Britain Has Already Answered

America supplied the deadline. Britain supplied the evidence. They arrived one day apart and almost nobody put them side by side.

On 27 August, Elbridge Colby briefed the North Atlantic Council in Brussels on what the Americans mean by NATO 3.0. Europe takes primary responsibility for its own conventional defence. America stays interested, but reprioritised — Western Hemisphere first, the Pacific second.

None of that is new. He coined the phrase in February. Hegseth opened the Europe posture review at the June ministerial and the Pentagon started it in July.

What was new that day was a document. Reuters obtained the review’s terms of reference, and they have dates on them. At least four sets of options for Hegseth by 6 November. Completion in December. Around eighty thousand American personnel are stationed across Europe, and by Christmas there will be a paper on a desk in Washington saying how many stay and where.

That is not a threat. It is a schedule. And everyone already knows it is coming, which is why it is not the interesting part.

The interesting part happened in London the following afternoon.

John Healey resigned as Defence Secretary on 11 June over the Defence Investment Plan. His letter was unusually specific: the Prime Minister had been unable and the Treasury unwilling to commit what the country needed, and what Britain had to set was a firm 2030 date for three per cent. The Ministry of Defence had asked for eighteen billion pounds over four years. The Treasury would go to thirteen and a half. Al Carns went with him.

A month later Healey became Chancellor.

On 28 August — thirty-nine days into the job, seventy-eight days after the letter — he confirmed he will not commit to three per cent of GDP on defence by 2030. The pathway goes to next year’s spending review, which has no date. Current plans reach 2.7 per cent by the end of the decade. His first Budget, on 28 October, will concentrate on the hole already sitting inside the existing plan.

I want to be careful about what that is. It is not hypocrisy, and reading it that way is the lazy version of this argument. Healey at the Treasury is looking at the same arithmetic that beat Healey at the Ministry of Defence, from the other side of it, with the whole budget in front of him instead of one department’s share. The constraint that produced his resignation is the constraint he now administers. The man who quit rather than accept the number has not found the number.

What would falsify this. If the December review lands soft — a substantial American presence retained, no material reduction — the pressure I am describing does not arrive on the timetable I am describing, and this piece is early rather than right. And if the 28 October Budget produces a specific, funded route to three per cent, the central claim collapses outright, because the claim is that Britain keeps declining to pay and never aggregates those decisions into a position. A funded number would be Britain paying. I would have to say so.

Here is what is documented.

The Defence Investment Plan was published on 30 June, three weeks after Healey left. It adds an average of £3.8 billion a year over four years. The Institute for Fiscal Studies looked at it the same day and found that roughly a third of that increase — about £1.2 billion a year — has no identified funding source. That was pushed to the Budget. It is still there.

So the plan Healey resigned over was published part-unfunded, and he now has to fund it.

Then there is the other track, which almost nobody in British politics discusses.

On 19 May 2025 Britain and the EU signed a Security and Defence Partnership. Canada signed its equivalent five weeks later. Under the SAFE regulation, a partnership of that kind is the legal precondition for a third country negotiating access to the €150 billion instrument — competitively priced, long-maturity loans for defence procurement. In September the Council authorised the Commission to open negotiations with both countries on the same day, with the same mandate.

Canada’s negotiation concluded in December, the agreement was signed at Munich in February and formally concluded in June. Initial contribution: ten million euros. Canadian firms may supply up to eighty per cent of the value of a SAFE procurement, against a thirty-five per cent ceiling for other non-European countries. Canada is the first non-European participant.

Britain’s talks collapsed on 28 November 2025. The dispute was the size of the contribution. Reported figures ranged from around four billion euros to nearly seven, a British proposal for pay-as-you-go went nowhere, and the chair of the Lords European affairs committee said publicly that a figure that large suggested some member states did not want Britain in at all. Ministers afterwards said the access did not represent value for money and the effort was not restarted. John Healey was Defence Secretary throughout the negotiation.

I should be precise, because the numbers are doing real work. The entry price was reported, disputed and moving. What is not in dispute is that Britain and Canada started from the same legal position on the same day, and only one of them finished.

Now the strongest objection, granted in full, because it deserves better than being argued away.

The Treasury is not being stupid. Britain carries high debt, high borrowing costs and stretched public services. Every pound found for defence comes from a queue where other people are also waiting. A multi-billion contribution to a European instrument, in exchange for procurement access of genuinely arguable value, is not obviously the best use of scarce money. And a government that promised to end austerity cannot casually reopen it for a defence line. All of that is true, and none of it is bad faith.

Accept it. The conclusion still does not follow, because the objection assumes the alternative is free.

Staying outside the European defence-industrial architecture is not the cautious option or the neutral one. It is a position with its own costs, and here is what some of them currently look like.

As of August 2026 the Royal Navy’s escort force is eleven ships: six Type 45 destroyers and five Type 23 frigates. Two frigates were retired in July before any replacement had entered service. The first Type 31 is expected in 2027, the first Type 26 in 2028 — which is to say, after the window in which the American drawdown is scheduled to happen.

And the dependency is not confined to ships. Ninety-nine per cent of Britain’s international data travels through undersea cables, and two of them carry three-quarters of transatlantic capacity. The Joint Committee on the National Security Strategy reported that it could not be confident Britain would prevent a deliberate attack on that infrastructure, or recover from one in an acceptable time.

So: a nuclear deterrent, a substantial but increasingly thin conventional force, and no European economic architecture through which to share the burden of rebuilding it. That is the position. It is not an accident and it is not a plot. It is the accumulated result of decisions each of which had a reason.

The Hedge and the Wall named the shape of this. A dependency stops being defended the moment it stops being recognised as a dependency and starts presenting as a background condition. American cover was never a wall Britain built. It was the direction the weather was assumed never to come from. What the December review does is turn that assumption back into a line item.

The distinction worth holding is between agenda-setting power and response power. Britain has very little of the first. It did not set the American calendar and cannot move it. The review reports in December regardless of anything said in London.

Response power is different, and Britain still has it. Constraint is not the absence of choice. The choices available inside a constraint are real choices, and they are being made. Walking away when the reported price was in the billions was a choice. Publishing a plan with a third of its increase unfunded was a choice. Deferring three per cent to a spending review with no date is a choice. Each was defensible on the day. Nobody hid any of them.

That is the point, and it is worth being exact about it. Nobody is concealing these decisions. What is missing is the moment when they become a decision.

The Treasury knows what SAFE would have cost. The Ministry of Defence knows what three per cent costs. Officials know what European procurement access is worth, and what eleven escorts mean in a bad winter. Every piece of it is on the record. But no one has ever had to stand in front of you and say: taken together, these amount to this version of Britain’s defence, chosen deliberately, with these consequences accepted.

This is where the vocabulary cost bites. The language you would need to hold that conversation in public — interdependence, European alignment as practical national interest, paying now rather than more later — was toxified eight years ago by being coded as the enemy’s position. So the conversation does not happen. The decisions get made anyway, one at a time, on price, by people doing their jobs properly.

I have no neutral vantage point on this and you should know it. I spent over a decade working Alpine seasons, and I know what it is like to be handed the bill for a change of arrangements years after the argument about it finished, by which point nobody is interested in relitigating the argument. That is the pattern I would ask you to watch for, not the grievance. The cost of the other option is real, it arrives later than the debate, and it is paid by people who were never shown the invoice.

The Americans have produced a timetable. Britain has produced a series of reasons. A series of reasons can add up to a strategy even when nobody announces one — and if nobody announces it, nobody can be asked to defend it.

December is not the decision. December is the receipt.

