Pauline Hanson Gilds the Lily

Manufactured consent, real anxieties, and the Rinehart ledger behind One Nation’s numbers

A Ledger of Bots

Over the past six months I have blocked more than five hundred accounts on X for aggressively targeting people who criticised Pauline Hanson and One Nation. I want to be precise about what that sentence does and does not claim, because precision is the whole point of this essay. It is not a leaked internal memo from One Nation’s digital operation, and it is not a forensic audit conducted by a cybersecurity firm. It is the record of one moderator – me – looking at the timeline of every account before deciding to block it, across roughly half a year of ordinary use of the platform. On that basis, and on that basis alone, a very large majority of those five hundred-plus accounts showed what looked, to my eye, like the pattern of automation rather than genuine political engagement. I ran no coding frame and no inter-rater check, I have no comparison sample of the pile-ons that greet Labor or Greens figures under the same conditions, and I cannot cleanly separate cheap engagement farms from genuine cranks recycling the same three talking points from actual automation. A precise percentage would claim more rigour than one moderator’s impression can support. What I can defend is the pattern, not a rate.

What did that pattern look like? Overwhelmingly, the same stock phrasing, deployed with minor variations, built around the words “communist” or “commie.” That detail matters more than it might first appear. Communism is not a live political fault line in Australia the way it is in the United States. It carries no real charge in a country where the Australian Labor Party has governed for a century without triggering a Red Scare, where the Communist Party of Australia was never more than a marginal presence, and where “commie” is more often a term of dry mockery than a genuine accusation. When an account replying to an Australian post about an Australian senator reaches reflexively for an American culture-war insult, it is telling you something about its provenance that has nothing to do with the argument it is pretending to make.

None of this proves that One Nation’s head office is running a bot farm, and independent reporting bears out both the caution and the underlying pattern. AAP FactCheck has repeatedly traced fabricated quotes and AI-generated stories about Hanson – including a December 2025 claim, since debunked, that she was suing the prime minister in a US court – to a cluster of Facebook pages run out of Vietnam and built to farm engagement rather than serve any coherent Australian political agenda. Separate reporting by The Klaxon has documented sham “news” pages pushing pro-Hanson propaganda at scale; Hanson herself barred The Age, The Guardian and the ABC from the Melbourne press conference where she launched One Nation’s Victorian campaign on 3 August 2026, having already excluded the ABC from party events in May and June. None of that tells us who, if anyone, directs the specific accounts I blocked, and coordinated inauthentic activity of this kind trails most high-salience populists – Donald Trump and Nigel Farage attract the same weather without it being the cause of their support. What the independent record does establish is that the phenomenon I was seeing was not an isolated impression: engagement-farming built around Hanson’s name is documented, foreign-run, and real, even where its relationship to her actual vote is looser than a block list alone could show.

The Gilded Lily

That last caveat cuts against the cleanest version of the argument, and it is worth stating plainly rather than burying it: an organic movement can still attract synthetic amplification, precisely because it is large enough to be worth farming, not only when it is weak. “If her support were real she wouldn’t need it manufactured” is too neat an inference to rest a case on. What survives that concession is narrower but still real: genuine Australian supporters replying under their own names do not reach for “commie” as a first instinct, and a movement whose loudest online defenders skew synthetic tells you something about the discourse surrounding it, even when it tells you little about the ballot box.

And the facts around the ballot box have become considerably more complicated over the course of 2026 than they were when bot-spotting alone might have settled the question. One Nation’s primary vote in national polling has, at various points this year, sat in the high twenties and low thirties – a Redbridge/Accent Research poll published in early August had the party on 31 per cent nationally to Labor’s 29, a lead the pollster itself noted sat inside the survey’s margin of error; a Roy Morgan survey in June, taken after Hanson’s National Press Club address, put her within two points of Labor on a two-party-preferred basis. In Victoria, ahead of the state election on 28 November, One Nation has been polling in the mid-twenties, an extraordinary result for a party that has never previously threatened to win Legislative Assembly seats in that state. These are not bot numbers. They are drawn from randomised polling samples of a thousand-plus respondents, conducted by reputable firms for mastheads with every commercial incentive to get it right. Whatever inflation the bot networks are contributing to the discourse around Hanson, they are not manufacturing a thirty per cent primary vote out of nothing.

