Xi Summit Pause

USTR holds back its Section 301 excess capacity findings on China days before a Washington summit, leaving importers with a 7.5 percent question mark and a truce that expires on November 10

By the US Trade Desk

WASHINGTON, September 20, 2026. The Office of the United States Trade Representative has postponed the release of its long awaited findings on structural excess capacity in Chinese manufacturing, shelving a recommendation for an additional 7.5 percent tariff on Chinese goods until after President Donald Trump and President Xi Jinping meet in Washington this week, according to reporting by Bloomberg News published on September 17 and picked up across Asian and European outlets over the weekend.

The delay, first reported by Inside U.S. Trade and confirmed by Bloomberg, removes what had been the single largest near term tariff risk hanging over importers of Chinese origin goods. It also hands Beijing a quiet concession on the eve of a state visit that both governments have declined to frame in ambitious terms. USTR and the White House did not respond to requests for comment from Bloomberg, and neither agency has published a revised timetable.

For companies that source from China, the practical effect is narrow but real. Nothing changes at the border this week. The Section 301 duties already in force stay in force. What has changed is the calendar, and with it the odds that a new layer of duty lands before the fourth quarter shipping season closes.

What was on the table

USTR initiated the excess capacity investigations on March 11, 2026, publishing the initiation notice in the Federal Register on March 17 at 91 Fed. Reg. 12,886. The investigations cover sixteen economies: China, the European Union, Japan, South Korea, India, Mexico, Taiwan, Singapore, Switzerland, Norway, Indonesia, Malaysia, Thailand, Vietnam, Cambodia and Bangladesh. The notice named twenty one sectors, among them aluminium, automobiles, batteries, cement, chemicals, electronics, glass, machine tools, non ferrous metals, paper, plastics, robotics, satellites, semiconductors, ships, solar modules, steel and transportation equipment.

The legal theory is that the sixteen economies, in the words of the initiation notice, “are producing more goods than they can consume domestically,” conduct that “displaces existing US domestic production or prevents investment and expansion in US manufacturing.”

Bloomberg reported on August 24 that the administration had settled on a 7.5 percent additional rate for China. The figure is not arbitrary. China already carries a 12.5 percent Section 301 forced labor duty imposed on July 24, 2026. Adding 7.5 points brings the second term Section 301 total to exactly 20 percent, the aggregate ceiling that Chinese officials have publicly described as consistent with the trade truce struck at Busan in October 2025 and reaffirmed when Trump visited Beijing in May.

That arithmetic is the clearest single clue to the delay. Analysts at the Center for Strategic and International Studies noted in an August 10 commentary by Scott Kennedy and Claire Reade that a credible damage calculation for Chinese excess capacity “could easily far exceed 7.5 percent.” A number chosen to fit a diplomatic ceiling and a number derived from an evidentiary record are not necessarily the same number, and the gap between them is precisely the sort of thing plaintiffs’ lawyers look for.

Statutorily, USTR has until roughly March 2027 to complete a discretionary Section 301(b) investigation, twelve months from initiation. The agency’s own working target had been July 24, 2026, the date the Section 122 surcharge expired. That target was missed by nearly two months before this week’s further slip.

The summit

Xi Jinping is due in Washington from September 23 to September 25, with the summit itself expected on Wednesday. It is his first visit to the American capital since 2015. Ahead of it, Treasury Secretary Scott Bessent and Trade Representative Jamieson Greer met Vice Premier He Lifeng in Manhattan on Sunday, September 20, in talks USTR announced on September 18 and that Reuters reported were expected to run most of the day.

In announcing the trip, Greer said: “Building on the strong relationship between President Trump and President Xi, the United States and China are deepening their engagement to facilitate mutually beneficial trade between the two largest economies in the world.” He added that the administration “will continue to pursue balanced trade with China by monitoring the implementation of recent commitments, optimizing trade in non sensitive goods, and improving market access for American farmers, manufacturers, and workers.”

