The High Court Is Deciding Whether Australia Needs a Power It Already Rejected at a Referendum
I want to be clear about where I'm standing before I make any argument, because it would be easy to read what follows as though I'm defending what this organisation says.
I'm not. I don't like what they're saying. I'm not going to pretend my disagreement with it is a technical one that's rather the point of the case.
But I believe they should be able to say it. Not because their views deserve protection, but because the moment the state decides which beliefs are speakable, the thing I've been relying on disappears, and the person doing the deciding changes with the government.
And to be concrete about the line: incitement to violence, threats, and damage to property are already criminal under the existing Criminal Code. They don't need a new power. The case before the Court isn't about whether that conduct is punishable. It's about whether a separate mechanism should exist that makes belonging to an organisation a criminal offence in its own right. That's a different question, and it's the one I care about.
I've been through all seven of the parties' documents on the case page. This is what they say.
The question nobody noticed
White Australia Party Inc. & Anor v The Commonwealth of Australia & Anor (M50/2026) was referred to the Full Court as a Special Case by consent, heard on 8 and 9 September 2026, and judgment is reserved.
The plaintiff is a Victorian incorporated association registered on 20 December 2025 with 1,778 members. It applied to register as a political party with the AEC on 25 April 2026. On 15 May 2026 the Governor-General in Council made the Criminal Code (Prohibited Hate Group White Australia) Regulations 2026, and the following day it became an offence to be a member of it.
Most of the coverage has treated this as a free speech case. Read the special case questions Gageler CJ settled on, and the first one isn't about speech at all:
- Is Pt 5.3B of the Criminal Code invalid on the basis that it is not a law with respect to any head of Commonwealth legislative power?
- Is Pt 5.3B invalid on the basis that it impermissibly burdens the implied freedom of political communication?
- Is Pt 5.3B invalid on the basis that it infringes Ch III of the Constitution?
Question 1 is the load-bearing one, and it is a much more fundamental challenge than "this restricts too much speech." It asks whether the Commonwealth Parliament has the power to do this at all.
The Commonwealth pleads exactly two sources. Section 51(xxix), the external affairs power, read with the ICCPR and the Convention on the Elimination of All Forms of Racial Discrimination. And an implied power of nationhood, said to arise from s 61 read with s 51(xxix).
Two. That's the entire defence to question 1.
And the plaintiffs put the constitutional position in terms that are hard to argue with:
"There is no affirmative power under the Constitution s 51 for the Parliament of the Commonwealth to make laws for the formation, banishment or dissolution of a political party which is not a corporation, including the Australian Communist Party and White Australia Party Inc. That was the decision of this Honourable Court in Australian Communist Party v The Commonwealth, and remains the position."
They also pre-empt the obvious rejoinder, which is worth quoting because so much of the coverage has reached for it:
"The point is not answered by reference to the opinions, objects or alleged associations of the First Plaintiff. Constitutional power is not enlarged by the unpopularity of the political party to which the law is applied."
That's a better version of my own argument than I managed, and I want to be honest that the organisation's lawyers got there first.
The stream cannot rise above its source
The most interesting thing in the plaintiffs' case isn't about hate groups at all. It's about separation of powers, and it works structurally.
Part 5.3B doesn't define a offence that a court then applies. It lets a Minister decide, by regulation, whether an organisation has crossed a line. So the facts that bring the law within constitutional power are the very facts the Executive is being asked to find.
The plaintiffs' formulation:
"Parliament cannot confer upon the Executive a power to determine the facts which bring the law within constitutional power. The stream cannot rise above its source. If the constitutional connection depends upon whether an organisation has engaged in conduct of a kind sufficient to enliven power, that connection must be found in the law itself and, where necessary, adjudicated by a court. It cannot be supplied by ministerial satisfaction expressed in a regulation."
And in their reply, a concession recorded from the Commonwealth itself:
"neither Defendants nor Intervenors take issue with the characterisation that Part 5.3B does not provide for any objective standard of liability relevant to the subject matter of the power, nor rule of conduct or corelative duty, nor for any legal or political rights or duties in anyone"
Read those two together and you have the case in a sentence. There is a power that says the Commonwealth may legislate against certain conduct in external affairs, and a power that says the Commonwealth may legislate to maintain itself as a nation. Part 5.3B asserts that the connection to both of those powers exists in every case where a Minister is satisfied. And the Commonwealth accepts there is no objective standard against which that satisfaction can be tested.
What the listing actually does
Gageler CJ set out the scheme in June. Once the Association is specified, a person commits an offence if, knowing the Party to be a prohibited hate group, they:
- "intentionally direct the activities of" the Party;
- "intentionally is a member of" the Party unless they prove they took all reasonable steps to stop being a member once they knew;
- "intentionally recruit" someone to join or participate; or
- receive funds from, or make funds available to, the Party intending them to help it "expand or to continue to exist" unless received solely for legal representation in Pt 5.3B proceedings.
