Category: Federal Politics 2022-2025

The first and best tranche of data from the 2021 census was released yesterday by the Australian Bureau of Statistics, prompting an article by me in Crikey yesterday in which I examined how the demographics of various electorates had changed over the past two five-year census cycles, with an inevitable focus on the teal independent seats (which have actually changed very little in demographic terms, some reporting and conservative rhetoric to the contrary) and other seats that turned against the Coalition. The latter tended to be notable for having few old people and, in many cases, large Chinese populations.

For the purposes of this exercise, I supplemented the new census data for federal electorates with comparable figures from the 2011 and 2016 census, aggregating finely granulated Statistical Area 1 results to produce numbers based on current boundaries. From this I further offer the tables below, the first two of which identify the fastest and slowest growing electorates in population terms over the past ten years. The majority of the former are on the outskirts of Melbourne and, in one case, Geelong — although as will be noted below, the underlying population boom this reflects has hit the skids since the onset of the pandemic. The latter include fully urbanised seats in the big cities and remote electorates that tend to be stagnant at the best of times, but have particularly come off since a resources development boom that peaked over a decade ago.

The tables in the Crikey article show which electorates have changed the most in terms of age, income and multiculturalism. Those below simply list those which rank highest and lowest on these measures, and where they placed on the rankings based on the 2011 data. It is notable that all ten of the bottom ranked seats by household income, as well all being in regional areas, are held by the Coalition – this was not the case ten years ago, when Lyons, Richmond and Gilmore featured. Labor’s near lock on the most multicultural seats, marred only by the loss of Fowler, has been assisted by the gain of Reid, which in turn was symptomatic of the swing against the Coalition among voters of Chinese heritage. Bennelong and Chisholm are placed twelfth and fourteenth on the list, and first and second for Chinese language speakers. The “growth” figures are as compared with the 2016 census.

Talk of federal electorates and population growth naturally leads on to the important question of how House of Representatives seats will be apportioned between the states and territories after the next election, which will be determined on the basis of the latest available population figures in the middle of next year. This is a little hard to call at the moment given growth hit a wall in the year after the onset of the pandemic, the impact of which fell so heavily on Victoria that its population actually fell by 1.5% over 2021, but a compensating recovery is now projected by people whose business it is to project such things. With that in mind, the following table shows how the determination would have looked based on population figures from the end of the past three years.

So steep has been the fall of Victoria’s share of the national population that it would appear to be headed for an unprecedented loss of two seats, although I would presume that with the return to normal conditions it will now bounce back and in fact lose only one. New South Wales is also presently teetering on the cusp of 46.5, below which it will lose a seat. By contrast, population growth in Queensland and Western Australia proceeded apace over the past three years, such that Western Australia looks like it will recover the sixteenth seat it lost last time — a fact that would once have boded ill for Labor, but seemingly no more — while Queensland would gain a thirty-first if its quota growth rate were maintained. It’s also not impossible that South Australia will arrest over three decades of declining representation to gain an eleventh seat. Constitutional limits and vagaries of the calculation formula ensure the status quo will be maintained in Tasmania and the Northern Territory.

American pollster Morning Consult’s current read on various international leaders’ domestic approval credits Anthony Albanese with an approval rating of 57%, up six on his debut showing last month, with disapproval up one to 26% and the balance accounted for by a drop in the uncommitted. It seems this poll is conducted on a daily basis and its published numbers are seven-day rolling averages – I’m not sure how often updates are published, but this one came out a week ago, from polling conducted between June 15 to 21.

In the absence of anything to tell us on voting intention or leadership approval, the most interesting finding of the fortnightly Essential Research survey for mine is that 67% support the Fair Work Commission’s decision to increase the minimum wage by 5.2%, with only 15% opposed. It appears Essential Research now has a regular question on whether Australia is headed in the right or wrong direction, the latest figures of 47% and 29% differing little from the result a fortnight ago, which registered a post-election surge of optimism.

The survey also features questions on the gas crisis and emissions targets, which to my mind are flawed by a lack of response options capturing anti-renewables climate skeptic sentiment. Forty-five per cent blamed the gas crisis on “years of neglect and of successive governments” when given a choice between that and “factors that couldn’t have been predicted, like the war in the Ukraine and the pandemic” and the “fossil fuel lobby and the LNP” having “deliberately fought against the transition to renewables”, which scored 35% and 20% respectively. Forty-nine per cent felt the government should implement the emissions reductions target it took to the election and 30% felt it should go further, with “unsure” the only option for those of neither opinion.

There were two questions on foreign policy, one of which found overwhelming majorities felt it important to have close relationships with the United States, Pacific nations and European Union nations, with a more modest 58% feeling the same way about China and 33% doing so about Russia. Sixty-two per cent believed “Australia should take a more assertive role in protecting our national interest”, compared with 38% who favoured the alternative option of “Australia should look for opportunities to increase global cooperation”. The poll was conducted Thursday to Monday from a sample of 1087.