Expectations are being managed downward on both sides. Edgard Kagan, a senior adviser in China studies at CSIS, told Euronews on September 18 that “the visit is the message,” adding that “there’s been a surprising lack of signalling by either the White House or the PRC on major goals for the visit.”

The most concrete item under negotiation is a mechanism the two governments have taken to calling a Board of Trade, under which each side would cut duties on roughly 30 billion dollars of non sensitive goods. The American opening proposal draws on List 4A from the first term Section 301 China action, which is to say footwear, apparel, kitchenware and other low end consumer products. Chinese negotiators have pushed for a broader basket.

Scaled against 2025 trade flows, CSIS calculates that 30 billion dollars in each direction equals about 28.3 percent of American goods exports to China but only 9.7 percent of Chinese goods exports to the United States. The asymmetry is the whole negotiation in miniature.

On the duration of the truce, Scott Kennedy of CSIS told Euronews: “The Chinese have wanted an agreement to extend the trade ceasefire through the end of the Trump administration. The Trump administration just wants to extend it for a few months to maintain leverage on China. I think they’ll end up somewhere in the middle, probably at about a year.”

Why November 10 matters more than September 24

The summit is the event. November 10 is the deadline.

Under the Busan understanding reached on the sidelines of APEC in October 2025, the United States suspended heightened reciprocal tariffs on China until November 10, 2026, cut fentanyl related duties by ten percentage points, extended certain Section 301 exclusions and delayed retaliation under the Section 301 maritime and shipbuilding action. China, in exchange, suspended its rare earth and dual use export controls until the same date, resumed agricultural purchases and ended retaliation against American semiconductor firms.

Both suspensions lapse on the same day, three weeks after the midterm elections on November 3. Rare earth flows are already running roughly 50 percent below pre restriction levels, and Greer has given Beijing only what one client briefing characterised as a passing grade on critical minerals commitments. China controlled 59 percent of world rare earth mining and 91 percent of refining as of 2024, according to International Energy Agency data.

Read against that calendar, a delayed tariff determination is not restraint. It is a card held back. Politico Pro reported in August that the administration was weighing a structure under which additional duties would be formally imposed but their implementation suspended, preserving the 20 percent understanding while banking leverage for the autumn. Section 301 expressly contemplates exactly that arrangement through agreements on reciprocal trade.

Beijing’s position

China’s Ministry of Commerce has been consistent and unusually direct. At a Beijing press conference on August 27, ministry spokesperson Huang Ling said: “Capacity issues should be viewed in a comprehensive and fair manner, and should not be used as a pretext for protectionism.” She added that China “will continue to closely follow and comprehensively assess subsequent US moves, and reserve the right to take all necessary measures.”

On July 28 the ministry published a formal position paper titled “China’s Position on the So called Excess Capacity Issue,” arguing that Chinese export growth reflects economies of scale, innovation and global demand for the energy transition rather than subsidised overbuilding.

Beijing has also taken a step that complicates the American framing. From September 2026 China introduced new consumption taxes on batteries and photovoltaic cells, explicitly aimed at curbing manufacturing overcapacity and pushing producers toward higher efficiency technology. A country that is taxing its own solar and battery output is a harder target for an overcapacity case than one that is not.

The economic backdrop on the Chinese side is soft. Retail sales grew 0.4 percent in August. Fixed asset investment contracted 7.2 percent in the first eight months of the year against 2025. Beijing injected 360 billion yuan into state banks and insurers. The official growth target of 4.5 to 5 percent is the lowest in decades.

Where American tariffs actually stand

Importers navigating this should be careful which tariff number they use, because two credible sources publish very different figures on different methodologies.

The Penn Wharton Budget Model, which divides collected duties by import value using USITC DataWeb, put the American average effective tariff rate at 6.7 percent as of July 2026, against 2.3 percent in January 2025. On that basis China sits at 22.8 percent, the highest of any major partner, though Penn Wharton describes it as a marked decline from earlier months. Steel and aluminium carry the heaviest product level rate at 40.5 percent. Automotive vehicles are at 13 percent.