The more defensible version of the argument, then, is not that Hanson’s support is fake, but that it is structurally overstated by the way primary-vote polling gets reported and understood. Thirty per cent in a pre-election survey does not translate into thirty per cent of the seats in a chamber elected under single-member preferential voting. This is not a partisan talking point; it is arithmetic that Australian election-watchers have observed for decades in a different context. The Greens have, election after election, out-polled the National Party on primary votes nationally, yet the Nationals routinely hold more than a dozen House of Representatives seats to the Greens’ one, four, or occasionally zero – because Green votes are efficiently spread across a handful of concentrated inner-city seats plus a diffuse and ultimately wasted vote everywhere else, while National votes are concentrated exactly where they need to be, in regional seats without strong competing incumbents. One Nation’s vote has historically shown the same inefficiency in reverse: broad, shallow, and evenly spread across marginal and safe seats alike, which is precisely the distribution that preferential voting punishes. A party can lead the primary count in published polling and still walk away from an election with fewer seats than that lead implies, because the primary vote share and the seat share answer two different questions. The Demos AU federal poll from May 2026, for what it is worth, illustrates the uncertainty without needing exaggeration: its own seat-projection range for One Nation ran from forty-seven to fifty-eight seats, depending on assumptions about preference flows that nobody can pin down this far from polling day. An eleven-seat band inside a single poll’s own modelling is not a rounding error, and seat estimates for the party have moved by dozens of seats between polling houses and across the year. Whatever the true number turns out to be, nobody serious is claiming to know it yet.

The Bondi Complication

Steelmanning the other side means naming the thing that makes this argument harder to make than it was six months ago: the Bondi Beach massacre of 14 December 2025, in which fifteen people were killed, reshaped Australian public opinion on immigration and multiculturalism in ways that long predate and extend well beyond anything Pauline Hanson’s office could engineer. The surge in One Nation’s polling through the first half of 2026 tracks that event closely, and it would be dishonest to attribute it primarily to bots or to Rinehart’s chequebook. Real people, frightened and angry after a real atrocity, moved toward a party that had spent three decades saying “I told you so” about immigration and social cohesion before it was politically comfortable to say so. Hanson’s National Press Club speech on 17 June, in which she called for Australia to abandon multiculturalism in favour of what she called a “monoculture,” drew a genuine, organic spike in support in the days that followed – the Roy Morgan primary vote of 31.5 per cent recorded immediately afterwards was not manufactured by a bot farm; it reflected an actual, if troubling, shift in sentiment among actual Australian voters responding to an actual political moment.

This is the honest caveat that any evidentiary account of Hanson’s 2026 has to carry: not all of her rise is astroturf, and pretending otherwise would be its own kind of dishonesty, however satisfying the bot-farm story is to tell. The bots are real. The Rinehart money is real, as the next section sets out in detail. But so is a portion of the underlying sentiment, and treating a genuine, if unwelcome, movement in public opinion as though it were entirely synthetic risks the same error Hanson’s own supporters make when they dismiss inconvenient facts as conspiracy: it substitutes a comforting story for an uncomfortable one.

