2PP Aggregate 51.1 To ALP (2022 preferences) (-0.1 since end of last week)
With One Nation adjustment 50.5 to ALP
If polls are accurate, Labor would most likely win election "held now" (probably in minority but with a significant chance of majority)
If polls are accurate, Labor would most likely win election "held now" (probably in minority but with a significant chance of majority)
----------------------------------------------------------------------------------------------------------------------
Donations welcome!If you find my coverage useful please consider donating to support the large amount of time I spend working on this site. The sidebar (scroll down and click on "view web version" if viewing via mobile) has Paypal or PayID instructions or email me via the address in my profile for my account details. Please only donate if you are sure you can afford to do so.
----------------------------------------------------------------------------------------------------------------------
This was intended to be a rolling Budget week poll roundup but over the Budget poll release weekend I was in Melbourne and too busy to write much, so it will have to be a retro one instead. I was in Melbourne for primarily non-election reasons, staying in the electorate of Deakin with trips to several surrounding electorates. I didn't see quite as much electoral activity as I expected at this early stage, though while on a bus through Warrandyte I did happen to look out the window and see the Member for Menzies at a fair.
It's becoming better known in media circles that the once much anticipated "budget bounce" in polling is an elusive beast that is seldom sighted outside Coalition government election years, and there was less speculation about it this year than usual. My own aggregate has found very little movement in the post-Budget polls, and such overall movement as it has found (Labor losing 0.1% since the end of last week) is because of the way it is responding to Morgan's recent strong results for Labor.
Again I don't propose to run through primary votes that are covered elsewhere but just to mention how my aggregate has handled the recent batch of polls for 2PP purposes:
* YouGovMRP (entirely pre-Budget) with a 2PP of 50.2 to Labor by YouGov's modified (respondent and "past elections" model). I got 51.4 by last-election preferences, aggregated as 50.3 because of recent strong numbers for Labor in YouGov. Note that YouGov has now increased its estimate of Greens and One Nation preferences to Labor to 80% and 33% respectively.
* Redbridge (mostly pre-Budget) 51-49 to Labor by last-election preferences. I got 51.3 by last-election preferencess, aggregated as such.
* Morgan (partly pre-Budget) 53-47 to Labor by respondent preferences, 53.5 by last-election preferences. I got 53.5 but my aggregate is increasingly adjusting Morgan for its recent divergence from the other polls; aggregated at 51.7
* Newspoll 51-49 to Labor by previous election preferences with a One Nation modification. I got 51.5 by last-election preferences, aggregated as such.
* Freshwater 49-51 (Coalition ahead) by unknown preference method. I got 50.2 by last-election preferences, aggregated as 50.6 as Freshwater has been slightly stronger for the Coalition off the primary votes than other pollsters over time but the difference seems to be reducing. Freshwater has been in hot water over news it is working with an astroturf group on top of being the Coalition's internal pollster, but says it has adequate measures in place to manage potential conflicts of interest. While that may very well be the case, this highlights the need for Freshwater to be far more transparent about its polling methods and for the AFR and other media publishers to insist on this.
* Resolve 50-50 by respondent preferences, 51-49 to Labor by last-election preferences. I got 50.5 by last-election preferences, aggregated as 51.2 as recent Resolves have been good for Coalition (especially their previous poll which was an outlier)
* Essential 48-47 (equivalent to 50.5-49.5) to Labor by primarily respondent preferences. I got 51.2 by last-election preferences, aggregated as 50.8 as recent Essentials have been slightly good for Labor.
* YouGov public data poll 51-49 to Labor by YouGov modified preferences. I got 52.1 by last-election preferences, aggregated as 51.0 as recent YouGovs have been strong for Labor.
