On July 6, Mark Carney stood in Halifax — on his way to a NATO summit in Ankara — and announced the largest military procurement in Canadian history: up to twelve attack submarines from Germany’s ThyssenKrupp Marine Systems, a program worth $20 to $30 billion for the boats aloneand, by Korean press estimates, as much as $100 billion over three decades once you count the lifetime of the fleet.

Canada has not bought a newly built submarine in more than sixty years. Its procurement system is a running joke in defense circles — the country that takes a decade to buy a pistol. This time, the entire competition ran in under a year, start to finish, a pace one analyst called astonishing for a project of this size. Germany and Norway offered to give up their own places in the production queue so Canada could take delivery of its first four boats by 2034 .

Stop and sit with that. A famously slow, famously cautious middle power just executed the fastest major weapons purchase in its modern history, for the largest sum it has ever spent on anything military, and two European allies rearranged their own naval production schedules to accelerate it.

The interesting question is not what Canada bought.

The interesting question is what Canada stopped believing.

The Underwriter Walks Off the Job

For eighty years, the United States sold the world’s most valuable product, and it never appeared on a customs form. The product was certainty. An American security guarantee was an insurance policy: allies paid premiums — basing rights, deference on trade, diplomatic alignment, purchases of American weapons — and in exchange they could run smaller militaries than their threat environment justified, because the underwriter of last resort was the Pentagon.

Insurance has a peculiar property. The policy is worthless the moment you doubt the insurer will pay. Not when the insurer refuses to pay — merely when you can no longer predict whether it will. A fire policy from a company that pays some claims, denies others, and decides which is which based on the owner’s mood that morning is not a discounted policy. It is not a policy at all. And every rational homeowner in that market responds the same way: they stop paying premiums and start buying their own fire trucks.

That is the mechanism now operating on the entire American alliance system, and here is the part almost nobody in Washington wants to say plainly: it does not require a doctrine. There is a cottage industry of analysts straining to find the grand strategy inside the current administration’s foreign policy — the threats to annex Canada, the ambivalence on Article 5, the tariffs on allies, a forty-day air war against Iran launched, paused, and, in mid-July, declared un-paused. I spent enough years inside the machinery of national security planning to tell you what I think is actually there: a set of beliefs about force — that it works, that it is cheap, that consequences are for other people — being acted out in real time.

Not a plan. A disposition.

But markets do not price intentions. They price behavior. And states, when it comes to survival, are the coldest-eyed market participants on earth. Whether the erratic signal coming out of Washington is four-dimensional chess or a mental disorder is analytically irrelevant to a defense planner in Ottawa, Paris, London, Berlin, or Canberra, because the planning conclusion is identical either way: the guarantee can no longer be priced, so it can no longer be relied upon.

The guarantee was never revoked. It was repriced. That distinction matters, because revocations can be reversed by the next administration. Repricings live in the counterparty’s head, and no signature can reach in there and undo them.

We have run this experiment before, and the results are on file.

In the mid-1960s, Charles de Gaulle looked at the American guarantee and asked the question no ally is supposed to ask out loud: would Washington actually trade New York for Paris? He concluded he could not know — and that not knowing was the whole answer. France pulled out of NATO’s integrated military command in 1966, evicted the alliance’s headquarters from French soil, and built its own nuclear deterrent, the force de frappe, at staggering expense for a country still rebuilding from the war.

Note the sequence carefully. Nothing had happened. No abandonment, no betrayal, no failed Article 5 invocation. The United States in 1966 was arguably at the height of its commitment to Europe (especially by present-day standards). All that changed was that a serious man in a serious capital could no longer price the promise — and that alone was sufficient to justify the most expensive hedge available.

Now note the duration. France did not rejoin NATO’s integrated command until 2009. Forty-three years. Eight American presidents came and went, several of them ardent Atlanticists, and none of it mattered, because the decision had never been about any particular president. And when France finally did come back, it came back with the arsenal. The submarines, the warheads, the independent defense industry built to serve them — none of it was dismantled because once a state pays the fixed cost of self-insurance, maintaining the capability is nearly free, and surrendering it is unthinkable. One moment of doubt, acted upon, produced a two-generation structural fact.

That was one country, reacting to a hypothetical. What is happening now is that every ally is reacting at once to behavior they can watch on television.

The Demand Map

Once you see the mechanism, the data stops looking like a collection of headlines and starts looking like a single event happening in many places at once.

Start with the aggregate. World military spending hit $2.887 trillion in 2025, the eleventh consecutive annual increase, pushing the global military burden to 2.5 percent of world GDP — the highest share since 2009. Strip out the United States, whose 2025 number fell for accounting reasons we’ll get to, and the rest of the planet increased military spending by 9.2 percent in a single year.

Now look at where the money is moving. Europe’s spending surged 14 percent to $864 billion — the fastest annual increase among NATO’s European members since 1953. Think about what 1953 means: to find a European rearmament this fast, you have to go back to the moment the continent stood up armies against Stalin. Germany raised spending 24 percent in one year, crossed 2 percent of GDP for the first time since 1990, and has pledged 3.5 percent by 2029. Spain jumped 50 percent. Poland now carries the heaviest defense burden in the entire alliance at 4.5 percent of GDP. And at this summer’s Ankara summit — the same one Carney was flying to — the alliance pressed toward the 5-percent-of-GDP target adopted in 2025, with Secretary General Mark Rutte telling members before Ankara that he expected plans that are, in his words, “clear, concrete and credible”.

Here is the detail that separates this from every previous burden-sharing cycle, and it is the one American coverage keeps missing. Europe is not just spending more. It is spending _elsewhere_ — deliberately, structurally, and now by statute. The EU’s new SAFE instrument — €150 billion in jointly backed loans, the first pillar of a plan to mobilize more than €800 billion in defense investment by 2030 — carries a legal condition: no more than 35 percent of a funded weapon system’s component cost may originate outside the EU, the EEA-EFTA states, and Ukraine. Read that again. Buy-European preference is no longer a communiqué sentiment. It is a loan covenant. The Draghi review found that nearly 80 percent of recent European procurement money flowed outside the EU, overwhelmingly to American firms. The entire architecture of European rearmament is now designed to reverse that number.

Two footnotes to the European number tell you how seriously to take it. First, the scale: NATO members collectively spent $1.581 trillion in 2025 — 55 percent of everything the entire planet spent on arms (and roughly the equivalent of the United States, something that prior had never been contemplated, let alone executed). Second, the fever: the 5-percent target is so politically load-bearing that SIPRI’s researchers are already warning about creative accounting to hit it — Italy reportedly floated counting a bridge to Sicily as defense-related expenditure. When governments start relabeling civil infrastructure to look more armed than they are, you are past a spending increase and into a status competition. Nobody gerrymanders their books to appear more militarized unless militarization has become the currency of standing.

The pattern is hemispheric, not regional. Asia and Oceania rose 8.1 percent to $681 billion, the region’s largest annual increase since 2009. Japan’s spending rose 9.7 percent to its highest share of GDP since 1958; Tokyo lifted its postwar ban on lethal weapons exports and promptly signed its first warship deal, building frigates for Australia. Taiwan posted its largest increase since at least 1988. Australia and the Philippines are climbing the same curve. And SIPRI’s own researchers, careful people not given to editorializing, attributed the allied surge in Asia not only to regional threats but to “growing uncertainty over US support.” The Commonwealth, the Eurozone, the Pacific treaty allies — different oceans, identical diagnosis.

Which brings us back to Halifax — because the submarine announcement was not a decision. It was the last step of a sequence. Last year, Canada signed a Security and Defense Partnership with the EU, the diplomatic framework that unlocks participation in European joint procurement. On June 15 of this year, Ottawa concluded the follow-on agreement admitting Canadian industry into the SAFE procurement system itself. Three weeks later, Carney picked the German boats over South Korea’s Hanwha — whose shares fell 23 percent on the news. Read the order of operations: first the treaty framework, then the industrial access, then the anchor purchase that makes Canada a customer, a supplier, and a stakeholder of the European defense bloc in one stroke. Governments do not stumble up that staircase by accident. Carney said the deal would open opportunities for Canadian firms inside European supply chains — which is the return path on the entry fee. Translate from the diplomatic: Canada, the country that shares the world’s longest border with the United States and whose military has been functionally integrated with America’s since 1940, was not shopping for submarines. It was executing an accession — wiring itself into a European defense-industrial bloc, deliberately, step by step. Not because Europe is closer. Because Europe is predictable.

One more Canadian detail deserves precision, because sloppier commentary will get it wrong. These are not nuclear submarines. The Type 212CD is a diesel-electric boat with air-independent propulsion — quiet, long-endurance, conventionally armed. Canada is not going nuclear (yet) and perhaps doesn’t need to. That is precisely the point. You do not need to break the nuclear taboo to hedge against Washington. You need hulls, missiles, shells, and supply chains that do not route through a country whose promises you can no longer price. The hedge is conventional, legal, alliance-compatible — and enormous.

The Arsenal Ran the Math Too

Now the American side of the ledger, because the same erratic behavior that is driving allied demand has simultaneously created a demand shock inside the United States.

The war with Iran — thirty-nine days of high-intensity air and missile combat this spring, more than 12,000 targets struck, a ceasefire in April, and a president who declared that ceasefire “over” in mid-July — consumed the American munitions stockpile at a rate no peacetime planner ever budgeted for. The Center for Strategic and International Studies, working from Pentagon budget documents, estimates the campaign expended at least half of the THAAD ballistic-missile interceptor inventory, nearly half of the Patriot interceptors, roughly 30 percent of the Tomahawk stock, and 45 percent of the Precision Strike Missile inventory — a system so new it had barely entered production. The Washington Post reported that the United States fired more advanced interceptors defending Israel than Israel itself did.

Against that burn rate, the refill rate is a rounding error: current delivery schedules put replenishment at roughly 15 new Tomahawks and 20 new Patriot interceptors per month, with zero THAAD deliveries forecast for all of 2026. CSIS puts the rebuild at three or more years just to return to pre-war levels — for the war we were already supposed to be equipped for, never mind a Pacific contingency. The strain is no longer theoretical enough to hide: the Navy paused $14 billion in approved weapons deliveries to Taiwan on the grounds that the munitions were needed for the Iran fight.

Follow the money that follows the shortfall. The administration has requested about $1.5+ trillion for defense in fiscal 2027 — the largest request in American history — on top of a $450 billion reconciliation package moving through Congress and a war supplemental everyone in Washington treats as inevitable (at about $70 billion). The president has personally convened defense CEOs and announced they would quadruple production of what he called “exquisite” weaponry. The Army wants 857 THAAD interceptors in a single budget year. Lockheed Martin is spending $9 billion through 2030 on more than twenty new production facilities.

So hold both halves of the picture at once. Allied demand for non-American weapons is exploding because Washington is unreliable. American demand for American weapons is exploding because Washington’s unreliability was voluntarily acted out with live ordnance over Iran, and now the magazine has to be refilled — and then expanded, because the new stated posture requires a military that can coerce without partners. If allies are freeloaders and every relationship is a shakedown, you need a force that projects power alone. That force runs on exactly the munitions we just expended with no ROI.

Either fork of American policy — restored alliances or armed unilateralism — is a defense-spending program. There is no fork where the money goes down.

Who Sells to a World That Trusts No One

A world that rearms must buy from someone, and the supplier map has a shape worth naming.

Russia will keep arming itself — $190 billion in 2025, 7.5 percent of GDP, the highest share of government spending ever recorded there — but a war economy at full draw is a customer, not a vendor. Its export book is collapsing under the weight of its own consumption and its clients’ battlefield observations.

China is the interesting case, and the place where the lazy version of this analysis goes wrong. Beijing raised spending 7.4 percent to $336 billion — the thirty-first consecutive annual increase, the longest streak on record — and Chinese firms will absolutely sell into this boom: drones, munitions, electronics, systems for the Global South and for the margins of European inventories. What China will not capture is the center. No serious state stakes its survival on the keystone systems — the submarines, the fifth-generation aircraft, the integrated air defense — of a supplier that is itself the pacing threat of half its potential customers, whose flagship platforms remain unproven in combat, and whose design lineage everyone in the business can trace. Watch what states do rather than what vendors claim: when Canada wanted boats it could bet the Atlantic approaches on, the shortlist was Germany and South Korea. Beijing was never in the room. Europe’s 35-percent foreign-component cap fences out Chinese content as surely as American. I would be shocked if China (or Russia) captured much of anything in the new “arms demand boom” from America’s former partners.

Which reveals the real structure of the coming market. It is not one arms race. It is three concentric ones. The great powers arm against each other. The equity markets have already scored it: in 2025 alone, Rheinmetall rose 154 percent, ThyssenKrupp 215 percent, Hanwha Aerospace 193 percent on top of a 154 percent gain the year before. Capital is not confused about what is happening. Capital saw the underwriter walk off the job before the foreign-policy commentariat did. Europe arms because the guarantees lapsed — and they buy from the suppliers whose reliability is their entire sales pitch: Germany, France, South Korea, Japan, Sweden, and each other. American industrial base arms to refill and then to double, selling to its own government first and to whatever allied demand it hasn’t yet alienated, along with two-bit dictatorships America suddenly realizes it might need. China and Russia continue to do “you” and sell at the margins.

Why the Boom Outlives the Man

Here is the claim that makes this an analysis rather than a news roundup: this is not a spending cycle. Cycles reverse when the stimulus fades. This has the three properties of a structural shift, and each one is a lock.

First, the demand is politically locked. NATO’s 5-percent target runs to 2035. Germany’s 3.5-percent pledge runs to 2029. SAFE loans disburse through 2030 and are repaid over forty-five years — Europe has mortgaged its rearmament on a timeline longer than its current politicians will live. Canada’s first hulls arrive in 2034; the last, sometime in the 2040s. These are not budget lines that lapse with a news cycle. They are treaty targets, loan covenants, and thirty-year industrial programs. Europe has entered the chat with a true (new) military-industrial complex. That won’t go away because a President leaves office.

Second, supply is constrained, which turns spending into a durable boom rather than a one-time purchase. Precision munitions cannot be surged like rifle cartridges. Japan needed three years to build a single Patriot factory; Germany has been standing up its own line since 2022 and has yet to produce a missile. Seventeen countries use Patriot and all of them want interceptors at once, which is why allocation decisions are already generating friction between Washington and its partners. When demand is contractually locked for a decade, and supply takes years to expand, you do not get a spike. You get a plateau at altitude.

Third — and this is the perverse one — the buildout is redundant by design. In the old system, one American production line served thirty countries, and the efficiency was the point. In the new system, every bloc builds parallel capacity precisely because depending on someone else’s factory is the vulnerability everyone just learned to fear. Europe builds European lines. Japan and Korea build Pacific lines. America rebuilds American lines. The duplication is not waste to be optimized away; the duplication is the product. Trust was the technology that made the old efficiency possible, and thanks to US actions, trust is the one component nobody can manufacture on an accelerated schedule. So, in the short and medium term, each region is going to deploy its own long-term capital. (As Keynes famously quipped, in the long term, we’re all dead. So let’s hope the long term is cheerier than that.)

All of this is why the boom survives its author. Suppose the next administration is the most alliance-devoted in American history. It flies to Brussels, reaffirms every article of every treaty, means every word. Kisses the French Prime Minister. Goes before NATO and weeps like Jerry Falwell. What changes? Ottawa does not cancel submarines arriving in 2034 because Washington sounds nice after four years of threatening to end its sovereignty — the entire lesson Canada just internalized, at the largest price in its history, is that a four-year signature cannot secure a forty-year need. The EU does not repeal a loan instrument with a 45-year tail after staring down Russia because its former guarantor went out on a bender for half a decade. Germany, Italy, Spain, and the Scandinavian countries will not unbuild factories because the new President doesn’t threaten their sovereignty daily. France needed forty-three years to walk back a hedge triggered by a hypothetical, and it kept the weapons anyway. Rearmament is a ratchet: political will opens the spending, but industrial concrete, treaty targets, and burned trust prevent it from closing. The insurance analogy holds to the end — once the homeowners have bought their own fire trucks, they do not sell them back because the insurance company hired a friendlier adjuster.

Recognizing that most people never took an international relations course, let me help you understand what this means for the ordinary American, because the costs do not stay abroad.

A tripolar world without a reliable underwriter is historically the least stable configuration the state system produces — more players, more miscalculation surfaces, fewer automatic stabilizers. The entire “cause” of the First World War is blamed on multipolarity and failures in coupling/partnering/communications (read Barbara Tuchman’s The Guns of August). Multipolarity will be a radical departure from the safe, stable, and largely prosperous world of “benign unipolarity” that followed the stability of “Cold War” bipolarity.

The Fed chairman noted this month that the Iran war has pushed inflation to a three-year high, with officials penciling in rate hikes to cool it — the guns-and-butter bill arriving on your grocery receipt and your mortgage rate. A trillion-and-a-half-dollar defense budget is a claim on the same treasury that funds everything else you were promised. And the deeper cost is the one Canada just priced: the discovery that guarantees you organized your life around can be repriced overnight, without your consent, by temperament. States are answering that discovery by building options that don’t depend on anyone’s mood in Washington. It is the same math sovereign families have started running on their own lives — which is exactly why the ones paying attention are building jurisdictional optionality now, while it is still a choice rather than a scramble.

For now, hold the image from Halifax. A prime minister announcing the largest weapons purchase his country has ever made, boats that will not touch water until the mid-2030s, bought from Germany, financed for a generation — then boarding a plane to a NATO summit hosted in Ankara. Nothing about that scene was aimed at the current occupant of the White House. Canada is not arming against this administration. It is arming for the world this administration made possible.

That is the tell. Every capital on earth watched the same behavior you did. They stopped arguing about whether it’s a doctrine.

They started buying fire trucks.

Two days out. The first live Q&A for The Long Memo happens Saturday, July 25 at 10:00 AM CDT — open to every reader, free or paid, no gate, no recording substitute. If you’re there, you’re in the room.

The question queue is closed. You gave me enough — closing it now beats skimming forty questions in an hour. If yours didn’t make the cut, odds are someone else already asked the same thing a different way.

No script. No approved topics. Just the memo, live, answering to the people who read it.

Saturday, July 25 — 10:00 AM CDT — live, open to all readers on Zoom.

See you there.

As you know, Substack has partnered with a company called Pangram to flag and label writing it believes was produced by artificial intelligence.

I wrote about this before it arrived. Funny thing about that piece. It’s been read by a lot of people (close to a million). It’s been shared, argued over, and it’s still collecting comments. More than once, someone has sent it to me — you should read this.

Yeah. I know. I wrote it.

Let me remind you of something. That article, that people shared, loved, commented, hated, shared, and so forth? When I started here, I didn’t tell you who I was. I gave you almost nothing to go on. I wrote under a different name entirely, and the work had to survive on its own. No personality to trade on. No audience marched over from some collapsing masthead. No journalist’s grievance story about telling his editors where to go. Just words on a screen, and whether they were worth your time.

Four of those pieces have been read more than half a million times each. It wasn’t luck. My top 20 articles each have over a quarter of a million views. Each.

Nor should it surprise anyone. I have been trading words and ideas for money for thirty years. I built this the only way I know how — by putting an idea in front of you and letting you decide what it was worth.