Stupid Is As Stupid Does

Where the record does vindicate the original instinct – that sustained media attention exposes Hanson’s limitations rather than consolidating them – is in what happened after that same Press Club speech. The initial surge did not hold. By the second week of July, a Resolve Political Monitor survey for Nine newspapers had One Nation’s support down three points to 26 per cent, and Hanson’s own standing as preferred prime minister had collapsed by eight points in a single month, from 33 per cent in June to 25 per cent in July. A month-on-month poll swing this size rarely has one clean cause, and this essay won’t pretend to isolate it. But one visible flinch sat right in the middle of the window: her Press Club remarks about paid parental leave, which she suggested amounted to being “paid to take time off work,” landed badly even among voters otherwise sympathetic to her immigration message, and she spent the following fortnight explaining that the comment had been “taken out of context” – the standard argot of a politician who has said, in public, on the record, exactly what she meant to say, and then discovered that meaning it out loud was a mistake.

This is the pattern that gives the phrase in this essay’s working title its bite. Hanson thrives at a comfortable distance – in a thirty-second grab, a tabloid headline, a viral clip stripped of the surrounding argument. She struggles under sustained scrutiny, where a single unscripted answer at a National Press Club lectern can undo weeks of favourable polling in the space of one bad news cycle. The correct forecast is not that attention will destroy her; forecasting is a mug’s game in a year this volatile. It is narrower and more defensible than that: the more oxygen Hanson receives, the more chances she gives herself to say something that costs her the very voters her surge depends on, and 2026 has already supplied one clean case study of exactly that dynamic playing out in real time.

The View From Victoria

The Victorian test arrives first, on 28 November 2026, and it illustrates the seats-versus-votes problem in unusually sharp relief because Victoria elects its two houses by two different methods. The Legislative Assembly, like the House of Representatives, is eighty-eight single-member seats decided by preferential voting – the terrain on which a broad, shallow vote share converts poorly into seats, for exactly the reasons set out above. The Legislative Council, by contrast, is elected by proportional representation across eight multi-member regions, which is far more forgiving of a party whose support is wide but thin. A RedBridge/Accent poll for the Australian Financial Review taken in the first two weeks of September had One Nation on 25 per cent of the state primary vote, essentially level with the governing Labor Party under its new leader Ben Carroll and not far behind the Liberal-National coalition under Jess Wilson. On those numbers, One Nation should expect to translate a meaningful bloc of Legislative Council seats into its first real parliamentary foothold in Victoria – but should expect nothing like a proportionate return of Legislative Assembly seats, because winning twenty-five per cent of the vote spread evenly across eighty-eight electorates typically means running second in dozens of them and first in almost none.

This is not a reason to dismiss the Victorian numbers, and it would be exactly the kind of motivated reasoning this essay has tried to avoid if it were used that way. A quarter of the Victorian electorate telling pollsters they intend to vote One Nation is a serious political fact, whatever chamber it eventually shows up in, and a Council bloc of four, five or six members would hand Hanson’s party genuine leverage over Spring Street for the first time. The point is narrower: watch the Council result and the Assembly result separately when they land, because a headline that reports only the statewide primary vote – the number every wire service will lead with on election night – will systematically overstate what the result means for who actually governs Victoria, and understate what it means for who holds the balance of power in its upper house.

The Rinehart Ledger

It has become fashionable to summarise Hanson’s finances in a single line: a vote for One Nation is a vote for Gina Rinehart. It is a good slogan and a bad argument, and it is worth saying so plainly before setting out what the documented record actually supports. Voters casting a One Nation ballot are not, in any literal sense, purchasing Rinehart’s aircraft; most will be voting on immigration, housing, the aftermath of Bondi, or plain exhaustion with the major parties, exactly as the earlier section of this essay conceded. Collapsing all of that into an equation with a mining billionaire is the same kind of motivated compression this essay has criticised elsewhere. The defensible claim is narrower than the slogan, and it does not need the slogan to matter.