The Peterson Institute for International Economics, using statutory trade weighted rates as of December 31, 2025, put average American tariffs on China at 47.5 percent, Chinese tariffs on American goods at 31.9 percent, American tariffs on the rest of the world at 18.4 percent and Chinese tariffs on the rest of the world at 6.5 percent.

Both are correct. They measure different things. Effective rates fall below statutory rates because trade shifts away from the most heavily taxed lines, and because exemptions bite.

On revenue, Penn Wharton counted 298.5 billion dollars in gross customs receipts from new tariffs between January 2025 and July 2026, with a further 59.7 billion dollars foregone as importers front loaded and re sourced. Of the gross figure, roughly 166 billion dollars was collected under the International Emergency Economic Powers Act before the Supreme Court invalidated that authority on February 20, 2026. Customs and Border Protection had certified about 107 billion dollars of that back to Treasury by August 21, roughly 64 percent. Net of assumed IEEPA refunds, the revenue figure falls to about 132.5 billion dollars.

Sector exposure

Semiconductors are the live wire. Section 301 rates on Chinese chips rose from 25 percent to 50 percent in 2025, and semiconductors are named explicitly in the excess capacity notice. The policy concern is concentrated in mature or legacy nodes, where Chinese share has grown rapidly and where officials fear a repeat of what happened to European solar manufacturing. An additional 7.5 point layer would push combined charges on some Chinese semiconductor lines toward 70 percent once other measures stack.

Solar cells and modules already carry 50 percent under Section 301, doubled from 25 percent in 2024, and are also named. Electric vehicles went from 25 percent to 100 percent in 2024 on an explicit overcapacity rationale. Batteries and automobiles both appear in the March notice.

Steel and aluminium sit at the top of the effective rate table at 40.5 percent, reflecting Section 232 duties raised from 25 percent to 50 percent in June 2025, and USTR’s notice alleges excess Chinese steel production directly.

Shipbuilding is the anomaly. The Section 301 maritime measures aimed at Chinese shipyards were suspended for a year under the Busan truce and remain suspended, even though ships are named in the excess capacity notice. That suspension, like the rest of the truce, runs to November 10.

Textiles and apparel are exposed from the other direction. They carry the heaviest machinery under the July forced labor action, including 1,737 tariff codes in the apparel exemption annex and tariff rate quotas directed for Bangladesh, Cambodia, Indonesia and Malaysia tied to their purchases of American cotton and textile inputs. They are also the leading candidates for relief in the 30 billion dollar Board of Trade basket. Importers in this category are simultaneously the most taxed and the most likely to benefit from a deal.

Industry reaction splits along familiar lines

Domestic textile manufacturers, represented by the National Council of Textile Organizations, have generally supported additional duties on finished apparel from Asia. American fashion brands, retailers and importers have opposed them. The American Apparel and Footwear Association and the Footwear Distributors and Retailers of America both testified at the Section 301 hearings urging USTR not to proceed with new tariffs.

No major industry association issued a statement responding specifically to the September 17 delay, and no congressional reaction to it has been reported. That silence is itself informative. A postponement is not a decision, and trade associations rarely spend credibility on a pause.

The skeptics

The most substantive critique of the whole exercise comes from CSIS. Kennedy and Reade argue that the United States is “moving back toward a high tariff scheme unseen since the 1930s” and that none of the claimed benefits has materialised. Most IEEPA revenue is being refunded. The overall American trade deficit has shifted geographically rather than shrunk. Manufacturing employment continues to decline both absolutely and as a share of the economy. Federal debt has risen from roughly 36 trillion dollars at the start of the administration to over 39 trillion.

They are equally sharp on the bilateral numbers. The goods deficit with China fell from 297.1 billion dollars in 2024 to 202.7 billion in 2025 and dropped a further 43 percent in the first five months of 2026. But American exports to China fell 35.2 percent between 2024 and 2025 and another 1.1 percent in early 2026, and much of the import decline reflects rerouting through Southeast Asia rather than genuine substitution.