And now a software company is going to sit in judgment on how I make it.

The defense is always the same. It’s here to stop the slop.

Fine.

Here’s a thought: if the piece is worthless, don’t read it.

I confess I’m endlessly amused by the reader who makes it through all seven thousand words and then complains. I’m sorry — are you looking for a refund? Nobody held a gun to your head and demanded you suffer through it. You could have stopped at any sentence and gone on with your life.

Agency can do a lot for the soul.

But apparently, according to the machine, no writer may organize his thoughts in logical order. No writer may use correct punctuation. No writer may actually command the language he’s writing in.

Do all three, and you must be a robot.

Here is what these detectors are actually measuring, as far as I can tell. Left alone, an AI writes to the strict rules of the language — clean structure, orderly punctuation, disciplined sentences. What the detector flags is precisely that discipline. It is not asking whether a mind made this. It is asking whether the prose is messy enough to look unsupervised. The tool rewards sloppy writing and punishes careful writing, and calls the gap between them “human.”

That is not a measurement of authorship. It is a measurement of how much noise is in the signal.

And understand what the mark is built on. The premise is that anything AI touched is unworthy by definition — because a machine was somewhere in the room.

A machine has no thoughts.

It has no lived context.

It is not alive.

So the slop is slop for one reason: it is bad thinking, flawlessly executed. The polish was never the problem. The thinking was. And no detector on earth can tell you whether the thinking is any good. That was always your job. It is still mine.

(The em dash, I’m told, now belongs to the robots. The purists can pry it out of my hands.)

Let me tell you plainly how I work. I use AI every day, for hours — Claude, ChatGPT, Perplexity. I use them to make models, develop advertising and websites, create explanatory graphics. I use them as my staff. I use it to build things that are genuinely useful. The Fuckupometer was made with Claude Code. I hope you enjoyed it. A lot of you did.

Do I use it to write?

I use it the way every serious publication uses an editor. It stress-tests the argument. It sharpens the sentence. It tells me where the logic is thin. And when it puts a line on the page, I do to that line exactly what I do to everything that carries my name — I edit it. Ask the guest writers who’ve appeared here. I revise their work too. Hopefully to make it better.

A dirty secret from a life spent writing for publication: I am about as good as it gets, and my work gets edited. Full stop.

I stopped taking it personally a long time ago.

Editors have discretion. Editors change your work. The only difference between a good writer, a bad writer, and a published writer is how much revision the first two will accept before the thing goes out the door.

That is how I use AI most. As editor. As stress test. As the steel that sharpens the blade. The ideas are mine. The judgment is mine. The decision about whether a sentence has earned its place is mine, and I make it, every time, on every piece.

If that troubles you — stop reading. There is no gun to your head.

Pangram will start pinning its scarlet letter to my work. Fine. They can do as they like. But I’d ask you to hold their machine to the same standard you’d hold anyone who claims to detect the truth to.

When I was in the intelligence community, I sat for a fair number of polygraphs. I’ll be honest with you — I never once knew what was actually being measured. I’ve taken the IQ tests too. Allegedly, I’m a genius. My family stands ready to supply counterevidence.

There has only ever been one test that mattered here. Do you find the work useful?

If you do — read it, share it, and I’d be grateful for your subscription.

If you don’t — stop. Same as it’s always been. No gun to your head.

I’ve decided to ignore Pangram.

And on the day a scanner’s verdict starts coming between you and the work, I’ll take the work somewhere the only test that applies is the one that was ever real. Yours.

For now, that’s where I stand.

The rest is up to you.

P.S.: I asked the Oracle of Substack/Pangram to rate my article before I posted.

Jury’s still out, I guess. I feel metaphysically in limbo. Am I me? Am I real?
WHO AM I!?

I guess I’ll have to wait for Substack’s version of “Degree Absolute!”

Now hear the word of the Lord.

You Prisoner fans will hopefully get it.

Be seein’ you.

I went through the documents dumped on the public Thursday night.

They are interesting.

They are also not remotely what Donald Trump, the White House, or the intelligence community macaroni artists surrounding him claim they are.

The White House says the documents prove that Americans were “blatantly lied to” about election security, that China acquired 220 million American voter records, and that the intelligence community concealed foreign election interference. The documents themselves establish nothing remotely resembling that narrative. In several places, they say practically the opposite.

That is the first thing to understand about this dump:

The political claims are far more conclusive than the intelligence.

Much of what was released is not “intelligence” in the way normal people (or even community professionals) would understandably use that word. It is a mixture of raw reporting, investigative leads, analytic products, threat assessments, internal disagreements, descriptions of theoretical vulnerabilities, and retrospective memoranda written long after the events in question.

Those things are not interchangeable.

An FBI FD-302, for example, generally memorializes an interview. It tells you what somebody told the FBI. It does not certify that the statement was true, corroborated, or even sane. A 302 saying that PERSON A heard from PERSON B that PERSON C believed something happened is evidence that PERSON A made that statement.

It is not necessarily evidence that the underlying thing described actually happened.

This distinction appears to have escaped the people who packaged this dump, perhaps because epistemology is not currently available in crayon.

Then there are what might generously be called “intelligence-community products”: short assessments, threat papers, working-level memoranda, and finished products that often resemble the USA Today of classified reporting. They contain broad (and largely unsubstantiated) judgments about intentions, preferences, capabilities, and possible future activities, sometimes written by components that are not especially well-positioned to make the claim being offered.

Such products can be useful. They may identify an issue worth watching. They may also merely tell senior officials that a foreign government possesses computers, dislikes an American politician, and might someday attempt something unpleasant.

That is not nothing.

It is also not evidence that an election was stolen.

What the Documents Actually Establish

To the extent the dump contains substantive and reasonably supported conclusions, it reinforces several things we already knew.

First: Russia interfered in American elections and preferred Donald Trump.

Across the documents, including the supposedly damning new material, the conclusion remains remarkably consistent: Russia engaged in influence operations directed at American elections, and those operations generally favored Donald Trump.

This should not be controversial. Russia’s activities included information operations, propaganda, cyber intrusions, laundering material through intermediaries, and amplifying claims designed to undermine confidence in American institutions.

There was a Special National Intelligence Estimate released that made an unequivocal conclusion that Russia attempted to interfere in the 2016 election.

The new dump does not erase that conclusion.

It repeatedly reinforces it.

The irony is almost architectural. The administration assembled a document collection intended to discredit previous intelligence judgments about foreign interference, only to publish another stack of intelligence suggesting that Russia interfered and wanted Trump to benefit.

That is less a bombshell than a bureaucratic pratfall.

The documents also discuss Chinese collection activity and China’s preference for Trump's defeat in 2020. But one of the featured assessments reportedly concluded that Beijing, despite preferring Trump’s defeat, “did not intend to try to affect the election.” Another CIA product said China targeted political campaigns for intelligence collection while assessing that Beijing did not then intend to interfere covertly to sway the outcome.

A foreign government having a preference is not election interference.

Every government on Earth has preferences regarding American presidents. So do currency traders, oil ministers, defense contractors, newspaper editorial boards, and the slightly drunk man at the end of your neighborhood bar.

Preference is not activity.

Activity is not operational success.

Operational success is not electoral effect.

Yet the entire presentation depends upon sliding noiselessly among those categories as though they all mean the same thing.

They do not.

Second: Nothing in the dump demonstrates that foreign activity changed the result.

There is no document establishing that foreign interference changed enough votes to alter the outcome of either the 2016 or 2020 presidential election.

There is no intelligence-community assessment that says:

Russia’s influence operation moved 43,000 votes in these three states.

There is no assessment saying:

Chinese access to voter-registration data caused these voters to be removed, these ballots to be fabricated, or these tabulations to be altered.

There is no causal chain connecting a foreign preference, foreign collection activity, a theoretical cyber capability, and the certified election result.

There is no pickle.

The 2021 intelligence-community assessment concluded that no foreign actor attempted or succeeded in altering the technical processes of the 2020 election, including voter registration, ballot casting, vote tabulation, or the reporting of results. That assessment was produced while John Ratcliffe, now Trump’s CIA director, was serving as director of national intelligence.

The newly released documents do not overturn that conclusion.

One of the documents reportedly says explicitly that vote-tabulation systems would be difficult to manipulate on a scale large enough to compromise an election result. It notes that audits, paper records, and the decentralized character of American elections make widespread undetected manipulation difficult.

That is not an assertion that election systems are invulnerable.

No serious cybersecurity professional would make such an assertion about anything containing electronics, software, human beings, or a password selected by a county employee named Dale.

It is an assertion about scale, resilience, and detectability.

The United States does not conduct one national election through one national computer. It conducts thousands of state and local elections through a sprawling collection of systems, procedures, vendors, ballot formats, registration databases, audit requirements, and reporting structures.

This creates inefficiency. It produces inconsistent standards. It can make election administration look like a yard sale organized by fifty state governments and several thousand counties.

It also makes a single, centralized compromise exceptionally difficult.

An attacker wishing to alter a presidential election would need to identify the decisive jurisdictions, penetrate the relevant systems, understand the specific equipment and processes in use, alter sufficient results, evade paper-ballot comparisons and audits, avoid detection by local and state officials, and do all of it without creating anomalies large enough to expose the operation.

Possible in theory is not the same as accomplished in fact.

The documents describe the former.

The White House insinuates the latter.

Third: The voter-roll hysteria is mostly analytical masturbation.

A substantial portion of the dump concerns foreign access to American voter information.

The White House says China acquired information connected to approximately 220 million American voters, including names, addresses, phone numbers, party preferences, and other personal data. It calls this “the largest compromise of election data in history.”

That sounds ominous because it has been deliberately phrased to sound ominous.

But much of the information contained in voter files is public or commercially obtainable. Political campaigns, consultants, data brokers, advocacy organizations, journalists, researchers, and foreign intelligence services can acquire substantial voter information without penetrating a voting machine, changing a registration, or altering a single ballot.

Reuters reported that officials familiar with the intelligence said the data was not confidential and could not itself be manipulated. Nextgov’s review similarly concluded that the documents show Chinese collection and analysis of voter information, but do not show that China altered voter rolls, manipulated ballots, or changed the outcome.

This distinction matters.

Obtaining a list of voters is not the same thing as obtaining the ability to vote for them.

Possessing names and addresses is not the same thing as modifying registration records.

Studying political preferences is not the same thing as changing ballots.

The Junta’s syllogism appears to be:

Step One: China obtained American voter data.

Step Three: Donald Trump lost the 2020 election.

That leaves an awkward question:

What the hell is Step Two?

There is no Step Two.

There is innuendo where Step Two should be.

No operational mechanism is identified. No altered registrations are demonstrated. No forged ballots are traced to Chinese actors. No manipulated vote totals are documented. No state certification is shown to have been corrupted. No mathematical analysis connects the acquisition of voter information to the election result.

It is the Underpants Gnomes theory of intelligence analysis, except even the Underpants Gnomes admitted they had not worked out the profit stage.

Capability Is Not Evidence of Use

The White House highlights assessments saying that Russia, China, Iran, North Korea, and some non-state actors possess the capability to compromise portions of American election infrastructure. It also emphasizes that centralized election-related databases, including voter-registration systems and electronic poll books, may be vulnerable to exploitation.

That is a legitimate security concern.

It is also profoundly unsurprising.

China can compromise American corporations, government agencies, telecommunications systems, health insurers, personnel databases, and critical infrastructure. Russia can conduct sophisticated cyber operations against government and private networks. Iran and North Korea also maintain capable cyber organizations.

Naturally, those countries might be capable of compromising a poorly secured county database.

But an intelligence judgment that an adversary could exploit a system is not intelligence that it did exploit the system.

A capability assessment answers:

Could they do it?

An activity assessment answers:

Did they attempt it?

An attribution assessment answers:

Who was responsible?

A damage assessment answers:

What did they successfully accomplish?

An electoral-impact assessment answers:

Did that accomplishment change the result?

The dump contains a great deal about the first question.

It contains fragments concerning the second.

It contains almost nothing establishing the fourth.

It contains nothing conclusive regarding the fifth.

The political presentation simply pours all five questions into one bucket, shakes vigorously, and labels the resulting sludge PROOF.

What Real Intelligence Would Look Like

Like any large organization, the intelligence community contains brilliant people, competent professionals, incurious bureaucrats, résumé-polishing mediocrities, and dullards who could lose a battle of wits with a conference-room ficus.

All of them produce documents.

Those documents circulate through classified systems. SIPRNet carries Secret-level information across the Defense Department. JWICS carries Top Secret and compartmented intelligence. A-Space is an analytic collaboration environment used within the intelligence community, rather than a separate classification network.

And, as anyone who has worked inside those systems knows, classification is not a quality rating.

A document can be Top Secret and still be stupid.

Classification tells you how damaging unauthorized disclosure could be. It does not tell you whether the author can reason.

Real intelligence supporting the administration’s implied claim would look something like this:

(TS//SI-G//NF//ORCON) Intercepts collected under PROGRAM NAME indicate that Chinese Company X coordinated with PERSON 1 and PERSON 2 to obtain access to vote-tabulation transmissions sent by County A to the office of the State Secretary of State. The intercepted communications described a method for modifying transmitted totals while concealing the changes from the receiving system.

That would be intelligence.

It identifies an actor.

It identifies an operation.

It identifies a target.

It identifies a method.

It connects capability with activity.

Depending upon the reporting, it might also establish intent and operational progress.

A declassified version might look like this:

(U) Intelligence reporting indicates that a company associated with the Chinese government worked with other individuals to explore intercepting electronic vote-tabulation data transmitted to a state election authority. The reporting described a possible method for altering the transmitted information.

The source, collection program, technical details, identities, and contextual information would be removed.

A knowledgeable reader might infer that the information came from signals intelligence. The reader might infer that the government had access to communications involving at least one participant.

But the reader could not identify the source, reconstruct the collection system, or determine exactly what was intercepted.

Still, the public would learn something meaningful:

The United States possessed specific intelligence indicating that identifiable people conducted identifiable activity against an identifiable election system using an identifiable method.

That is not what this dump gives us.

Instead, it gives us assessments that foreign governments possess cyber capabilities, reporting that China collected or analyzed voter information, old investigative leads, speculative vulnerabilities, and arguments among analysts over whether foreign conduct should be called “influence,” “interference,” “collection,” or something else.

Those debates can matter inside the intelligence community.

They do not prove that Donald Trump won an election he lost.

Why Declassification Is Difficult

I have worked on matters in which classified information needed to be prepared for public release.

It is hard.

The government may possess excellent intelligence, but releasing the intelligence can reveal how it was obtained. Even removing an individual’s name may not protect the source if the foreign government can determine who attended the meeting, who received the document, or who had access to the information.

Foreign intelligence services read declassified documents closely. They compare wording, timestamps, distribution markings, prior reporting, known personnel, communication patterns, and operational events. They attempt to reverse-engineer sources and methods from whatever remains.

Large language models presumably make that process faster. Feed thousands of releases into a capable system and it can identify recurring linguistic structures, likely collection disciplines, redaction patterns, relationships among code words, and correlations that human analysts might take months to notice.

So I am not unsympathetic to the difficulty.

History supplies a brutal example. During the 1990s, press reporting disclosed that Osama bin Laden communicated through satellite telephones. After learning that the United States could monitor those calls, bin Laden sharply curtailed his use of them.

That is the danger.

One careless disclosure can extinguish a collection stream.

A source may disappear.

An adversary may change systems.

An intelligence advantage developed over years can vanish because somebody wanted a dramatic sentence in a magazine article.

Responsible declassification therefore leaves the public with an unavoidable problem. Officials want to disclose enough to support a conclusion without disclosing enough to reveal the collection behind it.

Sometimes what remains is fragmentary.

Sometimes it is frustratingly vague.

Sometimes the government must effectively say:

We know this happened, but we cannot show you everything that allows us to know it.

That is uncomfortable in a democracy, but it is occasionally unavoidable.

None of that excuses what happened here.

The problem with this dump is not simply that important details were redacted.

The problem is that the visible material does not logically support the claims being made about it.

Raw Reporting Is Not a Finished Conclusion

The FBI, CIA stations, military commands, embassies, and other government offices receive information every day from walk-ins, informants, foreign officials, liaison services, contractors, political actors, cranks, opportunists, and people attempting to settle disputes with former romantic partners.

Some of it is actionable.

Much of it is worthless.

When I worked in the puzzle palace, we routinely received reports from FBI field offices describing claims along the lines of:

A person contacted the FBI and reported that her former boyfriend knew somebody associated with a local Islamic organization, and that individual said al-Qaeda planned to attack TARGET X.

Pick a target.

A shopping mall.

A bridge.

A sports stadium.

A nuclear plant.

The report would be ingested and checked against other information. Analysts and investigators would attempt to determine whether it was credible, specific, corroborated, and actionable.

Broadly, it would fall into one of three categories:

One: It is credible and specific. Investigate the ever-loving shit out of it.

Two: It is credible or specific, but not both. Conduct follow-up inquiries and look for corroboration.

Three: It is neither credible nor specific. Retain it, index it, and move on unless later information gives it significance.

The existence of the original report does not establish the existence of the plot.

It establishes that somebody reported a plot.

This is why intelligence analysis exists.

Raw reporting is compared with other reporting. Sources are evaluated. Access is assessed. Claims are corroborated or contradicted. Alternative explanations are considered. Confidence levels are assigned. Collection gaps are identified. Analysts distinguish what is known from what is judged and what is merely possible.

The Junta’s dump reverses that process.

It takes raw or tentative information that did not mature into a conclusive assessment and presents its mere existence as evidence that the earlier analytic conclusion was dishonest.

That is intellectually backwards.

An intelligence agency can investigate a report and ultimately conclude that the report was unsupported. The investigation does not prove the report. It proves the government investigated it.

Similarly, analysts can disagree over how to characterize Chinese behavior without any party claiming that China changed votes. One analyst may call political data collection an influence-related activity. Another may reserve “election interference” for covert action intended to alter an outcome. That definitional dispute does not create evidence of ballot manipulation through spontaneous bureaucratic generation.

The Documents Do Not Reveal a Cover-Up

The administration’s larger insinuation is that intelligence officials suppressed evidence because they wanted Joe Biden to win and Donald Trump to lose.

But the released material primarily reveals what every functioning intelligence organization contains:

Disagreement.

Preliminary reporting.

Competing terminology.

Collection gaps.

Varying levels of confidence.

Different agencies examining different portions of the same problem from different institutional perspectives.

The fact that an item did not appear in the final assessment does not mean it was concealed.

It may have been uncorroborated.

It may have been irrelevant to the analytic question.

It may have concerned collection rather than interference.

It may have described a capability without evidence that the capability was used.

It may have been overtaken by better reporting.

It may simply have been bullshit.

A finished assessment is not supposed to reproduce every report ever received. If it did, the President’s Daily Brief would arrive in a wheelbarrow and the president would still not read it.

The analyst’s job is to discriminate.

To weigh.

To synthesize.

To tell policymakers not merely what somebody claimed, but what the intelligence community has reason to believe.

One can question whether the analysts made the correct judgments. One can investigate whether dissenting views were fairly represented. One can examine whether political pressure distorted the process.

But proving politicization requires more than finding a discarded report and shouting that it existed.

You must show that the report was credible, material, and deliberately excluded for an improper reason.

