Vancouver – The Federal Reserve is expected to raise interest rates this week. It will not do so because the American economy has suddenly overheated, nor a quarter-point move will produce a barrel of oil. It will do so because Chair Kevin Warsh made 2% a test of the new central bank, declined to move after talking tough in July, watched the bond market treat that as a bluff, and then received an August inflation report — and a war-driven jump in fuel prices — that left him no clean way to stand still. The funds rate does not drill crude. It raises the cost of the next payment. Wednesday’s meeting is less a fresh decision than the Fed following through on itself.
Warsh did not invent the 2% target. He turned it into a character test: firm, fixed, and not to be swapped for a friendlier index after the fact. Once a chair frames the number that way, holding rates is no longer patience. It is walking back the institution’s own story. Inflation is not fine. But a single ceiling on a lagged average is a poor constitution for policy. CPI and PCE arrive late. They lean on imputations, owners’ equivalent rent, and substitution. Warsh talked about better data, then pointed the committee back at the old print. The headline is carrying energy. Core is how official seriousness is measured. Neither describes the grocery cart, and neither describes the house that never broke ground.
What looks like strength in the aggregates is a mere illusion. GDP and mega-cap earnings are being held up by AI capital spending and the wealth that spending has marked higher — not by a country getting richer in hours worked and food taken home. The financing should be called what it is. In an August filing, Nvidia disclosed residual-value guarantees of up to $105 billion to help finance OpenAI infrastructure built around Nvidia’s own chips. That is vendor finance: the supplier underwriting the customer who buys the supplier’s product. It shows up in the national accounts as demand. It is not a raise. The boom is narrow and circular. The totals swallow it anyway.
Look through those totals and the economy is still K-shaped, in the same pattern set after the pandemic: assets at the top, prices at the bottom. Research from the Federal Reserve Bank of New York has been explicit. Since 2023, wealth gains have gone mostly to high-income households, while inflation has run hotter for low-income ones. Real hourly earnings have fallen about 0.7% since fuel prices jumped after the war. Mark Zandi’s reading of the accounts is starker. Households in the top fifth of the income distribution — those earning more than $175,000 — now account for nearly 60 percent of outlays. Their spending has outpaced inflation. Spending by the bottom 80 percent has not. National consumption can look resilient under those conditions. Most households still lose to the receipt.
The image is simple. Families buy one carton where they used to buy two, pay more for it, and the accounts record an increase in consumption. The New York Fed finds the same pattern in the large: real spending on necessities has been weak for most groups, while luxury spending has kept rising. A rise in dollars spent on fewer units is not growth. It is a smaller life at a higher price. Working households already tightened — fewer items, smaller tickets, repairs deferred — before the Federal Open Market Committee decided that “demand” needed restraint. The Fed is late to a contraction it will not name.
When that household picture gets uncomfortable, construction is the exhibit that still looks respectable. It shouldn’t. A homebuilder and a data-center contractor are not the same business, even if the forecasts treat them as one fog labeled “builders.”
Residential construction is the part attached to families, and it is sliding. Zillow estimates that residential permitting is running a record 19% below its pre-pandemic trend. Census data show private residential construction down more than 4% in the first seven months of the year, with single-family work leading the drop. Affordability was already broken.
The part attached to a press release is the data center. Through July, Census figures show that category running more than a third above the prior year and accounting for 57% of private “office” construction, even as the rest of office building fell. Associated Builders and Contractors reports the same split in backlogs. The largest firms are full. Smaller firms are at their weakest since 2021. Only about one contractor in eight has data-center work at all.
In a town next to a new campus, housing can look busy. That is usually overflow from the AI build — temporary crews, not a recovered first-time buyer. The National Association of Home Builders has warned that hyperscale projects bid land and skilled trades away from housing. A rate increase will lean on mortgage demand, which is already fading. It will not pause a pre-leased gigawatt project financed by the same circular spending that has been propping up GDP. The official story has been pointing at builders and hoping nobody asked which buildings. That story is not a housing recovery. It is a customer that does not need a mortgage.
A higher funds rate will not restock a dairy case. It will make solvency more expensive: credit cards, car loans, floating-rate debt, the mortgage that still does not close. That is demand destruction aimed at the households already absorbing the inflation in the basket. If stock prices fall, officials and strategists will call it a correction. That is what the channel does, not evidence of a plan. Credit spreads and the long end have already tightened. Official policy would be stacking itself on top.
Two systems are in play. In one, reserves, Treasuries, and dealers set the price of money. In the other, a paycheck still buys milk and almost never buys a house. The Fed’s tool reaches the first as a signal and the second as a bill.
The dual mandate is on the door. In a crunch, the Fed answers the market that can freeze the banks and the Treasury market. 2% is not a reading from the grocery aisle. It is a rule that the unit of account will not be allowed to follow war, oil, and a capital-spending boom concentrated in a few industries. The stress is already visible in ticket sizes, freight, credit, and housebuilding. The Fed will hike because declining to hike would mean the new regime had failed to enforce its own rule. The system comes first.
None of this is an argument for the policy of 2021. Cheaper revolving credit will not fix inflation sitting in fuel, food, and rent. It is an argument that this particular medicine is for the doctor. If the Fed needs to demonstrate independence from the White House, it can say so. It does not need to use the household that already cut the carton — or dropped the house — as proof of courage.
The Fed will hike the index and call it the economy. The economy that buys milk, and the one that builds a house for the person who buys milk, has already moved on.