China Plus One Tariff Costs and the Continuity Premium
The current U.S. tariff stack leaves only a narrow rate gap between China and key alternative origins for some products. Importers should price the current landed-cost spread separately from the cost of keeping non-China capacity ready.
Primary lensTrade policy
Sub-topicPolicy monitoring
Evidence base12 records used
Use casePolicy monitoring
China Plus One's latest problem is an existing factory that no longer wins the next purchase order. Publication 2026-07-29, The New York Times, Trump's Tariffs Are Sending Some Companies Back to China reported that Alliance Consumer Group's Southeast Asian production cost 12 to 15 percent more than Chinese production and that the flashlights at issue faced a duty gap of about one percentage point. The company had built enough capacity to make more than three quarters of its goods outside China. It was still making about two thirds of them in China.
The company figures describe one case and cannot supply a reusable tariff calculation. They still expose a decision that country-level sourcing plans tend to hide. A non-China line can fail the current landed-cost test and remain worth keeping because it preserves a way out of China during the next disruption. Importers now need to calculate that option explicitly, before a financially tidy plant closure turns a reversible sourcing shift into another costly relocation.
The July 24 action reset a shared tariff layer
The immediate policy change is specific. The USTR final notice in 91 FR 47318 imposed additional duties under 60 separate Section 301 investigations. Covered goods of China, Thailand, and Vietnam receive a 12.5 percent rate under this action. Covered goods of Cambodia receive 10 percent. The rates apply to covered entries on or after July 24, subject to the notice's transit rule and exemptions.
That did not make China, Thailand, and Vietnam equally taxed. China can still carry the older Section 301 duties, and every comparison must include the base rate, other Chapter 99 provisions, Section 232 treatment, and any product exclusion. The action imposed the same 12.5 percent layer on covered goods of China, Thailand, and Vietnam, so it did not by itself change China's percentage-point spread against those two alternatives. Cambodia's 10 percent rate made the difference under this action 2.5 points. Different customs values can still change the dollar burden per unit.
The final notice also warns against a country-average shortcut. Its Annexes exempt categories and tariff lines, and some economies receive a rate net of their most-favored-nation duty. China, Thailand, Vietnam, and Cambodia are not all treated the same. A buyer has to reconstruct the live stack for the imported article rather than paste four national percentages into a presentation.
The relevant number is the product's relative spread
Sourcing teams often begin with the China duty and ask whether another country avoids it. The cleaner comparison begins one step later. Calculate the full duty rate for the same imported article from each legally supportable origin, then subtract the alternative-origin rate from the China rate. That difference is the origin-relative tariff spread.
The spread is what can pay for higher conversion cost, longer inland transport, imported inputs, lower line utilization, more quality work, and duplicate management overhead outside China. If the non-China operating premium is 12 percent and the verified tariff spread is 15 percent, the move may still work on current unit economics. If the spread falls to one percent, the same factory needs another reason to remain active.
The newspaper's flashlight numbers make the mechanism visible, but they should stay inside the case. The article does not identify the imported models, their exact HTS classifications, the China list status, or the exemptions used in the calculation. The USITC Harmonized Tariff Schedule and applicable Chapter 99 notes are the starting records for a fresh calculation. The article's roughly 20 percent and 19 percent totals should not be copied to a different flashlight, much less to a different product category.
This is also why the phrase "China tariff" is too blunt for an investment memo. The current CBP Section 301 guidance administers the China measures by product and origin. A valid comparison needs the imported model, tariff classification, claimed origin, entry date, exclusion status, and every additional duty provision that applies. The spreadsheet should preserve the source and last verification time for each field.
One useful discipline is to show the comparison twice. The first view reports duties as a percentage of customs value. The second converts the difference to dollars per commercial unit. Factory decisions are made against purchase prices, margin, freight, yield, warranty cost, and financing. A percentage-point spread that appears meaningful at the border may fund very little production overhead on a low-value article.
Past investment does not justify the next order
Millions spent on equipment, certification, supplier development, and factory approval are sunk once paid. They cannot make the next unit cheaper. Sending volume to an uneconomic plant solely to vindicate the original China Plus One decision can compound the loss.
Closing the line because its current unit cost is higher can be just as mechanical. Some of the spending created a capability that still has a forward value. The relevant question is what it costs from today to keep that capability usable, and what it would cost and how long it would take to recreate it after a China-specific disruption.
That distinction separates sunk relocation cost from a continuity option premium. The premium includes the minimum orders needed to keep trained labor and quality systems credible, recurring factory and product audits, tooling maintenance, engineering-change control, supplier approval, customer approval, and the inventory needed for a restart. It excludes money that cannot be recovered regardless of the next decision.
Finance should compare that premium with a plausible re-entry cost, not with the original project budget. Re-entry may require new qualification runs, replacement tooling, revised certificates, a customer notification, fresh packaging, a broker review, and months of low-yield production. A line that can restart in four weeks is a different asset from a building whose equipment has been sold and whose supplier approvals have lapsed.
Use a fixed decision horizon to calculate the option rather than assigning it a slogan. For each approved disruption scenario, add the margin loss and qualification or restart spending the retained line would avoid, multiply that total by management's probability range, and compare the expected benefit with carrying cost measured over the same decision horizon. Show the range beside restart-time and available-capacity assumptions instead of collapsing it into one precise number.
The result will sometimes support closure. If critical inputs still come only from China, the alternative plant cannot operate during the disruption it is meant to cover. If restart requires the same time as building elsewhere, the supposed option is weak. If the product is near end of life, preserving the line can waste cash. The calculation does not create a permanent presumption in favor of diversification.
