Washington readies a 7.5 percent Section 301 overcapacity duty on Chinese goods that would lift second-term replacement tariffs to the exact 20 percent ceiling Beijing says it was promised, one month before a White House summit
WASHINGTON, Aug. 26, 2026
The Trump administration is preparing to impose a new 7.5 percent tariff on Chinese goods under a Section 301 investigation into structural excess manufacturing capacity, a move that would push combined Section 301 exposure on Chinese semiconductors to roughly 70 percent and arrive just weeks before President Donald Trump hosts Chinese President Xi Jinping at the White House.
Bloomberg first reported the plan on Monday, Aug. 24, citing people familiar with the deliberations. The news agency reported that officials are weighing an announcement timed around the September summit, and that one option under discussion would set a higher headline rate and then suspend a portion of it so the effective duty lands at 7.5 percent. The White House called the reporting speculative and neither the White House nor the Office of the U.S. Trade Representative confirmed the figure on the record.
The number itself is the most revealing detail. On July 27, China’s Ministry of Commerce stated publicly that the United States had committed during bilateral consultations to cap any replacement tariffs on Chinese goods at 20 percent. The Section 301 forced labor tariff that took effect on July 24 placed China at 12.5 percent in second-term replacement duties. Adding 7.5 points brings the total to precisely 20.0 percent. In other words, the administration appears to be filling the remaining headroom under a ceiling Beijing has described in public, rather than testing whether the ceiling holds.
For importers, that arithmetic is far more consequential than the headline rate suggests, because the new duty would not replace anything. It would stack.
What stacking actually means at the port
A Chinese-origin good entering the United States can now carry as many as four separate layers of duty, each applied on top of the others rather than in place of them.
The first layer is the ordinary most favored nation rate drawn from the Harmonized Tariff Schedule, which ranges from zero to roughly 32 percent depending on the classification. For most semiconductors and consumer electronics the MFN rate sits at or near zero, which is why the Section 301 layers dominate the landed cost calculation for technology importers.
The second layer is the original Section 301 tariff structure that emerged from the 2018 intellectual property investigation and survived the statutory four year review completed in 2024. That review left the framework intact and raised rates in several strategic categories. The current sectoral rates include 7.5 percent on List 4A consumer goods, 25 percent on most industrial products and electronics in Lists 1 through 3, 50 percent on semiconductors and solar cells, and 100 percent on electric vehicles.
The third layer is the 12.5 percent forced labor tariff imposed on July 24 under a separate Section 301 action covering roughly 60 economies. USTR concluded that those trading partners had failed to prohibit or effectively enforce prohibitions on imports produced with forced labor, and applied duty rates of 10 percent for one group of partners and 12.5 percent for a larger group. That action replaced the Section 122 global surcharge, which expired by operation of statute after 150 days.
The fourth layer is the overcapacity duty now expected before the September meeting.
Run the arithmetic by sector and the picture sharpens considerably. Chinese semiconductors classified under headings 8541 and 8542 would face 50 percent from the original action, plus 12.5 percent from the forced labor action, plus 7.5 percent from the overcapacity action, for approximately 70 percent in combined Section 301 charges before any MFN base rate. Solar modules would sit at the same 70 percent, with an additional Section 201 safeguard duty of roughly 14.75 percent still in force, taking total exposure toward 85 percent. Electric vehicles would reach approximately 120 percent in combined Section 301 charges, and about 122.5 percent once the 2.5 percent MFN rate on passenger vehicles is added. Standard electronics carrying the 25 percent List rate would land near 45 percent.
Those figures describe the same physical product moving through the same port on the same vessel. Nothing about the good changes. Only the duty stack does.
For a firm importing 10 million dollars a month in Chinese-origin chips, the shift from roughly 62.5 percent to roughly 70 percent in Section 301 exposure translates into an additional 750,000 dollars a month in duty, or roughly 9 million dollars annualized, assuming no change in volume, classification or country of origin.