In November last year, John Mearsheimer stood in front of a conference at the European Parliament in Brussels and told the people in the room that the American guarantee they had built their continent on was going away. Not might. Was. He gave them the reasoning, he gave them the timeline pressure, he gave them a forecast of what Europe would look like afterwards, and then he sat down.

He did not tell them what to do about it.

I want to be careful here, because there is a version of this argument that is cheap and I am not making it. Mearsheimer is a serious scholar working inside a coherent tradition. He has been saying the same thing since well before it was useful to anyone, which is more than can be said for most of the people now quoting him. And the central claim he makes about Russia — that it is not a potential regional hegemon, that its army is not what the mythology says, that a country grinding through a fourth year to take a fifth of Ukraine is not about to overrun the continent — I am going to grant you in full. Not soften it. Grant it. If you have been persuaded by that argument, you have been persuaded by something with real evidence behind it, and anyone who tells you otherwise is selling you a different kind of story.

So here is the model, and here is what would kill it.

The claim is that strategic realism has no prescription for Britain — not because its popularisers are lazy, but because the framework itself stops before it reaches us. If Mearsheimer, or any comparable realist, publishes a direct prescription for British policy that runs on structural interest rather than on inherited national temperament, this piece is wrong and I will say so. I looked before writing. It is not there.

Consider why it might not be there, because the reason matters more than the absence.

Mearsheimer’s theory is a theory of great powers. That is not a slight — it is the stated scope of the thing. On his own current count there are three: America, China, Russia. Britain is not on the list. A theory built to explain how great powers behave toward one another owes no advice to a state that isn’t one, in the same way that a book on how weather systems form owes you nothing about whether to take a coat. The silence is not a failure of the framework. It is the framework working as designed, and stopping where it was always going to stop.

Which leaves a gap. And the gap is where the trouble starts, because something always fills it.

Watch what fills it when Mearsheimer himself is asked. Three days ago, on an American podcast, he was asked why British policy toward Russia is what it is. His answer was Russophobia — a national disposition, traced back through the Cold War to nineteenth-century imperial anxiety about India, carried forward into the present as inherited feeling. He added that he suspects the new Prime Minister privately knows the policy should change but cannot say so.

Take a moment with that, because it is the whole argument in miniature.

Structural realism exists to explain state behaviour by position and interest rather than by mood, prejudice, or what leaders secretly believe. It was built specifically to replace explanations of that kind. And when the framework’s most prominent living exponent turns to Britain, he reaches for a cultural-psychological account of a nation’s feelings and a guess about one man’s private mind. Whatever that is, it is not the method. It is the thing the method was invented to get away from.

I should say plainly that this comes from a podcast transcript rather than from anything he has published under his own name, and it sits a tier below the Brussels address in what it can be asked to carry. But it is not an outlier. It is what the framework produces when it is pointed at a country it was never built to analyse.

So let me do the thing that is not being done, and be clear that this next part is mine. This is my inference from his premises, not his conclusion, and you should weigh it as such.

Start with his own foundational move: you can never know another state’s intentions, so you do not stake your survival on anyone’s goodwill. Applied to Russia, that produces caution. Applied honestly, it produces caution about everyone — including the state on whose intentions Britain has staked its deterrent, its intelligence architecture, and a good deal of its industrial base. Mearsheimer spends the Brussels address establishing that American interests have moved. He does not then ask what a dependent state ought to conclude from that. The premise is his. The question is the obvious next one. It is simply never put.

Now take his forecast, which is bleaker than the people quoting him tend to mention. He does not predict calm. He predicts European-Russian relations poisoned for as far as anyone can see, an ever-present possibility of war across a run of named flashpoints — the Arctic, the Baltic, the Black Sea — and a Russia with standing incentive to cause trouble inside Europe and to split it from America. That is his picture, offered unprompted, by the analyst you trust.

I suspect that is where you want to get off, so let me stay on it. The objection is that Russia has no such incentive — that it is a status quo power with a border problem, that it wants a buffer rather than a project, and that talk of mischief inside Europe is the threat inflation that got us here.

Look at what the objection has to assume. The argument you have been persuaded by is that Russia went to war over the European security architecture — over the alliance’s expansion, over Ukraine’s alignment, over where the line sits. I am not going to adjudicate that here, and I do not need to. I need only note that it is the argument you have accepted, and ask what accepting it commits you to.

It commits you to this: that Russia cares enough about that architecture to spend four years, an economy and an extraordinary number of its own people trying to change it.

You cannot then hold that it is indifferent to whether that architecture is strong or weak. A state that fought a war over the shape of a coalition has an obvious and entirely rational interest in that coalition being more divided, slower to decide, and less certain of its patron. It does not need to hate anyone. It needs only to prefer a weaker adversary bloc to a stronger one, which is the most ordinary thing a state can want, and to notice that some ways of pursuing it cost almost nothing and are hard to attribute.

That is not an accusation about intent. It is what the position itself produces.

Put Britain into it. Not an invasion — I am not making that argument and neither is he. Something far more likely and far cheaper: a cable cut, a seabed sensor, an approach contested, an infrastructure failure nobody can quite attribute. The escort fleet that would have to contest that stands at eleven ships, six destroyers and five frigates, after two more were retired in July. The first replacement arrives next year, the next in 2028.

Be careful with that number, because on its own it proves less than it looks like it proves. It invites a nostalgia I want no part of — Britain had around fifty escorts in 1990, and saying so establishes nothing except that the fleet was built for a different war. And a small fleet is not automatically an inadequate one. Fewer and better can be the right answer.

What turns eleven into a question is concurrency, and this next part is mine rather than a published finding. The same eleven hulls have to cover carrier escort, standing alliance commitments and home approaches, and a ship in refit is not available for any of them. So the question is not whether the fleet is smaller than it was. It is whether the defensive instrument can be in enough places at once to contest the offensive one at the tempo that one actually operates.

Which brings me to the part I would ask you to sit with.

Standing back does not buy you quiet. On Mearsheimer’s own forecast, it buys you the frozen conflict, the departed patron, the weakened alliance and the opportunistic neighbour — all of it, exactly as described, with Britain in it and less able to do anything about it. Withdrawal is not the absence of a choice. It is a purchase, and that is what is being bought.

I do not say this from a distance. I spent over a decade working in the French Alps, in kitchens and chalets and behind a wheel, and the arrangement that made that life possible is gone, and it was taken by a vote I lost. I followed that evidence somewhere uncomfortable — to the conclusion that the people who voted the other way had reasons, and that the settlement’s benefits were never as evenly spread as I wanted to believe. I am not asking you to do anything I have not done.

And there is one more thing, which is the part I find hardest to get past.

In Manila, three weeks ago, Elbridge Colby, the American Under Secretary of War for Policy, told allies that Washington is looking for “partners, not protectorates,” that alliances built on permanent dependency are not durable, and that countries should take responsibility for their own sovereign security. The administration promotes the doctrine under a name of its own choosing: flexible realism.

I want to be precise about what that is and is not, because the whole piece collapses if I get it wrong here. Colby is not John Mearsheimer, and a serving official’s policy statement is not a scholar’s conclusion. What the Manila speech establishes is not that realism prescribes something for Britain. It establishes that the state Britain has been relying on has itself concluded that the reliance should end, and has said so out loud, under that name.

Which closes the circle in an awkward place. Mearsheimer describes the environment. Washington describes what it now expects of allies within it. Neither of them tells Britain what Britain should build. That part nobody is doing, and it is the only part that is actually ours.

The concrete version is on the table now. Britain is weighing a fee, reported at between four and six and a half billion euros, to buy capped access to the European procurement instrument that nineteen other states are already inside. You can argue about whether it is worth the money. That is a real argument, it is not settled, and I am not going to settle it here — except to note that the alternative is not saving four billion euros. It is carrying the same requirement alone, which is a different bill and not obviously a smaller one.

