Data Desk

The Import Wall, Audited

America's bill for foreign steel has fallen by nearly half in five years, and monthly Census trade data lets anyone audit what the tariff wall actually did. Here is what a thinner import channel means for the people who buy steel.

The United States' monthly bill for imported iron and steel came to $2.21B in Jun 2026, up 1.1% from a year earlier, according to Census Bureau trade data. Set that against the start of the window and the story tells itself. At year-end 2021, the country was buying foreign steel at a $39.3B trailing-12 pace. Over the latest twelve months, the pace is $22.1B. That is a 43.7% collapse in the dollars crossing the border, a gap of $17.2 billion a year, roughly $1.4 billion a month that no longer flows to foreign mills. The import wall that Washington began stacking with national-security steel duties, and has kept adding courses to since, is not a talking point. It is a line item in the trade accounts. This is its audit.

What the wall was built to do

The steel tariffs had one stated purpose: make foreign metal expensive enough that American buyers would source at home and American mills would run hotter. Most trade policy gets judged in press releases. This one can be judged in a monthly Census series, because every dollar of iron and steel that clears customs is counted. The audit standard is simple. If the wall works, the import line shrinks. If it merely taxes, the line holds and buyers eat the duty. On the crude test, the wall is doing what its architects promised: the dollars are down by nearly half over the five-year window, and the slide has been persistent since the June 2022 peak rather than a single step down. A plant manager does not have to like the policy to need the number. Whatever your politics, your next steel quote is being priced on the far side of that wall.

U.S. iron and steel imports, Jun 2026: $2.21B. Ranged from $1.59B in November 2025 to $2.41B in May 2025 across the archived window.

What the dollar series proves, and what it cannot

Honesty requires a caveat before anyone takes a victory lap. This is a value series, dollars rather than tons, and dollars conflate two different things: how much steel came in and what each ton cost. The monthly peak of $4.477B in June 2022 landed when steel prices, not just steel volumes, were running unusually hot, so part of that peak was price rather than tonnage. By the same logic, part of the fall since then is deflation in the cost of each imported ton, not only fewer tons on the water. A dollar series cannot separate the two on its own, and this audit will not pretend it can. What the series does establish is narrower and still useful: the money actually leaving the country for foreign iron and steel has fallen 43.7% in five years, and the monthly low of $1.594B in November 2025 shows just how thin the channel can run. Peak month to trough month, the swing spans a factor of 2.8.

Where does the latest month sit? At $2.21B, the reading stands at the 21st percentile of the archive, which begins in May 2021, and 50.6% below the June 2022 peak. In plain terms, the import channel is running in the lower half of its recorded range. Steel still crosses the border every month. The months when the pace lifts off the floor are the months when import quotes start disciplining domestic list prices again, and the months when it sinks are the months when domestic mills quote with the least fear of being undercut. That swing, not the tariff headlines, is the thing worth watching.

A worked example: your book, thinned by the same fraction

Scale the national number down to a shop floor. Take a fabricator buying $250,000 of steel a month, with 20 percent of it, $50,000, bought at import-linked prices, either directly through a trader or through a service center whose own mix leans on foreign coil and plate. If that slice thinned in line with the national 43.7% decline, roughly $21,850 of monthly material would have to be re-sourced, re-quoted, or renegotiated on domestic terms. Over a year, that is $262,200 of spend moving from the competitive side of the wall to the protected side. The direct cost is whatever premium the domestic quote carries. The indirect cost is worse and harder to see: every import quote that disappears is one less bid keeping the mill's next price letter honest. Competition is a supply chain input like any other, and this series says the supply of it has fallen by nearly half.

Every import quote that disappears is one less bid disciplining the next domestic price letter.

What to do with the number

First, re-run make versus buy on anything with heavy steel content, and run it against current domestic quotes rather than the file copy from the last negotiation. A sourcing answer computed when the import pace was $39.3B a year is stale on its face now that the pace is $22.1B. Second, treat the import channel as a pricing instrument even when you never intend to buy foreign: solicit the quote anyway, because the number in your hand changes the conversation with the domestic mill. Third, watch this series the way you watch scrap surcharges. A sustained climb off the November 2025 floor of $1.594B would say the wall is leaking and negotiating room is returning; continued slippage toward that floor says lock in domestic capacity before your competitors do. The chart above refreshes with the data, so the audit repeats itself every month. Its verdict today is that the wall stands. Your job is to price what standing behind it costs you per part.

Put your part's material, labor, and overhead into the make versus buy calculator and test whether the answer survives a steel quote priced from behind the wall. Run the sourcing math

Published 2026-08-18.