The legal theory behind the duty
USTR initiated the overcapacity investigation on March 11, 2026, nine days after the Supreme Court ruled in Learning Resources, Inc. v. Trump that the International Emergency Economic Powers Act does not authorize the president to impose tariffs. That ruling invalidated the emergency-based tariff architecture the administration had built during 2025 and early 2026, and forced a rapid migration of trade policy onto statutory authorities with firmer footing.
The overcapacity probe covers 16 major trading partners and names 22 manufacturing sectors, including aluminum, automobiles, batteries, cement, chemicals, electronics, energy goods, glass, machine tools, machinery, non-ferrous metals, paper, plastics, processed food and beverages, robotics, satellites, semiconductors, ships, solar modules, steel and transportation equipment.
The legal theory rests on a distinction between cyclical and structural overcapacity. Cyclical excess capacity corrects itself when demand recovers. Structural overcapacity, the argument runs, is manufactured by policy and persists regardless of demand because the subsidies, state-owned enterprise preferences and regulatory forbearance that created it do not switch off. Under that reading, the resulting export surge is not the ordinary operation of comparative advantage but an unreasonable practice that burdens U.S. commerce, which is the statutory trigger under Section 301 of the Trade Act of 1974.
The evidentiary record supporting that theory has grown considerably. An Organisation for Economic Co-operation and Development report published June 2, 2026, drawing on the organisation’s MAGIC database of industrial subsidies, found that Chinese manufacturing firms across 15 key sectors received three to eight times more government support than competitors based in OECD member countries between 2005 and 2024, including approximately 108 billion dollars in 2024 alone.
OECD Secretary-General Mathias Cormann put the finding in blunt terms at the organisation’s June ministerial meeting. “Just like doping in sports, the risk is that subsidies help less productive players win unfairly at the expense of better, more innovative and more efficient ones,” he said.
The solar sector offers the clearest illustration. Chinese annual module manufacturing capacity reached approximately 1,200 gigawatts in 2025, close to double total global demand. Average factory utilization ran at roughly 44 percent for polysilicon, 54 percent for wafers and 47 percent for modules. The three largest Chinese producers, Tongwei, LONGi and TCL Zhonghuan, were projected to post combined losses exceeding 10 billion yuan, approximately 1.49 billion dollars, in the first half of 2026. An industry-led consolidation plan under which six polysilicon producers proposed raising roughly 50 billion yuan, about 7.4 billion dollars, to purchase and retire a third of excess capacity was suspended by China’s antitrust regulator in January 2026 on monopoly grounds, leaving the imbalance unresolved.
Analysts at the Center for Strategic and International Studies have noted that a rigorous economic assessment of the injury from Chinese overcapacity could support a rate well above 7.5 percent, which raises an uncomfortable question about whether the figure now under consideration was set by the injury analysis or by the diplomatic ceiling.
Beijing’s response and the summit calendar
China’s embassy in Washington rejected the premise of the action, saying economic and trade disputes should be resolved through bilateral consultation and disputing the characterization that China has an excess capacity problem.
The timing is deliberate. Trump has confirmed Sept. 24 as the date for Xi’s visit to Washington, the first Chinese state visit to the U.S. capital in more than a decade and the second meeting between the two leaders this year, following Trump’s May trip to Beijing. That earlier meeting produced the Board of Trade mechanism, a bilateral framework for product-by-product tariff relief covering up to 30 billion dollars in Chinese goods classified as non-sensitive. Secretary of State Marco Rubio, speaking after talks with Chinese Foreign Minister Wang Yi in Manila on July 22, indicated the mechanism could become operational before the September meeting rather than being announced at it.
Behind the summit sits a harder deadline. The one year truce negotiated in Kuala Lumpur in October 2025 by Treasury Secretary Scott Bessent, U.S. Trade Representative Jamieson Greer and Chinese Vice Premier He Lifeng expires Nov. 10, 2026. Both governments have signalled interest in an extension, and Bloomberg reported that extension talks are already under way.