The Free Speech Union's submissions add the penalties: ten to fifteen years for directing, recruiting, training or funding, and seven years for membership. They then draw the inference that matters:
"The listing mechanism therefore transforms conduct that would otherwise constitute ordinary political association, advocacy or even nominal membership, into a category of serious criminal offending punishable by severe penalties."
Three features of the trigger deserve attention.
The threshold is remarkably wide. Section 114A.4(1)(a) is satisfied if the organisation engaged in, prepared or planned to engage in, or assisted the engagement in conduct constituting a hate crime or advocated engaging in one. Advocacy alone. Not commission. s 114A.4(4) says in terms that a person "does not need to have been convicted of a hate crime" for the Minister to be satisfied.
It reaches backwards and reaches abroad. Section 114A.4(2) allows retrospective conduct to be considered, going back, in the plaintiffs' submission, to the Federation. And s 114A.4(1)(a)(ii) counts advocacy for engaging in a hate crime "whether or not in Australia." The definition of "organisation" captures bodies based outside Australia, comprised of non-citizens, or forming part of a larger organisation.
There are no procedural safeguards before the listing. Section 114A.4(5) states that the Minister "does not have any obligation of procedural fairness in deciding whether or not such a state of satisfaction is reached." Judicial review comes afterwards. The Parliamentary Joint Committee on Intelligence and Security did review both listings White Australia and Hizb ut-Tahrir, the only two and found no reason to disallow either.
The plaintiffs put the consequence bluntly:
"Upon specification an organisation does not merely lose particular privileges or face additional regulatory controls. It ceases to exist for practical purposes. Its meetings cease. Its campaigns cease. Its publications cease. Its members must withdraw from participation on pain of criminal liability. Communication protected by the Constitution is thereby silenced at its source."
The Commonwealth is right that this is distinguishable
Now the part that cuts against the case, because the Commonwealth's distinction is good and I don't want to pretend otherwise.
Dixon J's judgment in the Communist Party Case set out a test at 83 CLR 1 at 192, and the Commonwealth simply recites it and says Part 5.3B satisfies every limb. The law is "addressed to suppressing violence or disorder or to some ascertained and existing condition of disturbance", it does "take the course of forbidding descriptions of conduct or of establishing objective standards or tests of liability upon the subject", and it is preventive, operating "on the basis of their activities". The plaintiffs say the distinction is that the 1950 Act named a single organisation and this one doesn't.
That's a real argument and it may well succeed. Dixon J's judgment is genuinely difficult to read against the Commonwealth here.
What's also true is that the Commonwealth answered question 1 with two heads, one of which the external affairs power it supports with a substantial body of argument. It runs the Victoria v Commonwealth "reasonably capable of being considered appropriate and adapted" test; it leans on CERD article 4(b), which genuinely does obligate states to "declare illegal and prohibit organizations… which promote and incite racial discrimination" and to recognise "participation in such organizations… as an offence punishable by law"; it cites the Racial Discrimination Committee's General Recommendation 34, the travaux, and Toben v Jones. The agreement of South Australia is worth noting, which embraces the Commonwealth's own submission that "meeting that obligation is not unconstitutional."
The best structural answer from the Commonwealth is at written submissions [32], and it doesn't engage the merits at all:
"The Plaintiffs do not challenge the Regulations. The Plaintiffs therefore assume the burden of showing that the empowering provisions have no valid operation. If the empowering provisions could be validly exercised, then the Court should not strike them down. At least because the empowering provisions could be exercised to specify groups with no discernible interest in communicating on government or political matters, they are capable of being justified across a range of potential outcomes of their exercise."
If that's right, the case ends at the threshold and the membership offences never get considered. It also draws a jab the plaintiffs won't enjoy, about the judicial preference for invalidating things that would let people escape consequences they've chosen to incur.
The precedent that stopped short of this
There's something almost perfect about the 1950 interlocutory history, which Gageler CJ set out in June.
The plaintiffs rely heavily on Dixon J's unreported interlocutory orders of 1950. Gageler CJ held the authority distinguishable in three ways, and the third is the one that matters. Dixon J granted injunctions against the appointment of a receiver, against declarations that affiliated bodies were unlawful associations, and against declarations that members were prejudicial to national security. He declined to widen those injunctions to restrain prosecutions including for offences constituted by "becoming, continuing to be, or performing any act as an officer or member" of the Party.
Structurally identical to Division 114B. The precedent the plaintiffs invoke stopped at exactly the point they're now asking the Full Court to go past.
