The NSW Labor Party has again hit rock-bottom in a new state Newspoll out today in the Australian. The poll gives the ALP only 26% of the vote, which is their current record low from a poll last year following the removal of Morris Iemma. Another remarkable figure is the performance of the Greens, who have jumped 5 points to a record 17%. The Greens’ previous best in NSW was 15% in the first poll of this year and have consistently polled in the 13-14% range for most polls in the last two years. The Liberal-National coalition is on a dominant level of 44%, unsurprisingly.
The Australian reports that the two-party-preferred split is 59-41, although they sensibly bury this statistic in the ninth paragraph, as such a calculation has little bearing on results in a state like NSW with optional preferential voting and 30% of voters favouring a minor party or independent. In such circumstances pendulums lose some of their value in predicting the seat changes with such swings. As such, I’ve attempted to construct a seat calculator which attempts to change primary vote levels as a proportion of the swing. I’ve plugged the latest poll figures into this calculation and produced the following figures:
- Liberal – 43
- National – 16
- Labor – 21
- Independent – 8
- Greens – 5
I must stress how broad and loose such a calculation is, but it gives us an idea of the scale of defeat Labor could suffer. I have assumed low levels of Greens preferences to the ALP and strong preference flows from independent candidates to the Coalition, meaning Labor pretty much doesn’t win any seats except where they lead on primary votes.
It’s also important to stress that this is a single poll and Labor is extremely likely to perform better than this, and I expect the Greens will return to lower, but still very high, levels, such as the 13-14% range they have occupied for most of this term. This isn’t a prediction, but it is a projection of how bad an election could be for the ALP on 26% of the vote. Think of it as a thought experiment. Like a pendulum, this assumes a uniform swing across the state and assumes that only independent candidates who ran last time will run again, which is obviously a flawed assumption. This assumption produces a result of four independents being elected in the Hunter, up from the current one.
Maps posted below the fold.
I know how much you love maps, but can’t you at at least provide the pendulum list as well? 🙂
Are you suggesting the Greens could/will win Blue Mountains and Coogee? In both seats, the increased Liberal vote would surely give them a better chance than the Greens? I’d submit that Kiera (safe Labor seat where the Liberals don’t really register) is more likely for the Greens.
Also reckon any general rise in the Coalition vote would put Tamworth under threat.
Ben,
What are the five seats your are predicting for the Greens. Such a result obviously would help in the Upper House as well – would that mean 4 seats for the Greens? You would expect a primary of 17% in the lower house would be translated to a higher vote in the upper house; is that normally true?
Simon
Simon – the LA figure wont transfer directly to the LC because their will be more tickets running in the LC – for instance, last time their were 20 columns – 19 groups plus ungrouped. There are 18 registered parties at present so I would expect most if not all to lodge an LC ticket. This will disperse some of the LC Green vote (ie; the Democrats, Socialist Alliance and the Human Rights Party will be there). That said, last time the Greens LA result was 8.9%, with the LC result being 9.1% – the correlation might be that it will translate as a 17% into the LC. But I don’t seriously expect that as a primary vote…more likely 12-13%. Still, that’s 3 LC seats.
MDM – Coogee will depend if the Greens get past the ALP and ALP voters preference to the Greens – I suspect that unless there is some sort of preference exchange this wont happen (ALP voters are pretty disciplined lot, so no recommendation will equal just a first preference. Blue Mtns would require the ALP vote to plummet and the same scenario to be played out as in Coogee. I still would be thinking 2 seats to the Greens and then a couple of maybe’s.
So, by the logic of the NSW ALP this poll result should mean the powerbrokers will now be rushing to dump Keneally and replace her with someone else?
Notice Heffron (Keneally’s seat) is one of the seats going to the Greens.
I assume the Blue Mtns numbers are affected by the 10%+ vote for an independent there last time. I think I’ve noted before that it’s not entirely implausible that the Greens could win it if a strong independent who mainly takes votes off Labor polls in the 15-20% range, falls just behind the Greens, and a decent share of their vote flows to the Greens in prefs, putting Greens ahead of Labor, with Labor prefs then pushing ahead of Libs. Difficult, but worth considering.