So the question I would put back to you is not the one you are being asked. You are being asked whether Britain should stand back. The question the situation actually poses is what Britain needs to be able to do if the American guarantee has become conditional rather than assumed — and neither the scholar nor the official is going to answer that for us.

What you cannot do is quote realism at the question and then leave the room before it is answered. The chart runs out somewhere off our coast. That is not the same as there being no water there.

Sources and acknowledgements

The Mearsheimer material comes from his address to a European Parliament conference in Brussels, delivered 10 November 2025 and published under his own byline, and from a Judging Freedom podcast interview of 25 August 2026, cited here via transcript. Elbridge Colby’s remarks were delivered at the Stratbase Institute in Manila on 10 August 2026. Royal Navy figures follow the July 2026 retirements of HMS Iron Duke and HMS Richmond. Pew Research Center’s international survey was published on 23 June 2026. SAFE figures and the reported UK participation fee come from Council of the European Union material and subsequent reporting.

This piece builds on earlier work here on the Dead Spot and on Britain’s naval asymmetry, and on the distinction between an argument that happens to serve an interest and a person producing content for one. Mearsheimer belongs firmly in the first category, and the piece would be dishonest if it implied otherwise.

On evidential status: the Brussels address is a published text. The podcast material is reported speech at a lower tier and is flagged as such where used. The argument that realism’s own premises point toward capability rather than retreat is my inference from Mearsheimer’s stated positions, not a conclusion he has drawn, and it is labelled that way in the piece.

THE REMOVABILITY TEST

Written on the evening of Thursday 27 August 2026. The dates in this piece matter more than they usually would, so it seemed worth telling you where I am standing.

In March, the President signed an executive order instructing Homeland Security to compile a national list of eligible voters, and the Postal Service to deliver ballots only to people whose names appeared on it.

On 25 June, a federal judge in Boston blocked the key parts of it, for the twenty-three states and the District of Columbia that had sued.

On 11 August, the same judge issued a second injunction in a separate case brought by the League of Women Voters. That one went further. It barred the Postal Service from even finishing the rule.

On Friday 21 August, the Postal Service finished the rule anyway, and published it, and said it would not put it into effect while the injunctions stood.

On Monday, the Supreme Court lifted the first injunction. Six to three, unsigned, on the emergency docket. It did not rule on whether the order is lawful. It ruled that the challenge had come too early, because courts review final rules rather than proposed ones. If the final rule harms the states, the majority said, they can challenge the final rule.

On Tuesday, the judge found that the Postal Service had violated her standing injunction by finishing that rule while it was in force. She imposed no penalty.

On Wednesday — yesterday, as I write this — she vacated her own second injunction, writing that the Supreme Court’s reasoning compelled her to. In the same order she described the executive order as likely unconstitutional and warned of the disorder it could cause. Twenty-four attorneys general filed a fresh lawsuit the same day, with the governor of Pennsylvania joining as a separate plaintiff. The voting rights groups amended their existing complaint. The Democratic campaign committees moved as well. Three challenges now run in parallel, and the new case has been assigned to the same judge.

The Postal Service had said it would not apply the rule to this election unless both injunctions were lifted. Both have now been lifted. None of the three new challenges stops it in the meantime.

The first ballots of this election go out on 4 September, eight days from now, to military and overseas voters in North Carolina.

Read that sequence again, because there is a shape in it that the coverage has not picked up.

Nobody in it broke the law and got away with it, exactly. Nobody defied a court. The judge ruled. The higher court reviewed. The judge complied with the higher court while saying plainly, on the record, that she thinks the thing she has just unblocked is probably unconstitutional. More than two hundred thousand people filed comments on the proposed rule before it was finalised, which is the notice-and-comment process doing precisely what it exists to do. Every part is working.

And here is the part I keep coming back to. The Supreme Court’s reason for saying the challenge was premature is that there was no final rule to review. There is a final rule now. It became final because the government completed it while an injunction forbade completing it — a violation a court has now formally found, and declined to punish.

I want to be careful about the next step, because it is the piece’s most important claim and it is inference rather than documented fact. Nobody has established that the sequence was designed. What is documented is the sequence itself: the breach, the finding, the absence of consequence, and the fact that the completed rule is now the only thing the states can challenge, on a fresh clock, with eight days until the first ballots are posted. Whether anyone intended that outcome, I have no evidence. What it did is not in dispute.

Consider the box on the wall of your house with the switches in it. When too much current runs through a circuit, a breaker throws and cuts the power before the wiring in the joists gets hot enough to set the house alight. The whole point of it is that it is designed to fail. It gives up the circuit to save the building. A house wired without one does not feel any different to live in, right up until the evening it does.

There are two ways to lose that protection. Someone can tape the breaker shut. Or the breaker can be perfectly functional and take ninety seconds to trip on a fault that starts a fire in thirty.

The first is the story everyone is watching for. The second is the one in front of us.

Here is what would prove me wrong, and you should hold me to it, because it is checkable in weeks rather than years. If the fresh challenges produce a block before ballots go out, or if November’s elections are run without measurable disruption from this order, then the machinery beat the clock and the latency I am describing is smaller than I think. That is an observable outcome with a date attached. And a second thing would kill it just as completely: if the equivalent failure on the left is not named here with the same specificity as the failures on the right, this is not a test, it is a preference wearing a test’s clothes. I have tried. Judge whether I managed it.

Karl Popper spent The Open Society and Its Enemies arguing that political philosophy had asked the wrong question for two thousand years. The question was who should rule — the wise, the many, the strong. Popper’s move was to say that this guarantees bad answers, because you eventually get a bad ruler however you select them, and a theory built on choosing well has nothing to say about what happens next.

So he proposed a different question. Can the ruled get rid of the rulers without bloodshed? Where they can, he called it democracy. Where they cannot, tyranny. He was explicit that this was a stipulation rather than a description of the people ruling themselves.

What makes it unusual is that it says nothing about policy. It does not ask whether a government’s programme is wise, or fair, or good for you. It asks only whether the arrangement that produced it can be undone by ordinary means. That is a procedural test rather than a test of content, which is why it can be run in both directions without collapsing into whatever you already believed on Sunday.

I am extending it further than Popper stated it — from the removal of rulers to the correction of error more generally, and then to the speed of that correction. I would rather say so plainly than let the extension pass as his.

Run the test and most of what dominates political argument turns out not to be the concern.

Reform’s housing platform and its welfare platform are contestable policy. You may think they are wrong. I do. That is not what this test measures. A party can stand on them, win on them, enact them, and be turned out by people who decide they don’t like the results. The circuit is live and doing what circuits are for. Disagreeing with a policy is not the same as identifying a threat to the mechanism that lets you reverse it, and treating those as one thing is how a procedural argument quietly becomes a partisan one.

What the test catches is narrower.

The day after Todd Blanche was sworn in as Attorney General, the Justice Department published a twenty-one page opinion extending executive privilege to the President’s communications with private advisers, defining a private adviser as anyone the President consults outside the executive branch. It does not carry his signature. It is not binding law. A former Office of Legal Counsel attorney described it as a significant instrument for deflecting congressional investigation of private parties.

The documented fact is the date. It was published in August, before the November midterms, and before any subpoena it might be used against. The reading that this is anticipatory positioning rather than coincidence is inference — widely shared among legal commentators, but inference. I do not need it to be more than that. A legal position established before the confrontation arrives does not have to be sinister to lengthen the trip time. It only has to exist, so that when a committee eventually issues a subpoena, the argument is already on the record and the fight starts from there rather than from nothing. How long that fight then takes is the whole question, and it is not a question anyone can answer in advance.

Same shape as the ballot case. Not the breaker taped shut. The trip time lengthened.