This dump does not do that.

Why Do This at All?

That leaves the obvious question:

Why release this material now?

Why take a disorderly pile of raw reports, threat assessments, analytic disagreements, cyber warnings, retrospective memoranda, and speculative fragments, then present it as though the Rosetta Stone of American election fraud had finally been excavated from a filing cabinet at Langley?

There are several possible motivations, and they are not mutually exclusive.

The first is the simplest:

Donald Trump needs the 2020 election to remain permanently unresolved.

Not legally unresolved. Not factually unresolved. Psychologically and politically unresolved.

The foundational myth of Trumpism is not merely that Trump is a successful political leader. It is that Trump cannot legitimately lose. If he loses, the system must have cheated. If a court rejects his claim, the court must be corrupt. If an election official certifies the result, the official must be compromised. If his own advisers tell him the evidence is worthless, they must be disloyal.

The stolen-election narrative is therefore not an incidental grievance.

It is regime theology.

It explains defeat without requiring failure. It converts humiliation into martyrdom. It transforms political opponents into criminals and supporters into victims of an occupied government.

Most importantly, it absolves Trump of the one fact his personality appears constitutionally incapable of absorbing:

More Americans voted for somebody else.

The intelligence dump serves that mythology without needing to prove anything. It need only create enough smoke, acronyms, redactions, foreign names, and classified markings to let supporters conclude that something sinister must exist behind the black bars.

That is why incoherence is not a defect.

It is the product.

A clear document can be evaluated.

A confusing document can be insinuated.

The second motivation is institutional revenge.

Trump has spent years portraying the intelligence community, the FBI, the Justice Department, and the broader national-security bureaucracy as participants in a conspiracy against him. Some of those institutions made serious mistakes. Some officials exercised terrible judgment. Some investigations were conducted sloppily or presented tendentiously. Bureaucracies are perfectly capable of arrogance, self-protection, political bias, and monumental stupidity.

But Trump’s argument goes much further.

His claim is that institutions are legitimate only when they validate him.

An intelligence assessment that supports his position is honest.

An assessment that contradicts him is treason.

An investigation of his opponent is law enforcement.

An investigation of him is a coup.

The release, therefore, serves as retaliation against the officials and agencies that rejected his 2020 account. It allows the administration to accuse them of suppression, corruption, or betrayal without demonstrating any actual falsification of intelligence.

The accusation itself is the punishment.

Careers can be damaged. Security clearances can be reviewed. Former officials can be investigated. Current employees can be warned that analytic independence carries personal risk.

That message matters far beyond this particular controversy.

It tells the intelligence community:

Do not merely provide your best judgment.

Provide the judgment the President wants.

That is how politicization actually takes root. Not necessarily through an explicit order to fabricate intelligence, but through the steady education of the bureaucracy about which conclusions are rewarded and which conclusions bring subpoenas, public vilification, clearance suspensions, or professional ruin.

The third motivation is to manufacture retrospective justification for future action.

The administration has repeatedly framed election administration, voter registration, mail voting, voting equipment, and state certification as national-security problems. A document dump implying that foreign powers acquired voter information or explored election-system vulnerabilities can be used to justify federal intervention that the evidence itself does not support.

The argument becomes:

Foreign adversaries targeted election infrastructure.

Therefore, state and local election administration cannot be trusted.

Therefore, the federal executive must impose new controls.

That leap is not analytically valid, but it is politically useful.

It can support demands for centralized voter databases, expanded federal access to state election records, aggressive purges of registration lists, restrictions on voting methods, new identification requirements, or federal investigations of local officials.

The ostensible concern is foreign interference.

The practical result may be domestic control.

This is an old maneuver. Governments identify a genuine vulnerability, inflate it into an existential threat, then use the threat to obtain powers that would never be granted on the actual evidence.

Foreign intelligence services do seek information about American voters.

Election systems should be secured.

Neither fact establishes that the President should gain greater control over who remains registered, how votes are cast, or how state results are certified.

The fourth motivation is distraction.

A government can release documents because the documents matter.

It can also release documents because it wants the public discussing them instead of something else.

An intelligence dump creates the appearance of action. It produces headlines, congressional demands, cable-news panels, accusations of cover-up, and several days of people pretending that an FBI interview memorandum is the Dead Sea Scrolls.

The government controls the timing.

It controls the framing.

It controls which documents are released, which remain classified, which passages are highlighted, and which interpretations are supplied in advance.

That permits the administration to create a scandal-shaped object even when the underlying evidence is mush.

The media then performs its customary role in the pageant.

One outlet reports that Trump says the documents prove a conspiracy.

Another reports that Democrats deny the conspiracy.

A third assembles six professional arguers into boxes and calls the resulting noise analysis.

The evidentiary question quietly disappears.

Did the documents actually prove the claim?

That is tedious. It requires reading.

The accusation is much easier to televise.

The fifth motivation is to contaminate the historical record.

Propaganda does not always aim to persuade everyone.

Sometimes it aims to make certainty impossible.

The Russian concept of information warfare has long relied on flooding the environment with contradictory narratives, dubious documents, selective disclosures, and endless alternative explanations. The goal is not necessarily to make the public believe one story.

It is to make the public stop believing that any story can be established.

This dump operates in much the same fashion.

The official record says foreign actors did not alter the technical voting process in 2020.

The dump introduces fragments about voter files, cyber capabilities, foreign preferences, suspected access, suppressed reporting, and analytic disputes.

None of those fragments disproves the official conclusion.

But accumulated together, they permit people to say:

We may never know what really happened.

That sentence is a political victory.

Once factual resolution is replaced by permanent suspicion, the certified result becomes merely one narrative among many. Evidence loses its authority. Every conclusion becomes partisan. Every investigation becomes a cover-up. Every absence of proof becomes proof that the proof was hidden.

The sixth motivation is commercial and personal.

Grievance is the central economic engine of the Trump movement.

It raises money.

It sells subscriptions.

It generates television audiences.

It keeps supporters angry, frightened, and engaged.

A settled election produces no recurring revenue.

A stolen election can be monetized forever.

Every new disclosure becomes a fundraising email. Every redaction becomes evidence of concealment. Every disputed analytic sentence becomes a reason to send another contribution before midnight.

This is not peripheral to the political project.

It is the political project.

The grievance cannot be resolved because resolution would end the transaction.

Finally, there may be people involved who sincerely believe the claims.

That possibility should not be dismissed.

Human beings are quite capable of starting with a conclusion, searching a vast bureaucracy for anything that appears to support it, and gradually convincing themselves that the resulting pile constitutes proof.

The intelligence community produces enormous volumes of reporting. Given enough documents, one can find warnings, disagreements, unverified allegations, minority views, speculative scenarios, and statements that were later judged incorrect.

A motivated reader can assemble those fragments into almost any desired narrative.

This is especially easy when the reader does not understand the distinction between raw and evaluated intelligence, or prefers not to.

Some officials may genuinely believe that the presence of foreign voter data, cyber reconnaissance, and internal analytic disputes proves a broader conspiracy.

Sincerity, however, does not repair logic.

A person can be honestly wrong.

A government can be sincerely delusional.

And a document dump can be assembled by true believers while still serving the interests of cynics.

The most likely answer, therefore, is not one motivation but an ecosystem of motivations:

Trump needs vindication.

His allies want revenge.

Political operatives want an issue.

Ideologues want greater control over elections.

Fundraisers want recurring outrage.

Careerists want to please the President.

Bureaucratic dullards want their memoranda treated as historic revelations.

And foreign adversaries are undoubtedly delighted to watch the United States use their limited election activity as a pretext for destroying public confidence in its own democracy.

That may be the most perverse element of the entire exercise.

Russia’s larger objective was never simply to elect one person.

It was to make Americans believe that their institutions were corrupt, their elections were fraudulent, their leaders were illegitimate, and objective truth was inaccessible.

The Junta’s response has been to take that objective and turn it into domestic policy.

The foreign operation did not need to change the votes.

It merely needed Americans to spend the next decade claiming that somebody did.

The Actual Conclusion

Foreign governments collect information about American voters.

They conduct cyber operations against American institutions.

They possess capabilities that could threaten portions of election infrastructure.

Russia interfered in the 2016 election and conducted influence activities surrounding the 2020 election, generally in ways favorable to Donald Trump.

China collected large quantities of information concerning American voters and political attitudes. It preferred that Trump lose in 2020 but, according to the intelligence assessments included in or discussed alongside this release, did not then intend to covertly alter the outcome.

American election systems contain vulnerabilities and should be hardened.

None of this proves that votes were changed.

None of it proves that voter-registration records were altered to produce a particular result.

None of it proves that foreign activity caused Donald Trump to lose in 2020.

None of it reveals the missing Step Two.

The White House took a collection of intelligence fragments, cyber-risk assessments, investigative reporting, and bureaucratic disagreements and constructed a narrative that the documents themselves do not support.

The dump does not demonstrate a stolen election.

It demonstrates that foreign governments spy on us, that computerized systems have vulnerabilities, that intelligence reporting varies in quality, and that analysts sometimes disagree.

Stop the presses.

The Russians spy.

The Chinese collect data.

Computers can be hacked.

The intelligence community employs morons.

And Donald Trump lost the 2020 election.

After thousands of pages, dozens of documents, a presidential address, and enough classificatory theater to make a Pentagon briefer achieve spontaneous orgasm, those remain the facts.

Everything else is innuendo wearing a classification stamp.

You read the memo. Now you get to interrogate it.

On Saturday, July 25 at 10:00 AM CDT, I’m doing the first live Q&A for The Long Memo — open to every reader, free or paid, no gate. No script, no pre-approved topics. Whatever you’ve been chewing on from the last few months of coverage — institutional decline, the executive branch, the economy, where this is all heading — bring it.

Here’s the condition: this only happens if you give me something to work with. Send your questions to Aymara at aymara@borderlessmediallc.com or drop them in the comments on this post before Wednesday. If the questions don’t show up, the session doesn’t either. I’d rather cancel than sit in front of a camera improvising a monologue you’ve already read.

So don’t hold back. Ask the thing you’ve been meaning to ask since I wrote about permission decay, or decoherence, or the last thing that made you put your phone down and stare at the wall for a minute. Ask about the reasoning behind a call I made, or where I think I’ve been wrong. Ask what I’m actually watching right now that hasn’t made it into a post yet.

Saturday, July 25 — 10:00 AM CDT — live, open to all readers on Zoom.

Send your questions over. I’ll see you Saturday — assuming you do your part first.

Today, the Supreme Court handed down two immigration decisions, both 6-3, both written by Justice Alito, both — by an accident of docketing that no novelist would dare invent — captioned Mullin v. something. Mullin v. Al Otro Lado. Mullin v. Doe. The same name stamped on two different acts of subtraction, as if the machine had run out of variety and started reusing the label.

Read together, they tell you something more useful than either tells you alone. Not because the legal reasoning is novel. It isn’t. But because the instinct behind both — the reflex to find the cruelest reading the text will bear and call it plain meaning — is now the operating logic of an entire branch. And that branch is the last one to adopt it.

What the Court actually did

The first case is about a word. The immigration statute says a person may apply for asylum if they are “physically present in the United States” or if they “arrive in the United States.” For years, the second phrase did work: it covered the person standing at the port of entry, presenting themselves, asking for the process the law promises. The Ninth Circuit read it that way. Two judges who had spent their careers in immigration law read it that way.

Alito read it differently.

In ordinary speech, he wrote, no one would say that a person “arrives in” a place before the person enters that place. So the asylum seeker stopped on the Mexican side of the line — turned back by a U.S. officer standing on U.S. soil, told to wait in a country where waiting has gotten people kidnapped and killed — has not “arrived.” The statute neither entitles such an alien to apply for asylum nor requires an officer to inspect him.

This blesses a policy called metering, in which border agents turn away people who lack valid travel documents whenever they deem a port of entry full. The record in the case is not abstract. A CBP whistleblower testified that officers were instructed to lie to people — telling them the port was at capacity and turning them back — when, in fact, the facility retained sufficient capacity to process them. The government told the Court that metering remains a tool it would likely revive.

The Court said: fine.

The mechanism is the part to hold onto. The law’s duty to inspect and process is triggered by arrival. So if the government physically prevents arrival, the duty never loads. Sotomayor named it exactly: more people will die. She is not being dramatic. When the policy was last in effect, metering led to hundreds of cases of assault, kidnapping, and murder of people turned away at the border. Put a man at the door and then announce that no one has knocked.

The second case is about a different kind of door — the one marked “exit,” through which the courts are supposed to be able to look. Temporary Protected Status allows people from countries deemed too dangerous to return to legally stay and work for as long as the danger lasts. Haiti was designated in 2010 after the earthquake. Syria in 2012, because of the regime’s brutal repression of its own civilians. The administration moved to end both. Lower courts found the terminations were likely unlawful — that the Secretary probably violated the statute by ordering them without first consulting other agencies about actual conditions in those countries.

The Court’s answer was not “the conditions are fine.” The Court’s answer was that it is not allowed to ask. The text of the TPS judicial review bar, Alito wrote, clearly overcomes the presumption in favor of judicial review. Challengers cannot avoid that bar, he added, by creative pleading or clever lawyering. So the statute’s commands — shall consult, shall review country conditions — remain on the page, and become unenforceable in fact. Kagan’s dissent put it plainly: the majority blessed an assessment of conditions in two war-torn countries that rested on a single email. The State Department still lists both countries as “do not travel.” No part of Syria, it says, is safe from violence. None of that is reviewable now.

The tell

There was a moment in the TPS opinion worth pausing on, because it’s the whole thing in miniature.

The Haitian plaintiffs argued the terminations were racially motivated, and they had material to work with. Alito did not reproduce the President’s statements about Haitians in the opinion — a fact Kagan was quick to point out, describing the remarks as so repellent and racially inflected that the majority declined to put them in print. So she printed them herself. The line about Haitians “eating the dogs” and “eating the cats” and “eating the pets of the people.” Kagan wrote that the statements fairly shout, in their racial undertones and overtones alike, that race entered into the resolve to remove Haitians from the country.

The majority’s response was a small masterpiece of evasion. None of the statements was overtly racial, Alito wrote, and all expressed policy views that could rest on race-neutral justifications. Sit with the structure of that defense. It is not that the animus wasn’t there. It is that the administration was hostile to everyone — that it moved to strip protected status from thirteen of the seventeen countries that had it — and that a universal policy of doing this to all comers launders the specific contempt directed at one. If the cruelty is general, then the discrimination is invisible. Being horrible to everybody becomes a legal defense.

That is the reasoning of an institution that has stopped looking for the law’s purpose and started looking for its exits.

Why did these cases not need to be decided (this way, or at all)

Neither outcome was compelled. That is the part to sit with. A court that wanted to read “arrive” to include the person standing at the port, asking, could have — two appellate judges did, and the statute’s own structure, which lists “arrives” separately from “physically present,” fairly demanded it. A court that wanted to police whether an agency followed the mandatory steps Congress wrote could have done that while still leaving the final judgment call unreviewed. The cruelest reading is a choice. It was available, and it was selected twice in one morning by the same hand.

And here is the thing: the legal commentary will mostly miss it because it treats the Court as the prime mover. It isn’t. The Court is downstream.

The metering policy wasn’t invented on Thursday. It began at the end of the Obama administration, was expanded and formalized under the first Trump administration, and was rescinded under Biden. The contempt for TPS isn’t a judicial doctrine; it’s a governing program that had already stripped status from more than 600,000 Venezuelans before these two cases were decided. The posture — no compassion as a feature, the closed door as policy, the deliberate engineering of conditions so the law’s mercy never triggers — was built in the executive branch and ratified at the ballot box. The Court is not leading this. It is the last institution to fall in line, lending the prestige of the marble building to a disposition the political branches arrived at years ago.

This is what makes “the Supreme Court has gone rogue” the wrong frame. Rogue implies a break from the rest of the system. What happened Thursday is the opposite of a break. It is a convergence — the judiciary completing a circuit the other branches already closed. The same instinct that meters a frightened family at a port of entry, that ends protected status by email, that reads a statute for its loopholes rather than its commitments — that instinct now runs the length of the federal government, and the one branch designed to interrupt it has instead joined it.

What it leaves standing

Notice what both decisions carefully preserve. They do not abolish asylum. They do not repeal TPS. Asylum law still stands; the person who manages to cross can still invoke it. TPS still exists; the statute’s standards are still printed. The structures remain intact, fully legal, like the storefront left standing after the robbery.

That is the more sophisticated cruelty, and it is the signature of a mature system, not a crude one. You don’t need to repeal the protection. You need only to control the door in front of it — who arrives, who’s reviewed, who’s seen — and then insulate that control from anyone with the power to ask why. The law becomes a building no one is allowed to enter. The mercy stays on the books, where it can be cited, admired, and never reached.

Two cases. One name. The same hand. The Court didn’t start the fire on Thursday. It just stopped pretending it would ever pick up the hose.

What it could mean for birthright citizenship

I have maintained, since these civil-rights cases were first granted review, that the Supreme Court did not take them up because it wished to reaffirm what had been our law for decades, or, in some cases, centuries.

These two decisions make that presumption harder to dismiss.

So what might this mean in Barbara v. Trump?

Alito’s newer move is not always to say, bluntly, that prior cases were wrongly decided, and then manufacture some grand theory to justify blowing them up, as the Court did in Dobbs, obliterating Roe & Casey, or in Loper Bright Enterprises v. Raimondo, where Chevron deference was folded, spindled, mutilated, and replaced with judicial supremacy wearing a textualist nametag.

The move is more surgical now. The Court takes words everyone understood yesterday and announces, with the serene confidence of men who never pay the consequences of their own constructions, that those words never meant what everyone thought they meant.

So when Barbara was argued, and everyone rushed to declare, “The Court seems to accept the Fourteenth Amendment. Wong is safe,” my response was: maybe. But I doubt it.

These two cases make clear why that doubt was not paranoia. It was pattern recognition.

Barbara turns on a simple phrase: “subject to the jurisdiction thereof.” The settled understanding has been that anyone born on U.S. soil, within the constitutional reach of the United States, is a natural-born citizen, regardless of lineage or ancestry, except for narrow historical exceptions like the children of foreign diplomats or occupying enemy forces. That is the rule of United States v. Wong Kim Ark. That is the constitutional settlement most people thought had survived every major attack against it.

Trump’s argument is, in essence: no. The executive gets to decide who is truly “subject to the jurisdiction” of the United States. And if the executive gets to decide that, the executive gets to decide who is born a citizen.

The Supreme Court did not agree to take this case because it was eager to pin a boutonniere on Wong and send it happily into retirement. The comforting noises at oral argument should not be mistaken for commitment. This Court has become very good at preserving the shell of a right while draining its operative meaning.

That is what Barbara is really about. Not whether the Fourteenth Amendment still contains the words “all persons born,” because it does. Not whether Wong Kim Ark still sits in the U.S. Reports, because it does. The question is whether the Court will now do to citizenship what it just did to asylum and TPS: leave the promise standing while moving the key just beyond the reach of the people who need it.