And of those 21 remaining Labor seats, one could see independents potentially picking up a few more of them. Cessnock might be vulnerable given the caning Labor took in the council elections there last year (indeed, it must be close to falling to the Nats anyway). Wollongong and Keira must be strong chances of seeing independent challenges, particularly in the wash-up from the corruption scandal. Shellharbour perhaps as well, but that is much more solid Labor voting territory.
And Ben, what happens in Kiama? It’s only on a 12% margin, it must come ultra close to falling to the Libs?
I just can’t see Heffron falling, especially with Keneally as Premier. While the Labor vote will almost certainly fall, she has a very rusted-on support base, especially in the Southern suburbs and Waterloo. She’s a pretty popular local member (certainly given that this is not a particulalrly popular Gov).
I also expect Keneally to gain at least some traction over the next 15 months; she’s done pretty well in the media so far, even with the Herald and Tele blatantly campaigning against her. As Ben said, this is probably the high water mark for the Greens and the low water mark for Labor. Still, if the Greens were offered 2 seats and 13% at the 2011 election a couple of years ago, there wouldn’t be a single person who wouldn’t have taken that.
Hi all,
I only used the calculator to determine primary votes, I then made predictions on the result in seats where no-one gained a majority. In some cases I made brave predictions. For example, Kiama came out as Lib 38, ALP 35.5, Green 16, Ind 10, and I predicted in that seat the ALP could win on preferences. Maybe that was an incorrect assumption now that I think about it more.
In Keira the ALP came out ahead on primaries and the Greens were third on 23.7%. So the ALP held on.
In Coogee the Liberals polled 38.7%, the Greens 33.4% and Labor 23.9%. I figured in that case Labor preferences would put the Greens over the top.
In the Blue Mountains the Liberals led on 31% on primaries with both Labor and Greens on 26%, but Greens slightly ahead. I figured that this would give the Greens the seat.
Again, it’s just a thought experiment to show the general scale of the disaster they are facing, not individual seat results, but for Tim’s benefit the list is posted below.
INDEPENDENT
Charlestown
Lake Macquarie
Maitland
Newcastle
Northern Tablelands
Port Macquarie
Sydney
Tamworth
GREENS
Balmain
Blue Mountains
Coogee
Heffron
Marrickville
NATIONALS
Ballina
Barwon
Bathurst
Burrinjuck
Clarence
Coffs Harbour
Dubbo
Lismore
Monaro
Murray-Darling
Murrumbidgee
Myall Lakes
Orange
Oxley
Tweed
Upper Hunter
LABOR
Auburn
Bankstown
Blacktown
Cabramatta
Campbelltown
Canterbury
Cessnock
East Hills
Fairfield
Keira
Kiama
Kogarah
Lakemba
Liverpool
Maroubra
Mount Druitt
Parramatta
Shellharbour
Toongabbie
Wallsend
Wollongong
LIBERAL
Albury
Baulkham Hills
Bega
Camden
Castle Hill
Cronulla
Davidson
Drummoyne
Epping
Gosford
Goulburn
Granville
Hawkesbury
Heathcote
Hornsby
Ku-ring-gai
Lane Cove
Londonderry
Macquarie Fields
Manly
Menai
Miranda
Mulgoa
North Shore
Oatley
Penrith
Pittwater
Port Stephens
Riverstone
Rockdale
Ryde
Smithfield
South Coast
Strathfield
Swansea
Terrigal
The Entrance
Vaucluse
Wagga Wagga
Wakehurst
Willoughby
Wollondilly
Wyong
Given that an election will have a number of 3-way contests in traditionally safe Labor seats, will either the Liberals, Labor or Greens be suggesting preferences be given to the other two parties? How could the HTV cards look in these contests?
Eg. for Balmain could it be:
Greens
1. Greens
2. Democrats
Labor
1. Labor
2. Democrats
Liberal
1. Liberal
2. DLP
We won’t know until closer to the date. The Libs threaten to upset some in their base by preferencing either Labor or the Greens (which I understand is why they didn’t last time), but these things change.
There is a disturbingly high number of National seats in that list.
Ben
What do you think about the situation in the LC? Do you agree with my conclusion that I posted on Facebook that a Coalition government will need the support of the Greens to be able to pass any legislation opposed by Labor?