While there was a 4% spread in the headline 2PPs my aggregate therefore thought it could be mostly explained by temporary house effects and preference calculation method differences, and converted every poll to a slim Labor lead on 2022 preferences with a 1.4% range from weakest to strongest. The lead becomes slimmer if one assumes a weakening in Labor's preference flow but it's notable that the gap between respondent and last-election preferences doesn't seem to be as large lately as early in the term, usually around half a point to a point lately.
The Budget
The Budget didn't poll very well in Newspoll, with 22% thinking it would be good for the economy vs 32% thinking it would be bad. The net -10 result is the fourth worst in budget Newspoll history, trailing only the three Budgets between 1991 and 1993. On personal impact 16% said they would be better off and 35% worse off (net -19 which is actually better than most ALP Budgets). Here it is on my increasingly over-scribbled graph showing all Budgets since 1986 (yes 1986 Simon Benson, you keep wrongly saying Newspoll Budget polling only goes back to 1999):
Overall in polling terms the Budget appears to be a nothingburger and its presence in the news cycle is already being swamped by Donald Trump's latest tarriff warring. As the government's 2PP is identical to the previous edition, there's not much to add to the assessment there of what it all means, except to note the latest YouGov MRP as a voice against the idea that the Coalition will overperform for a given 2PP.
Leaderships
The most popular (or least unpopular) leader wins federal elections about three quarters of the time so it could be bothering the Coalition that their man's ratings are going south in some of the polls. In the latest batch we had:
* Newspoll Albanese net -9 (43-52) Dutton net -18 (37-55, his worst since before the Voice referendum), Better PM (favours incumbents) Albanese leads 49-38 (first double-digit lead since last May)
* YouGov Albanese net -6 (44-50) Dutton net -15 (38-53, term worst from this pollster and down ten points in a fortnight), Better PM Albanese leads 45-38
* Resolve Albanese net -10 (39-49) Dutton net -10 (37-47) with the two leaders having the same netsat after Dutton's netsat was 27 points better than Albanese's in their previous (outlier) poll, preferred PM Albanese leads 42-33 after being four behind last time
* Freshwater Albanese net -12 (37-49) Dutton net -10 (37-47) preferred PM Albanese leads 46-45
* Essential Albanese net -6 (41-47) Dutton net -2 (44-46)
So three of the five polls in this cycle show movements that are significantly not in the Coalition's favour.
YouGov MRP
Said YouGov MRP was an eye-opener in that it had Labor knocking on the door of a majority with a median 75 seats off a 2PP of a mere 50.2%. That's two more than my model gives it, whereas previous MRPs have often had Labor underperforming by at least three. However, the two more are because Labor wins seats from the Greens, not because Labor does especially well in the classic seats for the given swings.
YouGov even had Labor holding steady in Victoria (swapping Deakin for Aston), which has not been a position with many takers. This is interesting mainly as an "ain't necesssarily so" following previous MRPs that generally had the Coalition overperforming its 2PP, but would the Coalition's underperformance vs overperformance dynamic flip that strongly and that quickly? In Victoria this YouGov has got something like the 2022 Victorian election, where the government made a seat gain despite a minor 2PP swing because the swing against it was wasted in NW Melbourne area seats on safe margins. In particular the latest YouGov seems to have a more acute rich seat/poor seat dynamic than previous models that tended to find that Labor was just doing badly in Victoria generally. Anyway the PM has already shown up in Deakin in a sign that Labor either thinks they really are competitive there or at least wants to provoke the Coalition into defending it.
The model has Labor losing Aston, Bennelong, Gilmore, Robertson (bellwether status RIP), Werriwa and Lyons and gaining Deakin as well as Brisbane and Griffith from the Greens. It also has the Coalition dropping Cowper and Wannon to the independents who ran there last time. I have not yet had time to unpick this MRP to see which seats (if any) would change with a less generalised preference flow than YouGov is using. Similar issues occur as with the previous model in terms of some apparently excessive One Nation support levels in particular seats, though that shouldn't amount to much on a 2PP basis. I will be much more interested to see what we get once individual seat candidates are known (if the MRP will still function on that basis).
