Category: Federal Politics 2022-2025

I have a live results feature standing ready for service this evening, which will presume to offer projections according to a somewhat experimental method that will be explained further below, together with individual results pages by electoral division featuring booth results displayed in table and map form (click on “activate” at the bottom of the individual seat pages for the latter). It may be the only place where the latter data will be available online, not counting the Australian Electoral Commission media feed from which it will be extracted.

Some final polls are in, with pollsters collectively offering a fairly wide spread that means the contest for post-result bragging rights is wide open:

• The Australian has a Newspoll result that concurs with others in finding movement back to yes – though by an entirely insufficient three points to 37%, with no down one to 57%. The 1413 sample from this poll has been combined with 1225 sample poll from last week to produce state breakdowns with no leading 54-41 in New South Wales, 51-43 in Victoria, 65-30 in Queensland, 65-28 in Western Australia, 60-33 in South Australia and 55-38 in Tasmania. The new batch of polling means we are also treated to a second set of Newspoll federal voting intention numbers in a week, in this case putting Labor ahead 54-46 (53-47 last week) from primary votes of Labor 36% (up two), Coalition 35% (down one), Greens 12% (steady) and One Nation 6% (up one).

• A JWS Research poll in the Financial Review has no at 52% and yes at 39%, converting to 57-43 after exclusion of the undecided. The poll was conducted Friday to Monday from a sample of 922.

• The Australian related yesterday that a poll conducted from October 1 to 9 by DemosAU, whose director George Hasanakos had a polling analysis website back in the day called Poliquant, had no leading 57% to 30% in both Queensland and Western Australia.

The projection model in my results will use the seat-level estimates from Focaldata’s multi-level regression with post-stratification exercise as a baseline for measuring the results as they are reported. However much the results that are in differ from what Focaldata predicted will then be projected on to Focaldata’s overall results at state level. Doubtless this will be noiser than the booth-matched swings methods that can be applied at elections, but it should at least go some way towards correcting for the peculiarities of the early numbers.

My attention this evening will be focused on the referendum, but I will have some sort of a post up following the progress of counting in New Zealand, a dedicated thread for which is here.

Roy Morgan has an Indigenous Voice poll of 905 respondents conducted last Monday to Sunday showing a relatively modest lead for no of 50-45. The accompanying release relates that we will shortly see “full results of surveying over the last two weeks, including detailed state-by-state, gender, age and party support breakdowns”.

There’s a fair bit of this sort of thing around at the moment, notably a Resolve Strategic document running to 100 pages covering every imaginable detail of this week’s 4728 sample poll, barring a few redacted details that are being held back to provide material for Nine Newspapers reports. The difference between Resolve Strategic’s 56-44 to no and Newspoll’s 58-34 comes largely down to the 18-to-34 cohort, which is difficult to poll and accordingly tends to get upweighted. Resolve’s has it breaking 62-38 in favour of yes, while the Newspoll has it at 49% no, 46% yes and the rest uncommitted.

The main story from the poll is that it’s the first one to provide any indication of Indigenous support since early in the year, after going to particular effort to secure a sub-sample of 420 (the report stresses that this “will not adequately cover remote communities”). The result is a split of 59-41 in favour of yes, a good deal narrower than much-disputed figures of 80-20 from earlier in the year. Kevin Bonham knows something I don’t in relating that “there is another one of these coming from another pollster that is mid-high 60s but smaller sample”.

There is also some further detail in The Economist in the other day’s poll from British outfit Focaldata, emphasising an age gap that looks more like Resolve’s than Newspoll’s. The accompanying multi-level regression with post-stratification exercise that produced demographically related estimates for each House of Representatives seat was discussed the other day on Twitter by pollsters Kos Samaras and Shaun Ratliff, the former saying it “missed demographic nuances, including education levels and ethnic make-up of regional seats” (more recently arrived groups being stronger for yes), the latter saying at least double the 4000 sample would be needed for duly robust results. UPDATE: More from Focaldata here.

UPDATE (13/10): There is now also a result from YouGov that has no up three points from its poll last week to 56%, with yes steady at 38%. The poll was conducted Friday to Tuesday from a sample of 1519. Roy Morgan has also expanded on the poll reported above as being conducted Monday to Sunday from a sample of 905, with a field work period now extending to today (Thursday) and a sample up to 1419. It has no leading 51% to 44%, breaking down to 49% to 46% in New South Wales, 64% to 30% in Queensland, 54% to 44% in Western Australia, 51% to 39% in South Australia and 52% to 47% in Tasmania, with yes leading 54% to 42% in Victoria (with due regard to small sub-samples here, particularly in the smaller states).