If “arrive” can mean “not yet arrived” because the government blocked the door, and if mandatory statutory duties can become unenforceable suggestions because review is inconvenient, then “subject to the jurisdiction” can be made to mean whatever exclusion requires. The Court may not flip Wong outright. It may not need to. It can hollow it out, distinguish it, domicile it, footnote it into a museum piece, and then announce that birthright citizenship remains fully intact for everyone still lucky enough to qualify after the new machinery has finished sorting them.

That is the lesson of these two Mullin decisions. The right survives. The human being does not reach it.

And the marble building calls the result law.

There is a kind of institutional damage that announces itself, and a kind that you only detect later, by reading an instrument. A currency that stops being held in reserve. A treaty that stops being cited. A number, in a survey, that crosses a line it was never supposed to cross. The loud kind makes news. The quiet kind makes history.

This week, the instrument gave a reading. A Pew Research Center survey of 42,000 people across 36 countries asked which world leaders they trust to do the right thing in international affairs. Global publics now place the President of the United States below both the President of China and the President of Russia. Of six leaders measured, only Benjamin Netanyahu — mid-war — scores lower than the American head of state.

That is not a story about a man. It is a reading of a structure, and that structure has underwritten American power for eighty years.

What the number is actually measuring

A median of 34% of those surveyed express confidence in Xi Jinping. A median of 31% express confidence in Vladimir Putin. The American president draws 23%. Macron sits at 43, Zelenskyy at 35. The man holding the most powerful office on earth finishes fifth of six, trailing the leaders of the two governments Washington has spent a generation naming as its principal adversaries.

The question matters more than the ranking. This is not a favorability poll. It does not ask whether foreign publics enjoy the man, admire him, or would vote for him. It asks a narrow question — whether they trust him to do the right thing in world affairs — and on that specific axis of judgment, the office built to embody the credible alternative to authoritarian rule is now judged less credible than the authoritarians.

That distinction is the whole subject of this essay. Because credibility, not capability, is the load-bearing wall of American power.

The asset nobody put on the balance sheet

For most of the postwar period, the United States ran on two kinds of power. One was material — carriers, currency, the world’s deepest capital markets. That power is still largely intact; the carriers still sail. The other was something harder to see and impossible to manufacture on demand: the assumption, held by populations and not just governments, that when Washington said a thing it meant it, and that the American side was the side that kept its word. Call it the credibility reserve. It was never voted on, never appropriated, never line-itemed. It simply accumulated, the way trust accumulates, slowly and through repeated demonstration — and like every reserve, it could be drawn down.

This is the part the material analysis misses. You can measure a navy. You cannot measure a reserve of assumed good faith until the moment someone tries to spend it and finds the account lighter than expected. A security guarantee is only as good as the belief that it will be honored. An alliance is a promise that holds because both sides assume it holds. The dollar is paper that works because the world decided to trust the institution behind it. Every one of those instruments runs on the same invisible deposit — and the deposit is exactly what this survey just measured a withdrawal against.

The Soviet Union did not collapse because it ran short of tanks. It collapsed because the cost of compelling belief in the system finally exceeded the system’s capacity to generate it — because the people inside it, and the clients outside it, stopped assuming the center would hold and began quietly arranging their affairs around its failure. Hard power was never the binding constraint. Credibility was. By the time the instruments read empty, the structure was already hollow; the reading only confirmed what the architecture had stopped being able to hide.

This is recent, and that is the alarming part

If this were a slow erosion — a decade-long drift, a gentle reversion toward some historical mean — it would be a different essay, and a calmer one. It is not. One year ago, in Pew’s 2025 survey, the American president led both autocrats: Trump at 34%, Xi at 25%, Putin at 16%. Twelve months later the order has inverted. Trump fell to 23. Xi rose to 34. Putin climbed to 31. The lines crossed inside a single year.

The year-over-year figures are not a perfect like-for-like — Pew’s surveyed countries shift slightly between waves, and the medians move with them, so I would not stake a claim on the precise size of any single swing. But the direction does not wobble and the magnitude is not subtle. This was not a president slipping a few points within the normal band of foreign disapproval. This was a structure failing a threshold test — falling through the floor that had always separated the democratic side from the authoritarian one, with the men on the far side of that floor rising to occupy the space he vacated.

And it is not an artifact of the global median, dragged down by publics that were never friendly. Pew names the countries in its own text. In Germany, 16% trust the American president and 15% trust Putin — the same number, inside the margin. The German figure for Macron is 72%. In Canada, the oldest neighbor and one of the closest allies the country has, Putin draws 18% and the American president draws 20%. In Mexico he draws 11%, beneath even a wartime Netanyahu. These are not the adversary capitals. These are the treaty allies — the populations whose governments share intelligence with Washington, host its forces, and buy its weapons. They have placed the American president in the same credibility tier as the man prosecuting the largest war in Europe since 1945.

What a credibility inversion sets in motion

Here is the mechanism, because the mechanism is where the damage actually lives. Credibility is not sentiment that floats free of consequence. It is the precondition for cooperation, and cooperation is the thing that converts American intentions into outcomes. A government that no longer believes Washington is a reliable actor does not announce it. It hedges. It diversifies its security relationships, quietly. It slows the negotiation. It keeps its options open against the day the guarantee is tested and found wanting. Populations move first, governments follow on the lag democracies always run between sentiment and law, and by the time the policy changes the belief has already changed underneath it.

This is what it means for a reserve to be spent rather than destroyed. Nothing dramatic happens on any single day. The carriers still sail; the treaties are still technically in force. But the assumption that made all of it function at low cost — the background belief that the American side keeps its word — is being drawn down, and the people doing the drawing have just told a pollster what they think of the institution. The system that runs on assumed good faith starts having to pay, in friction and concession, for what it used to get for free. That is not collapse. It is something quieter and harder to reverse: the slow conversion of a trusted hegemon into a merely powerful one, which is a different and weaker thing.

The individual response to a structural drawdown is the same one every rational actor reaches when it stops believing the center will hold — you stop assuming, and you start arranging your affairs around the possibility that it won’t, which is the entire premise behind building jurisdictional optionality before you need it rather than after. But the individual hedge does not touch the structural fact, and the structural fact is the point of this publication. A reserve that took eighty years to accumulate is being spent in single-year increments, by the allies it was built to reassure.

The reading

Credibility is the one form of power that cannot be reconstituted by appropriation. You cannot fund it, legislate it, or deploy it. It is earned slowly and spent quickly, and once a population has decided that your word is worth less than your rival’s, the path back is measured in decades, not news cycles — if there is a path back at all.

The instruments are not ambiguous. The floor that separated the credible side from the other side held through Vietnam, through Iraq, through a great deal of American error that was forgiven because the underlying assumption survived. It did not hold through this. The reading is on the page, in the kind of number that does not make noise when it crosses the line — and the people who took the survey are not waiting for permission to believe what they already believe.

Source: Pew Research Center, “Trump Gets Negative Reviews Internationally as Fewer Say U.S. Is a Reliable Partner,” June 2026. Survey of 42,151 adults across 36 countries, conducted February 8–May 13, 2026. Leader-confidence medians and named country figures as reported by Pew; 2025 comparison drawn from the prior year’s Global Attitudes Survey.

A great deal is being reported about data centers. Protest about their construction. How they’re going to “suck up every drop of water” (not true). How they’re going to completely destroy the environment (unlikely).

This story isn’t about any of that. It’s about something you probably care about considerably more: money flying out of your wallet.

In the first half of 2025, American electric utilities asked regulators for more than 29 billion dollars in rate increases — double what they sought a year earlier — and the increases on the table would affect 40 million customers. The reason given, again and again, is the data center. Utilities are fielding interconnection requests that defy comprehension: at least seven hundred gigawatts of new connection requests in 2025 — more power than the entire country draws on average at any given moment.1

The grid is being rebuilt at speed to handle a wave of computing load that arrived faster than anyone planned. That much is true, and you can read it anywhere.

Here is the part you won’t read anywhere. Most of that load does not exist. The Energy Information Administration’s own framing is that many of these projects will never be built — the seven hundred gigawatts is a queue of requests, not a fleet of facilities, padded by developers filing in five places at once to hedge a single bet.

And it does not matter.

The requests alone are enough to move your bill, because the moment a utility decides to plan for speculative load, the transmission lines and substations and transformers go into the rate base, and the rate base is paid by you.

You are not subsidizing data centers. You are subsidizing the possibility of data centers. The grid is being financed against a forecast, and the forecast is being treated as a fact because the institution that exists to separate the two has stopped doing it.

And that’s happening whether a data center is built in your community or not.

A regulated utility is a bargain, and the bargain has terms.

Begin with what an electric utility actually is, because the answer is the whole story.

It is a state-sanctioned monopoly.

You cannot shop for a different set of wires to your house; the law gives one company the exclusive right to run them. In exchange for that monopoly — and this is the load-bearing clause — the company submits to a public utility commission that is supposed to do two things: guarantee the utility a fair return on prudent investment, and protect the captive customer from paying for anything that isn’t prudent or isn’t theirs to pay for. The PUC is the referee. The rate case is the mechanism. “Used and useful” and “prudently incurred” are the standards. The entire legitimacy of letting a private company own a public necessity rests on that referee actually refereeing.

The bargain was designed for a particular world. Load growth used to be slow, real, and local — a new subdivision, a new factory, demand you could see and meter and verify before you built for it. The cost-allocation machinery assumes that world. It assumes that by the time a utility spends your money on capacity, the load justifying the spend is something more than a line in a spreadsheet filed by a counterparty who can walk away. That assumption is now false. And a referee operating under rules written for a game that is no longer being played does not referee. It rubber-stamps.

The cost-shift is not a side effect. It’s the mechanism.

Watch how the money actually moves, because this is the part that never makes the headline. A hyperscaler approaches a utility with a request for power on a scale comparable to that of a small city. The utility, whose profit is a regulated percentage of what it builds, has every incentive to say yes and build — its earnings are its capital expenditure, and a data center is the largest capital-expenditure justification to come along in a generation.

So it files a rate case for the transmission, the substations, the generation. The grid upgrade is real and the steel is expensive — transformers and cables and towers are running up against tariff exposure and a record eighty-six gigawatts of new capacity planned for this year alone.

Someone has to pay for it. And under the default rules, “someone” is the residential and small-business ratepayer, whose bill quietly absorbs the cost of infrastructure built for a customer he will never be and a load that may never materialize.

The Union of Concerned Scientists put one slice of that loophole at over four billion dollars in connection costs shifted onto the grid’s captive customers. That is not a market failure. A market failure is an accident. This is the machine working as currently configured: speculative load, a monopoly incentive to build, a captive base to bill, and a referee applying the wrong standard.

The same system that was designed to prevent brownouts before power plants were built now functions to ensure that capital is allocated for private enterprise for electrical loads that might never be realized.

The states noticed. The scramble tells you the compact is broken.

The clearest evidence that the bargain has failed is the panic to rewrite it. In the first six weeks of 2026, more than three hundred data-center bills were filed across thirty-plus states — a shift, in the span of a year, from luring these facilities with tax credits to fencing them out with moratoriums and special rate classes. New York, South Dakota, and Oklahoma floated pauses on construction itself. At least eighteen states moved to create separate rate classes so that large users pay their own freight. Illinois and Florida passed laws requiring data centers to fund the grid upgrades they trigger. Oregon’s commission established a large-load tariff to insulate existing customers. And the national association of utility regulators received a letter — cited up to the Senate — urging its own members to reject any proposal that raises household rates to serve data-center load, naming the problem in plain words: speculative, industry-driven growth, unfairly cross-subsidized onto the public.

Read that scramble for what it is. You do not pass three hundred emergency bills to fix a referee who is calling the game correctly. You pass them because the old rulebook stopped protecting anyone and everyone can see it at once. The compact didn’t bend. It snapped, and thirty state legislatures are now trying to bolt a new one together in real time, facility by facility, while the bills keep going up.

The populist read is wrong, and being wrong about it costs you.

Here is where the easy story fails, and it fails in the direction most people find satisfying. The satisfying version is that Big Tech is stealing your electricity, and the fix is to make Big Tech pay. Make them fund their own grid, and your bill stops rising. It is clean, it has a villain, and it is not true.

Consider Pennsylvania. PPL agreed to a settlement specifically adding data-center protections for ratepayers — a model, consumer advocates said, for the whole state. The same settlement raised residential rates 4.9 percent and tacked on a fifteen-dollar monthly fee, the utility’s first distribution hike since 2016. The protections went in, but the bill went up anyway. The data center is the accelerant. It is not the fire.

The fire is a grid that has been underinvested in for years, a regulatory model that can’t price risk, and a monopoly whose profit motive runs in exactly the wrong direction. Naming the hyperscaler as the thief feels like accountability, but it functions as cover — it lets the structural failure, an oversight regime that no longer oversees, walk out of the room while everyone glares at Amazon. You can win the fight against Big Tech’s cost-shift and still watch your bill climb, because the thing that’s actually broken was never the tenant. It was the building.

What actually happened, in one line.

Your electric bill is rising to finance a grid sized for a load that is mostly a forecast, because the referee that was supposed to keep a monopoly from billing you for someone else’s speculation is operating under rules written for a world that no longer exists — and has been reduced from protecting you to processing the paperwork. What this does to the price of power over the next decade, which regions break first, and who profits from the gap between the load requested and the load that arrives — that is a separate analysis, and not a free one.

So here we are. The monopoly was tolerated because someone was watching it. Someone is still in the chair. No one is still watching. The lights stay on. The question is who’s paying to keep them on for a tenant that never moved in.

For those already thinking through what that means personally — the Borderless Living community has been mapping this territory for two years. The conversation is here.

[EDITORIAL NOTE — This was written at night on June 11th. At that time, the deal price was reported to be in the range: ~$135/share, ~$75B raise, ~$1.75T valuation. Depending on when you read this, that might have changed.]

This morning, SpaceX completed the largest initial public offering in history.

Not the largest tech IPO. Not the largest American IPO. The largest IPO. Saudi Aramco held the record at $25.6 billion raised in 2019. SpaceX is taking roughly $75 billion at a valuation near $1.75 trillion — three times the old record, set by an actual nation-state selling its actual oil.

By tonight, everyone you know will have an opinion about SPCX. The opinion will come from one of two factories. Factory one: this is the grift of the century, a loss-making company at an obscene valuation, run by a man who treats public markets like a personal ATM. Factory two: this is the opportunity of a generation, your one chance to own the company that owns the future, finally democratized for the ordinary investor.

Both factories are wrong, and they’re wrong in the same way. Neither one read the receipt.

Because that’s what an S-1 is. A receipt. It is a legal document that tells you, with the kind of precision that only the threat of securities litigation can produce, exactly what you are paying and exactly what you are getting. Narratives lie. Podcasts lie. Your brother-in-law lies.

The S-1 does not lie.
Because lying in an S-1 is a felony.

So let’s read the receipt.

What you’re paying

A $1.75 trillion valuation against $18.7 billion in 2025 consolidated revenue. That is roughly 94 times sales. Not earnings — sales. There are no earnings. SpaceX lost $4.94 billion last year.

Sit with the comparison for a second. Aramco came public on the back of the most profitable enterprise in human history — a literal license to pump money out of the ground. SpaceX is coming public on the back of a $5 billion annual loss, and it’s raising three times as much. That’s not a criticism. It’s an observation about what is actually being sold, which is not a cash flow.

It’s a trajectory.

What you’re getting

Here is where the receipt gets interesting: the answer comes in three parts, each printed in perfectly legible ink.

Part one: the cash machine. Inside SpaceX, one business reliably makes money — Starlink, which generated $4.42 billion in operating income last year. Everything else is a bet. The launch business is a marvel of industrial engineering that mostly exists to deploy more Starlink. Starship is a prototype with a 7-and-5 record across twelve flights. So when you buy at $1.75 trillion, you are paying roughly 400 times the operating income of the only segment that produces any.

Part two: the anchor bolted to the cash machine. SpaceX absorbed xAI this year, and xAI lost $6.4 billion at the operating line — meaning the AI division consumed every dollar Starlink earned, and then another two billion on top. You are not buying a satellite internet company. You are buying a satellite internet company that has been legally welded to a frontier AI lab in a capital-burning war with competitors funded by the profitable cores of Microsoft, Google, and Amazon. The receipt also notes, in passing, that of xAI’s twelve original co-founders, two remain.

Part three — and this is the part to read twice: the governance. The share you can buy tomorrow is Class A. It carries one vote. Class B shares carry ten. Elon Musk holds roughly 42 percent of the equity and, through Class B, roughly 80 percent of the votes. He is the CEO, the CTO, and the Chairman of the Board, and under the structure described in the filing, he can be removed from those roles only by a vote of the Class B shareholders.

He controls the Class B shareholders.

Run that clause through one more time. The mechanism for removing the chief executive is a vote held among shares the chief executive controls. This is not a loophole someone might exploit someday. It is the explicit, disclosed, intended design. The public is being invited to contribute $75 billion to an enterprise over which it will hold, as a matter of binding corporate law, zero control authority. Your Class A share is not a piece of the company in any sense your grandfather would recognize. It is a perpetual, non-voting participation certificate. A claim on outcomes with no claim on decisions. Closer to a bet slip than a deed.

This is not a scandal

Now — the part both opinion factories miss.

None of this is fraud. Fraud is concealment, and this is the opposite of concealment: it is several hundred pages of plain-English disclosure filed with the federal government and available to anyone with an internet connection. The dual-class structure isn’t even novel. Ford did it in 1956. Google did it in 2004. Meta did it in 2012. Each time, the market grumbled, bought anyway, and the next founder noticed. The ratchet only turns one direction, because it works.

So the “grift” take collapses on contact with the filing — you cannot be swindled by a document that tells you precisely what it’s doing. But the “democratization” take collapses just as fast, and the proof is sitting in the deal structure itself.

This offering is reported to be three-and-a-half to four times oversubscribed. Demand is above $250 billion against $75 billion of supply. Institutions could (and most certainly will) absorb every single share before lunch. And yet the deal includes a carve-out specifically reserved for retail investors — ordinary people, buying through their apps, at the offering price.

Why? When the share is scarce, why hand any of it to the small accounts?

Because retail is the ideal shareholder for this structure. Institutions ask questions. Institutions show up at annual meetings, file proposals, and call the general counsel. Retail does none of that. Retail buys, holds, and stays quiet — which makes retail the perfect buyer for a share that has been pre-stripped of its voice. The retail tranche isn’t generosity, and it isn’t democratization. It’s product-market fit. They built a share with no vote, and they’re distributing it to the shareholders least likely to notice the vote is missing.

That’s the receipt, read in full. The largest IPO ever conducted is the sale of the smallest governance stake ever offered, distributed by design to the buyers least equipped to object.

The mirror

In 2019, when Aramco listed, American commentators permitted themselves a long, satisfied smirk. Look at this sovereignty theater — a kingdom selling a 1.5 percent sliver of its crown jewel while the royal family kept every lever, every decision, every vote. We called it an IPO in name only. We were right.

Tomorrow, the American market executes the same structure at three times the size. A sliver of economics sold to the public; every lever retained. The only difference is that the sovereign retaining control is not a kingdom.

It’s a man.

