A family earning $42,000 a year, roughly the 25th percentile of U.S. household income, lost about $1,700 in purchasing power over the past 12 months to a CPI running at 4.1%. That is $142 a month that used to buy groceries, gas, or a kid's prescription. The 2.1% real spending growth figure that anchors every soft-landing argument does not describe this family. It describes the average of this family and a dual-income household in Scottsdale putting $900 a month into a brokerage account. Averages are comfortable. They are also how you miss a recession until it is already here.
Economist Saidel-Baker's point about bifurcation is the most important data footnote of 2026, and almost nobody is pricing it in. The bottom quartile of earners is not "slowing." They are contracting. When your wage gains run 3% and your cost of living runs 4.1%, you do not decelerate. You fall behind. The Q4 2025 consumption downshift that J.P. Morgan flagged did not come from nowhere. It came from the part of the consumer base that runs out of buffer first.
The 2007 Problem
In the summer of 2007, aggregate consumer spending was still positive. GDP printed 2.5% in Q2. The S&P 500 hit an all-time high in October. None of that prevented a recession from beginning in December, because the stress was concentrated in a segment, subprime borrowers, that the averages smoothed away. I am not arguing the mechanism is identical. I am arguing the pattern is: broad metrics stay green while a specific cohort quietly breaks, and by the time the aggregate catches the signal, the damage is structural.
The 16% S&P 500 earnings growth number is real. I will grant that. Corporate America is generating cash. But earnings quality matters more than earnings growth at this stage of the cycle, and a significant share of that growth is coming from pricing power, which is another way of saying companies are passing costs to consumers who are already stretched. That works until it doesn't. Revenue beats built on price increases rather than volume gains are the financial equivalent of burning furniture to heat the house.
20 Million Barrels a Day Through a Chokepoint
The Strait of Hormuz carries roughly 20% of global oil supply. The IMF's base case already assumes a 19% energy price increase baked into 2026. Their severe scenario, a prolonged shutdown, drops global growth to 2% and pushes inflation above 6%. Larry Fink's $150-per-barrel warning is not a forecast, true. But it is not a fantasy either. The distance between $80 oil and $150 oil is one military miscalculation in a conflict that has already produced several.
For that family earning $42,000, a 30% spike in gasoline prices means roughly $75 more per month at the pump. Stack that on top of the $142 they already lost to inflation. Now you are looking at $217 a month in vanished purchasing power. No tax refund covers that for long.
The fiscal stimulus that J.P. Morgan expects to drive a first-half rebound is real, but stimulus checks do not fix a supply shock. They add demand into a system where the constraint is supply. We ran this experiment in 2021. The result was the inflation we are still paying for at 4.1%.
A 35% recession probability from J.P. Morgan means their models see a 1-in-3 chance the expansion ends. That is not reassurance. That is the odds of rolling a 1 or 2 on a die. You would not board a plane with those failure rates.
The soft landing requires every assumption to hold: the Hormuz stays open enough, oil stays below $100, lower-income spending stabilizes instead of collapsing, and 16% earnings growth survives a volume slowdown. Miss on 1 of those and the landing gets harder. Miss on 2 and the word "landing" stops applying. The consumer who makes this work, the one spending 2.1% more in real terms while absorbing 4.1% inflation and a geopolitical energy shock, is an abstraction. The real consumers are splitting apart, and the fracture is where recessions begin.
