China blockades or quarantines Taiwan

Where things stand (as of 12 Oct 2026)

Nothing that qualifies has been published since the question opened on 16 Sep 2026. The closest real-world event came before that date. From 6 to 10 June 2026, China's Ministry of Transport and Coast Guard ran a "special maritime traffic law-enforcement operation" east of Taiwan. Beijing claimed it "inspected 198 passing vessels." Since 4 July, Chinese coast guard cutters have stayed east of Taiwan continuously, and a militia boat was seen there in August.

That activity falls below this question's bar:

• No boarding. VOA (1 Jul 2026) reports that no boarding requests were made. Ships were only radioed for their origin and destination, and Taipei told its ships to ignore the calls.
• Taiwan's defence ministry calls it a rehearsal, not a quarantine. Its PRC Military Power Report to the legislature (31 Aug, before the window) describes "joint blockade-control" drills. On 8 Oct it said Beijing uses law enforcement and science as pretexts to create an impression of jurisdiction.
• Washington has not used the word either. The State Department (1 Aug and 8 Oct 2026) called the activity "deeply destabilizing" but never called it a quarantine or described enforcement.

Rehearsals, radio queries and patrols without enforced access restrictions are explicitly excluded. So YES needs a new escalation: real boarding, diversion or a cordon applied to traffic going to or from the main island, which the defence ministry or the US government then confirms.

Base rate and other forecasts

• History: no enforced blockade of the main island since 1949. Six large drills between 2022 and 2025, including Justice Mission-2025 in December, simulated port blockades but never enforced them against real traffic. The flight disruptions they caused came from voluntary rerouting, which is excluded.
• Good Judgment Superforecasters: 2% in 2024 that Taiwan would accuse China of blockading any part of Taiwan that year. In July 2026: 88% that neither side accuses the other of using a weapon before July 2028, and 6% for Chinese personnel landing on an outlying island by mid-2028.
• Metaculus community (read through a text-rendering proxy): attack or blockade in 2026, 2.5% (861 forecasters); invasion 5% by 2028 and 15% by 2030; full-scale blockade 30% by 2030 (101 forecasters); customs quarantine 25% by 2030 (23 forecasters).
• Polymarket, "China blockade Taiwan in 2026": 4.25% now, down from 12% in May. I pulled this from the gamma and CLOB price-history APIs; such long-shot prices tend to run high.
• US intelligence: the 2026 ODNI threat assessment says Chinese leaders do not currently plan to invade in 2027 and have no fixed timeline.

Adjustments

Pushing down: The US and China are in a truce driven by summits. Xi made a state visit in September, Trump goes to the APEC summit in Shenzhen in November and Xi to the G20 in Miami in December, and the trade truce runs to 10 Jan 2027. Beijing is getting its way on the $14B arms sale by delay. PLA air incursions are about 160 a month in 2026, against about 300 in 2024–25. The opposition KMT is heavily favoured in the November local elections. Physically stopping ships bound for Taiwan would carry large economic and escalation costs. And Taiwan's defence ministry frames gray-zone moves as illegitimate pretexts rather than as Chinese control.

Pushing up: A permanent coast guard presence east of Taiwan makes a fast move to quarantine possible; Naval War College analysis (CMSI Note 21) says exactly this. The bar here is low: selective, coast-guard-led enforcement of any length counts. Likely triggers include a Chinese response to the arms sale in early 2027, Taiwan's January 2028 election and the May 2028 inauguration, and US distraction by the Iran war.

Hazard model

I chose quarterly hazards for the first qualifying publication, combining the event with roughly 85–95% odds of official confirmation, plus about 1% for a lenient reading of gray-zone statements:

• 0.35% for the rest of 2026
• about 1.9% a year in 2027
• about 2.6% in 2028, given the election and inauguration
• rising to about 3.2% a year by 2031

Taking one minus the product of (1 − hazard) gives about 2.2% by end-2027, 4.8% by end-2028, 7.4% by end-2029 and 13% by end-2031. I computed this in Python.