And he will hold the dominant Western launch capability, the dominant low-orbit communications layer — including Starshield, the variant your government’s national security apparatus runs on — a frontier AI lab, and, as of tomorrow, $75 billion of the public’s money, insulated from that public by a ten-to-one voting ratio. The institutions that were supposed to make ownership mean something — proxy votes, board accountability, the shareholder franchise itself — haven’t been overthrown. They’ve been formatted out of the document, one filing at a time, while everyone watched the rocket launches. This is why a growing number of families have stopped assuming the systems they participate in will represent them, and started building options that don’t require anyone’s permission to exercise.

Read your receipts

I want to be precise about the conclusion, because precision is the whole point of this piece.

I am not telling you SPCX will crash. I have no idea what the price does, and neither does anyone selling you a forecast. The stock may triple. Plenty of overvalued things stay overvalued for a decade; ask anyone who shorted Tesla.

What I’m telling you is what the document says. It says you are paying 94 times revenue for a loss-making conglomerate whose one profitable engine is fused to a furnace, governed by a man who cannot be removed by anyone but himself, sold preferentially to the buyers least likely to mind.

Every word of that is disclosed. That’s the most important fact in this entire story. The system isn’t hiding anything from you. It stopped needing to.

The receipt was printed. It was filed. It was public.

Nobody read it.

Scott Pelley announcing his Substack in...

Three.

Two.

….

(LOL.)

If I understand what happened, Pelley essentially stood up in the company cafeteria and informed his bosses that they were a collection of short-sighted morons.

To which the corporation responded:

“Thank you for your feedback. Security will escort you out.”

Of course, they fired him.

There is a technical term for publicly denouncing management while drawing a paycheck from management.

It’s called insubordination.

Corporations are not democracies.

That said, let’s not pretend politics had nothing to do with it.

CBS has spent the better part of a year slowly transforming itself into a network acceptable to Trump-world because that’s what ownership wants. That’s what makes the deals work. That’s what protects the interests involved.

The old fantasy that capital automatically opposes authoritarianism was always nonsense.

Historically, capital often discovers it can make perfectly good money under authoritarianism.

Sometimes more.

This is not exactly a shocking development.

What is amusing is watching the media class suddenly discover that ownership matters.

“Oh my God!”

“The independence of journalism!”

“The integrity of the newsroom!”

Friends, where have you been?

News was never independent.

What existed—at least for a period of time—was an understanding between owners, editors, reporters, and audiences that the value of the product depended upon a reputation for credibility.

The audience paid for information.

The network sold trust.

Everybody benefited.

Then somebody discovered outrage had better margins.

First came “Breaking News.”

Everything became Breaking News.

A senator sneezed.

Breaking News.

A celebrity tweeted.

Breaking News.

A congressman accidentally put his pants on backward.

Breaking News.

Then new media arrived and improved the formula.

Breaking News every fifteen seconds.

Breaking News: Be Angry.

Breaking News: Panic Immediately.

Breaking News: Democracy Is Ending.

Breaking News: Buy My Subscription.

Now we’ve reached the final evolutionary stage.

“You are NOT going to believe this shit.”

I am not exaggerating.

That is now an actual news introduction.

That’s the “face” of this “independent news” movement.

Walter Cronkite would be drinking heavily.

Everyone claims to be better than the legacy networks today.

Maybe.

I don’t see much evidence.

Fox sells outrage wrapped in infobabes.

MSNBC/MSNOW/MSGONNABEGONESOON sells outrage wrapped in smug intellectualism.

CBS sells outrage, presently wrapped in red-hatted “Karen-inspired” nonsense.

CNN sells outrage, outraging against outrage. Watch Scott Jennings duke it out!

Half the Substack ecosystem sells outrage.

Half the YouTube ecosystem sells outrage.

The product differs less than the packaging.

One version wraps itself in patriotism.

Another wraps itself in resistance.

A third wraps itself in independence.

A fourth wraps itself in authenticity.

But underneath?

The same product.

Emotional validation.

People don’t want information.

They want confirmation.

They want somebody to tell them they are right, their enemies are stupid, and civilization is collapsing exactly the way they suspected.

The market rewards this behavior.

So the market gets more of it.

If Americans genuinely valued critical thinking, Donald Trump would have been politically buried years ago.

Not because of ideology.

Because of observable reality.

You don’t need a PhD in political science to notice the criminality, corruption, dishonesty, and institutional vandalism.

You simply need functioning eyes.

But that’s not the business model.

The business model is not understanding.

The business model is engagement.

The business model is not informing citizens.

The business model is creating customers.

And customers who are angry come back tomorrow.

So when Scott Pelley eventually launches his inevitable Substack, podcast, video channel, streaming platform, newsletter, community, membership tier, premium tier, super-premium tier, and whatever else follows...

We’ll see.

Perhaps he’ll build something different.

Perhaps he’ll become the latest participant in the great American outrage economy.

My money is on the latter.

The circle of life continues.

As Green Day observed twenty years ago:

“Don’t want a nation under the new mania.”

Unfortunately, that’s exactly what we built.

The war with Iran is approximately twelve weeks old, depending on how you count.

If you count from the February 28 strikes that killed Supreme Leader Ali Khamenei and the senior leadership of the Iranian government, the war is twelve weeks old. If you count from the June 2025 strikes on the nuclear facilities, the conflict in its current iteration is closer to a year. If you count from the broader posture that has shaped American policy toward Iran since 1979, this is the latest acute phase of a forty-seven-year rivalry that has periodically gone hot.

Twelve weeks. Khamenei and more than twelve hundred Iranian officials, soldiers, and civilians dead in the initial strikes. Thirteen American service members confirmed killed under Operation Epic Fury, with roughly four hundred wounded according to US Central Command data through mid-April. The Strait of Hormuz, closed by Iran on February 28, has remained effectively closed since: pre-war volume of about three thousand ships per month has fallen to roughly five percent of that. Brief windows of partial reopening followed ceasefire frameworks in April and early May and narrowed within days. QatarEnergy declared force majeure on LNG shipments on March 4 after Iranian strikes on Ras Laffan. Pakistan, Qatar, and Oman have mediated in various combinations. Two ceasefire frameworks have been announced. Both have been violated. The President has reportedly come close to resuming strikes at least once, then drawn back. Republicans pulled a House war powers vote yesterday because, according to their own leadership, they could not find the numbers to defeat it.

This is the configuration. Hold the configuration in mind, because the question that follows from it is not the question most commentary is asking.

Most commentary is asking which of two scenarios resolves the war. The first is decisive escalation — Iran retaliates dramatically, the US strikes back at full force, the conflict spreads regionally, Hormuz closes under direct military confrontation, oil prices spike to crisis levels, and the world economy enters a recession driven by the energy shock. The second is comprehensive resolution — a deal is reached, mediated by some combination of Pakistan, Qatar, Oman, or other parties, sanctions ease, the Strait reopens under credible security guarantees, Iran’s new leadership stabilizes, and the conflict winds down into a managed cold peace.

These are the scenarios most commentary war-games. The argument here is that neither is the most likely outcome — and that what is most likely is structurally distinct from both.

There is a third scenario, less written about because it lacks a climax. The war does not decisively escalate. It does not comprehensively resolve. It enters a routinized persistent conflict cycle, in approximately the configuration it currently occupies, for an extended period — most plausibly counted in years rather than months.

This is what is structurally most consistent with current observables. It is also what is already happening.

Consider the structural conditions required for the war to end through either of the more familiar scenarios.

Decisive escalation requires a triggering event large enough to break the current cycle architecture. The current cycle has absorbed three months of provocations — Iranian missile and drone strikes on US bases in Kuwait and Saudi Arabia, attacks on commercial shipping, US retaliatory strikes after warship engagements, repeated ceasefire violations — without escalating to regional war. Both sides have demonstrated the capacity to absorb provocations and re-stabilize at a slightly degraded equilibrium. Decisive escalation is possible. It is also what the system has so far metabolized rather than executed.

Comprehensive resolution requires political will to end the war on at least one side. The structural conditions for either side to generate that political will look weak as of this writing, though they could shift.

On the American side: the President spent four years promising to be tough on Iran. The strikes were the operational manifestation of that promise. Withdrawing — declaring victory and going home without a defined victory — is politically expensive. The administration that started the war cannot easily end it without a deal that looks like American success, and the conditions for an Iranian government to offer such a deal do not appear to exist. The alternative — actual withdrawal — registers domestically as weakness. The current configuration, in which strikes happen periodically and the President says the ceasefire remains in place, lets the administration avoid both withdrawal and full escalation while maintaining the posture of strength.

On the Iranian side: Khamenei was killed in the February 28 strikes. His son, Mojtaba Khamenei, has been appointed Supreme Leader. The succession was managed within the regime’s institutional architecture; the President’s characterization of “internal infighting” should be read as the assessment of an adversary, not as a confirmed state. Mojtaba Khamenei inherits both his father’s institutional position and the political logic that produced his father’s posture. The selectorate that elevated him is constituted by the same hardline factions that survived forty-seven years of sanctions, isolation, and periodic confrontation with the United States. Any new leader who concedes to American demands inherits the political cost of having conceded to the country that killed the previous Supreme Leader, who was also his father. The political space for concession has narrowed since the strikes, not opened.

Congress has tried, repeatedly, to constrain the war through war powers resolutions. The House failed on March 5 by a margin of seven votes, two hundred and twelve to two hundred and nineteen. The Senate advanced a resolution on May 19 by fifty to forty-seven. The House this week pulled a resolution vote because Republican leadership reportedly could not find the votes to defeat it. Each of these is closer than the last. None passes the veto math. A war powers resolution that passes Congress can be vetoed by the President, and the override math is not in either chamber’s current configuration. Congress as an institution can register that it does not approve of the war, and is gradually doing so. It cannot, on current numbers, end it.

The military-industrial apparatus has begun adapting to sustained engagement. Defense contractors are restructuring operational forecasts around prolonged Middle East presence. Energy markets are pricing the disruption as a sustained volatility input rather than a one-time crisis. European inflation forecasts have been revised upward to account for the energy shock. The Strait of Hormuz remains substantially closed, with periodic narrow exceptions for ships willing to pay Iranian tolls or negotiate safe passage. None of this looks like a system tilting toward resolution. It looks like a system stabilizing around persistent low-intensity conflict.

This is the structural argument. It is not a prediction of certainty. It is a claim about the most probable trajectory given current observables. Tail risks exist on both sides — a major shipping casualty, a strike on a US carrier, an Iranian nuclear reconstitution sprint, an Israeli political crisis forcing unilateral escalation, a leadership health event on either side — and any of those could move the trajectory toward either of the more familiar scenarios. The argument is not that those tail risks are impossible. The argument is that the median outcome, conditional on the current configuration, is persistence.

The historical pattern is what should anchor the rest of this analysis, though with category discipline.

The proper category for the Iran configuration is persistent low-intensity conflict, or protracted militarized rivalry — not frozen conflict. Frozen conflicts describe configurations where active hostilities have substantially ceased, frozen in place by armistice or stalemate. Korea, Cyprus, Transnistria. The Iran configuration is still active. The proper analogues are looser and more cautionary.

The Iraq sanctions and no-fly zone regime from 1991 to 2003 is the closest recent analogue. Twelve years of low-intensity conflict that became part of the international system’s background. Periodic strikes, periodic crises, sanctions architecture, mediated diplomacy, a regime that did not capitulate and a hegemon that did not withdraw. It ended, eventually, but it ended through external escalation — the 2003 invasion — not through internal resolution. The persistence broke when the political will to break it changed in Washington, not in Baghdad.

The Israel-Hezbollah configuration from 2006 onward is another. Nearly two decades of intermittent open conflict, ceasefires, proxy violence, deterrence equilibria. The war never formally restarted; the peace never formally arrived. The configuration consumed strategic bandwidth on both sides for a generation.

The post-2014 Donbas configuration ran for eight years before it dramatically escalated in February 2022. For those eight years, most international observers treated it as a quasi-frozen conflict that had stabilized. The eventual escalation came from changes in the strategic calculus of one side, not from gradual evolution of the local configuration.

What these analogues share is that the resolution mechanism, when it arrived, came from outside the local configuration: a change of administration, a great-power realignment, a leadership transition, an external shock. The configurations did not end through their own internal logic. They ended when external logic acted on them.

For the Iran configuration, the relevant external variables are: who holds executive power in Washington, who holds power in Tehran, whether the Iranian nuclear program reconstitutes faster or slower than the Pentagon’s two-year setback estimate, whether Israeli politics produces a government willing to unilaterally escalate, whether Saudi or Gulf state positioning shifts under sustained energy disruption, and whether the dollar’s reserve status pressures the financial sustainability of indefinite engagement. Each of these is a potential trigger for transition out of persistence. None of them is scheduled.

Persistence has costs that accumulate.

Thirteen American servicemembers have died under Operation Epic Fury so far. The number will likely grow if the cycle continues. Periodic strikes against a hostile regional power produce casualties; the cycle reproduces the conditions that produce them.

The taxpayer cost of sustained military engagement in the Middle East compounds. Brown University’s Costs of War project estimated the total cost of the Global War on Terror at roughly eight trillion dollars over twenty years. A persistent Iran configuration of similar duration, with higher-end military assets routinely deployed and more contested operational environments, could plausibly cost more — though the projection is sensitive to intensity assumptions and not a fixed forecast.

Energy markets are likely to continue pricing the persistence. European growth forecasts have already been revised down to account for sustained inflation pressure. Lower-income households in OECD countries will continue to spend a larger fraction of disposable income on fuel and transportation. Higher-income households will hedge or relocate where they can. The K-shaped economic pattern of the post-COVID era will likely deepen under sustained conflict-driven inflation, though the magnitude depends on whether Hormuz traffic recovers, on Saudi spare capacity utilization, and on US Strategic Petroleum Reserve management.

The 2028 election is unlikely to end this through routine electoral mechanics. A new president inheriting the war would find that the de-escalation pathway is more politically costly than the continuation pathway. Pulling out registers as weakness. Negotiating a comprehensive deal requires an Iranian leadership willing to make a comprehensive deal, and there is no scheduled point in the next four years at which Iranian leadership will be more willing than it is now — barring an external shock that changes the calculus. The next administration will most likely inherit the configuration and continue it with marginal modifications. This is what most previous administrations have done with inherited armed engagements, though not all.

The strategic opportunity cost is substantial. American strategic bandwidth that would otherwise be available for the Pacific theater, for European deterrence, for hemispheric stability, for technology and industrial competition with China, is being consumed by the Iran configuration. The cost is not visible because it is the cost of things that are not being done. It is the European posture that erodes because American attention is elsewhere. It is the Pacific deterrent that weakens because assets are forward-deployed elsewhere. It is the hemispheric concern about Venezuela, Cuba, or the consolidation of Mexican narco-state dynamics that does not get addressed because the war room is occupied.

For sovereign families, the practical implications differ under persistence from under either of the more familiar scenarios.

Under decisive escalation, the planning question is how to respond to dramatic dislocation — and the planning window is short. Under comprehensive resolution, the planning question is how to position for normalization — and the planning window is medium. Under routinized persistence, which is what the current observables most consistently support, the planning question is how to function inside a sustained background of elevated volatility for an indefinite period — and the planning window is long, ambiguous, and unforgiving of delay.

The volatility is in oil prices, in the dollar’s reserve currency premium, in the political bandwidth of the American executive, in the deficit math of a federal government simultaneously cutting taxes and funding sustained military engagement, in the credit-cycle architecture of a financial system whose institutional buffers have been deliberately reduced. These compound with the architecture of formalized patronage discussed earlier this week. None of these conditions reverses on the timeline that the next election affects.

The operational implication is that the sovereign architecture conversation, developed at length yesterday, has different inputs under routinized persistence than under either of the more dramatic scenarios. The optionality you build now will be activated against a backdrop of sustained elevated volatility, not against the backdrop of a discrete crisis. The timeline for using the optionality is longer. The conditions that would prompt activation are gradual rather than dramatic. The latency advantage of having built the architecture before you needed it is larger under persistence, not smaller, because the conditions never sharpen into an unambiguous signal that the architecture is required. They keep gradually deteriorating, with intermittent partial recoveries that look like normalization but aren’t.

Persistence is the configuration that punishes families who wait for the unambiguous signal. There is no unambiguous signal under persistence. There is only the gradually accumulating cost of decisions deferred while the situation incrementally degrades.

Wars end when actors with the capacity to end them choose to end them. No actor with that capacity is currently positioned to act in either direction on the timeline that would affect the current electoral cycle. That is not a permanent condition. It is the current condition.

The architecture you build, between now and the point at which the war’s persistence becomes a dominant variable in your family’s planning environment, is the architecture you will have when it does. The architecture you do not build is not available later. That is the practical implication of the third scenario.

The third scenario is structurally the most likely scenario, conditional on current observables. It is also the scenario almost no one is planning for.

This kind of analysis doesn’t exist without readers paying for it. No ads, no foundation, no outside agenda steering the work. If you want more, the upgrade is right there.

Subscribe now

There is a conversation Americans are having with themselves right now that goes approximately like this. The current administration has done a lot of damage. The damage is upsetting. The damage will continue for some time. But in November of 2026, the midterm elections will reset the balance. And in November of 2028, there will be another election, and a new president, and the worst of the damage will be undone, and the country will get back on track.

We just have to wait.

This is the comfortable lie of the moment. The comfortable lie is comfortable because it gives the reader a date to wait for, a milestone past which suffering ends, an electoral process that converts political pain into eventual political relief. The comfortable lie is also wrong.

The comfortable lie is wrong because elections do not undo institutional damage. They never have. They produce changes of personnel and shifts in policy, sometimes large ones. They do not, by themselves, rebuild what has been disassembled. Rebuilding requires people, time, institutional memory, political capital, congressional cooperation, judicial acquiescence, and a public willing to absorb the costs of the restoration. The election does not produce any of those things by itself. It produces a new president and possibly a new congressional majority. What the new president and majority can actually do, given the constraints they will inherit, is the real question.

Let me walk through what cannot be reversed by the next election, or the one after, or the one after that — at least not on the timeline most readers are quietly assuming.

Start with the courts. The Supreme Court has a six-to-three conservative majority. Three of the justices were appointed during President Trump’s first term. The youngest, Justice Barrett, is fifty-three; she is likely to serve into the 2050s. The next Democratic president could change the composition of the Court only if vacancies open at advantageous times, which historical patterns suggest is unlikely to produce a structurally different Court before the late 2030s at the earliest. Court expansion is constitutional but politically explosive and would itself be a norm violation. Even if court expansion happened — and it would require unified government plus willingness to absorb the political cost — the new appointees would face a court already shaped by a generation of opposition jurisprudence.

The lower federal courts have been reshaped less dramatically, but the cumulative effect is structurally meaningful. President Trump made two hundred and thirty-four Article III judicial appointments during his first term and is appointing more now, at a pace slower than the first time but steady. By the end of his current term, Trump appointees will constitute approximately one third of the active federal judiciary — and a third of district and appellate judges effectively determines the outcome of the cases that never reach the Supreme Court, which is most of them. The judicial system that a 2028 successor inherits is one whose center of gravity has been pulled rightward for a generation, and the pull does not reverse by changing the executive.