There is a second place correction machinery lives, and it is not institutional.

You already know this one without needing the research, and I would rather you accepted it on its own logic than on borrowed authority. Once changing your mind about something would cost you your sense of who you are, you have lost the ability to change your mind about it. The academic work is real — Atran and Ginges on values that behave differently from preferences, so that offering someone compensation to trade one away hardens their position rather than moving it; Swann on the state where a person’s sense of self and their sense of the group stop being separable. Neither body of work has studied the movements I am about to name, and I am not going to pretend otherwise. It illuminates the mechanism. It does not establish it.

Which brings me to the part of this piece I have least enjoyed writing.

There is a specific argument circulating on the British left that has become resistant to falsification, and I want to name the argument rather than the tendency, because “the anti-imperialist left believes” is exactly the kind of category-making this piece is warning about. Plenty of serious left analysts reject particular uses of the imperialism framing while accepting Russian imperial conduct in other senses. That is ordinary disagreement and it corrects normally.

The argument I mean is narrower: that documented Russian conduct in Ukraine — the killings in Bucha, investigated on the ground by the United Nations Independent International Commission of Inquiry, and the conditions under which the Crimea referendum was held — cannot be raised without the person raising it having changed sides. Watch what happens when the evidence is put. It is not usually contested on the facts. It is sorted into a different category altogether. The response is not “your evidence is weak,” it is “I see where you’ve ended up.” That is the same failure I described above, running in the other direction, reached through entirely incompatible reasoning. Both maps. Same destination. And on this test it is not the smaller instance, because an institution has a higher court above it and a belief of that kind has nothing above it with compulsory authority. Evidence can still move it. So can experience, and consequence, and time. Nothing can compel it.

If your response to the last three paragraphs was that they were the price of admission for the rest of the piece, then I have written it badly, because they are the piece.

Let me show you what I left out, since a test you cannot see the working of is not a test.

I cut a positive contrast case — a commentator I intended to cite as evidence that public self-correction still happens on the populist right — because I could not source it twice and would not run it once. I cut the Christian Zionist material, which belongs to a different argument. I cut a line about Thiel and the Crusades, which was good texture and nothing more.

And I have now cut two versions of this piece. The first was built on the reading that Monday’s ruling had cleared the way for the executive order, which is what most of the coverage said. Checking it before drafting is the only reason that version does not exist. The second was built on the fact that a nationwide injunction still stood, which was true when I wrote it on Tuesday and false by Wednesday afternoon.

I mention it because a piece about the machinery of correction that had quietly skipped its own would not have been worth your time. And because it is the only demonstration I can offer you of the thing itself — an error, found, and corrected, fast enough to matter. That is what we are trying to keep.

The breaker has not been taped shut. Nobody is arguing about whether it works.

Check the trip time.

SOURCES AND ACKNOWLEDGEMENTS

The litigation sequence is drawn from Votebeat’s national election coverage, SCOTUSblog and Axios on the Supreme Court’s 24 August order, and Associated Press and WBUR reporting on Judge Talwani’s rulings of 25 and 26 August, including her characterisation of the executive order and her finding on the Postal Service’s rulemaking. The Solicitor General’s timing argument and the dissents of Justice Sotomayor and Justice Jackson are taken from those accounts. The Office of Legal Counsel opinion of 11 August, its definition of a private adviser and its length are as reported by Axios, Truthout and Heather Cox Richardson; the assessment of its practical effect is that of Jonathan Shaub, a former Office of Legal Counsel attorney. Rick Hasen’s expectation that the litigation would return to the Supreme Court within days is from the Election Law Blog. The volume of public comment on the proposed rule, and the 4 September date on which North Carolina sends ballots to military and overseas voters, are from US News. The composition of the new suit — twenty-four attorneys general with the governor of Pennsylvania as a separate plaintiff, filed as State of California v. United States Postal Service and assigned to the same judge — is from Reuters, CNBC and the California Attorney General’s own announcement, as is the existence of the two parallel challenges. The findings on Bucha are those of the United Nations Independent International Commission of Inquiry on Ukraine, which visited the town and reported to the General Assembly.

The removability argument is Karl Popper’s, from The Open Society and Its Enemies. The sacred values research is Scott Atran and Jeremy Ginges; the identity fusion work is William Swann. This piece builds on the Two Clocks framework developed here across earlier pieces, and on the argument about convergent conclusions reached through incompatible routes set out in The Elephant in the Room.

On evidential status. The litigation sequence, the court findings, the Postal Service rule and the Office of Legal Counsel opinion are documented fact. The reading of that opinion’s timing as anticipatory positioning is inference, shared by several legal commentators but not established. The suggestion that a breach of an injunction produced the final rule which then made the original challenge unreviewable is my own inference from the documented sequence; no intent has been demonstrated by anyone, including me, and I have not claimed any. The application of sacred values and identity fusion research to the movements named here illustrates a mechanism rather than evidencing it — neither body of work has studied them.

Richard Tice stood up in Westminster on Monday and said something largely true.

There are about 1.3 million households on social housing waiting lists in England. Young people in work cannot get a social home, cannot afford to buy, and cannot afford to rent a flat on their own. They stay in their childhood bedroom, or they are thirty-four and sharing a kitchen with strangers. That is an accurate description of a real country. It is not a manufactured grievance, and anyone who treats the people nodding along as fools has already lost the argument.

Then he set out what Reform would do about it, and the second half did not follow from the first.

Monday was presented as an answer to the shortage. Its centre of gravity is a different question: who stands where in relation to the homes that already exist. There is a supply component, and I will come to it, but as of Monday it is an uncosted assertion — and it sits inside the same programme as a mechanism with a documented record of removing homes from the stock faster than they are replaced. That is the claim of this piece. Reallocation dressed as supply.

Here is what would prove it wrong. Reform has committed to fifty thousand extra affordable social homes a year for a decade, and says this requires no new taxation and no new borrowing. If that figure arrives with a costing an auditor could hold — a named funding source, a grant rate, a land strategy — and if the party’s simultaneous revival of Right to Buy is structured so that disposals do not outrun replacements, then this is a supply programme with an ugly allocation clause bolted onto it, and my central charge is wrong. As of Monday, no such costing has been published. The claim rests on the assertion.

THE SIZE OF THE HOLE

Take the arithmetic seriously.

In 2024/25, all providers in England between them delivered 12,198 homes for social rent. The two-year average is a little over eleven thousand a year. In 2009/10 the figure was 33,491, so we are running at roughly a third of what this country managed within living memory. Across 2023/24 and 2024/25, sixty local authority districts delivered no new social rent homes at all. Not few. None.

Shelter’s calculation is that at the current rate it would take 119 years to house the households on the list today. Their estimate of what is actually needed is 90,000 social rent homes a year for a decade.

Against that, fifty thousand a year would be a serious number if it were real. It is roughly four times current delivery, with no public costing attached to it.

THE ROOF

I have spent a lot of my working life in places where shelter is a fixed quantity and the weather is not.

Picture a bothy with sixteen bunks and forty people outside in bad weather. You can argue about the order of the queue at the door. You can decide the people born nearest get in first, and the argument will feel like it is doing something, because it is loud and it involves a decision. The bunks do not multiply. Twenty-four people are still outside. The only thing that changes is which twenty-four, and how the ones inside feel about it.

Now imagine someone up on the roof taking the slates off to sell.

That is the part of Monday that has had the least attention. Alongside the fifty thousand, Reform proposes to reinvigorate Right to Buy: a thirty per cent discount after five years in the property, forty per cent after ten. Tice is not unaware of the history. He said plainly that Right to Buy worked in the eighties and nineties and that the mistake was not building replacements.

