Import Prices Fell 0.4% in July. The Index Doesn't Count the Tariff.

There are two numbers attached to the same shipment, and they are not close to each other.

The first is what the supplier in the other country charged you. The second is what it cost to get that container off the dock and into your possession — the first number, plus the duty, plus freight, plus insurance, plus the broker's fee for filing the paperwork that calculates the duty. At 8:30 this morning the Bureau of Labor Statistics published a report about the first number and called it import prices. Every outlet that covered it wrote some version of "import prices fell."

They did. The number is real and BLS is not hiding anything. It's just that the thing it measures stops at the foreign port, and your money doesn't.

The headline is one commodity wearing a trench coat¶

This desk has now watched the same trick three times in seven days, and it works every time.

July CPI cooled because gasoline fell. July PPI was flat because energy fell harder. And this morning, import prices fell 0.4% because import fuel fell 7.2% while everything that isn't fuel went up 0.4%.

Three reports, three different points in the price system, one component quietly holding the headline down in all of them.

Capital goods0.9%

Foods, feeds, and beverages0.9%

Nonfuel imports, overall0.4%

Automotive vehicles and parts0.2%

Consumer goods, excluding autos0.0%

Nonfuel industrial supplies-0.5%

Headline, all imports-0.4%

Fuel-7.2%

Look at the shape of that. One bar is doing all the work, and it's the one that has nothing to do with most people's business. Unless you import crude, the fuel line is somebody else's news.

The line that describes you is the one BLS put in the release in so many words: "From July 2025 to July 2026, nonfuel import prices rose 4.5 percent, the largest over-the-year advance since the index increased 4.6 percent for the year ended June 2022."

Four years. The prices of the imported things that aren't fuel are climbing at a rate not seen since the supply-chain era, and this morning's headline number was negative.

The part where the number admits what it isn't¶

Here is the structural fact this piece is built on, and it's the reason the headline can't be read the way people read it.

The Import Price Index does not include the tariff. This is not a gap in the data or a lag in the methodology. It is the stated design, and BLS says so in plain English in its own Beyond the Numbers explainer: "Tariffs are not included in the prices used to calculate the U.S. Import and Export Price Indexes."

The valuation point is the tell. BLS's Handbook of Methods describes the import price basis as f.o.b. foreign port — which the handbook defines as "the price at the foreign port of exportation before insurance, freight, or duties are added." Before duties are added. The measurement is taken on the dock in the origin country, at the moment the foreign seller names a price, and everything that happens to that price on its way to you happens after the survey has already recorded it.

“

The index is measuring what a foreign supplier charged. Your bank account is measuring what a foreign supplier charged plus what the U.S. government charged you for the privilege of accepting it.

”

And the reason is almost aggressively boring, which is how you know it isn't a conspiracy. Import price indexes are used as deflators — they convert nominal trade flows into real ones for GDP. Tariffs are taxes. In national accounts, tax revenue shows up as government receipts, not as the cost of the imported good. BLS puts it this way: "Tariffs are taxes, and taxes are not included when calculating net trade. Instead, taxes enter the GDP calculation as a component of government expenditures."

Count the duty in the import price and in government revenue and you've counted the same dollar twice. So the statisticians made a defensible choice, wrote it down, and published it openly.

The choice is correct. The problem is what happens after it leaves the building.

What the index does still see¶

It would be wrong to say tariffs are invisible here, and the honest version matters.

BLS is explicit that the duty itself is absent but its shadow isn't: "Although the index calculation does not include the tariff paid, tariffs can have an impact on price trends before the tariff is imposed and when the tariff is in place." The release notes three channels — importers stockpiling ahead of an announced tariff, foreign suppliers cutting their own prices to stay competitive once one lands, and buyers substituting toward cheaper sources.

That's a real and useful signal. It tells you something you genuinely want to know: whether the foreign seller is eating any of it. If a supplier's f.o.b. price falls after a duty is imposed on their category, they have absorbed part of the cost to keep your business. If it holds or rises, they haven't.

What it will never tell you is your landed cost. Two different questions. The index answers the first one well and the second one not at all, and the headline that runs on top of it doesn't distinguish between them.

The pipeline, with a caveat attached¶

Line up the three price reports this desk has covered in the last week and there is a gradient.

The further you stand from the cash register, the hotter it is. That's a genuinely interesting shape, and it's the same story the CPI basket piece told from the other end: the number the public argues about is the last one in a chain, and it is the most cushioned.

Now the caveat, because without it that chart is a lie of the kind this desk exists to catch.

These are three different indexes measuring three different populations, not three readings of one thermometer. Import prices cover goods and some services crossing the border. PPI covers what domestic producers received for output sold in the U.S. CPI covers a weighted household basket. The baskets do not nest inside each other, the weights are unrelated, and a dollar of imported capital goods is not a dollar of consumer spending. You cannot subtract these to get "the markup," and anyone who does is selling something.