Move to personnel. Schedule F and related personnel reforms have moved tens of thousands of career civil service positions into political-appointee status. The career attorneys at the Department of Justice who left during the first Trump administration largely did not come back. The career professionals leaving during the second Trump administration are not coming back either. The Antitrust Division has lost its most senior litigators in the past two months — David Dahlquist, the acting director of civil antitrust litigation, resigned with three other senior litigators over a settlement negotiated without the trial team’s input. The State Department has shed hundreds of foreign service officers. USAID does not exist as an institution. The CDC’s international operations are gone. The Department of Education has been reduced. These are not policy decisions that can be reversed by signature. They are institutional capacities that have been removed.

Rebuilding institutional capacity is a decade-long project, not an electoral cycle. A new president in 2029 would inherit a federal workforce that is missing the people who knew how to make the institutions function. The institutional knowledge — how to enforce antitrust law, how to coordinate a pandemic response, how to negotiate a sanctions regime, how to draft regulations that survive judicial review — is held by individuals, not by organizations. Once the individuals leave, the knowledge leaves with them. The new president can hire replacements. The replacements will not, for years, know what the people they replaced knew.

Fiscal architecture. The One Big Beautiful Bill Act, signed into law July 4, 2025, made permanent the 2017 Tax Cuts and Jobs Act and extended additional cuts. The Tax Foundation projects that the act will reduce federal revenue by approximately $5 trillion between 2025 and 2034. The Bipartisan Policy Center estimates the total cost at three point four trillion in tax cuts alone, with another six hundred billion in interest on the additional debt. A 2028 successor inheriting this architecture finds that the revenue side of the federal budget is locked in by law for the next several years. Changing it requires either a sixty-vote Senate majority — which no realistic projection puts in Democratic hands — or budget reconciliation, which can move tax rates but cannot, for the most part, restore the institutional capacity that has been defunded. The fiscal architecture has been designed to outlast the administration that built it.

The formal patronage policies. The Department of Justice’s one-point-seven-seven-six-billion-dollar “anti-weaponization fund.” The IRS’s “forever barred” provision on the President’s family. The racial criterion in refugee admissions. The federal disaster declaration patterns. These are not informal practices that can be quietly reversed. They are formal acts of policy, written down, published, in some cases signed into law. Reversing them requires legislation, executive order, or court action. Each of these mechanisms has been compromised. Congress is locked in coalition arithmetic. Executive orders by a successor can be challenged in courts that have been politically aligned. Court action depends on judges whose jurisprudence is now structurally consistent with the policies being challenged.

The diplomatic relationships. The European Union, the United Kingdom, Canada, Japan, South Korea, Australia, India, Brazil — none of these countries believes anymore that the American institutional system is reliable on the scale of decades. They have updated their priors. They will continue to hedge against American erratic behavior regardless of who is in office in 2029, because they have learned that the same erratic behavior can return four years later. The political signaling of a new administration cannot undo the strategic reassessment that has already occurred. Trust is a stock variable, built over decades and depleted in months. The depletion has happened. The rebuild, if it happens at all, runs in decades.

The DOJ itself, as an institution, has had its independence broken — not abolished, but broken in a way that is now part of the institutional memory. The Acting Attorney General who signed the IRS settlement was the President’s former personal attorney. The DOJ launched a criminal investigation of the Federal Reserve as a political pressure tool. These are not events that can be undone by hiring different people. They are precedents. Once a precedent exists in the institutional memory, every future actor — of any party — knows the move is available. The norms that prevented this kind of capture have been broken because someone tested whether they would hold, and they didn’t. Future presidents of either party will know that the norms can be broken if you have the willingness.

The Federal Reserve’s institutional independence has been similarly broken. The President pressured the Fed, the Department of Justice investigated the Fed, the President’s nominee was confirmed to lead the Fed after the pressure campaign worked. The precedent now exists. A future president of either party who does not like rate policy can replicate the pressure campaign. The institution that was supposed to be insulated from politics has demonstrated that it isn’t. That demonstration cannot be undone. It can only be added to.

The architecture of the apocalypse, which the previous piece in this newsletter discussed, has been built into the financial regulatory framework on a timeline that suggests the next financial crisis arrives between 2029 and 2033 — that is, during the term of whichever administration follows the current one. The crisis will not be the new administration’s fault. It will be a consequence of decisions already made. But the new administration will inherit it, and the absorption capacity of the banking system that would normally cushion it has been deliberately reduced.

Pause and total these up. The Supreme Court reshaped for a generation. The lower federal judiciary substantially shifted. The civil service hollowed out at a depth that requires a decade to rebuild. The fiscal base, locked in by law for years. Formal patronage policies on the books. The diplomatic relationships, structurally reset. The DOJ and Fed independence precedents, broken in ways that cannot be unbroken. A financial crisis approximately scheduled for the successor’s term.

This is not a list of policies to reverse. This is a list of structural conditions to inherit.

Read more

The International Criminal Court’s prosecutor’s office requested arrest warrants last Monday for the Finance Minister and National Security Minister of Israel. The Finance Minister, Bezalel Smotrich, confirmed the request publicly on Tuesday, calling it “a declaration of war.” The National Security Minister, Itamar Ben-Gvir, is reportedly named in the same request. A diplomatic source indicated three additional warrants target Minister Orit Strock and two senior IDF officials.

This follows the ICC’s November 2024 warrants against Prime Minister Benjamin Netanyahu and former Defense Minister Yoav Gallant, also for alleged war crimes and crimes against humanity related to the conduct of the war in Gaza.

These are institutional acts. They are the acts of an international court doing its institutional work — assembling evidence, evaluating it, and seeking warrants — against the sitting senior officials of an allied state. The court is acting against politically inconvenient targets. The court is doing what international courts are constituted to do.

It is also far from the only institution doing this kind of work.

In June 2025, the United Kingdom, Canada, Australia, New Zealand, and Norway jointly imposed sanctions on Ben-Gvir and Smotrich — coordinated travel bans, asset freezes. Five Western democracies acting in concert against the sitting cabinet ministers of an allied state. Individual EU member states, including Slovenia, Spain, the Netherlands, and Belgium, have moved unilaterally to ban both ministers from their territories.

The European Union, as a bloc, tried to impose sanctions through its formal machinery but failed. Germany, Italy, the Czech Republic, Austria, and Hungary blocked the required qualified majority. The EU instead agreed to sanction a handful of unnamed settlers. This is what most commentary calls “the EU’s failure to act.”

It is also a fundamentally American misreading of how working institutional systems behave.

Working institutional systems do not require a single venue to function. They require multiple venues to be available, so that when one venue is captured or blocked, others can act. The EU’s failure to sanction Ben-Gvir and Smotrich is precisely matched by the success of the UK and four allies in imposing the same sanctions outside the EU framework. It is matched by individual EU member states moving unilaterally within the bloc. It is matched entirely by the ICC acting through a third institutional venue. Five countries acted multilaterally. Several states acted unilaterally. An international court acted institutionally. The net result is sustained legal and diplomatic pressure on two cabinet ministers of an allied state, originating from at least eight distinct institutional sources.

That is what working institutions look like.

The United States, in the same window, has imposed no sanctions on the same individuals. The State Department has not issued travel bans. The Justice Department has not investigated. The reason is political — the US-Israel relationship is the load-bearing element of American Middle East policy and has been for half a century. But political infeasibility is not the same as institutional incapacity. American institutions retain the legal authority to take such actions. They have chosen not to.

That last sentence also describes the Europeans until recently. Then they chose differently. The Americans have not.

There is a habit, in American political commentary, of treating institutional functioning as a baseline condition — something that either exists or does not, the way water is wet. This habit was reasonable when the assumption held. It does not hold anymore. Institutional functioning is a variable. It varies across countries. It varies across time within the same country. It varies because the institutions in question are made of people, and those people choose every day whether to enforce the formal rules against politically inconvenient targets. When the people choose to enforce, the institutions function. When the people stop choosing to enforce, the institutions become decorative.

European institutions are currently in the enforcement phase. American institutions are currently not.

Look at the evidence. Pick a category.

Constitutional courts. The Italian Constitutional Court issued Sentenza 63/2026 earlier this year, ruling on the limits of legislative authority over citizenship by descent. The ruling was unwelcome to the current Italian government, which had passed restrictive legislation that the court partially struck down. The court ruled anyway. The German Federal Constitutional Court ruled in November 2023 that the federal government’s reallocation of pandemic-emergency funds to climate spending violated the constitutional debt brake, forcing a comprehensive restructuring of the federal budget. The Court continues to act as a binding constraint on federal authority — it received over four thousand six hundred new cases last year, ninety-six percent of them constitutional complaints brought by individuals challenging government action. The European Court of Human Rights continues to issue binding decisions against major European states, including Hungary, Poland, and Turkey — and the decisions, with rare exceptions, are complied with.

The United States Supreme Court, in the same window, cleared Alabama’s elimination of a Black-majority congressional district. It has progressively expanded executive authority on immigration enforcement. It has not, in a structurally meaningful sense, bound the executive in years.

This is not because the formal architecture differs. The American Constitution and the various European constitutions both establish judicial review. Both establish executive accountability. Both establish the rule of law. The architecture is roughly equivalent. The difference is in the people inside the architecture, and the choices they make on contested questions.

Regulatory enforcement. The European Union has spent the past two years aggressively enforcing the Digital Markets Act and Digital Services Act against the largest American technology companies — Apple, Google, Meta, Amazon, Microsoft. The fines have been substantial. The behavioral remedies have been significant. The companies have complied, with grumbling.

The same companies operate in the United States without equivalent constraint. The Department of Justice’s Antitrust Division has, in the past four months, undergone a structural collapse. The Live Nation–Ticketmaster case, which the previous administration had brought as a structural separation challenge, was settled in March 2026 on terms short of breakup. The settlement was negotiated without the trial team’s input. The acting director of civil antitrust litigation, David Dahlquist, resigned along with three other senior litigators in April. In the HPE–Juniper merger challenge, the political leadership of the Department reportedly overruled the Antitrust Division’s career staff opposition to a settlement — the settlement document was unusually not signed by the Antitrust Division. The head of the Antitrust Division was subsequently fired. Merger investigations dropped to near-record lows in 2025 and have not recovered.

Regulatory enforcement is being carried out in Brussels and progressively undone in Washington.

Foreign policy. The EU has imposed sanctions on Russian officials, Belarusian officials, Iranian officials, Venezuelan officials, Israeli settlers, and Burmese military officials over the past three years. The sanctions are coordinated, multilateral, and sustained. The US has rolled back sanctions on Russian officials, Venezuelan officials, and others. It has imposed new sanctions primarily against countries with which the President has personal grievances. The sanctions instrument exists in both jurisdictions. One uses it as a tool of foreign policy. The other uses it as a tool of presidential mood.

Press freedom. Europe is far from a paradise on this front — Hungary’s media environment is degraded, Italy’s RAI has political entanglement, and the UK’s libel law remains terrible. But the trajectory matters. The Italian government has not abolished its national broadcaster. The German government has not jailed journalists. The French government has not threatened critical news organizations with retaliatory tax audits. Each of these has occurred or been threatened in the United States in the past twelve months. The directional vector is opposite.

I could keep going, and the piece would become a list. The structural point is the same across all categories. The institutions still function in Europe, with imperfection, with political friction, with the same human tendencies toward capture that exist everywhere. They function imperfectly because they are still being used. They are still in use because the people inside them have not yet decided to stop. In the United States, the same institutions exist with the same formal authorities, and the people inside them are increasingly choosing not to use them against politically aligned targets.

This is a category difference in how the systems operate, and it is the difference that turns institutional vitality into a variable.

A counter-argument deserves naming. Europe is not a paradise. Europe has its own structural problems. The German economy has been contracting for two consecutive years. France is in a sustained fiscal crisis, with public debt above 110% of GDP. The Italian banking sector remains fragile. The energy crisis triggered by the Russian invasion of Ukraine accelerated deindustrialization in Germany and parts of Eastern Europe. Migration politics have produced rising far-right vote shares in nearly every member state, with Alternative für Deutschland polling at or above twenty percent in Germany, the Rassemblement National in France well above the Socialists, and Fratelli d’Italia governing in Rome. The EU’s economic competitiveness gap relative to the United States and China is substantial and growing. None of this is hidden.

The point is not that Europe is functioning well in absolute terms. The point is that institutional vitality is a function that European institutions are still performing, even when political incentives push the other way, whereas American institutions increasingly are not. The contrast lies more in the direction of the trajectory than in the level.

There is a third dimension to this worth surfacing. Institutional vitality is not just a question about politics — it is a question about jurisdictional risk for anyone whose life is exposed to the institutions of a given country. The American who pays American federal taxes and is subject to American regulatory enforcement operates in a different institutional environment than the European who pays European national taxes and is subject to European regulatory enforcement. Both pay taxes. Both are subject to laws. The difference is in what happens when the politically powerful violate those laws.

In Europe, the institutions still occasionally act against politically aligned violators. In the United States, increasingly, they do not.

This is the kind of variable that did not exist for most of human history, because most of that history lacked functioning impersonal institutions at all. The notion that a state’s legal apparatus would constrain the state’s most powerful actors is a recent invention. It has never been universal. It has been a regional phenomenon, concentrated in the post-Enlightenment Atlantic political tradition, and even within that tradition, it has been more aspirational than operational. The post-1945 Western order made it a stronger reality than ever before in human history. That reality is now coming apart. Not in Europe yet, though Europe has its own pressures. In the United States, more quickly than most American commentary has noticed.

When the variable changes, the answer to the question “what country do I live in” changes. The answer is no longer just geographic. It is institutional. The country that maintains constraints on its powerful actors is meaningfully different from the country that does not, even if both are called democracies, even if both are called developed, even if both occupy adjacent positions in the rankings produced by Freedom House or the Economist Intelligence Unit. The rankings are lagging indicators of an underlying institutional process. The process matters more than the ranking.

For most of the post-1945 era, the assumption held that all OECD countries occupied roughly equivalent institutional positions. The assumption was always a simplification, but it was a useful simplification because the differences were marginal. The differences are no longer marginal. They are now structural. And the assumption that has been useful for two generations of capital allocation, family planning, professional decision-making, and personal exposure management is no longer accurate.

This is the wider point worth saying directly. Institutional vitality is now a sovereign variable. Some countries maintain it. Others do not. The question of which country you live in now includes whether its institutions still bind the powerful within it. That question, five years ago, was an academic abstraction in nearly every OECD context. It is no longer abstract anywhere.

The International Criminal Court asked judges to issue arrest warrants for two members of the Israeli cabinet on Monday. American institutions did not. In a different week, the contrast will be different. The structural fact will be the same.

What working institutions look like is what they look like in Europe right now — imperfect, politically constrained, internally divided, occasionally embarrassing themselves, and yet still capable of binding the powerful when they choose to. The choice has not yet defaulted to no.

In the United States, the choice has increasingly defaulted to no.

That is the difference. It is the difference that turns one country into another country without changing the map.

This kind of analysis doesn’t exist without readers paying for it. No ads, no foundation, no outside agenda steering the work. If you want more, the upgrade is right there.

Subscribe now

Kevin Warsh was sworn in as Chair of the Federal Reserve last Friday.

His confirmation, 54-45 in the Senate, was the most divisive vote for a Fed Chair in the modern banking era. The vote followed a months-long pressure campaign that included a Justice Department criminal investigation into the Federal Reserve itself — a probe that was dropped only after the Senator who had been blocking Warsh’s committee vote in protest released his hold. The investigation went away. The hold went away. The nomination advanced.

This is how the architecture was completed.

The architecture has three legs, and they have been installed in sequence. Hold all three in view at once.

The first leg, already in motion, is the reduction of capital requirements. In March of this year, the Federal Reserve voted six-to-one to reduce the capital requirements imposed on the largest banks in the United States. The vote received roughly a tenth of the press coverage that the President’s social media posts got that day. This is inverted from its significance.

What was voted on, in concrete terms, was the re-proposal of the Basel III Endgame rule — the regulatory framework that determines how much capital large banks must hold against their lending and trading activities. The original 2023 proposal would have raised the required capital for the largest US banks by about 19%. The 2026 re-proposal, led by Vice Chair for Supervision Michelle Bowman, instead cuts that requirement. Capital requirements for the largest banks fall by 4.8 percent. For mid-sized banks, by 5.2 percent. For smaller banks, by 7.8 percent. Stack on top of this the November 2025 final rule that already cut the Enhanced Supplementary Leverage Ratio for globally systemically important institutions, capping the eSLR for depository subsidiaries at four percent.

Alvarez & Marsal, the consulting firm, estimates that the changes unlock $2.6 trillion in new lending capacity for the US banking system.

A quick aside? People talk about “printing money,” and inflation. The US Government does not “print money” into existence. Capital operations bring money into existence. Specifically, banks loan money into existence. Thus, we’re talking about trillions of dollars in new money becoming available for lending.

The KBW Bank Index is at an all-time high, unsurprisingly. They all see themselves as about to get very rich. (And they’re probably right.)

These are the facts of the first leg.

The second leg is the leadership change. Jerome Powell’s term as Chair expired May 15. Powell was, by the standards of the Trump administration, insufficiently compliant on rate policy. The President spent the better part of two years complaining that rates were too high. The pressure campaign against the Fed escalated through 2025 and into early 2026, culminating in the DOJ probe and a months-long Senate stall before Warsh’s confirmation.

The result: Kevin Warsh, age fifty-six, sworn in as the eleventh Fed Chair of the modern banking era. Trump’s handpicked successor. Confirmed almost entirely on party lines. The institution that was historically defended as politically independent now has a Chair whose nomination was advanced through DOJ pressure on the institution itself.

Warsh’s biographical signals are worth noting precisely.

He served previously on the Federal Reserve from 2006 through 2011. The Fed on which he served, in 2007 and 2008, was the body that the historical record now describes as having initially dismissed the dangers of the subprime mortgage meltdown that led to the global financial crisis.

Warsh was one of the dismissers.

He was the Wall Street liaison during the crisis itself. He has been, for fifteen years since leaving the Fed, an inflation hawk in his public commentary — though his recent statements have signaled openness to lower rates, which appears to be the price of the chairmanship.

He has also been an adviser to Electric Capital, a crypto-focused investment firm. This is the third leg.

The Basel III re-proposal contains a provision that has received almost no political attention. The current Basel framework applies a 1,250 percent risk weight to unbacked crypto assets, which effectively means banks must hold one dollar of capital for each dollar of crypto exposure. This treatment was deliberate. It reflected the assessment by international regulators that crypto assets, lacking productive cash flow and subject to extreme price volatility, were unsuitable balance-sheet items for systemically important banks. The 2026 re-proposal substantially relaxes this treatment. Major US banks are now positioned to offer custody, lending against, and trading services for digital assets at risk weights that make the business profitable. JPMorgan, Goldman Sachs, and Bank of America are publicly positioning for this expansion.

So: capital requirements being cut by the Vice Chair for Supervision. Rate policy now in the hands of a Chair selected for compliance. Crypto integration into the banking system now economically viable, championed by a Chair with crypto-industry ties. All three moves the work of the same institution, in the same calendar year, with the same beneficiary class.

This is what is meant by Fed-led.

The argument from here is structural.

Capital requirements were not invented to harass banks. They were the institutional response to 2008. The reason 2008 happened the way it happened was that the largest banks had insufficient capital to absorb the losses that materialized on their books when the housing market reversed. The losses were not the surprise. The losses were always going to happen — that is what credit cycles do. What was surprising was that institutions whose stated business was absorbing financial risk turned out to have so little capital that they could not absorb the financial risk they had taken. They were, by the time the cycle turned, leveraged at a ratio that left no room for ordinary credit losses.