If the truce lapses without replacement, suspended measures on both sides snap back. On the Chinese side that includes export controls on gallium, germanium, antimony and graphite, inputs that matter directly to semiconductor fabrication, battery production and defense electronics. On the U.S. side it includes maritime tariffs targeting Chinese shipbuilding.
Publishing the overcapacity rate before Xi arrives removes it from the summit bargaining table. That reduces the risk of brinkmanship in the final weeks of preparation, but it also forfeits the option of holding the rate back as an inducement for Chinese movement on artificial intelligence governance, export controls or Taiwan.
Why Section 301 and not something else
The choice of statute is not incidental. Section 301 tariffs carry no statutory expiration and no rate ceiling, and they have accumulated a substantial record of judicial survival since 2018. On June 15, 2026, the Supreme Court declined to hear HMTX Industries’ certiorari petition challenging the original China tariffs, a development the administration treated as confirmation that the framework is durable.
The contrast with the alternatives is stark. The IEEPA tariffs lasted from February 2026 until the Supreme Court struck them down on Feb. 20. Section 122 expired by its own terms after 150 days. Section 232 requires a national security finding tied to specific articles and has drawn its own litigation. Section 301, by comparison, offers a wide investigative mandate, presidential discretion over remedy design and a track record of deference from the Court of International Trade and the Federal Circuit.
Greer has been explicit about using that authority for technology policy. At his Senate Finance Committee confirmation hearing in February 2025 he told senators, “Semiconductors are at the top of my list in terms of products that need to be brought back to the US,” adding, “Obviously, technologies like AI and quantum computing, we need to be ahead of the game here.” Speaking at a Micron Technology facility in May 2026, he framed the objective in terms of leverage: “We can’t have a situation where the Chinese keep this regime in place where they want to have veto power over the world’s high-tech supply chains.”
That said, the framework is under challenge. On Aug. 4, a coalition of 25 state attorneys general and governors, including officials from California, New York and Illinois, filed suit in the U.S. Court of International Trade arguing that the forced labor tariffs violate the Administrative Procedure Act and exceed the statutory boundaries of Section 301. The complaint contrasts the roughly three month duration of the forced labor investigation with the more than eight months USTR spent on the 2017 to 2018 China intellectual property investigation and the full year devoted to a Brazil investigation, and it uses public statements by administration officials to argue that the forced labor rationale was pretextual.
The overcapacity investigation carries a different profile. It has taken five months and counting, has not yet produced published findings, and rests on an independent evidentiary base in the OECD subsidy data. Greer said in July that the complexity of documenting sector-by-sector subsidy and capacity data across 16 economies accounts for the delay, and Bloomberg reported that finalizing the underlying report has proven legally challenging.
What it means for U.S. importers and exporters
For companies sourcing from China, the practical implications fall into four categories.
The first is duty modelling. Any landed cost model built on the current 12.5 percent forced labor layer is now out of date. Importers should re-run cost projections with a 7.5 percent addition across affected classifications, and should model a second scenario in which the November truce lapses and Chinese critical mineral export controls resume.
The second is classification discipline. When duty stacks reach 70 percent of entered value, the difference between two plausible Harmonized Tariff Schedule classifications becomes a material financial exposure rather than a compliance footnote. Importers with significant China volume should review binding rulings, confirm that classifications reflect current product configurations, and document the analysis. The same logic applies to customs valuation, where first sale treatment, assists and royalty treatment can move the dutiable base by percentages that were once immaterial and no longer are.
The third is country of origin. Substantial transformation analysis has become the central variable in China sourcing strategy, and it is also the area of greatest enforcement risk. U.S. Customs and Border Protection has expanded scrutiny of transshipment and minimal-processing arrangements routed through third countries. Firms shifting production to Vietnam, Thailand, Malaysia, India or Mexico need origin documentation capable of surviving a verification, not merely an invoice showing a new country of export.