For a lot more on the foreign policy front, the Lowy Institute has published its annual in-depth poll on the subject, which I haven’t had time to look at properly yet. It would seem declining confidence in Joe Biden is not a purely domestic affair, with 58% having confidence in him to do the right thing in world affairs, down from 69% last year. This places him effectively level with Boris Johnson on 59% and behind Jacinda Ardern on 87%, Emmanuel Macron on 67% and Japan’s Fumio Kishidia (who I’m guessing respondents weren’t required to recognise by name) on 65%. Vlaidimir Putin was down ten points to 6%, placing him on par with Kim Jong-un on 5%. The survey was conducted March 15 to 28 from a sample of 2006.

Still nothing from Newspoll; the fortnightly Essential Research should be along this week, but may not tell us anything too exciting if it’s still holding off on resuming voting intention; and who knows what Roy Morgan might do.

Recent news items relevant to the federal sphere and within the ambit of this site:

• John Ferguson of The Australian reports on Liberal plans to get Josh Frydenberg back into federal parliament, which one party source rates as “only a matter of how and when”. However, finding a vehicle for his return is a problem with no obvious solution. While some are reportedly urging him to win back Kooyong, another Liberal is quoted saying an infestation of sandals and tofu in Hawthorn means the seat is now forever lost. Another idea is for him to win Higgins back from Labor, supposedly an easier task since Labor will receive weaker preference flows than an independent. There is also the difficulty that the local party is dominated by a moderate faction of which Frydenberg does not form part, despite efforts to cultivate an impression to the contrary as he struggled to fight off Monique Ryan. Suggestions he might try his hand on the metropolitan fringes at La Trobe and Monash are running into concerns that he might go the way of Kristina Keneally. Yet another source says he might sit out two terms, the idea being that conditions are likely to remain unfavourable for the party in 2025.

• The Australia Institute has published results from a poll of 1424 respondents conducted by Dynata from the day of the election on May 21 through to 25 which found 86% agreed that truth in political advertising laws should be in place by the time of the next election, with little demographic or partisan variation. Sixty-five per cent said they had been exposed to advertising they knew to be misleading at least once a week during the campaign.

• A further study by the Australia Institute found that Labor led the field on social media advertising with expenditure of more than $5 million, after its 2019 post-election review found its social media strategy had been lacking. The Coalition collectively spent around $3.5 million and the United Australia Party $1.7 million.

Election analysis tools:

• Jim Reed of Resolve Strategic has developed a three-pronged “pendulum” to deal with the limitations of the traditional Mackerras model, which entirely assumes two-party competition. Labor, the Coalition and “others” each get a two-sided prong, with margins against the other two recorded on opposite sides.

• David Barry again provides Senate preference calculators that work off the ballot paper data to allow you to observe how each parties’ preferences divided among the various other parties, which you can narrow down according to taste. The deluxe model involves a downloadable app that you can then populate with data files, but there is now a no-frills online version that is limited to above-the-line votes.

• Andrew Conway has a site that allows you to do all sorts of things with the Senate results once you have climbed its learning curve, such as conduct a double dissolution-style count in which twelve (or any other number you care to nominate) rather than six candidates are elected in each state (on a relevant state page, click the “recount” link, enter 12 in the vacancies box towards the bottom, and click “recount”. Its tools can be used not only on each Senate election going back to 2013, but also on New South Wales local government elections at which councillors were elected under the Senate-style single transferable vote system last December.

• Mitch Gooding offers a tool that allows you to replicate how you filled out your Senate paper and calculates exactly how your vote was chopped up and distributed through various exclusions in the count and which candidates it helped elect, if any.

Roy Morgan has published the first poll of voting intention since the election, though in its typically unpredictable way it makes clear from an accompanying chart that it has continued conducting polling on a weekly basis. The primary votes from the poll are Labor 36%, which compares with 32.6% at the election and 34% in both Morgan’s poll last week and its pre-election poll; Coalition 37%, respectively compared with 35.7%, 37% and 34%; Greens 11%, respectively compared with 12.3%, 12.5% and 13%; One Nation 4%, respectively compared with 5.0%, 3.5% and 4%; and United Australia Party 0.5%, respectively compared with 4.1%, 1% and 1%. The two-party preferred result from the poll is 53-47 in favour of Labor, compared with about 52-48 at the election, 54-46 in last week’s poll and 53-47 in the final pre-election Morgan poll.

The two-party state breakdowns have the Coalition with an unlikely 53.5-46.5 lead in New South Wales, after losing there by 51.4-48.6 at the election; Labor with a scarcely more plausible 60.5-39.5 lead in Victoria, which they won by about 54-46 (here the two-party election count is not quite finalised); 50-50 in Queensland, where the Coalition won 54-46; Labor ahead by 50.5-49.5 in Western Australia, where they won 55-45 at the election; Labor ahead by 60.5-39.5 in South Australia, where they won 54-46; and Labor ahead 63-37 in Tasmania, where they won 54.3-45.7. It should be noted that sample sizes for the small states especially low, and margins of error correspondingly high. The poll was conducted online and by phone last Monday to Sunday from a sample of 1401.

