The AUKUS Blowout: The Bill in Brief
A five-page summary of the full assessment — the estimated cost and the worst case for each part of the AUKUS bill.
AUKUS is defended as one number — $368 billion. Its real cost is a stack of separate costs of different kinds: a lifetime bill, annual trade and fuel exposures, one-off losses, open-ended nuclear long-tails, and costs that cannot be priced at all. They measure different things and cannot honestly be summed, and the heaviest — sovereignty and lives — carry no dollar sign. Each is set out below with a base estimate and a worst case: the money already spent with no boat in the water, the trade risked by treating the largest customer as an enemy, the friendships spent, the capability that arrives only if a US president allows it, and the war that cannot be won. The full assessment, with all seventeen references, is linked throughout.
This is the bill in brief. The full assessment sets out each cost in depth with all seventeen references — read The AUKUS Blowout, the full assessment → (also available as a PDF). This summary is downloadable as a 5-page PDF.
The bill at a glance
| # | The cost | Estimate (base case) | Worst case |
|---|---|---|---|
| 1 | Buy and build | $368B (incl. 50% contingency) | $480–650B+ |
| 2 | China as the enemy | ~$10B+/yr (lost investment, eroding share) | ~6% of GDP (~$160B/yr) + supply collapse |
| 3 | Choosing America | ~$1–3B/yr (tariffs, no trade gained) | No shelter from the ally; tariffs escalate |
| 4 | Choosing Britain | ~A$4.6B+ to the UK industrial base | A partner that may not deliver |
| 5 | Sovereignty | Unpriced — the whole $368B+ is conditional | 0 boats unless US output doubles + sign-off |
| 6 | Costs no one totals | ~$10–20B+ (nuclear long-tail; toward ~5% GDP) | Open-ended nuclear liability |
| 7 | The wrong direction | The renewable & manufacturing nation foregone | Capacity locked to a depreciating asset |
| 8 | Lives | Unpriced | Tens of thousands, in a war that cannot be won |
Figures are order-of-magnitude. They measure different things and cannot be summed; the worst-case column is where the trajectory points, and for the trade and lives entries assumes the wartime rupture the programme is built for.
1. The cost to buy and build
$368 billion is a 2023 whole-of-life estimate to 2055 that already carries a 50 per cent contingency — the floor, not the ceiling. Comparable builds overran: the UK Astute class ran about 30 per cent over budget; France’s cancelled Attack-class climbed from about $50 billion toward $90 billion. About $11 billion will be spent by mid-2027 with no boat in the water, around US$3 billion of it non-refundable to US and UK shipyards. Seventeen years after the 2009 decision to replace the Collins class, zero boats have been delivered.
Estimate $368B · Worst case $480–650B+
Sources: Defence portfolio budget papers; UK National Audit Office (Astute); ANAO; Michael West Media / Rex Patrick.
2. The cost of declaring China the enemy
China takes about a third of all exports ($312 billion two-way, 2024) and is also Australia’s largest supplier — about a quarter of imports, the cheap goods the RBA judges disinflationary. The 2020 coercion disrupted about $20 billion of trade (net loss near $1 billion; iron ore was spared). Foregone Chinese investment now runs about $10 billion-plus a year. Iron ore is the shield, and China is spending about $20 billion on Simandou in Guinea to remove it; Lithuania shows China will cut a country off entirely over a Taiwan stance. A full rupture is the wartime case: the modelled income loss is about 6 per cent of GDP (~$160 billion a year) and 14 per cent per head — but the larger, unmodelled half is the simultaneous collapse of supply (goods, fuel, medicine), which breaks the economy.
Estimate ~$10B+/yr · Worst case ~6% of GDP (~$160B/yr) + supply collapse
Sources: DFAT trade data; J. Quiggin (rupture modelling); RBA (disinflation); ASPI / BITRE (seaborne shares); CSIS & Asia Society (Lithuania).
3. The cost of choosing America
Washington refused Australia a steel and aluminium tariff exemption in March 2025 (now 50 per cent / 25 per cent), added a 10 per cent import surcharge, and has proposed a 100 per cent pharmaceutical tariff — all under a free-trade agreement, and while the United States runs a trade surplus with Australia. AUKUS bought no trade relief. The US takes under 5 per cent of exports, so the direct dollar cost is modest; the signal — an ally that taxes a treaty partner when its own politics suit — is the cost that matters.
Estimate ~$1–3B/yr · Worst case No shelter when interests diverge
Sources: US executive orders / USTR; Australian Government statements; Lowy Institute.
