Trade & Sourcing
Changing Source Countries Offset 2.14 Points of the Hand-Tool Duty Increase
Complete 2024–2025 country boards show a 13.46-point within-country-rate component and a −2.14-point country-mix component for hand tools, leaving an 11.32-point increase in the family observed rate.
Editorial evidence cutoff: September 9, 2026. Published September 29, 2026. Observation periods are stated throughout; older figures are retrospective evidence.
A September 2026 sourcing review can learn from the completed 2024 and 2025 country records without treating them as current supplier quotes. This retrospective asks a specific question: did the changing distribution of origins amplify or offset the historical increase in the measured duty rate? All country observations used here were available by September 9.
The hand-tool duty rate rose sharply between 2024 and 2025, but the change in source countries worked in the opposite direction. Complete country records show a positive 13.46-percentage-point component from changes in observed rates within countries and a negative 2.14-point component from the changing country mix. Together they produce the family’s 11.32-point increase.
That is a more demanding question than identifying the largest supplier. It asks how the distribution of imports across countries is associated with the family-wide rate, while separating that distribution from changes in each country’s own recorded rate.
The result does not prove that switching suppliers saved buyers 2.14 points. It is an accounting attribution. But it shows why a sourcing review that looks only at the family headline misses two forces moving against each other.
THE FAMILY AVERAGE CONTAINS TWO STORIES
An observed family rate is a weighted average of country rates. Each country receives a weight equal to its share of the family’s customs value. The average can rise because rates within the countries rise, because more imports come from countries with higher recorded rates, or through both movements together.
For this analysis, every 2024 and 2025 country board was checked against its full family totals. Across all 57 families and both years, country customs values and calculated duties reconcile exactly. That matters: using a retained top-ten-country subtotal as the denominator would give a different set of weights and could change the conclusion.
The hand-tool comparison contains 95 reporting origins in 2024 and 90 in 2025. It uses the full boards from the saved USITC DataWeb records. Earlier country histories are incomplete in this repository, so they are not mixed into this two-year calculation.
THE MIX OFFSET IS LARGE ENOUGH TO MEASURE
The decomposition averages two orderings. It first changes country rates while holding country shares fixed, then changes shares. It also performs the changes in the opposite order and averages the results. This allocates the interaction symmetrically instead of giving it entirely to whichever factor was changed last.
Applied to hand tools, changes within country rates contribute 13.46 points. The source-share term contributes negative 2.14 points. Their sum is 11.32 points. The offset is substantial, even though it is not large enough to prevent the observed family rate from rising.
The family-wide increase conceals that offset if it is considered alone. A change in the leading source country can also distract from the much larger movement within country rates. Separating the components establishes how both changes enter the recorded average, while leaving their economic causes open to further investigation.
GLASS FIBER CONFIRMS THE PATTERN. FURNACES CHALLENGE IT.
Hand tools are not the only family with a negative country-mix component. Glass fiber shows a positive 7.49-point within-country component and a negative 1.80-point mix component. Its total observed-rate increase is therefore about 5.69 points.
Industrial furnaces go the other way. Their within-country component is positive 5.96 points, while country mix adds another 1.24 points. The total increase is about 7.19 points. A broader list of origins or a change in the leading country does not automatically imply a mix that lowers the recorded burden.
All 57 families were screened before choosing these examples. That prevents a convenient pair of cases from becoming a universal claim about diversification. The value of the three-family comparison is precisely that it contains a counterexample. The same calculation can produce an offset or an addition, depending on the observed shares and rates.
THE COUNTRIES THAT APPEAR OR DISAPPEAR
A country with no imports in one endpoint has no observed rate for that year. Assigning it a zero rate would create an artificial change. The main calculation instead carries its available rate into the zero-weight endpoint, placing the entry or exit contribution in the mix term.
That convention is explicit, but it still needs a sensitivity check. The analysis was repeated using only countries with positive customs value in both years, with their shares renormalized. For hand tools, 79 countries remain. They account for 99.9953% of 2024 value and 99.9900% of 2025 value.
The hand-tool mix term barely changes: negative 2.13796 points on the full board becomes negative 2.13793 points on common support. Glass fiber remains negative 1.80550 points, and furnaces remain positive 1.24219 points. The examples are not being driven by assigning rates to tiny new or disappearing origins.
WHAT THE WITHIN-COUNTRY TERM STILL HIDES
A country-family rate is itself an aggregate. Hand tools from one origin can shift among different classifications, specifications or eligibility conditions. Those changes can alter the observed rate without representing a uniform rate change on every hand tool from that country.
The calculation therefore removes one layer of composition, the distribution across countries, while leaving other layers inside the within-country term. It cannot establish that the legal treatment of a fixed item rose by 13.46 points. Nor does it establish that buyers deliberately shifted countries in response to duties.
It also cannot rank complete supplier costs. Freight, quality, tooling, delivery performance and comparable country-specific unit prices are absent from this rate identity. Those facts belong in a procurement decision, but inventing them would turn a reproducible comparison into a speculative sourcing recommendation.
A BETTER QUESTION FOR THE NEXT SOURCING REVIEW
The published hand-tool stories already describe China’s declining share and Taiwan’s leadership. Repeating that story would add little. The new contribution is a measured distinction between changes in where goods came from and changes in the burden recorded within those origins.
For a buyer, the practical next step is to construct the same table for a matched set of actual items: origin shares, verified duties, customs values and changes in product mix. A national family’s negative mix component is a reason to investigate the company’s own exposure, not a promised saving to put in a budget.
The records show that source composition mattered to the average. They also show that it did not dominate the increase. Keeping both facts together produces a more useful account than either a triumphant supplier-switch story or an unexplained higher family rate.
Source window: complete calendar 2024 and 2025 country boards. Calculated duties follow the Census statistical definition and need not equal cash duties paid. The analysis is descriptive, with all 114 family-year boards reconciled before decomposition.
Sources and evidence
Evidence period: Calendar 2024 versus 2025. 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
link.springer.com/article/10.1007/s10888-011-9214-z
Published 2026-09-29.