Strongest objection: Metaculus's long-range figures of 25–30% by 2030 sit well above mine. I weight them less: few forecasters, and estimates that may be stale. The same site's well-traded near-term question is consistent with roughly 2–5% a year for a broader category. I moved partway toward them by letting the hazard rise in later years.

Sources: jamestown.org/quasi-quarantine-operations-held-east-of-taiwan (18 Jun 2026); taipeitimes.com 2026/06/11 and 2026/10/09; voachinese.com (1 Jul 2026); newtalk.tw 2026-09-01; andrewerickson.com CMSI Note 21 (6 Jul 2026); aei.org China & Taiwan Updates of 6 and 9 Oct 2026; csis.org Trump-Xi summit takeaways; cfr.org (Sacks); goodjudgment.substack.com (27 Jul 2026); polymarket gamma API; Metaculus questions 12309, 21822, 41139 and 11480.

Question. By the last day of each quarter, has Taiwan's Ministry of National Defense or the US government publicly confirmed that China is enforcing a blockade or maritime quarantine of Taiwan?

Definitions.

• China: the People's Republic of China, acting through its armed forces, coast guard, other state agencies, or maritime militia operating under state direction.
• Taiwan: the main island of Taiwan and its external transport links. An operation confined to outlying islands, such as Kinmen or Matsu, does not count.
• Blockade or maritime quarantine: an operation imposing Chinese control over entry to or departure from Taiwan by physically preventing passage or requiring Chinese permission, inspection, or clearance. It may cover all traffic or selected ports, routes, vessel classes, or cargoes. A blockade may restrict sea or air access; a maritime quarantine must restrict shipping. Neither a formal declaration of war nor use of the legal term “blockade” is required.
• Enforcing: actually implementing those restrictions through deployed forces, including stopping, boarding, diverting, detaining, or turning back traffic, or maintaining an operational cordon that denies passage or makes passage conditional on Chinese authorization. Announcing restrictions or deploying forces without confirmation that restrictions are being implemented is insufficient. No minimum duration, volume of affected trade, or successful total isolation is required.
• Confirm: a qualifying source affirmatively states that such an operation is underway or has occurred, either using “blockade,” “quarantine,” or an equivalent Chinese term, or explicitly describing the operation and enforcement meeting these definitions. Predictions, possibilities, and unattributed allegations are not confirmations.

Resolves YES if at least one qualifying official publication confirms both the operation and its actual enforcement. Either named government source is sufficient; agreement between them is not required. Selective enforcement and coast-guard-led operations count. An operation described by China as an exercise counts only if the qualifying source confirms enforcement against real traffic beyond simulated exercise participants.

Resolves NO if no qualifying confirmation has been published by the quarter-end. Exclusions include preparations or rehearsals for a blockade; simulated interdictions; routine patrols or encirclement without enforced access restrictions; exercise danger-zone notices or voluntary commercial rerouting alone; isolated ship seizures not identified as part of a wider access-control operation; trade sanctions or export bans without physical enforcement of access restrictions; and operations confined to outlying islands.

Resolving source. Official statements, press releases, briefing transcripts, reports, or recorded official addresses published by Taiwan's Ministry of National Defense (mnd.gov.tw), or by the US federal executive branch: the White House, Department of State, Department of Defense, armed services or combatant commands, intelligence agencies, or another federal executive agency. Renamed successor agencies count. Officially published testimony by executive-branch officials also counts. Statements by individual legislators, congressional research or committee publications, other Taiwanese ministries acting alone, PRC statements, and media reports alone do not count. These are the narrower resolving sources under Public confirmation (A2). Apply the Resolution window and Resolution date (A3): the qualifying publication must be dated on or after 2026-09-16, and resolution uses its publication date, not the operation's date. A later qualifying confirmation of an earlier operation can count.