Trade & Sourcing
China’s 2025 Import Value Fell. Its Tracked Duty Bill Almost Doubled.
Across 57 families in 2024–2025, China import value fell 17.31% while reported calculated duties rose 93.43%. Its observed effective rate rose from 18.12% to 42.38%. Its share of all panel duties nevertheless fell from 69.50% to 36.55%, because other countries’ duty burdens rose faster.
Editorial evidence cutoff: September 9, 2026. Published September 29, 2026. Observation periods are stated throughout; older figures are retrospective evidence.
China's share of the tracked manufacturing duty bill fell sharply between 2024 and 2025. That sounds like relief until the dollars are examined. Calculated duties on imports from China almost doubled, even as the customs value of those imports declined.
The apparent contradiction is useful for September 2026 sourcing discussions. A country can become less dominant in a burden that is spreading elsewhere while remaining substantially more expensive on the measure being tracked. A falling share does not mean a falling bill, and a falling import value does not establish a falling physical quantity.
This retrospective uses the same 57 manufacturing product families in both calendar years, from the annual observations available by September 9, 2026. It examines a defined industrial panel rather than all US imports from China.
THE BILL GREW AS THE VALUE BASE CONTRACTED
China's customs value across the panel fell from $72.11 billion in 2024 to $59.62 billion in 2025, a decline of 17.31%. Reported calculated duties moved in the opposite direction, from $13.06 billion to $25.27 billion, an increase of 93.43%.
Dividing calculated duties by customs value puts the two movements on the same footing. The observed effective rate rose from 18.12% to 42.38%. That ratio describes the entries in the historical aggregate. It is not a single legal tariff applied to every Chinese product, and it does not establish what a particular shipment would owe in September 2026.
The underlying series comes from USITC DataWeb. These are nominal customs dollars and statistical calculated duties. Their connection to actual collections, company invoices and physical quantities has to be examined separately rather than assumed from the headline.
A SMALLER SHARE OF A MUCH LARGER TOTAL
China accounted for 69.50% of the panel's calculated duties in 2024. In 2025 its share was 36.55%. The numerator nevertheless rose by more than $12 billion. The denominator expanded faster because calculated duties associated with other source countries also increased.
The complete panel's duty total rose from $18.80 billion to $69.13 billion. Comparing China's $25.27 billion with the second total answers a different question from comparing it with China's own prior-year bill. One measures where the aggregate burden sits. The other measures the change within a country's recorded business.
This distinction matters whenever a dashboard emphasizes percentages of a total. A manager looking only at China's declining share could infer diversification away from duty exposure. The dollar record does not support that conclusion. Exposure became less concentrated on that particular measure while the China component itself became larger.
IMPORT SHARE ALSO FELL, BUT IT MEASURES SOMETHING ELSE
China's share of customs value in the 57 families declined from 12.51% to 10.66%. That is a real change in the recorded country distribution. It is not the same as a count of shipments, factories, products or qualified suppliers.
A customs-value decline can reflect fewer units, lower unit values, a shift toward cheaper products, or a combination. Broad family labels can conceal substantial movement within classifications. Without matched quantities and sufficiently comparable products, the aggregate cannot separate those possibilities.
Nor does the decline identify who received any displaced business. A purchase could move to another foreign origin, to domestic production, into inventory, or disappear. Those are different events with different implications for capacity planning. Country totals alone cannot follow the transaction from one supplier to its replacement.
THE TEMPTING CAUSAL STORY GOES BEYOND THE RECORD
The combination of higher observed duties and lower import value is consistent with several possible mechanisms. It does not prove that tariff changes caused the value decline. Demand, inventory decisions, product prices, exchange rates and timing could also affect the entries that appear in either year.
There is an additional complication inside the observed rate. Because it is calculated from actual duties and actual import values, it can change when the product or entry mix changes. Even a country-level comparison leaves differences in classifications, preferences and other entry facts unresolved.
The Federal Reserve's April 2026 analysis of announced and implied tariff rates provides relevant context for that distinction. The manufacturing panel adds a focused historical view; it does not turn the accounting relationship into a policy experiment.
WHAT A SEPTEMBER SOURCING REVIEW SHOULD TAKE FROM 2025
The strongest practical lesson is to keep value, duty dollars and observed rates visible together. Each can move differently. Replacing all three with a single country exposure score discards the very divergence that makes this case consequential.
For a company reviewing its own sourcing, the next useful comparison is narrower than this national panel. Match the same specifications and quantities across available suppliers, then identify the classifications and entry facts that determine the applicable treatment. Historical country averages can identify questions worth asking; they cannot settle a purchase decision.
The difference between share and amount also affects progress reporting. A reduction in China's fraction of a company's duties could be caused by an increase elsewhere. A meaningful report would show the country's dollars alongside the total and explain whether the intended objective concerned concentration, absolute expense or both.
THE COMPARISON HOLDS ONLY WITH A CONSISTENT DENOMINATOR
This analysis uses all recorded origins in the complete annual country boards for the same 57 families. It does not divide China by a changing top-ten list. The country values and calculated duties reconcile to the family totals, and the reported duty coverage spans the customs values used here.
The Census trade definitions also distinguish calculated duties from duties paid. That limits any claim about government receipts or an importer's settled liability. The figures establish the behavior of a recorded statistical burden.
Viewed from September 2026, the older observations provide a warning about how to read the next sourcing dashboard. China's import value fell, its calculated duties rose, and its share of the panel duty total fell. All three statements are true. A sound decision starts by preserving that complexity rather than choosing whichever single percentage tells the most convenient story.
Sources and evidence
Evidence period: 2024 to 2025, 57 tracked families. The frozen evidence record lists the source files and verified hashes available September 9, 2026. Source revision: 5e4fb7726c3d40060c0151c086c903baae856cab. Later live-data updates do not alter the historical evidence in this article.
census.gov/foreign-trade/guide/sec2.html
Published 2026-09-29.