He is right, and the scale of it is documented. Just over two million social homes have been sold under Right to Buy schemes since 1980. Over the same period the social housing stock in England fell from 5.52 million to 4.07 million. Since the one-for-one replacement commitment was made in 2012, there have been 85,646 eligible sales and 28,090 replacement starts and acquisitions. That is part of the mechanism behind the shortage Tice was describing. The proposal is to run it again on easier terms — halving the qualifying period and raising the discount — and to trust that this time the replacement side keeps up.

I am not claiming anyone intends to deepen the shortage. Intent is not the standard here and I cannot see inside anyone’s head. What I can name is the condition. If disposals run ahead of replacements, the stock falls, and no rule about who queues where can put it back. That is not a prediction about what Reform would do. It is the arithmetic of the last forty-six years, and Monday contained nothing that changes it.

WHAT THE POLICY ACTUALLY DOES

The emotional register of Monday was abuse. Queue-jumping. People who have not paid in getting ahead of people who have.

The substance is different, and it is worth being precise, because the precision is the argument.

New social tenancies would be closed to foreign nationals entirely. Existing tenancies would be ended, on six months’ notice, for couples where neither partner holds UK, Irish or EU citizenship. Labour and the Liberal Democrats both put the same face to that category: NHS staff, long resident, paying tax throughout, in a home they have lived in for years. And then the genuinely new element, the one the announcement was built around: British citizens born abroad would be placed below British citizens born here.

That last one is not about foreign nationals at all. Reform had already pledged to remove them. Monday’s novelty is the sorting of citizens by birthplace, with one narrow exception carved out for Britons born overseas while their British parents were temporarily abroad. Reporting of the announcement notes that the rules could not be introduced without a change in the law. A carve-out that specific, requiring legislation to enact, is not a drafting accident. It is the point.

The rest of the framing does not survive contact with the existing rules. People who arrive illegally are not eligible for social housing. Migrants on student or work visas are not eligible. It is unlawful for asylum seekers to access it. The housing ministry stated in January that close to nine in ten social homes go to UK nationals.

That figure counts nationality, not birthplace. British citizens born abroad sit inside it — which is precisely the population Monday’s headline measure moves down the list.

Tice’s own number — over half a million social homes occupied by foreign nationals — is not a fabrication. But a figure quoted without its denominator, and without the eligibility rules that produced it, cannot show what it is being asked to show. It does not establish that anybody received a home instead of a British citizen. It is the use of the number I am contesting, not the number.

We also have a trial run. A near-identical UK connection test was consulted on in January 2024. The government’s own review of the responses concluded it would mainly increase the number of households in expensive temporary accommodation, which was already at record levels.

Which raises the question nobody on the platform answered: where do the people go? Citizens Advice and Crisis published research in July finding that two per cent of private rentals in Britain are affordable to someone relying on housing benefit. Two per cent. That is not a housing market. That is a closed door with a sign on it.

The Conservatives’ response was that the announcement is unfunded and that Reform is catching up with a Tory policy already on the table. The Liberal Democrats called it a cruel gimmick. Labour called it unserious and unworkable. Shelter’s answer was the correct one and the least dramatic: build homes rather than reorder the queue.

WHAT I HAVE LEFT OUT

Deliberately: Reform’s June 2025 housing essay, which used far uglier language than Monday’s technocratic register and would have made this piece easier and less accurate. The separate fight earlier this year over EU citizens’ benefits, which is a different policy in a different month. And the full evidential case for changing the voting system, which is a piece of its own and does not get to be smuggled into this one.

I will say only this much about it. A shortage of this size cannot be closed inside one parliament by one party. It needs a settlement that survives a change of administration. That is an argument for another day.

For now, hold the two things together, because they are the whole story. The bunks are not multiplying. And somebody is on the roof with a crowbar, telling you the problem is the queue.

Look up.

---

SOURCES AND ACKNOWLEDGEMENTS

The policy details come from Reform UK’s press conference at the Royal Institution of Chartered Surveyors in Westminster on 24 August 2026, and from contemporaneous reporting by LBC, the Morning Star, GB News, HuffPost UK and Estate Agent Today, including Richard Tice’s quoted framing and the responses from James Cleverly for the Conservatives, Gideon Amos for the Liberal Democrats, a Labour spokesperson, Siân Berry for the Greens, and Shelter. Delivery and waiting list figures are Shelter’s analysis of MHCLG live tables. Right to Buy sales, stock and replacement figures are from MHCLG accredited official statistics and the Housing, Communities and Local Government Committee’s 2022 report. The nine-in-ten figure and the eligibility rules are from the housing ministry’s own statement of 6 January 2026. The temporary accommodation finding is from the government’s published response to the 2024 social housing allocations consultation. That the birthplace criterion would require legislative change, and the terms of the exception for Britons born overseas to parents temporarily abroad, are from the Guardian’s reporting of the announcement. The two per cent private rental figure is Citizens Advice and Crisis, July 2026.

This piece follows on from The Queue Doesn’t Get Shorter and sits alongside the project’s standing work on the post-industrial arithmetic Britain has not yet built a political language for.

On evidential status: the policy mechanics, the figures and the quoted responses are documented. The claim that the announcement’s emotional framing does not match its actual target population is inference, built from the government’s own statement of existing eligibility rules rather than from any assertion about anyone’s motives. The reading of Right to Buy’s likely effect on stock is a conditional inference from a documented forty-six-year record, not a prediction about what any government would do. The identification of NHS staff as falling inside the affected category is a characterisation made by Labour and the Liberal Democrats, reported as such. Reform’s fifty thousand a year is reported here as a stated intention. It has not been costed publicly and is not treated as a fact.

You know why this one is difficult. So let’s deal with it at the top rather than hoping you’ll stay long enough for me to get to it.

Over the last two years the word “Zionist” has done two entirely different jobs in British political speech. It has described a political position, which is what it means. And it has worked as a swap-in for “Jew” inside stories about hidden control — the ones where a small group is quietly steering the country from somewhere you cannot see. Both uses are live. Both are current. And you cannot always tell from the sentence in front of you which one you are looking at, because anybody using it in the second sense has every reason to insist they meant the first.

That is not squeamishness. It is an accurate description of the ground, and it is why the caution around the word is earned. People who reach for it casually are asking Jewish readers to do the sorting work in real time, every time, and to be wrong occasionally at their own expense.

So the caution stays. What I want to argue is that we have misdiagnosed what it costs us.

The usual version runs like this. There is a real thing out there, it is doing real damage, and our discomfort with the only available word means we cannot talk about it. I have made a version of that argument myself. I now think it is wrong, and wrong in a way that matters.

The trouble with the collective noun is not that it is dangerous. It is that it fuses two things which are moving in opposite directions, and then hands you the total as though it were a single quantity. Put the word down, name the actual organisations, and the two come apart in your hands.

Here is what would prove that wrong, and I would rather put it here than at the close where it would read as insurance. The organisational strength here is not in dispute — I am not going to argue these groups are weak, because they are not. The claim rests entirely on the second quantity: the direction of travel among young American evangelicals. If that stabilises or recovers across the next two or three rounds of polling, then I am describing a dip inside an ascendant movement, and the argument fails. Policy wins will not settle it either way. A reading that counts every victory as strength and every setback as decline explains nothing at all. The base is the test.

Now the specific thing.

In September 2025 a company formed the previous month registered with the United States Department of Justice as a foreign agent of the Israeli government. The plan it filed involves geofencing. If you have not met the term: you draw a perimeter on a map around a building, and every phone that crosses into it starts receiving your advertising. The buildings here are megachurches and Christian colleges in California, Nevada, Arizona and Colorado. There is also a mobile exhibit about the seventh of October, designed to tour church car parks. The firm’s own filings bill it as the largest Christian church geofencing campaign in American history. The Foreign Ministry budgeted up to four point one million dollars, alongside separate contracts with a former Trump campaign strategist for social media work and with a firm building an influencer network, all of it routed through the same intermediary, Havas Media.