Start with the sequence, because it has been reported piecemeal and is easy to scramble. In December 2025 – the same month Barnaby Joyce defected from the Nationals to sit as a One Nation MP, and days after the Bondi Beach massacre – Hanson and Joyce attended a fundraiser aboard the cruise ship The World, in which Rinehart owns a penthouse apartment, docked in Brisbane. Guests bid $300,000 for a dinner with Donald Trump at Mar-a-Lago, with the proceeds going to One Nation; both Hanson and Joyce separately billed taxpayers more than $3,000 to attend, despite parliamentary guidelines that explicitly exclude fundraising and donor-soliciting events from legitimate travel claims. One of the fund managers who paid for access that night, GCQ’s Doug Tynan, later said he gave because he believed it “should be illegal to burn the Australian flag” – an indication of how directly the Bondi aftermath fed the party’s fundraising, not a separate sum on top of it.

In April 2026, a company owned by Rinehart donated a Cirrus G7 aircraft – Hanson’s office put its value at about $2.1 million – for campaign travel, unveiled days before the Farrer byelection with Hanson thanking her backers on social media for their “faith and investment.” The same announcement disclosed $2 million in cash donations: $1 million from stockbroker Angus Aitken and his partner Sarah, and $500,000 each from former Northern Territory chief minister Adam Giles and geologist Ian Plimer, who hold executive roles inside Rinehart-linked companies – Giles at Hancock Agriculture, Plimer at Hancock Energy. Treasurer Jim Chalmers’s response was blunt: he called Hanson “a wholly-owned subsidiary of Gina Rinehart.”

In July 2026, Rinehart personally funded Hanson’s flights between London, Rome and Sicily, where the two women were photographed together at a Dolce & Gabbana fashion event and a luxury Italian resort; Hanson, pressed on the arrangement, said only that “it didn’t cost the taxpayer one cent” – a defence to a question nobody had asked, since the issue was never the taxpayer but Rinehart.

None of this means a One Nation ballot is, in Chalmers’s or anyone else’s phrase, materially a vote for Rinehart, and this essay does not make that claim. What it means is narrower and, on the documented record, harder to dispute: a movement that markets itself as the unfiltered voice of ordinary Australians against a self-serving elite has, inside eight months, taken a private jet, two million dollars in cash from a mining billionaire’s inner circle, taxpayer-subsidised access to a cruise-ship fundraiser, and a European holiday, all from the single largest personal fortune in the country. That contradiction – not an equation between a ballot paper and an aircraft – is what a voter weighing One Nation’s self-image against its financing is entitled to hold against it.

What the Numbers Actually Mean

None of the three threads here – the bots, the genuine post-Bondi surge, and the Rinehart patronage – cancels the others out, and a responsible account has to hold all three at once rather than collapsing them into whichever story is most convenient. The bot networks are real and their linguistic tells are diagnostic, but they are not the whole explanation for Hanson’s 2026 polling, and treating them as though they were would be exactly the kind of evidentiary sloppiness this masthead exists to push back against, whoever it is aimed at. The underlying sentiment shift after Bondi is real, and pretending it is manufactured insults the people whose fear and anger are genuine, whatever one thinks of where they have directed it. And the Rinehart money is real, documented, and squarely at odds with the movement’s self-image, regardless of how many of its votes turn out to be organic.

What follows from holding all three together is not a confident prediction that Hanson will collapse in Victoria on 28 November or federally whenever the writs are next issued – this essay makes no such prediction, because the polling record from this year alone shows how quickly her numbers move in both directions. What follows is a narrower, more defensible claim: that thirty per cent in a pre-election survey is a measurement of sentiment at a moment in time, not a reservation of seats, and that a movement built partly on synthetic online amplification and substantially on a billionaire’s patronage has less claim to the mantle of authentic populism than its leader asserts. Gilding a lily does not make it a rose. It only tells you the person doing the gilding suspects, on some level, that the lily needed it.

– – –

Bakchos is the founder of Blak and Black, an Australian media and advocacy platform established in 2010. Bakchos writes from the intersecting perspectives of Wiradjuri heritage, Jewish identity, and humanism.

© Bakchos, September 2026

BLAK AND BLACK  |  MEDIA AND ADVOCACY  |  EST. 2010