Teal Seat Poll Detritus
We continue to be treated to some unconvincing and/or unsatisfactory seat poll reports at this election. A JWS poll of Goldstein and Kooyong commissioned by an energy group was published by the Herald-Sun with its usual breathless gloating about numbers that had the resident teals behind 46-54 in the former and, um, actually ahead 51-49 in the latter. The Goldstein numbers are obviously not right; the primaries are Liberal 44 Teal 24 Labor 21 Greens 5 Other 6; we are supposed to believe that there will be a ten point swing against Zoe Daniel on primaries but that there will be a ten-point swing to Labor in Victoria?
Even if these numbers were plausible the respondent preferences used to get to 46-54 would be highly doubtful anyway; hardly likely Daniel would get only 69% of preferences that were over 80% Labor or Green in origin when she got 73% of preferences in 2022 when only 74% were Labor or Green then. It would be useful to know more about what this poll did so I could comment further on what's gone wrong with this sample but as usual when you get a non-Polling Council member reported by one of the nation's most sloppy outlets good luck with that. The Herald Sun declares JWS to be a leading pollster when JWS has hardly been seen in voting intention polling in the last decade (being best known for an issue-poll series called True Issues that only I seem to pay much attention to). How on earth can JWS say that the state Labor government being on the nose in Victoria is driving federal voting when one of the two seats surveyed has the government doubling its vote in a seat it doesn't even have a candidate running dead in yet?
Climate 200 promptly said "hold my beer" by enabling The Australian to publish a supposed finding that they were actually ahead 54-46 in Goldstein with no further details other than that it was "uComms polling of 1225 voters in Goldstein between March 18 and 25", that Wilson was ahead on primaries and that the primary margin narrowed when initially undecided voters were prodded.
Betting
Betting is not reliably predictive but is something I like to keep an eye on, mainly because it is always interesting to watch how many times the view that the money never lies can be wrong without its adherents learning. Labor finally flipped into a narrow favouritism the day after my last article which has mostly lasted since with a brief flip back; currently the headline markets give Labor something in the range of a 58% to 61% chance of remaining in office.
This has been very slow to trickle through to the seat markets, which a few days ago had changed a little in months. But looking at them just now the first flip back to Labor has occurred (Tangney) and all the Labor seats where Labor was at longer than $3 to hold now have Labor at $3 or shorter at least somewhere, except for Aston. If Labor continues polling reasonably I expect there will be more flips. Also a number of expected Labor holds are firming with the Coalition pushing outside the $3 range: Eden-Monaro, Parramatta, Bruce and Shortland. In tealsville the Coalition is now favourite to take Kooyong from Monique Ryan, which I can only think is a reaction to the sign-stealing controversy involving Ryan's husband. However there is no movement so far in response to the latest hint that Bradfield might have been the teal seat that missed in 2022 for a reason. (Here I reference both Women's Agenda highlighting that while Boele is guilty of a double standard she's far from the only one, and also The (teals-obsessed) Australian pointing out that this is not an isolated case of weirdness of some sort or other from the electorate's "shadow MP").
Overall markets now have Labor as favourite in 68.5 seats, the Coalition 69.5, the Greens 2 and others 10 (there's a split market in Brisbane). After adjustment for close seats my read of what the seat betting markets "think" is ALP 68.7 L-NP 68.6 Green 2.5 Others 10.2. This is consistent with the headline numbers because in a parliament with both majors roughly equal and nowhere near a majority, the most likely result would be Labor remaining in office.
It is hard to believe the announcement of nominations is only a week away! When that happens I will be updating my Tasmanian Reps guide as well as rolling out a Tasmanian Senate guide, and other usual pre-election features will follow.
Update Saturday: The new Redbridge came in at 52-48 to Labor off very similar primaries to the last election. I got it as 52.1 by last election preferences, aggregated as such, which saw a new reading of 51.3.