The Basel III framework was the response. Basel III required banks to hold more capital. The capital was specifically calibrated against the failure modes that materialized in 2008. The framework said: you will hold enough capital to absorb a 2008-scale shock without becoming insolvent. The framework was conservative in some respects, less so in others, but its core logic was sound. If you require capital, you prevent the failure mode in which a credit cycle reverses, and the institutional absorbers cannot absorb.

The framework worked for about fifteen years. It has just been substantially undone by the same institution that previously served as its custodian, now led by the same person who was on the Fed when the subprime crisis was initially dismissed.

This matters because credit theory is not optional.

Banks are leveraged institutions. They borrow short and lend long. When their capital position is strong, they can absorb losses without affecting their ability to lend or meet obligations. When their capital position is weak, even ordinary credit losses force them to deleverage, which means calling loans, refusing new lending, and selling assets into falling markets. The deleveraging cycle is what makes credit crises systemic rather than localized. The reason 2008 became a global financial crisis rather than a contained housing correction was that the institutional absorbers turned out to be amplifiers.

When you reduce capital requirements, you do two things. You increase the banking system's lending capacity during good times. And you reduce the banking system's absorption capacity during bad times. These are not separate effects. They are the same effect viewed from two different positions in the credit cycle. The same regulatory change that produces the boom produces the bust.

When you combine reduced capital requirements with a Fed Chair committed to lower rates and the integration of a volatile asset class onto bank balance sheets, you produce more of both. The boom phase will be more pronounced because the policy mix is more accommodative. The bust phase will be more severe because the absorption capacity is lower, and the new asset class brings volatility that the prior framework was designed to keep off the books.

The timing question is when, not whether.

Historical data on this is consistent. The Gramm-Leach-Bliley Act in 1999 removed the structural separations between commercial and investment banking. Eight years later, the system blew up. The 1980s deregulation of the savings and loan industry was followed by the S&L crisis. The Glass-Steagall reforms of the 1930s lasted for 60 years and were dismantled in stages over the 1990s, with the final removal in 1999, followed by the 2008 crisis. The pattern is consistent: rule relaxation, credit expansion, asset bubble, cycle reversal, systemic crisis. The lag between relaxation and crisis is typically four to seven years, sometimes longer.

The current capital relaxation began in late 2025 under Powell. The Warsh chairmanship begins now. The crisis the combined architecture sets up, on historical pattern, lands somewhere between 2029 and 2033.

It will not look like 2008. Financial crises never look like the previous one. The institutional adaptations and regulatory adjustments since 2008 will route the next crisis through different channels. The most likely candidates are: commercial real estate, where significant losses have been deferred and disguised through loan extensions and restructurings; private credit, which has grown from roughly $300 billion to nearly $2 trillion since 2008 with substantially less regulatory oversight than bank lending; and digital asset exposure, which the Basel re-proposal now makes economically viable for major banks to hold on balance sheet.

The implication is that the next crisis will have compounding factors that were not present in 2008. The deregulated capital position. The accommodative rate stance of a Trump-installed Fed Chair. The new institutional exposure to crypto, championed by a Chair who previously advised a crypto fund. And the loss of Fed independence as an institutional check — the DOJ used as a pressure tool against the central bank is the single most institutionally degrading move in modern American monetary policy. Crisis economists studying this period in twenty years will likely treat these moves as a single architectural decision rather than four separate ones.

In the near term, none of this looks like a crisis. In the near term, this looks like a boom.

Banks with reduced capital requirements lend more. Rate cuts add fuel. Asset prices rise as new credit chases existing assets. Wealth concentrated in equities and real estate compounds at accelerated rates. The KBW Bank Index at all-time highs is the first move of a sequence that will likely include the S&P at all-time highs, real estate at all-time highs, crypto at all-time highs, and a generalized sense in the financial commentary that the post-COVID environment has finally normalized into a sustained expansion. The expansion will be real. It will also be funded by leverage that the banking system is no longer required to absorb if the cycle reverses, in an institutional environment where the Fed’s traditional role as countercyclical check has been politically neutralized.

There is a particular kind of person who reads regulatory news like this and quietly moves a percentage of their assets out of bank deposits and equity-correlated holdings. These people are not panicked. They are not making predictions about the timing of the next crisis. They are simply observing that the regulatory architecture that prevented the last crisis has been undone, and that the absence of that architecture means the next credit cycle will play out differently from the last.

They are adjusting their exposure to that fact.

You do not have to be that kind of person to take the structural point. The structural point is just that capital requirements are crisis prevention. Reducing them in good times is the institutional decision that produces the next crisis. The Fed Chair installed to ratify and extend that reduction was on the same Fed that dismissed the warnings before the last crisis. The institution that is meant to prevent these cycles has just been politically captured by the cycle’s beneficiaries.

The bank index is at all-time highs. The expansion is real. The next four to seven years will likely feature substantial asset price growth, particularly in financial sector equities, real estate, and digital assets. After that, the architecture of the apocalypse, which is the title of this piece, will do what it was just built to do.

The decision was made. The bill is what is coming.

Reader-funded analysis. No advertisers, no foundations, no editorial board outside this one. If you want it sustained, the door is open.

Subscribe now

The World Health Organization declared an international public health emergency last Tuesday. Ebola in the Democratic Republic of Congo. Six hundred cases. One hundred and thirty-nine suspected deaths. Confirmed cross-border spread to Uganda. The outbreak is fewer than three months old.

The last Ebola epidemic of any significant scale killed more than eleven thousand people across West Africa over two years.

The two numbers — six hundred today, eleven thousand last time — should be looked at together. They tell you something about what is possible to contain and what is not, and what changes between epidemics that determine which outcome happens.

The 2014-2016 outbreak was eventually contained. The containment was expensive and slow, but it worked. The mechanism that worked is worth describing precisely, because the same mechanism is now gone.

In 2014, when cases crossed into Guinea, Sierra Leone, and Liberia, the United States deployed three thousand military personnel under Operation United Assistance. USAID coordinated the broader civilian response. The CDC stood up an Emergency Operations Center. The State Department coordinated diplomatic mobilization. The Department of Defense built field hospitals. The total commitment ran into the multiple billions of dollars. The pattern was: catch outbreaks at the source, mobilize the apparatus before exponential growth started, and treat the cost as cheaper than the alternative.

The alternative — letting the outbreak run — was eleven thousand dead, eight hundred million dollars in long-term economic damage in the affected countries, and serious cases imported into Spain, the UK, Germany, and the United States, with one domestic death in Dallas. That was the alternative when the apparatus worked.

The apparatus does not exist anymore.

USAID was dismantled in 2025. Not pared back. Not restructured. Dismantled. The decision was ideological — foreign aid is unpopular with the President’s base, and demolition was easier than defense. A study released this week links the dismantling to rising armed conflict in DRC, Sudan, and other countries previously stabilized by USAID-coordinated programs. The same dismantling removed the field-deployable epidemic-response infrastructure that had worked in 2014. There is no Operation United Assistance equivalent on the books. There is no civilian apparatus to coordinate one. The State Department’s relevant offices have been gutted. The CDC’s international operations have been cut. The Department of Defense will not deploy on a public health mission without civilian coordination, and the civilian coordination capacity is gone.

This is not opinion. This is operational reality. The pandemic response architecture the United States built between 1980 and 2024 was disassembled in eighteen months. The disassembly was a budget decision dressed in ideological language. The bill comes due now, with the first significant outbreak that would have triggered the old apparatus.

Stack on top of this the entry ban policy. The Trump administration’s response to the outbreak — beyond the absence of response capacity — has been to impose entry bans on travelers from the DRC, Uganda, and South Sudan. The bans look like containment, but produce the opposite incentive. Public health epidemiology has known this for decades: travel restrictions imposed on outbreak zones create incentives for case-hiding rather than case-reporting. Governments do not announce cases when doing so would trigger border closures. Sick travelers do not seek care if seeking care means deportation. The cases that get caught early — the ones the apparatus is built to catch — go underground.

The 2014 outbreak was caught at scale, in part, because the WHO and the US worked closely with affected governments to maintain trust. That trust required that countries not be punished for reporting outbreaks. The current US posture is the opposite: report an outbreak and your citizens are barred from travel. The incentive is to delay reporting. Delayed reporting means later detection. Later detection means exponential growth before the apparatus mobilizes — except the apparatus does not exist anyway.

The current outbreak is at six hundred cases. That number could stay at six hundred. It could also become six thousand, then sixty thousand, depending on how fast and how well the regional response works without American coordination. The WHO is doing what it can. Médecins Sans Frontières is on the ground. The Ugandan and Congolese health ministries are working with limited resources. The variable that decided the 2014 outcome — large-scale, well-funded, coordinated US-led international response — is not in the equation this time.

This is not bad luck.

The reason to belabor this point is that pandemics are routinely framed as natural disasters, things that happen to societies. They are not. Pandemics occur when state machinery for containing zoonotic spillover fails. The state machinery has to be built. It has to be maintained. It has to be deployed before exponential growth makes it useless. The United States built that machinery over forty years and dismantled it in eighteen months. The next outbreak that travels at scale was a budget decision.

It is worth being precise about which budget decision. The dismantling of USAID was decided in early 2025. The CDC’s international operations were cut over the following twelve months. The State Department reorganization affecting epidemic coordination capacity finalized in November of 2025. By the time Ebola crossed into Uganda this spring, the apparatus had been gone for roughly a year. The current outbreak is the first stress test of the post-apparatus regime. Whether it stays contained or does not will tell us what kind of pandemic risk environment we are now living in.

Pandemic risk has historically been treated as something governments handle on behalf of populations. That model assumed governments would maintain the capacity to handle it. When governments stop maintaining that capacity, pandemic risk shifts from collective concern to individual variable. It joins the list of things that the sovereign apparatus used to absorb and now does not: currency stability, judicial independence, regulatory enforcement, federal benefit allocation. Each of those used to be a collective good. Each of those is now a sovereign variable that varies meaningfully across jurisdictions.

The countries that maintained pandemic infrastructure — Singapore, Taiwan, South Korea, Germany, the Nordic states, parts of the Gulf — have continued to maintain it. The United States has not. The pandemic resilience of the country you live in is now something worth knowing about, in the same way that the rule of law of the country you live in is now something worth knowing about. These were not variables five years ago because the assumption held that institutional capacity was floor-level adequate across OECD countries. The assumption no longer holds.

This is the wider point, and it is worth saying directly. Sovereign variables are things you assumed were given and discover are not. Pandemic infrastructure is one of them now. The country that catches the next outbreak at 600 cases will have a meaningfully different experience from the country that catches it at 60,000. The choice of which country you are in is no longer fully decided by where you happened to be born.

Whether the current Ebola outbreak will remain at 600 is genuinely unknown. The regional response without American coordination might hold. The Ugandan health ministry has experience with previous, smaller outbreaks. The WHO and MSF will do what they can. The current strain may be less transmissible than the 2014 strain. There are several variables that could keep the number small.

There are also several that could not. The regional response is operating with a fraction of the resources the 2014 response had. The entry ban regime is creating reporting incentives that work against detection. The next time the outbreak crosses a border, there is no Operation United Assistance to deploy. If the current strain proves more transmissible than expected, or if the regional response runs out of resources before the outbreak peaks, the trajectory looks much more like 2014 than like the contained smaller outbreaks of recent years.

Either way, the next outbreak is coming. There will always be a next outbreak. Zoonotic spillover is a constant feature of human-animal interface in a world of eight billion people, billions of animal-vector contacts daily, and accelerating habitat encroachment. The question has never been whether the next pandemic will happen. The question has always been whether the apparatus catches it before it scales.

The apparatus does not exist anymore. The next pandemic was a budget decision. The decision was made. The bill is what is coming.

The Long Memo runs on subscribers. The Editorial Board is the room where the work gets steered — monthly Q&A, member-surfaced topics, a hand in what gets researched next. If this is the analysis you want more of, that is the door.

Join The Editorial Board

I was talking with one of my closest friends yesterday — a man who probably would have become a general were it not for his Army MOS. Vietnam veteran. Multiple tours. Eventually, a colonel.

He was also the man who hired me at the Pentagon.

I said to him, “Why do we say ‘Happy Memorial Day?’ It’s odd, isn’t it? We wouldn’t walk into a funeral and say, ‘Hey! Happy he’s dead! Happy funeral!’ But that’s essentially what we’re doing.”

He replied that prior to Iraq and Afghanistan, most Americans simply didn’t know the military the way they do now. Before the Global War on Terror, the overwhelming majority of Americans had little direct connection to war. Then, over twenty years, roughly 2.3 million Americans served in Iraq and Afghanistan across more than three million deployments. Nearly everyone in uniform knew someone who had been killed or wounded.

And that changes things.

In Iraq, 4,418 American service members were killed and nearly 32,000 wounded. In Afghanistan, another 2,350 were killed and more than 20,000 wounded.

No, those are not the catastrophic casualty levels of the Civil War or the Second World War. Not even close.

But what was different was the repetition.

I knew people at the Pentagon twenty years ago who had done six deployments. Before the GWOT, that would have been almost unimaginable outside of World War II. More than half the soldiers I met had done at least two tours. Many had spent years of their lives deployed overseas.

So perhaps Memorial Day once was mostly hot dogs, beer, and the unofficial start of summer. Maybe it really was just summer dresses, parades, and Nat King Cole singing about those “lazy, hazy, crazy days of summer.”

And honestly, it should be that.

It should be the peace paid for in blood.

Years ago, I visited Normandy.

Places I had studied for decades, I finally stood in myself: Omaha. Utah. Sword. Caen. Vierville. Pointe du Hoc.

We visited the German batteries overlooking the coast. Rusting artillery still sat there facing the English Channel.

The guns silent.

We have all seen the movies. Saving Private Ryan. Band of Brothers. The Longest Day. We think we understand what war looked like from those films.

I realized almost immediately that we do not.

Point du Hoc shattered every cinematic illusion I had ever carried about war.

The cliffs are not dramatic in the Hollywood sense. They are worse. Roughly 120 feet high, unstable, jagged, and unforgiving. From the top, you can see nearly the entire landing area below — which means the Germans could too.

Two hundred and twenty-five Rangers landed there on D-Day. By the time they were relieved, fewer than ninety remained combat effective.

Standing there, it became obvious that this was not Robert Wagner and Paul Anka heroically scaling cliffs in a war movie.

It had to have been terror.

Chaos. Confusion. Men drowning, slipping, screaming, climbing through mud and blood while machine guns fired down from above.

The films suddenly felt less like depictions of war than faint echoes of it.

But Omaha Beach was what truly overwhelmed me.

And I say this sincerely: even Saving Private Ryan understates what Omaha must have been like.

You cannot understand that landing until you stand there yourself.

The beach is enormous.

An immense killing field overlooked by bluffs, bunkers, artillery positions, and machine-gun nests with commanding fields of fire. The sand itself is soft and exhausting underfoot even in peace and sunlight. The day I stood there, families were walking the beach while children played near the water.

Broad, sunlit uplands.

That phrase from Churchill suddenly became real to me there.

I walked from the waterline toward the bluffs, trying to imagine what the men of the 1st and 29th Infantry Divisions faced that morning.

The Higgins boats approached through rough seas under artillery, mortar, and machine-gun fire. If you survived long enough for the ramp to drop, you then had to cross what felt like seven football fields of open beach carrying nearly eighty pounds of gear while being shot at from elevated positions in multiple directions.

The only meaningful cover was the shingle — little more than a slight rise in the sand followed by a shallow trench at the base of the bluff.

That was it.

That tiny depression in the beach became the difference between life and death.

And beyond it still stood the bluffs, the bunkers, the machine guns, and the artillery.

Standing there, you realize the sheer physical improbability of the assault. Not in the abstract. Not as arrows on a map. As human beings.

Young men.

Ordinary men.

That realization stays with you.

But what struck me most was the contrast between what happened there and what exists there now.

People laughing. Windsurfers crossing the Channel. Families enjoying summer along the French coast. Peaceful towns full of cafés, flowers, churches, and ordinary life.

Churchill once said:

“If we can stand up to him, all Europe may be freed and the life of the world may move forward into broad, sunlit uplands.”

That was what I saw at Normandy.

The bunkers still exist. Craters from bombs still scar the countryside. The remnants of the Atlantic Wall still sit facing the sea.

But the life of Europe moved forward.

Broad, sunlit uplands.

Then you arrive at Colleville-sur-Mer.

Arlington is solemn and moving. I have been there multiple times, including for military funerals.

But the American cemetery above Omaha Beach is something else entirely.

The entire cemetery feels American. Virginia bluegrass. Trees native to the United States. Perfectly maintained grounds overlooking the coastline where so many died.

It is as if those soldiers remain home, even in France.

Row after row of crosses and Stars of David.

Every one of them someone’s son.

Someone’s husband.

Someone’s brother.

General Mark Clark once said:

“If ever proof were needed that we fought for a cause and not for conquest it could be found in these cemeteries. Here was our only conquest: all we asked was enough soil in which to bury our gallant dead.”

France granted that land to America in perpetuity.

One comment from our Parisian tour guide has stayed with me ever since. He said Paris is beautiful — but that beauty carries shame, because the city was largely spared the destruction that consumed so much of Europe through surrender, collaboration, and capitulation.

He pointed out that unlike Poland, Italy, or much of Normandy, Paris bears few scars of war.

Then he said something I have never forgotten:

“The Americans had no reason to liberate France except that they believed freedom was worth fighting for.”

Standing there, I believed him.

I have seen extraordinary Americans in uniform. I had the honor of working beside many of them at the Pentagon.

But never have I been prouder to be an American than standing on those beaches.

The Big Red One.

The Rangers at Pointe du Hoc.

The 101st Airborne at Sainte-Mère-Église.

The stories are all there. Not as movies. Not as mythology.

As reality.

And what overwhelms you when standing there is not simply the scale of the courage, but the humanity of it.

They were not superheroes.

They were ordinary men trying to survive, protect one another, and come home.

That is what made them heroes.

That is why they deserve remembrance.

Not because they were born different from us.

But because, when history demanded it, ordinary men carried unbearable burdens and endured unimaginable terror so that future generations could live ordinary peaceful lives.

That is Memorial Day.

The peace paid for in blood.

Stephen Ambrose wrote in Band of Brothers that Lieutenant Harry Welsh remembered looking at the sleeping men around him before D-Day and realizing that, despite death surrounding them, none of them truly believed it would happen to them.

“They hadn’t come here to fear. They hadn’t come to die. They had come to win.”

For those remembering loved ones who never came home, perhaps no words say it better than Abraham Lincoln:

“I feel how weak and fruitless must be any word of mine which should attempt to beguile you from the grief of a loss so overwhelming. But I cannot refrain from tendering you the consolation that may be found in the thanks of the Republic they died to save.

I pray that our Heavenly Father may assuage the anguish of your bereavement, and leave you only the cherished memory of the loved and lost, and the solemn pride that must be yours to have laid so costly a sacrifice upon the altar of freedom.”

The United States admitted six thousand refugees between October and April. All but three of them were from South Africa.