This post is intended as the open thread for general political discussion – if you have something more in-depth to offer on the results of the recent election, you might like to chime in on my new post looking at the Australian National University’s new study of surveys conducted early in the campaign and immediately after the election, or the ongoing discussion of the Senate results.

Despite the polls not having failed as such, in that they uniformly picked the right winner, it seems we’re having another post-election voting intention polling drought just like we did in 2019. This is unfortunate from my perspective, as it would be interesting to compare Labor’s strength during its honeymoon period with that of newly elected governments past. It also means I have to work harder on material for regular open thread posts. Here’s what I’ve got this time:

• The Reuters Institute last week published its international Digital News Report 2022, the Australian segment of which was conducted by the University of Canberra, which asked questions on media consumption and trust. Respondents were asked to rank their trust in various media brands on a scale of one to ten. Typically for such surveys, this found the highest level of trust in public broadcasters, with ABC News ranking first and SBS News ranking second; television networks and broadsheet newspapers in the middle; and tabloid newspapers, specifically the Herald Sun and the Daily Telegraph, ranking last. The survey was conducted online in January and February from a sample of 2038.

• In an address to the National Press Club last week, Labor national secretary Paul Erickson dated a shift in voter sentiment in Labor’s favour from the announcement of the Solomon Islands’ pact with China on April 1. Erickson said voters were struck by the contrast between the Coalition’s “immature” warmongering rhetoric and attempts to associate Labor with the Chinese Communist Party and Labor’s promise to “restore Australia’s place as the partner of choice” for Pacific Islands countries. He further noted that the rot set in for Scott Morrison amid COVID outbreaks in mid-2021, when Labor internal polling showed his net competence score fall by 14 points in two weeks over late June and early July. The Coalition was also damaged by cabinet ministers’ partisan attacks on state governments in Western Australia and Victoria, and it was rated lower by voters on housing and wages.

• Saturday’s Financial Review reported on the Australian Electoral Commission’s efforts to confront online disinformation about the election process head on, through the work of its election integrity assurance taskforce and a media unit that abandoned bureaucratic formality in engaging with social media on social media’s terms. Electoral commissioner Tom Rogers claimed they had a “70 to 80 per cent success rate in changing minds”, and that Twitter had been “a bit self-correcting as a result”: “Someone would say something and you’d see people say, ‘hang on, that doesn’t sound right, I heard the AEC say this or that’”.

• Tom Rogers also foreshadows possible changes to electoral laws to allow for faster counting of postal votes after election day by streamlining the existing process whereby ballots are sorted at a central location and then sent to the voter’s electorate before they are counted.

• Nominations for the South Australian state by-election for Bragg on July 2 closed on Thursday, drawing a field of six candidates who are listed on my by-election guide.

Other recent posts on the site:

• A post on the Queensland Senate result, which was confirmed on Thursday. The buttons will be pressed today on the results for New South Wales at 9:30am and, most interestingly, Victoria at 10am. That will just leave Western Australia – the post just linked to considers at length the remote possibility that Labor might not win a third seat, as is being generally assumed.

• Courtesy of Adrian Beaumont, a preview and live commentary of France’s legislative elections, plus news on British by-elections and American opinion polling.

• A post on Saturday’s Callide state by-election in Queensland, a safe conservative seat which the Liberal National Party has retained with a swing in its favour of 6.5% against Labor.

The Guardian reports on the latest fortnightly Essential Research poll, which it seems won’t be treating us to voting intention for the time being. However, it does provide us with leadership ratings, which record a bounce for Anthony Albanese impressive even by the standards of post-election honeymoon polling: his approval rating is now at 59%, up from 42% in the final pre-election poll, while his disapproval rating has plunged from 41% to 18%. However, nothing is reported on ratings for Peter Dutton or a preferred prime minister result.

The poll also finds an eight point increase since pre-election on the question of whether Australia is headed in the right direction to 48%, with the negative response down from 42% to 27%. It apparently shows “about a third” expect economic conditions will improve over the next year, which is up five points though I’m not sure when the previous question was asked, with 40% expecting things to get worse, with expectations evenly balanced with respect to personal finances. Thirty-five per cent thought the new government would be better for their personal finances compared with 18% for one led by Peter Dutton. Asked whether Australia should become a republic with an Australian head of state, 44% said yes and 34% no, the latter being six points higher than when the question was last asked in March last year.

More detail from the poll will become available when the full report is published later today (UPDATE: Full report here). I am unable to offer any insight as to when Newspoll will be back, when Essential Research will resume voting intention polling, or what the enigmatic Roy Morgan organisation might do. However, I can relate that Ipsos’s and Resolve Strategic’s contracts ended with the election, though that’s not to say they won’t show up again in some form at some point.