A second factory may preserve location without reducing China dependence
The production address is only one layer of continuity. The flashlight company described Chinese equipment and China-linked inputs in its Southeast Asian operations. The broader pattern appears in Federal Reserve staff research on Vietnam's export boom. The authors found that Chinese-owned firms' share of Vietnam's exports to the United States rose from 11 percent in 2018 to 2019 to 25 percent in 2020 to 2023. Their method excluded shipments identified as potential rerouting. Newly established foreign-owned firms relied more on Chinese imports, a pattern the authors said was consistent with relocation but not definitive.
The finding is economic evidence, not a customs origin decision. It shows why a second location can diversify final production while leaving upstream recovery tied to China. A continuity ledger should therefore identify the inputs that can stop both plants at once. Batteries, printed circuit boards, semiconductors, aluminum bodies, tooling, firmware, and testing equipment may create very different common points of failure.
Ownership is relevant for the same practical reason. A Chinese-owned factory in Vietnam or Thailand may have faster access to an established engineering and supplier network. That can lower ramp risk. It may also mean that both plants depend on the same managers, financing, technical data, or upstream vendors. The useful record lists those links rather than treating ownership as proof of evasion or as a rule of origin.
Customs origin remains a separate gate. Moving final assembly does not guarantee that the new country becomes the origin for a particular trade remedy. Publication 2026, Traverse Analysis, Vietnam Transshipment Exposure After the 40 Percent Tariff Ended explains the product and process record needed for that determination. The continuity decision should not count a tariff saving until the importer can support the origin on the actual model and manufacturing sequence.
Keep a dual-source continuity ledger by product
The operating file should put the China line and each alternative line side by side for the same imported model. Start with the economics. Record customs value, base duty, each additional duty, exclusions, freight, conversion cost, expected yield, warranty provision, and financing. The result is a current landed cost and an origin-relative tariff spread that can be updated when an operative notice changes.
Then add the forward-looking fields that an ordinary landed-cost sheet omits. Record the minimum sustainable order for the alternative line, annual tooling and certification expense, time to resume commercial output, capacity available after restart, and the approvals that would expire during dormancy. Identify the event that would trigger transfer and the person authorized to release volume.
Add a common-dependency map. For each critical input, show its origin, approved supplier, available substitute, validation time, and inventory coverage. Record whether the two final-assembly plants share equipment, technical staff, firmware access, or a parent supplier. A nominally diversified footprint with one irreplaceable Chinese board is not a two-source system for a board disruption.
Finally, keep the customs proof next to the commercial assumptions. Store the HTS classification, applicable Chapter 99 provisions, claimed origin, ruling or counsel analysis, process map, bill of materials, and last review date. The relevant Traverse Policy Signal for USTR's final 60-economy action should trigger a new rate review, not an automatic percentage update across every SKU.
This ledger creates a decision rule that can survive the next tariff change. Send current volume to the source with the better verified economics, subject to concentration limits. Keep enough volume or spending at the second source to preserve the agreed restart time when the option value exceeds the carrying cost. Close the alternative line when the option no longer covers its forward cost or cannot respond to the disruption it is supposed to hedge.
Policy durability belongs in the capital request
A factory has a longer life than many tariff measures. The relevant policy question is therefore not only the rate in force on the approval date. It is the legal instrument, review path, scheduled endpoint, and product scope behind that rate.
The termination of the specified IEEPA duties under Executive Order 14389, followed by the temporary Section 122 surcharge schedule, shows how quickly the comparison can move. By contrast, the existing China Section 301 duties have their own statutory review process and product-specific history. The July forced-labor action arrived through 60 investigations, a proposal, more than 1,600 written comments, a three-day hearing with more than 100 witnesses, and a final notice. Those procedural differences do not predict the life of each rate, but they belong in the investment case.
The Federal Reserve paper on tariff uncertainty and supply-chain structure finds that reducing the probability of a trade war promotes durable importer and exporter relationships that support higher-quality inputs. For a sourcing team, the practical implication is narrower. A tariff advantage that depends on a short-lived authority should receive less weight in a factory decision than the same advantage under a measure with a stable review path.
The capital request should show a sensitivity table rather than one tariff assumption. At minimum, test the live stack, the expiration or removal of the newest layer, a change in product exclusions, and an additional China measure. Show the spread at which the alternative plant breaks even before any option value, then show the carrying cost needed to preserve a defined restart time. Management can then see whether it is approving a cost-saving move, buying continuity, or doing both.
The record should identify what would change the decision
Several facts could change a decision quickly. A new USTR modification could widen or narrow the relative spread. A product exclusion could remove the July layer from one origin. A confirmed classification or origin ruling could change the applicable stack. A supplier could localize a critical board or battery outside China and turn a second factory into a genuinely independent source. A customer could shorten or lengthen the allowed requalification period.
Importers should set review triggers around those facts. Recalculate the ledger when an operative Federal Register notice changes a rate or exclusion, when a bill of materials changes a critical input, when a factory approval approaches expiry, and before annual volume allocation. Keep the current decision and the evidence behind it, including the reason for any minimum order sent to a higher-cost plant.
The current stack can leave a narrow product-level spread between China and an alternative origin even while the older China duties remain in force. The July action requires a fresh calculation because it changed a shared layer and its country tiers are not uniform. A China Plus One plan can no longer be judged by the existence of a second factory. The importer has to know the verified tariff spread today, the cost of keeping the second line ready, and the time that line would save when the next disruption arrives.
From reading to review
Run the numbers on your lane.
The duty calculator runs the current stack for any HTS code and origin. A free account opens full tool output, AD/CVD detail, Chapter 98 processing, and available exports.