Read the stated purpose, though, because that is where the argument actually is. The filings describe the campaign as a response to declining support for Israel among evangelical Christians — younger ones in particular, who no longer inherit their parents’ position automatically.

They are not imagining it. Motti Inbari and Kirill Bumin, at the University of North Carolina at Pembroke, commissioned Barna Group to survey over seven hundred evangelicals aged eighteen to twenty-nine, in the field between late March and early April 2021, with a margin of error under four points. The work was published by Oxford in 2024. In 2018, just under seventy per cent of that age group said they sided with Israel. By 2021 it was thirty-four. Those choosing neither side went from a quarter to over forty per cent. Shibley Telhami, working from separate University of Maryland data at Brookings, found the generational gap widening well before that. And the mechanism underneath it may not be mood. Bumin’s own summary of the findings is that nearly seven in ten of those young evangelicals held amillennial or postmillennial positions — theologies which do not treat the modern state of Israel as necessary to the second coming. Two honest caveats. Inbari himself warns this may be youthful rebellion that softens with age. And the series stops in 2021, so the last five years are inference.

You do not put a fence around a congregation that is not going anywhere.

So take the two quantities separately. Christians United for Israel has access, money, a claimed ten million members and a genuine record of legislative wins. That is real and it is not diminishing. Underneath it, the constituency that gives those numbers their meaning is thinning at one end and holding at the other, and the organisation knows it precisely enough to buy perimeters around car parks in four states. Say “Zionist” and you get one number, and the number goes up. Name the organisation and name the base and you get two, and they are diverging.

And when you look at the achievement most often held up as proof of reach, the mechanism turns out to be simpler than the story requires. Christians United for Israel claims the Jerusalem embassy move, and its members did send the White House more than a hundred and thirty-five thousand emails. But Brookings, writing at the time in 2017, read the decision as personal, taken against the advice of Trump’s own defence and state secretaries, and as the delivery of a campaign promise that polled well with Republicans. Three years later Trump explained it himself at a rally in Wisconsin: the capital moved, he said, for the evangelicals. Not because they steered him. Because they vote.

That is not a hidden hand. That is electoral arithmetic, stated out loud, by the man who did it.

Which is the whole case for precision, and it is an analytical case rather than a polite one. The collective noun does not merely offend. It performs a specific operation on the evidence: it takes organisational strength and popular support and adds them together into one impression of mass. Once they are added, you cannot see that they are separating, and if you cannot see that, you cannot see what is actually happening — which is a formation with its institutional reach intact, spending heavily to hold a base that is walking out from under it, while the government it supports absorbs public criticism from Washington. That last part is current. When Israeli jets struck the runway at Abu al-Duhur in the early hours of the eighteenth of August, the sharpest public rebuke came not from Ankara or Damascus but from Tom Barrack, the American envoy, who said Washington was deeply concerned and called the strikes an unnecessary escalation that did not advance regional stability.

Some things I have left out, and why.

I have not told you that Christians United for Israel’s claimed ten million members outnumber America’s Jewish population, though the comparison is available and gets made constantly. It sets a self-declared signup list against a census figure, and the shape of that sentence is older and uglier than the point it is usually recruited to make.

I have left out a widely quoted figure about evangelical money reaching West Bank settlements, because the study underneath it puts the total at under three per cent of all foreign donations to Israeli non-profits. A number only sounds enormous when you decline to print the denominator.

I have left out a very good story about one cabinet member pushing for the Iran strikes against a room of sceptics, because when I went to check it he turns out to have been the first to speak rather than the only one, and the loudest outside voices reportedly included Rupert Murdoch, who is not waiting for the end times.

And I have left out a set of Saudi polling figures showing public support for the Abraham Accords falling from forty-one per cent to thirteen since 2020. They are real, they are well sourced, and they point the same way. But they describe a different machine — public opinion narrowing a government’s room to move, with no organisation, no constituency and no money anywhere in it. A pattern that fits is not a mechanism that transfers. Two examples pointing the same direction is precisely how you talk yourself into a claim larger than the one you can carry.

Which is the risk in the end, and it does not go away. In 2007 two American academics made a careful, sourced, national-interest argument about lobbying and foreign policy, and they were accused of antisemitism anyway, and are still fighting about it. I do not think precision protects you from that. What I think it does is make the accusation wrong in a way that can be demonstrated, which is a smaller claim and the only one available.

But safety was never the reason. The reason is that the vague word gives you the wrong picture, and the wrong picture makes you brace for the wrong thing. If you believe you are facing something unified and rising, you prepare for a fight on those terms. If you can see that a great deal of money is being spent to hold a perimeter around people who are drifting out of it anyway, you are looking at something else entirely, and you would want to know which.

Fences go up when people start leaving. Watch which way they are walking.

Sources and Acknowledgements

The geofencing campaign is documented in filings made under the Foreign Agents Registration Act in September 2025 and reported by the Times of Israel and the Jewish Telegraphic Agency, including the budget ceiling, the targeting across four states, and the campaign’s own stated aim of countering declining evangelical support. The generational polling comes from Motti Inbari and Kirill Bumin’s Barna Group surveys of 2018, 2020 and 2021, published as Christian Zionism in the Twenty-First Century by Oxford University Press in 2024, with independent corroboration from Shibley Telhami’s analysis of University of Maryland Critical Issues Poll data at Brookings. The Brookings assessment of the Jerusalem embassy decision was published in December 2017; Trump’s own account was given at a rally in Oshkosh in August 2020. The Abu al-Duhur strikes and the American response were reported by the Associated Press and NPR.

This piece builds on the Vocabulary Cost and the Rear Guard Action, both developed at length in earlier work here.

What is documented above: the filings, the budget, the polling, the statements. What is inference: that a formation can hold its institutional position and lose its base at the same time, and that this is what we are looking at. That reading is mine, it is stated as a reading, and the paragraph near the top explains what would falsify it.

Before anyone reaches for it: this is not a piece claiming Trump has a secret grand strategy, and it is not a piece claiming he has no strategy at all. The argument is that the doctrine and the man are two different objects, and that allies have started planning around the second one. I’ve named what would falsify that, and I’ve said in the piece which examples I deliberately left out and why — Greenland, tariffs, the dollar and Hormuz were all available and all would have made the pattern look stronger than the evidence supports.

Narration and visuals generated by NotebookLM from my own essay. The analysis, and any errors in it, are mine.

The sourcing is at the foot of the piece. Check the working — particularly the National Defense Strategy wording, which is the load-bearing quote.

[Substack link]

Stay In The Room is a political analysis project published on Substack. It looks for the structures underneath the week’s events — the conditions that produce the news rather than the news itself.

WHOEVER’S ON DUTY

Every August, American and South Korean forces run a joint exercise called Ulchi Freedom Shield. This year’s was carrying something extra: a large part of the demonstration Seoul needs to make before wartime command of its own forces transfers back from Washington. Hours before it began, the President ordered it cut roughly in half — because South Korea had declined to help him over Iran. Six days earlier, his own policy chief had stood in Manila and set out a coherent strategy asking allies to do exactly the sort of thing that exercise was designed to prove they could do.

This piece is about the gap between those two facts. The 2026 National Defense Strategy does not instruct South Korea to take primary responsibility for deterring the North — it assesses that Seoul is already capable of it. Colby had publicly called South Korea a model ally in January. Europe shows the same process further along: a force posture review whose stated criteria are allied contributions, basing and overflight, and a NATO surveillance aircraft programme that collapsed after Washington cancelled its own.