What you can say — and this is the whole claim, no more — is that in the same month, prices are rising faster at the border than at the wholesale level, and faster at wholesale than at retail. Direction, not arithmetic. One month, three series, and the 5.9% at the top of that chart still doesn't include a cent of duty.

The turn: the statistic is fine, the sentence built on it isn't¶

I want to be careful about where the fault sits, because it's tempting to write this as "the government is hiding the cost of tariffs" and that would be false and lazy.

BLS documents the exclusion in the handbook, in the FAQ, and in a dedicated explainer article written for exactly this confusion. Nobody buried it. The methodology is public, consistent, decades old, and the same today as it was when nobody was arguing about tariffs.

The failure is downstream, and it's structural rather than anybody's fault in particular. A monthly release goes out. A wire service compresses it to a sentence. The sentence is "import prices fell," which is accurate about the index and misleading about the world, and it enters the discourse stripped of the one methodological footnote that determines whether it applies to you. By the time it reaches a business owner it has become "good news on import costs," which for an actual importer in a tariffed category may be the precise opposite of what happened to their margin.

The Federal Reserve's small business survey work gives a sense of how many people that catches. In the 2026 Report on Employer Firms — fielded September to November 2025, 6,525 responses, so a read on last year rather than this month — "Forty-eight percent of firms reported that they sourced at least some of their inputs from outside the United States in 2024, while 14% said they did so for more than half of their inputs." More than four in ten firms reported tariff-related cost increases as a financial challenge, concentrated in retail (69%) and manufacturing (62%). Of the firms facing higher foreign input costs, 76% passed at least some of it to customers and 60% absorbed at least some of it — those overlap, because most did both.

That is a lot of operators for whom "import prices fell" is a sentence about somebody else's dock.

What to do with this on Tuesday¶

1. 1

   Find out whether you are actually an importer

   A surprising number of businesses import without knowing it, because they buy from a domestic distributor who imports. You never see a customs entry, so you never think about duty — but it's in your unit cost, and it moves when trade policy moves. Ask a supplier directly where a key input is manufactured. The answer is often not where you assumed.

2. 2

   Read the nonfuel line, not the headline, every month

   The headline import price index is dominated by fuel in a way that makes it close to useless as a business cost signal. BLS publishes nonfuel imports as a separate series in the same release. If you track one number from this report, track that one — it moved +0.4% on the month and +4.5% on the year, and neither figure appeared in most coverage.

3. 3

   Rebuild your landed cost from the parts, since no index will do it for you

   Supplier price, plus duty at your actual HTS classification rate, plus freight, plus insurance, plus brokerage. The BLS series moves the first term only. If your costing spreadsheet has one line called "cost of goods," it cannot tell you whether a bad quarter came from your supplier, the tariff schedule, or the freight market — and those three have completely different responses.

4. 4

   Use the f.o.b. trend as negotiating information

   This is the one thing the index does better than your own invoices: it tells you what's happening to foreign sellers' pricing across a whole category, separate from the duty. If category f.o.b. prices are falling while your supplier holds firm, that's a real, sourceable data point to bring to a conversation. If they're rising 4.5% a year, your supplier is probably not inventing their increase.

5. 5

   Work out which costs you can actually pass through before you need to

   Among the Fed-surveyed firms that saw their foreign input costs rise, roughly three-quarters passed on at least some of it and three-fifths absorbed at least some — the figures overlap because the realistic outcome is a split, not a clean pass-through. Knowing your split in advance is a pricing exercise, not a reaction, and it's much easier to run in a quiet month than in the week a rate changes.

Frequently asked questions¶

Two numbers, one shipment¶

Back to the container.

The supplier's price is a fact about a negotiation between you and a company in another country, and this morning BLS told you a great deal about how that negotiation is going across the whole economy — including, if you read past the headline, that it is going worse than it has since 2022 for everything that isn't fuel.

The duty is a fact about a decision made in Washington, and it appears on your broker's invoice as its own line, and it is not in this report, and it was never going to be, and the methodology note explaining why has been sitting on the BLS website the whole time.

What consumers say versus what they actually bought has been this desk's theme all week, and here's the supply-side version of it: the official statistic and your bank statement are both telling the truth, about different things, and only one of them is going to be quoted at you.

It's just business — but read the second number, because it's the one you paid.

Sources

1. U.S. Bureau of Labor Statistics — U.S. Import and Export Price Indexes News Release, July 2026 (released August 18, 2026)
2. U.S. Bureau of Labor Statistics — U.S. Import and Export Price Indexes, full July 2026 summary
3. U.S. Bureau of Labor Statistics — 'How tariffs relate to BLS import and export price indexes,' Beyond the Numbers vol. 9
4. U.S. Bureau of Labor Statistics — Handbook of Methods: International Price Program, Concepts
5. U.S. Bureau of Labor Statistics — Producer Price Index News Release, July 2026
6. U.S. Bureau of Labor Statistics — Consumer Price Index Summary, July 2026
7. Federal Reserve Banks — 2026 Report on Employer Firms: Findings from the 2025 Small Business Credit Survey