Nine Newspapers have published the latest federal voting intention numbers from Resolve Strategic, which offer no indication that declining support for the Indigenous Voice has damaged the Labor government. Labor is credited with 37% of the primary vote, up a point on last month, with the Coalition down three to 31%. The Greens are steady on 12% and One Nation are up two to 7%. The pollster does not provide two-party results, but based on previous election preference flows, this comes out at around 57-43. Anthony Albanese’s combined very good and good rating is up four to 44%, and his combined very poor and poor rating is down four to 43%. Peter Dutton is respectively down five to 30% and up two to 45%. Albanese’s lead as preferred prime minister is 47-25, out from 43-28.

The voting intention numbers are from the same juiced-up sample of 4728 and extended field work period of September 22 to October 4 that produced yesterday’s Indigenous Voice result of 56-44 in favour of no, which reflected the voting intention in being more favourable to the government than the tenor of polling elsewhere. I might have hoped this would have meant more comprehensive state breakdowns than usual, but there is no sign of that to this point, with only the usual results for the three largest states provided on the Resolve Monitor display.

The sample for the leaders’ ratings was only 1604, which presumably relates to the 3116 sample size for separately published follow-up results today on the Indigenous Voice – evidently respondents were asked one set of questions or the other. Among many other things, the Indigenous Voice results offer the finding that 38% of respondents considered that colonisation had had a positive impact on Indigenous people compared with only 23% for negative and 41% for mixed or unsure.

The usual practice for Resolve Strategic is to follow up its national poll later in the week with state results for New South Wales or Victoria, alternating between the two with samples that combine results from two of the monthly polls. This month was due to be the turn of Victoria, but given the extended sample and the complication of the change in Premier from one polling period to the next, I’m not sure where things stand on this particular occasion.

As the campaign enters its final week, three new poll results of note:

• Nine Newspapers have a large sample poll from Resolve Strategic finding 49% for no and 38% for yes, respectively steady and up three since last month. Thirty-seven per cent were definite for no, while 25% were definite for yes. Excluding the undecided, the result was 56-44 for no, in from 57-43 last time. Broken down by state, no leads 52.3-47.7 in New South Wales, 54.2-45.8 in Victoria, 64.2-35.8 in Queensland, 61.2-38.8 in Western Australia and 55.5-44.5 in South Australia, with the small sample from Tasmania producing a 56.1-43.9 in favour of yes. The poll was conducted September 22 to October 4 from a sample of 4728.

• Newspoll in The Australian found support for yes continuing to deteriorate, with no up two since the last poll three weeks ago to 58% and yes down two to 34%.

• British pollster Focaldata has conducted a poll showing no leading 61-39, and while details of how it was conducted are currently scarce, it has yielded results at seat level using multi-level regression with post-stratification, similar to an exercise conducted by YouGov before the last federal election (which came in a bit high for Labor and undersold the teals). It has yes leading in 22 seats out of 151, which are overwhelmingly inner-city seats where the Greens are strong, plus all three seats in the ACT. The only teal seat with yes ahead is North Sydney, although Kooyong and Wentworth are lineball.

As we enter day three of the two-week early voting period for the October 14 Indigenous Voice referendum, the latest poll findings are as follows:

• This fortnight’s Essential Research poll contains an Indigenous Voice referendum result that is unusual in not finding yes in decline — no leads 49-43, which is in from 51-41 a fortnight ago. No includes 42% hard no and 8% soft no, while yes includes 30% hard and 13% soft. A question on whether respondents felt well informed about the referendum found effectively no change over the past month, with yes steady on 49% and no up one to 29%. Forty-nine per cent expected the proposal would fail, compared with 26% who expected it would pass.

• A RedBridge Group poll of 1500 respondents conducted from September 13 to 21 had no leading 62-38. Breakdowns for the three biggest states had no leading 58-42 in New South Wales, 59-41 in Victoria and 68-32 in Queensland.

• A Roy Morgan poll of 1511 respondents conducted from September 18 to 24 had no leading 44-39. Based on small samples, no led 42-40 in New South Wales, 49-31 in Queensland, 46-30 in Western Australia, 48-36 in South Australia, while yes led 46-42 in Victoria. The negligible sample of Tasmanian respondents broke 56-43 to yes.

• With all the latest numbers added, the poll tracker being conducted by Professor Simon Jackman for the ABC currently has no leading 58-42. Jackman’s highly sophisticated methods are explained in detail here.

UPDATE: And now a poll from YouGov, which is no longer involved with Newspoll but from which I am told we can expect a fair bit of independently conducted polling in future, a finding that no leads 53-38. It comes, furthermore, with voting intention results showing Labor leading 53-47 on two-party preferred, from primary votes of Labor 33%, Coalition 35% and Greens 13%. Anthony Albanese recorded a net approval rating of minus 3%, Peter Dutton recorded minus 17%, and Albanese led as preferred prime minister by 50-33. The poll was conducted last Monday to Friday from a sample of 1563.