Their conclusion is bleak for anyone expecting the excess capacity case to change corporate behaviour: given the 20 percent ceiling, the parallel negotiation to cut tariffs on 30 billion dollars of goods, and rising duties on everyone else, “the tariff gap between China and others seems unlikely to widen significantly, if it does at all. In that event, all of this activity may end up being more empty posturing than anything else.”

Writing in Asia Times on September 4, Bob Savic of ApacEuroTrade argued that the analytical category itself is unsound. “Subsidization, dumping, competitive advantage and excess productive capacity are different economic phenomena and should not be treated as interchangeable concepts,” he wrote. Savic notes that foreign invested enterprises accounted for 1.97 trillion dollars of China’s 6.8 trillion dollars in 2025 goods trade, roughly 30 percent, and that a vehicle built in China by a European or American company is statistically a Chinese export but economically part of a Western production system. Global electric vehicle output reached about 22 million units in 2025, up 25 percent on the year. In a market expanding at that rate, he argues, competition should not be automatically equated with overcapacity.

The counterargument from Washington is straightforward. Section 301 explicitly authorises tariffs, courts have consistently upheld its use, and the remedy does not have to be proportionate under subsection (b) in the way it does under subsection (a). Whatever else is contestable, the statute is not.

The legal shadow over any new action

Any excess capacity determination will land in an environment where the administration’s two previous attempts at broad tariffs have already been struck down, and where its third is under active challenge.

The Supreme Court invalidated the IEEPA tariffs on February 20, 2026, holding that the statute’s authority to regulate importation does not extend to imposing duties. The Section 122 surcharge that replaced it within hours was struck down by the Court of International Trade on May 7, though the Federal Circuit stayed that ruling on June 11, and the surcharge expired on its own terms on July 24 after its 150 day statutory life ran out. The Section 301 forced labour tariffs that took effect the same day face a three judge panel at the trade court, with oral argument scheduled for September 30.

Section 301 is the strongest ground the administration has stood on. Unlike IEEPA, it expressly authorises tariffs, and it does not on its face limit how many trading partners USTR may address in a single action. But the excess capacity case carries a specific vulnerability the forced labour case does not: if the rate is set to fit a diplomatic ceiling rather than to match an injury calculation, the record will show it. USTR’s own analysts have to write down a number, and any number that is conspicuously smaller than the harm alleged invites the argument that the action was not designed to eliminate the practice, which is what Section 301 requires.

That may be another reason the report is sitting on a desk rather than in the Federal Register.

What importers should do now

First, do not plan on the delay lasting. No source has reported a new target date, and no source has confirmed that the 7.5 percent figure survives. The reason for the postponement has not been stated on the record by anyone at USTR. A determination could be published within days of the summit.

Second, model both outcomes at the line level. If an additional 7.5 percent applies to Chinese origin goods on top of existing Section 301 duties, the landed cost effect is calculable now. Chapter 99 stacking rules matter, as does whether the new duty applies net of most favoured nation rates, as several rates in the July action do for particular partners.

Third, note the drawback asymmetry. Section 301 duties are eligible for duty drawback at up to 99 percent on exported, destroyed or returned merchandise. Section 232 duties are not. For companies with export or destruction volume, a new Section 301 layer is materially less punitive than an equivalent Section 232 layer, and drawback programmes that were marginal at lower duty rates may now pay for themselves.

Fourth, watch the contract clauses. Tariff allocation, price adjustment, change in law and force majeure provisions written before 2025 frequently do not contemplate a duty regime that changes statutory basis three times in six months. Counsel across the trade bar have been advising clients to review these, and the advice has not become less relevant.

Fifth, treat November 10 as the real deadline. The excess capacity determination is discretionary and can slip again. The expiry of the Busan suspensions is automatic. If both the American reciprocal tariff suspension and the Chinese rare earth export control suspension lapse without an extension, the tariff question becomes secondary to a supply question for anyone touching magnets, motors, or the downstream products that need them.

The summit will produce a communique. The truce is what will determine landed costs in the first quarter.