China bans rare-earth exports to the US

Where things stand (11 October 2026)

No qualifying measure has been published since the question opened on 16 September 2026. When I checked MOFCOM's homepage on 11 October, the newest announcements were No. 44 (an anti-dumping case against the EU) and No. 46 (sugar and wool quotas). Neither concerns rare earths.

Relations are in a managed truce:

• After the 8th round of talks, MOFCOM's Americas department said both sides agreed to extend the Kuala Lumpur/Busan arrangement to 10 January 2027. It said it expects a "positive solution" on further extension through high-level contacts before year-end (Xinhua, 28 Sep 2026).
• Trump is due in Shenzhen for APEC in November and Xi in Miami for the G20 in December (CSIS, Sep 2026).

China still squeezes supply, but only through licensing:

• August magnet exports to the US were 512 t, down 21% from July (InvestorNews, 4 Oct 2026).
• Some Chinese suppliers are informally refusing US orders (Reuters via Firstpost, 4 Sep 2026).
• Neither of these is an official blanket prohibition.

Why the bar is very high

The question needs a published central-government measure that bans, with no exceptions, either all 17 elements or the whole category of rare-earth magnets bound for the US.

Coverage is the first obstacle. Announcement No. 18 of 2025 (text) lists only 7 elements. Its magnet items are samarium-cobalt magnets and neodymium-iron-boron magnets containing terbium or dysprosium. Even with the suspended October 2025 additions, lanthanum, cerium, praseodymium, neodymium and promethium are not on the list. So even a ban on "listed rare-earth items" to the US would not qualify unless the list were widened first.

China's drafting style is the second obstacle. Its toughest US-specific measures leave room for licences:

• December 2024, gallium, germanium and antimony: "in principle, no licences" (No. 46/2024).
• October 2025: "in principle, no licences" for military users (No. 61/2025).
• August 2026, drones to the US: "strict case-by-case review" (No. 34/2026).

I read "in principle, no licences" as a presumption of denial, which the question excludes.

Japan is the clearest test. Even when punishing Japan in 2026, the formal measure banned only military users and uses. The near-total halt in supply came from licence denials (Geopolitechs, 29 Jun 2026; EU Institute for Security Studies, May 2026). The 2010 embargo on Japan was never formally announced (Foreign Affairs).

Reference class and base rate

In about 16 years and 64 quarters, China has never formally banned rare-earth exports to any country. That includes 2025, when US tariffs reached 145%. Licensing gives Beijing leverage, intelligence on buyers and deniability; a blanket ban gives up all three and speeds up diversification.

Formal blanket destination bans do exist in China's toolkit: the 2022 natural-sand suspension for Taiwan and the Japanese seafood import bans. So the option is not unthinkable.

Paths to YES

1. Taiwan crisis. A war or blockade with heavy US involvement or sanctions. I put this at about 10% by 2031. In that case I give about 30–40% that China formally embargoes strategic materials to the US; the 1973 oil embargo is the analogy. Contribution: about 3%.
2. Trade war without fighting. For example, the truce collapses or a more hawkish US administration arrives after 2029. I give about 35% for a severe escalation and about 3–5% that it produces a qualifying ban. Contribution: about 1–1.5%.
3. Lenient reading. A small allowance for a resolver accepting a near-blanket measure as a ban.

Hazard path

Quarterly hazards:

• Q4 2026: 0.10%, because of the summit calendar.
• Each quarter of 2027: 0.18%, covering the 10 January deadline, the PLA's 2027 goal and the Party Congress.
• Each quarter of 2028: 0.20%, covering the Taiwan and US elections.
• Each quarter of 2029–2031: 0.24%.

I accumulated these as one minus the running product of (1 − hazard), computed in Python. That gives about 0.1% by the end of 2026, 0.8% by the end of 2027 and 4.4% by the end of 2031.

Strongest case against my estimate

China is weaponising supply more and more often, and has already named the US in a formal measure. Its leverage shrinks as other suppliers grow, which gives it a "use it or lose it" motive, and a temporary suspension would count. That could argue for 6–8%. I hold near 4–5% because of the coverage and no-exceptions requirements, and because China consistently prefers licensing even under maximum pressure.

Question. By the last day of each quarter, has China adopted and officially published a measure prohibiting all exports to the United States of rare-earth elements or all exports to the United States of rare-earth magnets?

Definitions.

• China: the central government of the People's Republic of China; the prohibition must cover exports from mainland China. Coverage of Hong Kong or Macao is not required.
• Official measure: a promulgated law, regulation, order, customs notice, or other binding central-government measure. A final measure with a specified future effective date counts on publication; a draft, threat, proposal, or conditional promise does not.
• Rare-earth elements: the 15 lanthanides, plus scandium and yttrium. The elements branch requires coverage of all 17, as exported metals, alloys, oxides, other compounds, and mixtures marketed as rare-earth materials. Coverage of ores or finished goods merely containing these materials is not required.
• Rare-earth magnets: permanent magnets whose magnetic material contains rare-earth elements, including neodymium-iron-boron and samarium-cobalt magnets, regardless of grade or whether sintered or bonded. The magnets branch requires coverage of the entire category of separately exported magnets; coverage of magnets already incorporated into finished equipment is not required.
• All exports to the United States: a destination-wide prohibition covering every exporter, consignee, and end use in the 50 states and District of Columbia, with no authorized shipment exceptions for the qualifying category. A worldwide export prohibition that includes the United States counts; expressly naming the United States is not necessary. Coverage of indirect exports through third countries is not required.

Resolves YES if the published measure legally prohibits the entirety of either defined product category from being exported to the United States. Both categories need not be prohibited. A temporary prohibition counts, with no minimum duration. An explicit suspension of all exports, or a binding rule that no export authorizations may be granted and no shipments under existing authorizations may proceed, counts. Several concurrently applicable official measures may jointly establish the prohibition, with resolution on publication of the last necessary measure. Actual evasion or smuggling does not negate a legal prohibition.

Resolves NO / Exclusions.

• Controls limited to selected rare-earth elements, selected magnet compositions or grades, military end uses, named purchasers, or particular exporters.
• Licensing requirements, a presumption of denial, case-by-case permission, quotas above zero, tariffs, customs delays, or an observed cessation of shipments without an official blanket prohibition.
• Measures allowing exceptions, waivers, grandfathered shipments, or special licenses that can lawfully authorize exports within the otherwise prohibited category.
• Export restrictions on rare-earth processing technology or manufacturing equipment alone.
• The licensing controls in MOFCOM–General Administration of Customs Announcement No. 18 of 2025, dated April 4, 2025, and the mere expiration of the suspension of the October 2025 controls, do not qualify without a measure establishing the defined blanket prohibition.
• Absent a qualifying publication by a quarter-end, that horizon resolves NO.

Resolving source. The authoritative Chinese-language measure published by the Ministry of Commerce (MOFCOM, mofcom.gov.cn), General Administration of Customs (customs.gov.cn), State Council (gov.cn), or National People's Congress or its Standing Committee (npc.gov.cn), including their official gazettes. These official texts are the resolving sources, rather than Public confirmation (A2); press reporting or foreign-government characterizations alone cannot resolve the question. Apply the Resolution window and Resolution date (A3) conventions. Subsequent repeal, suspension, or expiration does not reverse YES.