There is a coherent American doctrine. There is also a principal who applies it according to mood. The question is what allies do once they have worked out which of the two to watch.

Read the full piece on Substack — link below.

[SUBSTACK URL]

Stay In The Room is written and read by Dave Preston.

#NATO #ForeignPolicy #Geopolitics

WASHINGTON’S LOUDEST CLAIMS HAVE THE LEAST DOCTRINE BEHIND THEM

Greenland, Nato, the euro and the Strait of Hormuz rest on radically different foundations. The gap between what the documents authorise and what the president says is where the story sits.

On Friday afternoon, at a police academy in Garden City on Long Island, Donald Trump told a room of law enforcement officers that he would soon be declaring the Hormuz Strait a territory of the United States. He chuckled. According to the Wall Street Journal reporter in the room, he then appeared to add that it was true. Later that day a senior White House official told the same reporter the president had been joking and had not discussed it with advisers.

The temptation is to file that as noise. It is more useful placed alongside three other American pressure campaigns running concurrently, asking a narrower question than the one usually asked: what kind of documentation sits underneath each of them?

The answer is not uniform, and the unevenness is the finding.

Consider first the case with the most paperwork. The 2026 National Defense Strategy, published on 23 January, mentions Greenland five times and commits the United States to guarantee military and commercial access to key terrain, naming the Panama Canal, the Gulf of America and Greenland. The operative word is access. It is a serious claim on another country’s territory and it is not a claim of ownership. The document does not authorise what the president has repeatedly implied.

The distance between the two shows up on the ground. Jeff Landry, the Louisiana governor appointed special envoy to Greenland in December, told Fox News in May that Greenland could be exporting two million barrels a day, with production plausibly beginning within ten months. On 29 and 30 July, a barge carrying an excavator and more than a dozen containers was landed at Nerlerit Inaat on Greenland’s east coast. The following day Greenland’s government stated that the licence holder had no approval to land it. On 12 August the companies announced drilling would not begin before winter 2027, against the envoy’s ten months in May, scaled back from two wells to one. Greenland’s permanent secretary for mineral resources subsequently told Newsweek that even that date was only tentatively realistic.

The proximate cause was not diplomacy or protest. It was a permitting authority applying its own procedure.

The Europe case has comparable documentation and a live instrument. The 2025 National Security Strategy states that American policy should prioritise cultivating resistance to Europe’s current trajectory within European nations, and speaks approvingly of the growing influence of what it terms patriotic European parties. That is a published objective, not an inference from tone: an American administration has committed on paper to working on the internal politics of allied democracies.

Last week Bloomberg obtained the instrument. A questionnaire circulated to Nato members asks governments whether they have been publicly supportive of American foreign policy; what restrictions they place on American use of their bases, with specific reference to Middle Eastern and Western Hemisphere operations, and whether those restrictions might be reduced; whether they buy from American defence contractors; and whether they have publicly opposed European procurement rules favouring local firms. It arrived weeks into a six-month review of American troop presence in Europe due to conclude in December.

What Washington appears to be discovering is that the European system it wants to reshape already has several independent trajectories running through it, most of which it did not start. Denmark extended conscription to women in July 2025, over a year earlier. Hungary has blocked Council foreign policy decisions twenty-one times since 2022, three years before the strategy document that supposedly cultivated it. Poland vetoed a forty-four billion euro defence loan in March. The European Defence Union reported this spring is a grouping of around thirty MEPs from eleven member states, formed in May — parliamentary advocacy rather than Council policy. None of this is resistance, and none of it is compliance. A questionnaire is a thin instrument to bring to it.

The currency case is where the documentation thins. The intellectual reference point is a forty-one page essay published in November 2024, before the administration existed, by an economist then at a private investment firm and now a Federal Reserve governor. Its thesis was that the dollar is structurally overvalued by its reserve status and that a coordinated arrangement could weaken it. The tariff half was implemented rapidly. The currency half was shelved. Whether the paper supplied intellectual scaffolding for trade policy or merely described what was going to happen anyway remains genuinely contested.

What is documented is that on 31 July the United States intervened alongside Japan to support the yen — the first joint operation of its kind since 1998 — and funded it by selling euros rather than dollars. According to the Financial Times, the European Central Bank learned of the sale only after the transactions had completed. Senior ECB officials described it as a break with a consultation convention that has held since the war. The Treasury’s position was that reserve allocation within its Exchange Stabilization Fund is not coordinated with foreign authorities. Four days later the Treasury secretary said he had been in close touch with European partners and had assured them it was a reallocation of reserves. Those two accounts do not reconcile, and both are on the record.

Note the direction of travel. The essay’s programme was dollar devaluation. The action taken was selling euros to defend a third currency — the opposite manoeuvre, chosen partly because selling dollars would have signalled retreat from a strong-dollar position the Treasury was publicly holding. This is the strongest available evidence against a single coordinated design. It is not conclusive: a plan could sit above the Treasury, or the department could be executing one component without sight of the rest. But a coherent currency programme would have to run through this department, and what it did on 31 July was not that programme.

Which returns us to Hormuz, where no comparable doctrine appears to exist. No dedicated Gulf doctrine surfaces in the public record. An absence found by searching is not an absence established. What can be said is narrower: Greenland receives five mentions and a specific commitment in the 2026 strategy, and nothing comparable turns up for the strait. That is a difference in the published record, and no more than that.

What exists instead is a sequence. On 13 July the president announced a twenty per cent charge on cargo transiting the strait. The same day, his own secretary of state said publicly that no country is allowed to charge tolls or fees in international waterways. Twenty-five hours later the charge was withdrawn, replaced — according to a social media post citing highly productive conversations with Middle East leadership — by unspecified Gulf investment commitments. At least one regional government said it had not agreed to increase existing pledges in exchange.

The Gulf states did kill the fee. Their reasoning was narrower than principle. Several had privately signalled to American and European officials that they were not opposed to paying a transit fee as such; what they will not accept is Iran controlling access. One regional official told Middle East Eye the payment itself was insignificant. They objected to who would hold the switch, not to its existence.

Meanwhile, before the war began on 28 February, roughly 130 to 140 vessels transited the strait daily. Kpler’s tracking put the five-day average at around thirteen last Tuesday. The UK Maritime Trade Operations counted 151 ships in or out across seven days; the US energy secretary maintains private trackers undercount and puts oil exports at nearly nine million barrels a day against roughly twenty million pre-war. Three counts, three methods, no reconciliation — which is itself a measure of how little is firmly known. On 12 August the president told reporters that America has total control, that America owns it.

Line the four up by documentary weight and something falls out. Greenland: formal doctrine, specific and limited, with rhetoric well beyond it. Europe: formal doctrine and a working instrument. Currency: a contested precursor and an action that contradicts it. Hormuz: no doctrine located, and the loudest claim of the four. The volume runs inversely to the paperwork.

A caution is owed here. These four cases were selected, not sampled. An administration containing multiple factions and a great deal of public improvisation will generate four episodes matching most frames brought to it. What argues against pure selection is that the cases resisted flattening when pressed: the four documentary foundations turned out to be genuinely different in kind, and the strongest evidence in the sequence — the Treasury acting against the essay, the secretary of state contradicting the president within the day — tells against the tidier reading rather than for it.

The proposition, then, is modest. Different parts of this administration, working from different documents and different incentives, keep arriving independently at the same structural move: identify a dependency, convert it into leverage, demand a better deal. Nothing in the record establishes that anyone designed this as a programme. What the record shows is a path worn by separate walkers rather than surveyed by a planner — which is a different thing, and calls for a different response.