The fourth is contractual allocation. Purchase agreements written before 2025 frequently allocate duty risk in ways that no longer reflect commercial reality. Incoterms selection, duty escalation clauses, price adjustment mechanisms and force majeure language all deserve review. So does the importer of record designation, which determines who bears liability for the duty and who has standing to claim a refund if a tariff is later invalidated.
For U.S. exporters, the exposure is indirect but real. A snapback of Chinese export controls on gallium, germanium, antimony and graphite would affect U.S. manufacturers who use those inputs and, by extension, the downstream products they export. American agricultural exporters, who have historically borne a disproportionate share of Chinese retaliation, will watch the September meeting closely for signals about whether Beijing intends to respond in kind or absorb a rate it has already publicly conceded.
The compliance burden nobody budgeted for
There is a second cost to the layered approach that rarely appears in tariff impact estimates, and it lands hardest on companies without dedicated trade departments.
Each duty layer carries its own administrative apparatus. The original Section 301 lists are organized by eight digit subheading and have been amended repeatedly since 2018, with exclusion processes that have opened, closed and partially reopened. The forced labor action applies at the country level but interacts with the Uyghur Forced Labor Prevention Act rebuttable presumption, which operates on a separate legal track with a separate evidentiary standard and a separate detention mechanism at the port. The overcapacity action, if finalized as described, would apply at the sector level across 22 manufacturing categories whose boundaries do not map cleanly onto Harmonized Tariff Schedule chapters.
The practical result is that an entry can be correct under one layer and incorrect under another. A classification that places a product outside the original List 3 coverage may place it inside the overcapacity sector definition. A supply chain that satisfies forced labor due diligence may still trigger detention if a subcomponent traces to a region under the statutory presumption.
Customs brokers report that entry preparation time for China-origin shipments has risen substantially since 2025, and that the number of post-entry amendments has grown with it. Post summary corrections and prior disclosures are administrative costs that do not appear in duty statistics but consume compliance capacity that smaller importers do not have in reserve.
There is also the matter of duty drawback, which becomes materially more valuable as rates rise but which does not apply uniformly. Section 301 duties are generally eligible for drawback, while Section 232 duties are not. For a manufacturer that imports Chinese components, incorporates them into a finished good and exports the result, the difference between recovering 70 percent of the duty paid and recovering none of it is the difference between a viable export program and an unviable one. Companies that dismissed drawback as administratively burdensome when rates were in the single digits should revisit that judgment.
The wider picture
The overcapacity tariff, if finalized as reported, completes a reconstruction of U.S. trade policy that began with the Supreme Court’s February decision. Emergency authority has been replaced by trade statutes. Broad, uniform surcharges have been replaced by layered, sector-specific duties with individual legal foundations. The result is a system that is harder to challenge in court and considerably harder to model in a spreadsheet.
It also marks a shift in what tariffs are for. A 7.5 percent duty calibrated to arrive exactly at a negotiated ceiling is not primarily a remedy for injury. It is a positioning instrument, designed to establish a floor before a negotiation rather than to correct a measured harm. Whether that is good trade policy is a matter of legitimate disagreement, and defenders of the approach would argue that leverage credibly applied is precisely what produced the Kuala Lumpur truce in the first place. Critics, including the state officials now litigating against the forced labor action, argue that using Section 301 as a substitute for authority the Supreme Court has already denied strains the statute past its design.
For the businesses paying the duties, the debate is academic. What matters is that the layers keep accumulating, that each new layer has a separate legal basis and a separate expiry profile, and that a product with no economically comparable non-Chinese substitute now carries a duty burden approaching or exceeding its own manufacturing cost. That is not a surcharge to be absorbed in a quarterly margin. It is a change in the structure of the sourcing decision itself.