Mildura, being overwhelmingly dominant on the small number of absent votes (575 to independent Ali Cupper’s 227), and reversing earlier form to win a batch of pre-polls 870 to 640. This slashed Cupper’s lead from 859 to 281. My earlier judgement was that the number of votes outstanding here was too small for the Nationals to close the gap, and that probably still holds, as I believe there are only a few hundred postals still to come.

Nothing today from Ripon.

Wednesday morning

Labor solidly outperformed my projections yesterday in Bayswater, scoring almost exactly half of 4559 pre-polls added, where they only got 46.1% of the first 8383. They also won 52.1% of 674 postals, after scoring only 39.6% of the 2217 counted on election night, did about as well on absents as anticipated, winning 1132-922. That gives Labor a lead of 165, or 0.2% – with not much of the vote outstanding, my projection has it coming down to 0.1%, but Labor will more likely than not continue outperforming its assumptions.

Better news for the Liberals from Ripon, where 889 postals broke 535-354 to Liberal (60.2% compared with 58.0% in the first 3735) and pre-polls went 393-304 (56.4% compared with 51.0% of the first 3302). My Liberal projection has gone from 49.9% to 50.2%, but here too the number of postals received has exceeded my projection, so if anything it might be understating their chances. That said, the margin is narrow enough that a good pre-poll batch or better than expected show on absents for Labor could up-end it. Swings and roundabouts in Nepean, where the Liberals went below par on yesterday’s postals (360-297 in their favour, or 54.8% compared with 59.2% in the election night batch of 2341), but above par on pre-polls (587-357, 62.2% compared with 53.0% of the first 13,959). Before I was projecting a 134 vote win for Labor, now it’s 26 votes for Liberal.

The Liberal lead in Hawthorn increased yesterday from 53 to 235, but only postals were added, and these were slightly less favourable to the Liberals than those counted on election night, bringing my projected final Liberal margin down from 1.1% to 0.8%. The election night postals went Liberal 1104 (60.4%) and Labor 725 (39.6%), but yesterday’s batch went Labor 1115 (54.3%) and Liberal 937 (45.7%). No further pre-polls have been added, and the outstanding ones may yet surprise in either direction. Then there are absents, which I am projecting Labor to do well on, though evidently not well enough.

The Greens are firming in Brunswick: they won a second batch of postals 442-426, after losing the election night count 950-699, and they won a batch of absents 811-537, exactly the proportion anticipated when I projected them to win by 1.0%.

Tuesday afternoon

Labor leads on the raw count with about a third done in Morwell, but my projection is that this will flip when the outstanding votes are in — Northe is on track to receive about 70% of preferences, in which case he wins 52-48 (I conducted a regression analysis to test whether the existing preference count was representative of the whole, and found that it was). Better news for Labor in Geelong, where Christine Couzens leads Darryn Lyons, and Pascoe Vale, where Lizzie Blandthorn leads Oscar Yildiz 59.0-41.0. In Shepparton, Suzanna Sheed looks seat to emerge with 54% to 55% against the Liberal candidate.

Tuesday morning

With very little counting done yesterday, the chief news is that the Victorian Electoral Commission announced it is conducting new preference throws to indicate the likely winners Morwell (independent versus Labor), Geelong (Labor versus independent), Pascoe Vale (Labor versus independent) and Shepparton (independent versus Liberal), and will publish the results later today. The removal of the two-party numbers from the media feed caused my results reporting facility to conk out, so the figures it show remain those from Sunday.

The only thing I know so far about the new preference throws is that the Pascoe Vale pre-poll count has broken 6059-6008 for Labor’s Lizzie Bladthorn over independent Oscar Yildiz, as related by Richard Willingham of The Age. This suggests the advantage to Yildiz on preferences is only 53-47, in which case Blandthorn would win handily with between 54% and 55%. Independents Russell Northe and Darryn Lyons will respectively need around 72% and 66% of preferences in Morwell and Geelong. The deal in Shepparton is that it’s the Liberals rather than the Nationals who finished second, but unless I’m missing something, it would seem to me that Suzanna Sheed is home and hosed in either case.

The only change in the seven seats where I felt the existing notional counts were following was in Ripon, where a batch of pre-polls broke 587-357 to the Liberals – 63% compared with their earlier 53%. This means my projection has gone from 0.1% in favour of Labor to 0.1% in favour of Liberal. The votes counted totals for the upper house have edged up from the forties to the fifties, but I’m still holding off looking into them in detail.

Sunday night

If you want real detail on the likely course of the late count, Kevin Bonham is your man. For starters, I will content myself with the following projections of how the undecided seats where the correct two candidates have been picked for the notional preference count stand to play out. As explained below, there are methodogical details that one might well think imperfect, but if nothing else, consider it a conversation starter.

This assumes that a) outstanding pre-polls will break the same way as those already counted, and the number outstanding is as indicated by the relevant figures from the Victorian Electoral Commission; b) postals will break the same way as those already counted, with the total number to be counted equal to the total in 2014 adjusted in proportion to the growth in enrolment since that time; c) absent votes will differ from non-absent votes in the same way they did in 2014, with the total number determined the same way as for postals. No account is made for provisionals, which should throw a handful of extra votes Labor’s way.