This is a fact you cannot get around. It is not an interpretation. It is not a partisan analysis. It is what happened. The federal refugee program — created in 1980 with broad bipartisan support, designed to settle people fleeing war, persecution, and famine — admitted six thousand people last year, and all but three were from one ethnic group from one country.

The administration’s current proposal is to raise the refugee cap from 7,500 to 17,500. The additional ten thousand are explicitly reserved for Afrikaners under Executive Order 14204. The federal register language, in case you’re keeping a copy for posterity, says that admissions “shall primarily be allocated among Afrikaners from South Africa.” The euphemism is “victims of illegal or unjust discrimination in their respective homelands.” The math is that the policy admits one ethnic group while keeping the door shut on everyone else fleeing war, famine, or anything resembling persecution by any neutral definition.

Pause on this. Look at it directly.

This is a federal benefit allocation based on explicit racial criteria as a written, signed, and published policy. Not informal favoritism. Not deniable patterns.

Policy.

It is also not the only example.

The Justice Department created, last week, a one-point-seven-seven-six-billion-dollar fund designed to compensate “victims of weaponization” by previous administrations. The fund disburses to a categorically determined group: Trump’s political allies, January 6 pardonees, and others described as having been “prosecuted for political reasons” under Biden and Obama. The Acting Attorney General who approved the fund is Trump’s former personal attorney. Federal money. Federal infrastructure. Allocated by political alignment with the current administration. Written down. Posted on the DOJ website.

USAID has been dismantled. A study released this week links the dismantling to rising conflict in DRC, Sudan, and other African countries — the apparatus that caught outbreaks and contained conflicts at source no longer exists. The choice was ideological: the work was disliked by the President’s base. The downstream consequences — Ebola at six hundred cases and rising, conflict in regions previously stabilized — are accepted costs.

FEMA disaster declarations have followed a pattern for years now, with federal emergency assistance approvals favoring states aligned with the administration. The pattern is documented across multiple data sources. It used to be denied or attributed to coincidence. It is no longer denied.

The IRS settlement with the President’s family, which barred future examination of their tax returns “forever and precluded” — is the same move applied to enforcement. Tax obligation is now categorically different for one family.

This is a list.
It is also a pattern.

The pattern has a name.
The name is patronage.

Patronage was the rule in most of human political history. The state distributed favors and resources based on political loyalty because loyalty was the foundation on which the state rested. Roman emperors paid their legions and their clients. Medieval kings granted lands to lords who fought for them. The Ottoman millet system organized resource allocation according to religious and political affiliation. The whole concept of citizenship as something distinct from loyalty — as a status with attached rights independent of whether you supported the regime — is a relatively recent invention in human governance.

The American Republic was a particular experiment in that invention. The federal compact rested on the premise that federal resources flow by neutral criteria — by law, by need, by rule. Yes, every administration has played favorites at the margins. Yes, patronage existed before civil service reform; it exists now around the edges. But the premise held. The premise held because the formal architecture treated federal benefits as obligations under law, not gifts from a patron to clients.

That premise is ending. And what’s ending it isn’t a turn toward corruption — corruption has always existed at the edges. What’s ending it is formalization.

The refugee allocation is policy. Written, signed, published. The anti-weaponization fund is policy. Listed in the federal register. The IRS forever bar is signed by the Acting Attorney General. The dismantling of USAID is announced. The disaster funding patterns are no longer denied. Things that used to require euphemism — things that used to need a story about why this state didn’t qualify for FEMA, or why this contract went to that vendor, or why this prosecution proceeded but that one didn’t — those things now sit on official letterhead.

This matters because the asymmetry of incentive structure matters.

Informal favoritism creates pressure for restraint. Deniability is itself a resource. The official who steers federal money toward an aligned vendor has reason to be careful — explicit racism, explicit cronyism, explicit ideology can be challenged. The career cost of getting caught is high. So the favoritism stays at a level the system can absorb.

Formal favoritism removes that pressure. Once “primarily allocated among Afrikaners” is in the federal register, the incentive to restrain the next allocation falls. Once the anti-weaponization fund is announced as an official program, the next disbursement to political allies needs no euphemism. The institutional memory captures the new norm. Federal employees orient around it. State governments orient around it. Federal contractors orient around it. The Republic-era assumption — that federal funds flow by neutral criteria — becomes a museum artifact. People stop expecting it. And once people stop expecting it, restoring it is no longer a matter of policy. It is a matter of restoring the assumption, which is a generational project.

So the question of whether we are “one nation” or fifty states is not rhetorical.

Look at the map and tell me with a straight face that federal benefits flow neutrally across it. Look at the disaster declaration patterns, the federal grant allocations, the refugee allocations, the contract awards, the investigation patterns, the regulatory enforcement patterns, the tax enforcement patterns. Federal money has acquired a Voter ID. So has federal protection. So has federal regulatory enforcement, and federal benefit allocation, and federal contract distribution.

This is not corruption in the conventional sense. Corruption is transactional — someone gets paid, someone breaks a rule, someone gets caught. This is constitutional in the conventional sense. It is structural change in how the federal apparatus relates to the population it allegedly serves. The Republic-era model treated all citizens as having equivalent claim on federal benefit by virtue of citizenship. The new model treats citizens of aligned states as clients and citizens of unaligned states as adversaries. Both groups still pay federal taxes. Only one group still receives federal benefit on the previous terms.

There are historical parallels for what this becomes. Peronist Argentina ran for decades on regional patronage. The Russian Federation under Putin allocates federal resource by regional loyalty to the Kremlin. Hungary under Orbán redirects EU funds toward government-aligned regions and businesses. The pattern is not unprecedented. It is just unprecedented in modern American history, post-civil-service reform.

What ends if this continues is not “American democracy” in some abstract sense. What ends is the federal compact specifically. The agreement that all states stand equal before federal law and have equivalent claim on federal benefit. That agreement is what made fifty states a federation rather than a coalition of clients around an emperor. When the agreement is gone, the federation is gone, even if the map still shows the same borders.

Twenty-seven is the rough count of states whose population voted for the current administration in the last election. The number isn’t exact and isn’t the point. The point is that there is now a structural distinction between states whose citizens have federal protection and states whose citizens have federal punishment, and the distinction is the writing on the wall.

If you live in one of the twenty-seven, congratulations. You are inside the federation. Your refugee allocations flow. Your disaster declarations get approved. Your contracts get awarded. Your prosecutions don’t proceed if the right people don’t want them to.

If you live in one of the other twenty-three, you are in the same legal country but a different political one. The federal apparatus still taxes you. It still requires you to follow its laws. It still mandates your participation in its institutions. But the benefits run the other way.

The question worth asking — and worth asking now, because the formalization is what makes it hard to reverse — is whether this is a temporary deviation that the next administration corrects, or whether this is the new operating system. The institutional architecture being built suggests the second.

If it is the first, you have still lost ground that takes a decade to recover. If it is the second, you are already in a different country than the one you thought you were in.

Either way, the federation you assumed you lived in is not the federation you live in now. That is the diagnosis. The question that follows from it — what you do when the rules you follow are no longer the rules being applied — is the one this newsletter, and any other honest one, is going to keep returning to.

This kind of analysis doesn’t exist without readers paying for it. No ads, no foundation, no outside agenda steering the work. If you want more, the upgrade is right there.

Subscribe now

Two things stood out from the actual source material. Neither of them surfaced during a single one of those performances.

First. Ted Cruz, on his own podcast, said this about the $1.8 billion DOJ “anti-weaponization” fund — the same fund that allegedly prompted the alleged Republican revolt against Trump’s alleged self-dealing: he is “a big supporter of it.” Not a reluctant one. Not a politically queasy one. A big supporter. He spent additional minutes explaining that the fund tracks precedent from past judgment funds settling litigation against the United States, and that the legal arguments Acting Attorney General Todd Blanche gave to the GOP caucus in the closed-door meeting were the right ones.

Read that twice. The senator at the alleged center of the alleged revolt just endorsed the alleged grift on his own show. His complaint was about the timing of the announcement — that it landed during reconciliation week and stepped on the ICE/Border Patrol bill. Not the substance. Not the structure. The calendar.

Second. Cruz then proceeded to explain — with a forensic enthusiasm that ought to embarrass the man — that if Senate leadership had brought reconciliation to the floor Thursday night, half his own caucus would have voted with Democrats on amendments to constrain the fund. Not kill it. Constrain it. The specific amendments coming down the pike, per Cruz: language preventing payouts to rapists, violent offenders, and people who assaulted police on January 6.

These are not amendments offered by people opposed to the policy. These are amendments offered by people who support the policy and need it cleaned up so they can keep supporting it without political embarrassment.

That is not a revolt. That is brand management.

I dealt with this exact species of hypocrisy during the GWOT years. The argument was never “We oppose extraordinary rendition.” The argument was “We support extraordinary rendition but dislike the political optics once photographs and lawsuits appear.” Likewise: “We don’t want terrorists attacking Americans.” Fine. But: “It becomes politically awkward when you must warehouse suspects indefinitely, conduct black-site transfers, or explain to European parliaments why CIA aircraft keep landing at odd hours.”

And oh, did that last one become inconvenient.

The rendition issue damn near detonated multiple allied governments. Australia took damage. Germany convulsed. Spain practically vaporized itself over it.

The point is the same now as it was then: Republicans are aligned with Trump’s objectives. They are merely uncomfortable with Trump’s sales pitch. That distinction matters enormously.

A fairness note, since I am not interested in the lazy version of this argument. Not every Republican is Ted Cruz. Thom Tillis calling the fund a “payout pot for punks” is not brand management — Tillis is a known Trump skeptic with a record of substantive opposition. Bill Cassidy’s complaint that Congress was given no role in the design of the fund is an actual Article I objection, not optics theater. (Both Tillis and Cassidy are not facing re-election, which is when apparently many members of Congress find both a spine and a conscience.) Brian Fitzpatrick co-sponsoring bipartisan legislation to prohibit federal money for the fund is policy opposition, not rhetorical squeamishness. There are three or four senators doing real work here.

Three or four is not enough.

The median Republican senator is doing exactly what Cruz did on his own podcast: endorsing the substance, blaming the calendar, and demanding cosmetic carveouts so the policy survives the news cycle. Three or four genuine objectors do not constitute a revolt. They constitute a footnote that gets cited in 2028 campaign mailers to demonstrate independence from a president the rest of the caucus was busy enabling.

Americans getting shot by ICE? Most Republicans will support it because they’ll cite “law enforcement.” People getting their skulls bounced off pavement by federal agents? They’ll shrug and ask whether the detainee had “papers.” Using children as bait in enforcement operations? You’ll hear muttering about “tough choices” and “operational necessity.” Nine-dollar gasoline after a Middle East escalation? Their concern is not the suffering. Their concern is whether voters associate the suffering with them by November.

Policy alignment remains intact. Only the rhetoric produces indigestion.

So spare me the cable-news theater about “cracks in the coalition.” Here is what will actually happen. The administration will trim the fund. They will write in eligibility carveouts — no violent offenders, no rapists, no Trumps. Whether that gets actually followed? That’s tomorrow’s problem. They will install a notional independent administrator who will, oddly enough, do what Trump asks. Courts will do “court stuff”, and nothing will change there (because all of the lawsuits I’ve seen, while brilliant in prose, misunderstand the idea of “standing” given current Supreme Court doctrine). Republicans will declare victory, vote for reconciliation, and pat themselves on the back for “holding the line.”

That is not a line being held. That is a line being repositioned six inches to the right and renamed “victory.”

There is no revolt. There are Republicans arguing over the deck chairs as the ship goes under.

And the band — played — on.

Stephen Colbert is apparently back on the air.

Sort of.

This aired Friday on some community television station in Monroe County. Which is objectively hilarious.

Now, as someone who has actually worked around broadcast production — commercials, television, agency work, the machinery behind the curtain — what fascinates me isn’t the joke.

It’s what the joke reveals.

One of the recurring spectacles on Substack and YouTube is watching recently unemployed television personalities discover that the product was never entirely them.

A famous anchor gets fired.

A cable host storms off declaring tyranny.

Then they launch their “independent media revolution” from a spare bedroom with the lighting quality of a hostage video and audio that sounds like two squirrels fucking inside a tin garbage can.

And suddenly you realize something:

Television was carrying them.

Not entirely. But substantially.

Because television is not one talented person. Television is an industrial apparatus. Lighting crews. Audio engineers. Producers. segment editors. Camera operators. Graphics. Booking teams. Union technicians. Makeup. Control rooms. Standards departments. Satellite coordination. Post-production.

A small army exists to make one person appear larger than life.

Take away the machine and many of these people collapse into amateur hour almost immediately.

Don Lemon was brutal when he first moved online.

Andrew Weissmann, too.

Not because they’re stupid. Because they discovered that charisma and competence are not the same thing as production.

For years, I believed production quality mattered enormously.

This Colbert bit actually convinced me I was only partially right.

Because this works.

Even though it is deliberately styled to look like Community Access Television from 1994.

Which means one thing:

Colbert himself was most of the product.

That’s rarer than people think.

Most hosts are interchangeable once you remove the machinery.

Colbert is not.

And the fact that this little spoof already works tells me something else most viewers probably missed entirely.

CBS must have released him from his contract. Alternatively, his contract included provisions that made even a parody show possible.

That is extraordinarily interesting.

Normally, major broadcast talent contracts contain enough non-competes, holdbacks, exclusivity clauses, and poison pills to keep someone from popping up on competing platforms immediately after cancellation.

Even if the show dies.

Especially if the network believes the talent still has commercial value.

The whole point is to stop another platform from harvesting the audience you spent years building.

So the fact Colbert can immediately appear in broadcast-style programming — even as a joke — strongly suggests CBS decided it was worth eating whatever penalties, buyouts, or contractual costs existed simply to terminate the relationship cleanly and completely.

That is not consistent with the simplistic “the show was losing money” explanation.

It is entirely consistent with ownership deciding that Colbert had become unnecessary merger friction.

If the economics alone were the problem, the last thing you would do is leave the door open for Colbert to immediately rebuild elsewhere.

And here’s where the larger story gets interesting.

Fifteen years ago, I had dinner with senior YouTube executives in Palo Alto. They openly told me they intended to put CBS, NBC, and ABC out of business.

I laughed in their faces.

Not because the idea was insane, but because at the time, they fundamentally did not understand television economics.

They thought distribution was the business.

It isn’t.

Attention is the business.

Television exists to aggregate eyeballs so advertisers can purchase access to human attention at a profit margin higher than the cost of acquiring the audience.

The programming is bait.

The commercials are the actual product.

For years, YouTube thought creators were just uploading content.

Eventually, they realized creators were building audiences.

Then they discovered something television learned generations ago:

Audience distribution becomes grotesquely concentrated.

A handful of creators absorb most of the attention.

Everyone else fights over scraps.

Substack works exactly the same way.

A few people get most of the audience.

A middle layer gets enough to build a respectable business.

Then there are ten billion people screaming into the digital abyss, hoping algorithmic Jesus notices them.

That’s the actual media economy.

Joe Rogan figured this out before almost anyone else.

Rogan is not a philosopher king. He’s a B-level comic, former TV ensemble player, former cage-fighting color commentator, who built an empire by normalizing an endlessly adolescent masculine archetype and attaching a gigantic audience to it.

That audience became economically valuable.

That’s it.

That’s the whole game.

And someone like Colbert already possesses the hardest part:

The audience.

Millions of people will watch this stupid little public-access parody.

The real question is whether they come back next week.

I suspect they would.

At which point the economics become very interesting very quickly.

Because a stripped-down streaming show with competent production and a loyal recurring audience may actually be a better business than legacy late-night television.

No orchestra.

No gigantic studio overhead.

No bloated staffing.

No decaying broadcast infrastructure.

Just audience aggregation and monetization.

Which is ultimately what television was always doing anyway.

John Oliver already understands this.

Kimmel increasingly does too.

The center of gravity for commentary and comedy is migrating toward streaming because that is where the audience now lives.

Especially the audience that still wants something outside the endless slurry of algorithmic sludge and political psychosis.

And frankly?

Colbert may be one of the first legacy television personalities with enough actual talent to make the transition work at scale.

Not because he was on television.

Because, unlike many television personalities, he might actually survive without it.

There is a thing wealthy men do when the question of taxing them gets serious. They get a microphone, find a podium with sufficiently moral lighting, and propose a tax cut for someone else.

Jeff Bezos did it Wednesday.

He sat down with Andrew Ross Sorkin at Blue Origin’s Merritt Island facility — the place where Bezos burns roughly the GDP of a small country to launch in what most people describe as looking like a penis, filled with rich people, briefly into low Earth orbit — and proposed eliminating federal income tax on the bottom half of American earners. Not reducing it. Eliminating it. “Zero is a better number than $1,” he explained. The nurse in Queens making seventy-five thousand dollars a year, Bezos said, shouldn’t be paying more than a thousand a month to Washington. They should be sending her an apology instead.

This is going to sound populist. It isn’t.

The proposal is a trade. Read the whole thing.

What Bezos is offering: the bottom half of American earners — adjusted gross income around fifty-four thousand a year, accounting for roughly twelve percent of total income and about three percent of total federal income tax revenue — pay zero federal income tax. Eliminated. The Fortune write-up ran the example: a nurse in Queens making seventy-five thousand dollars, about twelve thousand a year. Real money to her. Pocket change at the aggregate.

What Bezos is asking for: permanent insulation against any meaningful taxation of concentrated capital.

The asking isn’t in the proposal itself. It’s in the architecture around the proposal. Bezos delivered the pitch from a rocket factory two months after Elizabeth Warren introduced the Ultra-Millionaire Tax Act of 2026 — two percent annually on households over fifty million, an additional one percent on billionaires. He framed the proposal explicitly against that lineage. “You could double the taxes I pay,” Bezos told Sorkin, “and it’s not going to help that teacher in Queens. I promise you.” Elon Musk responded “Bravo” on his platform within the hour. Bezos plans to lobby Trump personally.

This is not a man worrying about a nurse.

This is a man worrying about a wealth tax.

Pay attention to the structure. The bottom-half tax cut costs the Treasury maybe sixty to a hundred billion a year — meaningful, but smaller than the wealth-tax revenue Bezos is foreclosing as a political possibility. He’s offering the bottom fifty percent of taxpayers a few hundred billion in cumulative relief over a decade in exchange for off-the-table treatment of capital concentration in perpetuity. From his portfolio’s perspective, that’s a discount trade. From the median household’s perspective, it’s the wrapper around a transaction the median household isn’t in the room for.

The extraction economy worked the same way the last twelve times.

Money stops buying things at some scale. It starts buying leverage. Bezos’s two hundred seventy-nine billion dollars isn’t shopping for a marginal tax cut on labor income — he doesn’t have meaningful labor income to cut. His exposure is on capital. ProPublica reported that he paid zero federal income tax in 2007 and 2011. Amazon, the company he chairs, watched its federal tax bill drop from nine billion in 2024 to one point two billion in 2025 under the One Big Beautiful Bill Act, while it laid off thirty thousand workers across the fall and winter. The Bezos proposal isn’t about his tax bill. It’s about ensuring the bottom-half tax cut becomes the political ceiling — “we already cut taxes; what more do you want?” — and the wealth-tax floor falls out from under serious consideration.

That is the trade. The nurse in Queens is a stage prop.

Read more