Two other things from that judgment. Gageler CJ recorded that the Commonwealth "do not dispute the existence of a serious question to be determined" on the implied freedom a significant concession at the interlocutory stage, limited to question 2 and not extending to head of power. And he refused relief on the ordinary principle from Castlemaine Tooheys that, "in the absence of compelling grounds, it is the duty of the Court to respect, indeed, to defer to, the enactment of the legislature." The only compelling grounds he contemplated were Dixon J's kind "irrevocable and irreparable" physical dispossession of property and loss of livelihood. Not a criminal record.
His reason for refusing was also the most quotable line in the judgment, from Mr Kirk SC, accepted by the Court:
"you can injunct people; you can't injunct an Act"
Vona is not what you were told
I want to correct something, partly because I got it wrong the first time I looked at this.
The Commonwealth's single strongest international authority is Vona v Hungary (ECtHR, 2013), which upheld Hungary's proscription of the far-right party Jobbik. I had been treating it as the anchor of the Commonwealth's case.
It isn't. Vona appears once, in a single sentence of the Commonwealth's oral outline, and not at all in the twenty-two pages of written submissions. The submission is that "an equivalent justification has been upheld by the European Court of Human Rights" an appeal to an outcome, not an adoption of the reasoning.
The "distinctive threat" idea that does the real work is the Commonwealth's own, drawn from its own evidence about group dynamics. And that evidence is the most serious part of their case: hate crimes in the United States cluster in counties with active white supremacist chapters, associated with 285 more hate crimes per 10,000 residents; 62% of lone-actor terrorists had contact with clearly radical individuals; and radicalisation is "almost universally acknowledged as a group phenomenon." Whether that justifies acting on the organisation rather than its members is exactly the question.
Three states are fighting the Commonwealth
Nobody is covering this, and it's arguably the most interesting thing in the file after the head of power question.
Queensland intervenes under s 78A of the Judiciary Act "in support of neither party." It supports the Commonwealth on the outcome but is against it on the nationhood power, and its reason is the federation argument: "That the States can and do enact laws to address hate crimes dispels any notion that laws like Part 5.3B are 'peculiarly adapted' to enforcement by the Commonwealth." If every state already legislates hate crimes, the Commonwealth hasn't shown this is uniquely a national function. Accepting a broad nationhood power would be "wholly subversive of the Constitution", in Queensland's words.
South Australia goes further and calls Davis v Commonwealth (1988) 166 CLR 79, where a majority rejected any such implication. Its significant move, though, is doctrinal: SA is the only party asking the Court to overrule McCloy and abandon structured proportionality, calling the suitability limb "perfunctory and superfluous" and replacing it with a reasonable necessity test drawn from Gageler J's own reasoning in Tajjour. South Australia quotes Gageler CJ: "Constitutional analysis … cannot be reduced to the application of some pre-determined all-encompassing algorithm." The Commonwealth says the question doesn't arise and McCloy needn't be reopened.
New South Wales takes the opposite position to Queensland on range-of-exercise, citing Palmer Gageler J's own judgment to argue that a discretion isn't invalid just because it might be unlawfully exercised.
The Free Speech Union, appearing only on question 2, goes the other way from South Australia. It wants McCloy developed, not abandoned, into what it calls calibrated scrutiny: a burden sustainable only where Parliament shows the law is "strictly necessary to protect a constitutionally indispensable interest", meaning one essential to the fundamental preservation of the constitutional system. Asked whether this law qualifies, the FSU's answer is no. Taken at its highest, it says, the law is for "the promotion of the general law and order."
Five parties, four positions, and the question of what proportionality is is being relitigated in the same proceeding.
The referendum
Everything above is in the seven court documents. What follows is not no party has mentioned it, and the only constitutional referendum in any of the documents is the 1977 Senate Casual Vacancies amendment.
The Communist Party Dissolution Act 1950 declared the Australian Communist Party an unlawful association, dissolved it and forfeited its property. The Party and ten trade unions challenged it and the High Court struck it down in March 1951 Dixon, McTiernan, Williams, Fullagar and Kitto JJ, Latham CJ dissenting.
The government then put it to the people.
On 22 September 1951 Australians were asked whether Parliament should be given the power to make laws about communists and communism:
| Votes | Share | |
|---|---|---|
| Yes | 2,317,927 | 49.44% |
| No | 2,370,009 | 50.56% |
Turnout was 95.58%. It failed on both limbs of s 128 the national majority and a majority of states carrying in Queensland, Western Australia and Tasmania, and failing in New South Wales, Victoria and South Australia.
The states that carried it were the three where the Communist Party had the least industrial presence. The states that rejected it were the union states, where the Party and the unions that grew out of it had members, meetings, factories and weight. The better organised a state was, the more it voted to protect the space in which to organise.