This makes it clear enough that the Liberals should get home in Hawthorn, Sandringham and probably Bayswater, but Nepean and Ripon will go right down to the wire. The Greens’ traditionally strong showing on absent votes should see them home in Brunswick, and it seems likely independent Ali Cupper will gain Mildura from the Nationals. I was circumspect about this in my post last night, as I expected the Nationals would do well on postals – but it turns out that, for whatever reason, very few postals are cast in Mildura. Indeed, it ranks last in the state for number of postal votes received, according to the VEC’s figures.

Then there are the in doubt seats for which the two-party count doesn’t offer an insight. Prahran will be won by whoever out of Greens incumbent Sam Hibbins and Labor’s Neil Pharoah survives the second last exclusion on preferences, which will be absolutely touch and go. Then there is my watch list of five seats (not counting Mildura) that could potentially be won by independents, as discussed in my previous post. Morwell could stay with Nationals-turned-independent member Russell Northe, and will go to Labor if it doesn’t; the Liberals might lose Benambra; Labor might lose Geelong, Melton and Pascoe Vale.

That leaves Labor with 49 seats nailed down, on top of which they might keep Geelong, Melton and Pascoe Vale, and gain Prahran, Nepean, Ripon and Morwell. The Coalition have 27 seats in the bag, including Hawthorn, Sandringham and (perhaps generously) Bayswater, on top of which they might keep Nepean, Ripon and Benambra. I’m pretty sure the Greens will have Brunswick in addition to Melbourne, and are lineball to keep Prahran. I’m giving Mildura as well as Shepparton to independents, to which it’s at least possible to add another five.

As for the upper house, we’re still at too early a stage in the count for me to be bothered putting my oar in – only election day votes have thus far been counted, and an increasing number of voters have finally got the message about the advisability of voting below the line (requiring the numbering of only five boxes in the case of Victorian state elections). However, it looks fairly clear that there will indeed be a spectacular array of micro-parties on the cross bench, and that the principal casualty of this phenomenon is the Greens.

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.

457 comments on “Victorian election: photo finishes”

  1. The Greens, while they have suffered a primary vote swing against them at this election (much of which, outside contest where an independent was challenging one of the other major parties, came back on preferences, indicating increased competition from smaller parties rather than a wholesale decline), are in a decent position to gain at the next election.

    The ALP will be in government at Commonwealth level (almost certainly with a majority), rather than the current highly unpopular Coalition government, which the laws of electoral gravity indicate is a time there will be more voters potentially willing to switch from the ALP to Greens as the ALP will make some decisions that will not be left-wing enough for at least some of these voters.

    The Victorian ALP government will have had 8 years in office and will be subject to greater electoral gravity, like it did in 2006.

    The redistribution will either draw Pascoe Vale District south to include more Green friendly territory or create a new seat in the Fitzroy area, due to strong population growth in Melbourne, Richmond and Brunswick districts.

    The Greens are likely to learn from their preselection mistakes this time and thus not have a better campaign next time.

    The Greens are now clearly not longer an upper house minor party in Victoria, most of their parliamentary representation is in the Legislative Assembly, so they look more competitive in the Legislative Assembly.

  2. Love your ‘logic’ there, the Greens lost four LC seats, so have more LA seats, therefore they are now a lower house party. What the….? The reality is that the Greens are prisoners of the whims of the Tories and whether they run and or whether they allocate preferences to Labor.

  3. The Greens are not at the whim of the Tories. The Greens won their seats in the Legislative Assembly without the Liberals directing preferences and with the Liberals running in every seat the Greens won. The Liberals making a song and dance about potentially not running was mainly about getting a preference deal with the ALP (In Lib versus Nat and maybe Lib versus Independent), which they did not. The Liberals not running, like happened in Richmond, helped the ALP. The Liberals are restricted in their ability to not run by their need for Legislative Council vote in Northern Metro, which not running reduces. Many Liberal voters would disobey the how to vote card if it said to put the Greens ahead of the ALP in an ALP versus Green contest. Prahran was also won with ALP preferences, after overtaking the ALP on Animal Justice Party preferences, beating the Liberal candidate.

  4. Considering their campaign disasters, the Greens’ result was OK, not great but OK.

    They have condolidared their hold in the inner city LA seats. Richmond was a disappointment for them but they will be in with a big chance when Richard Wynne will retires, presumably at the next election.

    They might also have a shot at Northcote but this depends on gentrification of the working class areas that voted strongly for Theophanous.

    And as Tom the first said another seat might be created for them.

    They will struggle though in the LC under the current system which is a good reason for the Labor Party not to change it. Labor will be quite happy to throw a few crumbs at whatever minor cross bencher to get their legislation through and they can play the cross bs off each othet, Labor does not want to be dependent on the Greens.

  5. If Labor do keep the current upper house voting method in place it will be interesting to see what happens with the micro parties and the ‘preference whisperer’. Many of the micros must be feeling like they were sold a pup as they watched all their preferences funneled into the Derryn Hinch Justice Party (which he was working for at the same time as preparing all their preference ticket flows) in half of the regions, and certain other ‘favoured’ groups in each of the other regions.

    Rumours abound that some people paid for ‘gold’ service and were thus rewarded with more favourable status in the labyrinthine world of 14 different micro group voting tickets per region. It would have been hard for most micros to realise this until every card was on the table. Most probably will only truly realise it once the results are finally tabled.

    So there may not be the same degree of co-operation next time, as some of the micros may feel they were the ‘bunnies’ in a very profitable (for one person, and one micro in particular) rort.

    “All micro parties are equal, but some are more equal than others.”

Leave a Reply

Your email address will not be published. Required fields are marked *