4. The cost of choosing Britain
Britain is a marginal economic partner — about £24 billion of two-way trade, Australia ranking as the UK’s 21st partner at barely 1 per cent of its trade, while Britain sells Australia roughly twice what it buys. Keating’s “dumped us” is literal: joining the European Economic Community in 1973, Britain ended the Commonwealth preferences that had made it Australia’s protected market. Yet AUKUS reverses the ledger: Australia has committed about A$4.6 billion over a decade to expand Britain’s submarine industry (Rolls-Royce at Derby, BAE at Barrow) and fund part of a British-designed boat whose reactors are built in England — from a partner whose own yards are stretched by the Dreadnought build and may slip the first AUKUS submarine.
Estimate ~A$4.6B+ to the UK · Worst case A partner that may not deliver
Sources: UK DBT / ONS trade factsheet; gov.uk Defence Nuclear Enterprise; US Studies Centre; Lowy Institute; ASA Senate Estimates.
5. The cost in sovereignty
The submarines transfer only if a US president certifies it will not degrade US capability — and that certification is gated on a build rate the US has never hit. The Navy says it must produce about 2.33 Virginia-class boats a year to spare any for Australia, but the actual rate is about 1.1 to 1.3 and only reaches two a year around 2032, the year the first transfer is due. The 30 May 2026 revision (three second-hand boats, no new hull) was set to suit the US industrial base, not Australian need. Australia also leased the Port of Darwin to a Chinese firm (now buying it back) and has twice had to force China-linked owners out of a heavy-rare-earths project.
Estimate Unpriced — the bill is conditional · Worst case 0 boats unless US output doubles + a sign-off
Sources: US AUKUS legislation (FY2024 NDAA); US Congressional Research Service; Senate Estimates (ASA); Pentagon ‘America First’ review reporting.
6. The costs no one totals
Nuclear stewardship commits Australia to high-level waste for the life of the boats, a naval-nuclear regulator, and indemnifying the US and UK — with no waste site chosen; the UK’s own submarine-disposal bill runs on the order of $10–20 billion. Sustaining the programme pushes defence spending toward about 5 per cent of GDP (~$140 billion a year). Each non-refundable payment makes exit dearer — a plan engineered to be “too big to cancel”. Hosting the boats and bases raises Australia’s value as a target.
Estimate ~$10–20B+ · Worst case Open-ended nuclear liability; “too big to cancel”
Sources: UK National Audit Office (submarine disposal); ANAO; defence budget projections.
7. The largest cost: the wrong direction
Wrong as defence: a submarine cannot lift a blockade — the contest that actually decides an island trading nation. Wrong as industrial strategy: it draws down scarce engineers, trades and factory capacity to build a foreign-designed boat, and in its early years to expand US and UK shipyards. The same base could build rail, energy and manufacturing that compound in value and feed, fuel and power the country; the submarine only depreciates. And AUKUS need not be cancelled so much as redirected — the same yards, undersea trades and skills shift to sovereign subsea transmission, service and cable-repair vessels and uncrewed undersea systems, a far broader scope than eight nuclear boats. The opportunity cost is the nation not built.
Estimate The nation foregone · Worst case Talent & capacity locked to a depreciating asset
Sources: MMA Memos 31–34 (the net assessment and the alternative).
8. The cost that cannot be undone: lives
The boats and the US bases on Australian soil make Australia a combatant. Wargames of a Taiwan invasion project tens of thousands of combatant casualties; a blockade kills civilians by shortage, as twentieth-century naval blockades killed hundreds of thousands. These would be Australian lives spent in a war that, on the analysis, cannot be won — one only the United States and China survive as functioning nations.
Estimate Unpriced · Worst case Tens of thousands
Sources: CSIS Taiwan-invasion wargame (2023); naval-blockade history (WWI; Japan / USSBS).
The MMA position
At this price the programme buys the wrong capability for the threat that actually decides the country, and surrenders the power to decide its own use. A plan that costs this much, across this many dimensions, must be reviewed. The same money would buy a defence that is owned, relevant and delivered sooner — and, because every other cost can be reversed but lives cannot, keep Australia out of a war that is not its own. Find peace, and build a united Asia.
References
This is a summary; the full assessment carries the complete reference list (17 sources). See The AUKUS Blowout — the full assessment. Principal sources: Defence portfolio budget papers; ANAO; UK National Audit Office; US Congressional Research Service (Virginia-class / AUKUS Pillar 1); DFAT trade data; RBA; ASPI / BITRE; J. Quiggin (rupture modelling); CSIS (Taiwan-invasion wargame; Lithuania coercion); Lowy Institute; US Studies Centre; gov.uk Defence Nuclear Enterprise.