The right-wing minor parties seem to take LC votes off the Coalition, and with the fracturing of the Christian right vote I tend to think 10 seats is the best that the Coalition and Shooters Party could get between them, but possibly an 11th between them and the christian parties. If the combined Coalition + right wing minors get 11 seats, that would give them 21 with the 8 Coalition + 1 Shooters + 1 CDP elected in 2007, but they need 22 votes for a majority.
I’m just comparing the Greens LC vote by division to the LA vote. How popular is James Ryan? Ok, so there were only 3 candidates in Cessnock, but he got 13.81% vs the 8.51% LC vote.
Ah, and here are the Nat-preferencing Green voters of the Hunter Valley again – of the 47% of James’ votes that allocated prefs, 53% went to the Nats.
The LC will be interesting. Please note that the below musings are strictly hypothetical.
LibNats – 40% = 8.8 quotas (9 seats)
Labor – 31% = 6.8 quotas (7 seats)
Greens – 13% = 2.8 quotas (3 seats)
So that’s 19 seats, leaving just two that could be one left, one right, but I’ll be kind to the Libs and give them two right minors, for these purposes we’ll say a Fred Nile and a Shooter, though if it is AAFI or the Fishing Party etc it’s much the same.
So, on these figures, that would make the 2011 LC
Labor – 16
LibNat – 17
Greens – 5
Niles – 2
Shooters – 2
Total – 42
It is my understanding that a party needs 22 votes to pass (correct me if I’m wrong). So basically a Labor/Green coalition could block any legislation in the LC, such as a 10 year moratorium on marine parks, as proposed by the Nats and Shooters (and supported by the Libs). This would also make the Niles and Shooters a pointless proposal for the LibNats because they would need 5 votes to pass legislation. Variations are that the Nile candidate doesn’t get up and a Climate Change Coalition candidate (for example) gets up, which wouldn’t change things that much. Another variation is that the Labor vote does tank and the Libs/minor rights can garner 22 seats in the upper house, in which case we could just say goodbye to any form of environmental protection for the next four years.
Labor has lost Kogarah and Parramatta in the past. Which are the Green/Liberal contests, to what extent can the Greens rely on Labor preferences?
Cold showers anyone?
In Coogee and other seats, if Labor crash so badly they come third, it might be an idea to look at the seats where Labor already come third in a Lib/Grn contest: Vaucluse and North Shore. In Vaucluse, Labor votes went {Grn, Lib, exhausted} = {56.6, 7.6, 35.8}. In North Shore, {Grn, Lib, exh} = {50.9, 8.1, 41.0} (bearing in mind about 6% of those distributed Labor votes came from other candidates). Net effect is that the Greens got 40-50% of Labor votes to make up a gap between them and the Libs, so they’d have to be fairly close behind the Libs to win on ALP preferences.
What did Labor’s HTV cards look like in those Lib/Grn seats, anyway?
@Bird of Paradox,
If the vote in a seat like Coogee or the Blue Mountains is 35-30-25 then the Greens would need 5% off the ALP, which turns out to be about 20% of their vote. Leaves quite a lot of room, unless Labor falls into the teens or lower.
Labor will still have a large minority of votes in Coogee and Blue Mountains, considering they have held them at least for the entirety of the Labor government.
Ah right. Antony Green pulls a few different figures out of the air which he gets from the general swing, with the Libs geting above 40% – he sees Coogee going to the Libs on those figures, very marginally by my calcs.
I asked this on Poll Bludger, but it’s a bit noisy there so I’ll repeat here: who was Robert Stock, the independent who got 11% in Blue Mountains last time? His preferences seem to have favoured the Greens.
Something else, too: what’s Dubbo doing on the Nationals list? Does that calculator of yours show Dawn Fardell losing there?
My recollection is that Stock had a small business background or something like that. I recall reading his campaign website at the time. His HTV exhausted. He subsequently contested Blue Mountains First Ward (which covers Katoomba and the upper mountains) at the 2008 council elections and got 11% of the vote, his preferences mostly exhausting, but the handful that did flow being split between the Greens and the two progressive independents who between them won all three seats in the ward, shutting out both Labor and Libs. From what I recall though my impression was that he was more conservative-leaning. From his council result it does seem he doesn’t have that much of a profile, so his state election vote would seem to suggest clear potential for a higher profile independent to do better.