And the public mood was nowhere near what the arithmetic suggests. Twelve of fifteen newspaper editorials backed the ban; only The Argus opposed. The official campaign leaned on fear, in the aftermath of the establishment of the People's Republic of China. The margin was 52,082 votes, and the ANU's archives record that 30,000 voters in South Australia or Victoria would have flipped it.
So the only time Australians have been asked this question directly, they answered no. Not narrowly, and not in ignorance.
The plaintiffs make the a fortiori themselves:
"If a party alleged to be revolutionary and subversive could not be dissolved by reference to protection of the Commonwealth in peacetime, a fortiori the First Plaintiff cannot be proscribed under an implied nationhood power triggered by executive opinion."
Why I'd rather win the argument than win the point
Karl Popper spent the war years in exile writing about this, and he was not arguing that tolerance is unlimited. His paradox of tolerance goes the other way from how it's usually remembered, and the qualifier is the whole point:
"In this formulation, I do not imply, for instance, that we should always suppress the utterance of intolerant philosophies; as long as we can counter them by rational argument and keep them in check by public opinion, suppression would certainly be unwise."
Suppression is unwise because you can argue with them. The line is crossed when they stop engaging in argument and start denouncing it and that is incitement, which is already against the law. You don't need a listing power to reach it.
Hayne J made the same point in Manis v The Queen:
"History, not only recent history, teaches that abuse and invective are an inevitable part of political discourse."
There's research on the mechanism, and I want to be careful about which part I lean on.
The part I'll lean on: exposure to counter-arguments genuinely changes minds. van der Linden and colleagues found misinformation degraded judgments of the scientific consensus (d = 0.48) and that brief and detailed inoculation protected them (d = 0.33 and d = 0.75), working across the political spectrum. The Bad News game and the review by Traberg, Roozenbeek and van der Linden have replicated this widely.
The part I won't lean on: the backfire effect, where correction hardens belief. It was reported in 2010 and has been revised substantially downward Wood & Porter, and Guess & Coppock, both failed to reproduce it, and Lewandowsky and van der Linden are "reluctant to expect backfire effects generally." A post that quietly ignored that would be committing the error I'm objecting to.
The mechanism is simpler and better supported. Traberg and colleagues cite Van Bavel and colleagues on how people become "directionally motivated to seek out further information in ways that confirm their social identity." Argument is what moves people. And you cannot be inoculated against a claim you've never heard.
A power that criminalises membership doesn't just make an organisation illegal. It removes the counter-argument from the room. That's the confirmation-bias engine, and it runs on the exact machinery the research says does the work.
The parts that are weak
I'm not a lawyer, and this argues about a power rather than about doctrine. But three things don't hold up as well as the rest.
The Association isn't currently fielding candidates. It applied to the AEC on 25 April 2026 and its Victorian registration is under challenge. Its written submissions [12] make electoral intent conditional on invalidity: "If Pt 5.3B is invalid, the First Plaintiff intends to endorse and field candidates at the next federal election." The Commonwealth will argue the implied freedom burden isn't engaged by an intention that hasn't crystallised. The plaintiffs answer on the capacity to aggregate, recruit and communicate, which is a real answer but not the same as a political party actually on a ballot paper.
The plaintiffs concede the easy part. They don't dispute that Parliament may enact valid conduct-based offences: "The submission is not that the Commonwealth lacks power to protect the community from violence, threats, incitement or crime." Their case is entirely about the organisation-level power. That's a well-drawn line, but it means the decision may be narrower than either side wants.
ASIO's own view was that the group doesn't cross the line. The plaintiffs put it in writing: the ASIO report summary "acknowledged that the First Plaintiff is aware of 'legal thresholds for criminality'; it 'deliberately' avoids or 'stops short' of them." Which is the plaintiffs' problem rather than the Commonwealth's, and it narrows the immediate injustice considerably.
The question
None of this is a defence of the White Australia Party, or of the Bondi attack, which killed fifteen people, or of the Christchurch attacker who killed fifty-one. Serious organised violence deserves serious responses and I'd support strong ones.
The question the Full Court has to answer is narrower and more important than the group's case: does the Commonwealth have any power to define an organisation as criminal, at the executive's discretion, on the advice of the security agencies?
If the answer is that external affairs can carry it, the CERD argument is a serious one and Part 5.3B may well survive. If the answer is that the Commonwealth has no enumerated power for this, then Dixon J was right seventy-five years ago and the statute is doing what he said it could never do.
In 1951 we decided that proscribing an organisation was a constitutional change, not an ordinary piece of legislation. We were asked directly, at 95.58% turnout, and we said no.
Part 5.3B arrived by a route that would not survive a referendum. Whether it is constitutional is for the Court. Whether the Court should be the only institution defending the answer we gave in 1951 is a separate question, and I don't think it should be.