Start with a number that has no drama in it at all. In the first eleven months of fiscal 2026 the United States collected $4.845 trillion and spent $6.811 trillion.
The gap is $1.966 trillion. The Congressional Budget Office now expects about $2.1 trillion for the full year.
That is not news. It has been roughly true for years, and nothing happened.
Here is what changed underneath it.
The line that crossed
Interest on the national debt will run about $1.0 trillion this fiscal year. That is 3.3 percent of everything the country produces.
It is also more than the United States spends on defense. All of it. Every ship, every base, every salary, every missile.
The largest single thing the federal government now buys is the past.
Six pressures, and what kind each one is
This framework asks one question of anything that pinches. Is the scarce thing money, or is it the thing itself?
Run the six through it.
| Pressure | Where it stands | Real or monetary |
|---|---|---|
| Oil | Brent at $104.32 on 25 September, up about 20 percent in a month. Transit through Hormuz is selective and tolled. | Real. A strait is not a credit condition. |
| The Iran war | Direct operations put near $42 billion, an $87 billion supplemental requested, $97.3 billion sought for munitions this year. | Real. Production runs 36,000 shells a month against a 100,000 goal. |
| The AI build | Capital spending near $630 billion this year. Roughly 8,200 projects wait in the interconnect queue. | Real. Transformers, switchgear, electricians, grid. |
| Labor | Unemployment 4.1%, near a four year high. The work week has not moved from about 34.3 hours in twenty years. | Real. Labor cannot manufacture more of itself. |
| The deficit | $1.966T in eleven months. Interest near $1.0T a year and rising. | Monetary, and the transmission belt. |
| The bond market | Ten year at 4.94%, having touched 5.016 on 16 September, a nineteen year high. | The scoreboard. Where the other five get totaled. |
Four of the six are physical. Not one of them gets shorter because money shows up.
That is the unusual part, and it is worth sitting with.
The site keeps an index for exactly this distinction. The Credit Counterfeit Index reads minus 1.24. The monetary side of it sits at minus 0.09, which is to say at nothing. Credit spreads are among the calmest in the whole series. Banks are not tightening.
Money is not scarce in America. Almost everything else is.
Where the five arrive
Now the mechanism, because these are not six separate stories.
A physical shortage raises prices. That is what a physical shortage does. Oil at $104 raises the cost of everything that moves. A grid that cannot connect 8,200 projects raises the cost of power. A labor force that cannot add hours raises the cost of work.
Rising prices raise the inflation a bond buyer has to expect. A thirty year lender is not pricing today. They are pricing the average of the next thirty years.
So the long rate rises. It is at 4.94 percent and it touched 5.016 this month, the highest in nineteen years.
And a government borrowing two trillion dollars a year pays that rate on every new dollar and every refinanced one. Interest becomes $1.0 trillion. Then more.
Which means more borrowing, which means more issuance, which means a buyer demanding a little more yield.
The one that looks like nothing
Artificial intelligence deserves a separate note, because the honest reading of it cuts against the noise.
There is no economy-wide AI jobs shock in the data. None. Three and a half years after the technology arrived, the softening in American and British hiring looks cyclical. The unemployment gap between exposed and insulated workers is small and statistically insignificant.
Anyone telling you the robots took the jobs is not reading the aggregate.
But look one level down and it is not nothing at all.
Stanford finds a 13 to 16 percent relative fall in employment for 22 to 25 year olds in the most exposed occupations. British data shows exposed job adverts down 38 percent. The damage is concentrated at the entry rung, in the jobs where output is easy to verify.
That shape should be familiar. It is the same defect this site has already published about its own instrument. An aggregate reports health for as long as the large majority is healthy. It says nothing whatsoever about the people for whom the thing binds.
So AI is not yet a labor story. It is a distribution story, arriving quietly, at the door people use to get in.
Three paths, and the marker for each
Forecasting a number is a mug's game. Naming the paths and the tell for each one is not.
The grind. The real constraints persist. Inflation holds somewhere between 3 and 4 percent. The Federal Reserve holds or raises again. The ten year stays near 5. Interest compounds and the deficit widens. Nothing dramatic happens for years.
This is the upper right of the phase plane, where prices rise and quantities do not. It is where the plane already places this economy. It is also the likeliest path on current readings, and the cruelest, because there is no moment anyone can point at.
The marker: inflation sticky near 3.5 while unemployment drifts up. Both at once, for several quarters.
The rotation. Something snaps in credit. The monetary bind turns positive and climbs. Then the counterfeit shortage arrives on top of the real one. A country already short of oil and labor becomes short of money too.
That is what the treatise calls the bind. Two scarcities, one instrument. The tool that fixes the second makes the first worse.
The marker: the monetary bind crossing zero while credit spreads widen past 400 basis points. Watch the gauge's money side, not the headline number.
The release. Hormuz opens properly. Brent falls under $80. The real bind recedes on its own, inflation follows it down, the long rate falls, and interest costs stop compounding so fast.
This is entirely possible and it is not an economic forecast. It is a diplomatic one, which is a different discipline and a worse-behaved one.
The marker: open transit, not a ceasefire announcement. The June memorandum was announced and did not hold.
What this dispatch is not saying
It is not predicting a crash, and the site's own instrument is the reason.
The gauge sits at minus 1.24 against a threshold of plus 1.0. That is a distance of 2.24. Getting there takes the money side rising by about a full standard deviation. Credit spreads at crisis levels. Banks refusing to lend at net shares in the tens rather than the low single digits. A real interest rate several points higher than today.
None of that is present. None of it is close.
A framework that shouted crisis whenever the news was bad would be useless. This one has spent the year saying the opposite of what a reader might expect.
The forecast here is duller and probably more accurate. The machinery keeps running. The bill keeps growing. And a country spends more each year on what it already owes and less on what it might build.
What would prove this wrong
Three things, all public, all dated.
Say the ten year falls below 4 percent while the deficit stays near $2 trillion. Then the bond market is not pricing these constraints the way this piece says. The transmission belt does not run.
If inflation returns to 2 percent with Brent still above $100, the link is much weaker than argued here. Physical shortage would not be setting the price level.
And say the entry-level gap for 22 to 25 year olds closes while AI adoption keeps climbing. Then the distribution reading is wrong. The aggregate was telling the truth all along.
Two cautions. The full-year interest and deficit figures are projections. The receipts and outlays are Treasury actuals through August. Those are not the same kind of number and should not be added as if they were. And the three paths are not exhaustive. They are the ones this framework can see, which is a narrower claim than the ones it can name.