Removing the King from the Australian Constitution is, technically, easy. You delete the references and insert a President. The hard part — the part that constitutional lawyers spend their careers on and that has quietly wrecked every attempt at a clean republican model — is what to do about the reserve powers. These are the powers a head of state can exercise on their own initiative, against or without ministerial advice. They are largely unwritten. They have been used to remove an Australian government exactly once. And the fiftieth anniversary of that event passed in November 2025, which is part of why this question is live again.
What the reserve powers are
Most of what the Governor-General does is formality: signing legislation, commissioning ministers, acting on the Prime Minister's advice. The reserve powers are the exceptions — the situations where the head of state may act independently. They conventionally include appointing a Prime Minister when no one commands a clear majority, dissolving Parliament, refusing a dissolution, and dismissing a Prime Minister.
Their defining feature is that they are not written down anywhere authoritative. They exist in convention, precedent and scholarly argument, which means their exact scope is genuinely uncertain. Anne Twomey's 2018 study The Veiled Sceptre is the standard reference on how these powers operate across Westminster systems, and one of its findings is instructive: reviewing the historical record, she identified essentially one case of a Westminster government operating for any significant period without supply — the McCulloch government in Victoria in 1865, which kept going via a loan arrangement with a bank.
That matters because supply is the trigger that produced Australia's constitutional crisis.
1975, briefly
In late 1975, the Coalition-controlled Senate deferred the Whitlam government's appropriation bills, following revelations in the Loans Affair. The government had a majority in the House of Representatives but could not get its money bills through the Senate. On 11 November, Governor-General Sir John Kerr dismissed Whitlam as Prime Minister and installed Malcolm Fraser as caretaker, on condition he call an election.
The conventional assumption underpinning Kerr's action was that a government cannot function without supply — that it is both legally impermissible and practically impossible. Twomey's research suggests the historical precedent for that assumption is thinner than it was treated as being at the time.
There's also a more recent and rather awkward comparison available. The United States is a presidential system, not a Westminster one, but it shares the feature that the legislature must appropriate funds for the executive. Since a 1980 opinion by Carter administration attorney-general Benjamin Civiletti, the American answer to a failure of appropriation has been the government shutdown — and shutdowns have become routine, with the 2025 shutdown among the longest on record. Whatever else that demonstrates, it suggests a government deprived of supply can limp along for a while. Not comfortably, not indefinitely, and not without real cost. But the 1975 premise that supply failure requires immediate resolution by the head of state looks less self-evident in 2026 than it did in 1975.
Why 1975 is a republican problem, not just a historical one
Here's the bind. Any republican model has to decide what to do with the reserve powers, and there are only three options, each with a serious cost.
Leave them uncodified. Transfer them to the President as-is, unwritten. This is the most minimal change, but it means a new office — potentially one with a popular mandate — inherits powers whose limits nobody can state precisely. The 1999 model was criticised on a related point: Kirby listed concerns about the ease with which a Prime Minister could dismiss a President as one of the reasons the model lost support. A President who can be sacked by the Prime Minister they might have to sack is not obviously an improvement on a Governor-General.
Codify them fully. Write them down. This has been attempted repeatedly and has proved elusive, because the powers exist precisely for circumstances nobody anticipated. Herbert Evatt argued for codification on two grounds that remain the best case for it: it prevents an over-cautious head of state failing to act because the scope of their power is unclear, and it prevents an over-zealous one acting where they shouldn't. The counter-argument is that any code will be incomplete, and the gaps will be exactly where the next crisis happens.
Cut them back. Remove or narrow the most dangerous ones. This is what the ARM chose.
What the ARM's 2022 model does
The Australian Choice Model, released in January 2022, explicitly circumscribes the head of state's powers. Proposed section 70A(v) removes the power to terminate a Prime Minister who retains the confidence of the House of Representatives. Proposed section 5 appears to restrict the power to dissolve Parliament. The ARM's stated intention is to prevent the head of state having "king-like" executive powers or a governing presidency on the American model — the office is meant to safeguard the constitutional order and break deadlocks, nothing more.
Note what this does to a 1975 scenario. Because the model doesn't remove the Senate's power to defer money bills, a government could lose supply and still not be dismissable. Kerr, under these rules, could not have acted unilaterally. The pressure would instead have fallen on the Senate to relent or on Whitlam to advise an election — which is arguably the correct democratic outcome, and is also, in the American experience, a situation that can persist for a long time.
One further consequence gets little attention: this codification extinguishes certain reserve powers at the Commonwealth level while leaving them intact at state and territory level, since state governors' roles would be unchanged. Australia would end up with meaningfully different rules for constitutional crises depending on which parliament was in crisis.
The critique, and how strong it is
The Centre for Independent Studies argued the ARM model turns the head of state into a rubber stamp with no independent authority even to resolve a crisis or remove a Prime Minister acting clearly unconstitutionally — and that this is worse than either alternative, because the model still creates an elected office whose holder will have campaigned on holding governments to account without any legal means of doing so.
There are two ways to read this. The critical reading is that a genuinely powerless elected head of state is an invitation to a symbolic office-holder picking fights they cannot win, which corrodes trust in the presidency and the government at once. The sympathetic reading is that this is precisely the design goal: 1975 is exactly what the ARM is trying to prevent, and an office that cannot repeat 1975 is doing its job. The Irish presidency is often invoked here as proof that a ceremonial elected head of state can hold real moral authority without destabilising anything.
Which reading you find persuasive depends heavily on whether you think 1975 was a system working under stress or a system failing. That's not a question with an agreed answer in Australian politics, and it is closer to the surface of the republic debate than the head-of-state question ever is.
The uncomfortable implication
If the reserve powers are the real problem, then a republic referendum is not primarily a vote about national identity, symbolism, or whether the head of state should live in Australia. It is a vote about what should happen the next time a government loses supply — a technical constitutional question that most voters have never had reason to think about, arriving inside a question that most voters think is about the King.
That mismatch may be the deepest reason republican referendums are hard. The case for a republic is emotional and easy to make in a sentence. The design of a republic is technical and impossible to make in a sentence. The No campaign in 1999 understood this asymmetry and exploited it, arguing that the details had not been settled and the public should not sign a blank cheque. It worked.
The question worth sitting with
Fifty years after the dismissal, Australia has never resolved whether Kerr was right. A republic forces that resolution: any model has to specify, in writing, whether a future Kerr could do what Kerr did. Some republicans think that clarification is the strongest argument for change. Others think it is precisely why a referendum keeps not happening — because the moment you write the answer down, you turn a broad coalition into two sides of an old argument. Is the reserve powers question something the republic debate has to settle, or something it would be better off avoiding entirely?
Sources
- Monash Lens — Michael Duffy, "Whitlam dismissal 50 years on: What it means for Australia's push towards a republic"
- Australian Public Law — "The Devil is in the Detail: The Reserve Powers under the Australian Choice Model"
- Australian Republic Movement — An Australian Choice for Australians
- Centre for Independent Studies — "The ARM's 'Australian Choice' falls short"
- Anne Twomey, The Veiled Sceptre: Reserve Powers of Heads of State in Westminster Systems (Cambridge University Press, 2018)
- Australasian Parliamentary Review — "Uncertainty and the Exercise of Reserve Powers: Revisiting a Governor-General's Code of Practice" (Evatt's arguments for codification)
- Wikipedia — 1999 Australian republic referendum (Kirby's assessment)