That proposition is testable, and December will test it. If the troop review concludes with allies having materially altered base-access restrictions or procurement rules in the direction the questionnaire requests, then this was a strategy, and it worked, and the reading here is wrong.

For now the results are instructive precisely because they are not uniform. Greenland’s regulator stopped the drilling — a permitting authority applying its own procedure, which is institutional refusal in the strict sense. The Gulf states killed the transit fee inside a day, which is a different mechanism altogether: five governments with enough leverage to be heard, acting on their own interest rather than on any principle. Europe’s outcome is unresolved, and December will resolve it. In the currency markets there was no refusal available at all. The euro was sold, its custodians were informed afterwards, and no institutional counterweight existed to be applied.

Which suggests that what does the work is not resolve but architecture: a procedure to invoke, or a bargaining position to hold. Where one of those existed, the move has so far been stopped. Where neither did, it was not stopped at all.

Dave Preston writes Stay In The Room.

Friday afternoon, a police academy on Long Island. Trump tells a room full of law enforcement officers he’ll soon be declaring the Hormuz Strait a territory of the United States.

Then he chuckles.

Then, according to the Wall Street Journal reporter standing in the room, he seems to add that it’s true.

By that evening a White House official is telling the same reporter it was a joke and nobody’s discussed it with him.

Forget what was in his head. Nobody knows and nobody’s going to.

Here’s the useful question instead: what’s underneath it? Not the mood. The paperwork.

Because four of these pressure campaigns are running at once, and the answer is different every time.

Greenland has the most paper

The 2026 National Defense Strategy, published in January, mentions Greenland five times. It commits America to guarantee military and commercial access to key terrain, naming the Panama Canal, the Gulf of America and Greenland.

Access. Not ownership.

That’s a real claim on somebody else’s territory and it is still not the claim the president keeps making.

Watch what happened when it hit the ground.

Jeff Landry, the Louisiana governor Trump made his special envoy to Greenland, told Fox News in May that Greenland could be exporting two million barrels a day, and production could start within about ten months.

End of July, a barge lands an excavator and a dozen-odd containers on Greenland’s east coast. Next day the Greenland government says the licence holder had no approval to land any of it.

On 12 August the companies announce drilling won’t start before winter 2027 — two wells cut to one. Greenland’s own mineral resources secretary then tells Newsweek even that date is only maybe.

Nobody protested it to a standstill. A permit office did that.

Europe has the paper and the instrument

The 2025 National Security Strategy says American policy should prioritise cultivating resistance to Europe’s current trajectory within European nations. It’s warm about what it calls patriotic European parties.

That’s not reading between the lines. It’s written down. An American government has committed on paper to working on the internal politics of its allies.

Last week Bloomberg got hold of the tool. A questionnaire sent round NATO members asking governments whether they’ve publicly backed American foreign policy. What limits they put on American use of their bases, and whether those limits could be reduced. Whether they buy American defence kit. Whether they’ve publicly opposed European rules that favour European firms.

It landed weeks into a six-month review of American troops in Europe. That review reports in December.

But look at what it’s landing on.

Denmark extended conscription to women in July 2025 — over a year before any of this. Hungary has blocked EU foreign policy decisions twenty-one times since 2022, three years before the strategy document that supposedly encouraged it. Poland vetoed a forty-four billion euro defence loan in March.

Europe isn’t resisting and it isn’t complying. It’s busy doing its own things for its own reasons, and a questionnaire is a thin tool to bring to that.

Currency

The idea people point to is a forty-one page essay from November 2024, written before this administration existed, by an economist then at a private investment firm and now sitting as a Federal Reserve governor. Argument: the dollar is overvalued because everyone holds it, and a deal could weaken it.

The tariff half got done fast. The currency half got shelved.

Then on 31 July, America intervened alongside Japan to prop up the yen. First joint operation of its kind since 1998. And it funded it by selling euros rather than dollars.

The European Central Bank found out afterwards. According to the Financial Times, they learned of it only once the trade had closed. Senior ECB people called it a break with a consultation habit that’s held since the war.

Four days later the US Treasury Secretary said he’d been in close touch with European partners and had assured them it was just a reallocation of reserves.

Both accounts are on the record. They don’t agree.

Now notice the direction. The essay wanted a weaker dollar. What actually happened was selling euros to prop up a third currency — which is the opposite move, chosen partly because selling dollars would have looked like backing off the strong dollar line.

That’s the best evidence in this whole thing against a single plan. It’s not proof. A plan could sit above the Treasury. But a currency programme would have to run through that department, and what it did on 31 July wasn’t it.

Hormuz has no paper at all

I looked for a doctrine on Iran and the Gulf. Nothing surfaces.

An absence you find by looking isn’t the same as an absence proved.

Here’s what I can actually say. Greenland gets five mentions in that January strategy document, with a specific commitment attached. Nothing like that turns up for the strait. That’s a difference in what’s been published. It isn’t evidence about what’s sitting in a filing cabinet.

What exists instead is this.

On 13 July Trump announces a 20% charge on cargo going through the strait. Same day, his own Secretary of State says publicly that no country is allowed to charge tolls or fees in international waterways.

Twenty-five hours later the charge is gone, replaced by unspecified Gulf investment. At least one Gulf government then says it never agreed to increase anything in exchange.

And the Gulf states’ reason for killing it wasn’t principle. Several had quietly told American and European officials they didn’t mind paying a fee. One regional official told Middle East Eye the money was insignificant. What they won’t have is Iran holding the switch.

They objected to who collects, not to collecting.

Meanwhile: before the war started on 28 February, roughly 130 to 140 ships a day went through. Last Tuesday the five-day average was about thirteen.

There are three different counts in circulation and they don’t match — Kpler says thirteen, the UK Maritime Trade Operations counted 151 across a week, and the US Energy Secretary says private trackers undercount. That disagreement tells you how little anyone actually knows.

On 12 August Trump told reporters America has total control and America owns it.

The pattern

Line them up by how much paperwork sits underneath.

Greenland: formal doctrine, specific, limited — and rhetoric miles past it.

Europe: formal doctrine and a working instrument.

Currency: a contested essay, and an action that runs against it.

Hormuz: nothing found, and the loudest claim of the four.

The volume runs inversely to the paperwork. The place with no doctrine behind it produced a claim to annex an international strait.

One honest caveat. I picked these four. I didn’t sample them. Any administration this factional will throw up four episodes that fit most patterns you bring to it. What makes me think there’s something here is that they refused to flatten when I pushed — and the strongest bits of evidence, the Treasury going against the essay and the Secretary of State contradicting the president inside a day, cut against the neat version rather than for it.

So the claim is small. Different parts of this administration, working from different documents and different incentives, keep separately finding the same move. Spot a dependency. Turn it into leverage. Ask for a better deal.

Nobody drew that path. People kept walking the same line because the ground rewarded it.

And it’s testable. If that troop review reports in December with allies having actually changed their base rules or their procurement to suit Washington, then it was a strategy, it worked, and I got it wrong.

What actually stopped it

Four attempts. Four different outcomes, and they’re not the same kind of outcome.

Greenland: a regulator with a procedure used the procedure. Stopped.

The Gulf: five governments with enough weight to be heard, acting on their own interest. Stopped in a day.

Europe: not tested yet. December.

The euro: nothing stopped it. It was sold, the people whose currency it was got told afterwards, and there was no rule to point at and no hand to play.

So it isn’t about who’s brave.

It’s about whether there’s a procedure to invoke or a position to bargain with.

Where there was one, the thing in front of them got stopped. A drilling programme. A transit fee. Not the whole campaign — the specific thing.

Where there was neither, nobody even got told until it was done.

Don’t ask whether he means it.

Ask what’s underneath.

Dave Preston writes Stay In The Room.