Fardell only won Dubbo by a slim margin. Since there was a swing towards the Coalition in the poll this was enough to flip the seat into the Nationals column.
I think the difference in approach between Antony’s calculations and mine is that I have assumed that the swing is bigger where the primary vote is bigger. So if the Greens have increased from 9% to 17% then each seat’s Greens vote is increased by 89%. This does mean that totals may not add up to 100% so in some seats I’ve had to proportionally increase or decrease results so that it adds up to 100%.
This means that the Greens swing would be higher in stronger seats than in smaller seats. I think this makes sense. I can’t see us gaining 8% in Campbelltown or Dubbo, which means that 8% swing has to happen somewhere else.
In essence, though, both calculations result in the Liberals coming first, Greens second and Labor third, I just had a higher Greens vote and lower Liberal vote. I tend to think the swing to the Liberals will be quite small in Coogee compared to the average.
Ben,
I think I see where your Coogee analysis is flawed.
The Liberals got 35% primary in Coogee last time. You’re predicting ZERO swing to them when polls suggest their primary vote up by double figures, with all the swing going to the Greens. That’s not going to happen. Surely you’d expect the Lib primary to be above 40%, which would probably get them home quite comfortably.
Assuming the polls continue in the same vein, of course.
But surely Coogee is the sort of area where more Labor votes will go to the Greens than the Libs. Even if the Libs get 40%, there should be enough Labor prefs to get Greens over the line.
After a few interruptions I’ve just finished putting together my own table, and it’s a very interesting exercise since there are so many different assumptions one can make as to how to calculate the potential swings, and then how preferences may flow.
I’ve obviously made some different assumptions to what Ben has, and I’ve got a few different results. I end up with Labor on 27 seats, the Coalition on 52, Greens on 4, and Independents on 10.
I won’t go into the full details, but a couple of things I did were, firstly, decided to leave the ‘Others’ unchanged, even though they are technically down 2 in the poll (did this partly to make the calculations consistent for the seats where there were no Others). Secondly, after calculating the vote changes in the same way Ben described, I used the following methods for adjusting the votes to bring them to a 100% total. If the total was over 100%, I said, well, the Coalition and Green vote can only rise by as much as the Labor vote has dropped, so, I subtracted the projected new Labor vote from the 2007 vote, and apportioned the difference between the Coalition and Greens (where this produced a total for one of those parties that was higher than the first calculation, I over-rode that and transferred the excess to the other party, such as in Blue Mountains where I got 34.67% for the Liberals, which was more than 119% of their 2007 vote (33.87%), so that surplus got transferred to the Greens). Where the total was under 100%, I essentially did the reverse, assuming the Labor vote can only drop by as much as the Coalition and Green vote had increased.
The different individual seat results I got were: Coogee does go to the Libs by less than 0.2% after prefs. Labor holds onto Bathurst, Macquarie Fields, Mulgoa, Oatley, Riverstone, Smithfield and Strathfield, but loses Kiama (by 0.2%). Fardell holds on to Dubbo, and there is an independent win in Swansea (which is a bit of a rogue outcome). Most of those differences though would come from different preference allocations.
I thought it might also be interesting to work out what could happen in Newcastle without that anomalously large vote for the independents, so I looked at the LC votes in Newcastle. Across the state Labor and the Greens got almost the same percentage of the vote in both houses, but the Coalition’s LA vote was 108% of their LC vote, so I plugged the LC results for Newcastle, with the appropriate adjustment of the Liberal vote, into my table, and the result was a comfortable win for the Greens. Cool.
How does Newcastle work out using the LC votes like that on yours Ben?
Nick, the Greens aren’t going to win Newcastle. From what I understand, it’s likely that a Hunter Valley Independents will run in a collection of seats in the Hunter, and they stand a huge chance, but if it’s not them it’s Labor.
Yes, I agree Hamish, the Greens won’t actually be winning Newcastle.
Do you think both the Coalition and Labor not obtaining a majority is a significant possibility?
Comments are closed.