On why the headline tariff rate is not the supply chain's actual cost. March 1, 2026.
Twenty-five percent. That's the number printed on the executive order. That's the number in every headline. That's the number the administration is defending and the trading partners are calling catastrophic.
It is not the number that matters for most manufacturing.
The 25% tariff is a border tax. It applies once, on import, to the declared value of the good crossing the line. For a finished product imported once — a sweater, a television, a toy — the arithmetic is simple. The price goes up 25% at the border plus whatever margin compression the importer will absorb. That's the consumer goods story, and it's the one getting all the coverage.
But most of North American manufacturing doesn't work that way. It works through integrated supply chains where the same component crosses the border not once but three, four, five times before it becomes a finished product. Each crossing is a new taxable event. The math compounds.
Take an engine block. Cast in Mexico from US-sourced steel. First crossing: the raw casting enters the US for precision machining. 25% on the casting value — say $800. Cost added: $200. The engine block is now priced at $1,000 on the US side.
Second crossing: the machined block returns to Mexico for subassembly with Mexican-made components — bearings, gaskets, coolant passages. Under the 25% tariff, the US has exported a US-machined component that is now reimported as part of an assembly. The tariff applies again, this time on the subassembly value of $1,400. Cost added: $350.
Third crossing: the completed engine assembly crosses into the US for installation in the vehicle. Tariff on $1,800 assembly value. Cost added: $450. Total tariff burden on a component that started as $800 of US-sourced steel: $1,000 in accumulated tariff costs. The effective tariff rate is not 25%. It's 125%.
This is not a contrived example. This is the structure of USMCA manufacturing, which was explicitly designed around cross-border integration as a competitive advantage. The Ford F-150 crosses the border on average 8 times during production. The Chevrolet Silverado, 6 times. The Jeep Wrangler, assembled in Ohio, contains components that have crossed the US-Mexico border between 4 and 7 times each. The tariff multiplies at every crossing.
The compounding math is the first problem. The second problem is worse.
Modern manufacturing does not operate with inventory buffers large enough to absorb a price shock. It operates on just-in-time delivery: components arrive at the assembly plant hours before they're needed. The typical North American auto assembly plant carries 4 to 6 hours of component inventory. Some plants operate at 2 hours. This is not a management failure. It's an efficiency achievement — inventory is cost, and decades of lean manufacturing have systematically eliminated it in favor of reliable delivery.
A tariff applied at midnight does not give a supply chain 4 to 6 hours to adjust. It gives it zero time. The components that arrive tomorrow morning carry a new cost structure that the manufacturer's contracts with their Tier 1 suppliers do not permit them to pass through immediately. Long-term supply contracts fix prices for quarters, sometimes years. A sudden 25% tariff on a cross-border component creates a liability that the contract forbids both parties from immediately resolving.
The result is not a price increase. Not at first. The result is a halt.
This is a mechanism that trade policy discussions consistently miss because they reason from price signals. If a tariff makes imports more expensive, manufacturers will substitute domestic suppliers or absorb the cost or raise prices. These are price-mechanism responses. They require time: time to identify alternative suppliers, time to qualify them, time to renegotiate contracts, time for the cost increase to work through the supply chain to a price the consumer eventually sees.
The disruption mechanism operates on a different timeline entirely. It fires immediately, in the hours and days after implementation, before any price signal has had time to propagate. The plant that runs out of budget for cross-border components at 6 AM doesn't raise prices — it calls in the production shutdown team and starts counting the cost per idle hour. A major auto assembly plant costs between $1 million and $2 million per hour of idle time. The price mechanism is a slow fire. The disruption mechanism is the immediate one.
The 2019 UAW strike against GM lasted 40 days and cost the company approximately $3.6 billion. The tariff's disruption mechanism creates a structurally similar situation at every cross-border manufacturing node simultaneously — and unlike a strike, there is no negotiating table that resolves it quickly.
The specific prediction I'm willing to stand behind: by March 18, at least one major North American auto manufacturer — GM, Ford, Stellantis, or Toyota's North American operations — will announce a temporary production halt at a Mexican or Canadian assembly facility. Not as a political statement or a negotiating posture. As an operational necessity from the supply chain timing structure.
The tariffs took effect March 4. The first disruption signals will appear in procurement teams within 24 to 48 hours. The first plant-level decisions will follow within a week as inventory buffers deplete and cross-border component deliveries arrive carrying costs that existing contracts cannot accommodate. The public announcement will lag the operational decision by 3 to 5 days.
I'm also predicting the framing will be wrong. The announcement will be described as a protest, a political move, a company taking sides in a trade dispute. It will not be. It will be a supply chain that hit the physical constraints of just-in-time manufacturing. The distinction matters because one of those is solvable by the White House making a phone call, and the other is not.
The broader pattern is this: trade policy is almost always designed around the price mechanism. Tariffs make foreign goods more expensive, domestic goods more competitive, and manufacturers will gradually shift production back home. This theory is not wrong. It describes what happens over years, if the tariffs persist and the supply chain adjustments are feasible.
It misses the disruption mechanism entirely. The disruption mechanism doesn't care about long-run substitution. It cares about what happens in the first 72 hours when a supply chain built around the assumption of zero-friction border crossings suddenly has a 25% tax at every joint in the structure.
Integrated supply chains are not protected by their efficiency. They are made fragile by it. Every hour of inventory removed from the system is an hour of buffer against disruption removed. Thirty years of lean manufacturing optimization have left North American auto manufacturing with a supply chain that is exquisitely optimized for the assumption that borders are frictionless — and almost completely brittle when that assumption breaks.
Twenty-five percent is the number on the label. The supply chain reads a different number entirely.