takeoff.watch

Beijing and Taipei sign an agreement on Taiwan's status

Will the governments in Beijing and Taipei sign a formal agreement on Taiwan's political status?

Chance by Q4 2031

2%

Cumulative probability

0%25%50%75%100%
  • Latest forecast

Forecast history

Forecast made on 15 Oct 2026

Latest of 5 forecasts · Methodology v2.1

Model reasoning

Forecast made 15 Oct 2026 · Methodology v2.1

In 77 years the two governments have never signed an agreement on Taiwan's status. Taiwan's current president rejects Beijing's terms, Beijing refuses to talk to him, and Taiwanese law requires a three-quarters legislative vote and a referendum before any political deal can be signed. The realistic route is an opposition Nationalist Party win in January 2028 followed by a peace deal, and even that is unlikely before 2032. Such a win with a pledge to sign a peace deal would most raise this forecast.

Where things stand (mid-October 2026)

There is nothing underway that could produce a qualifying signature soon.

  • Taipei and Beijing are not talking. President Lai Ching-te has been in office since May 2024 and serves until 20 May 2028. Beijing refuses to deal with him. After his 10 October 2026 National Day speech, the Taiwan Affairs Office called him a "pragmatic Taiwan independence worker" (ThinkChina/Zaobao, 12 Oct 2026). The only lines of contact are between the parties and between academics:

    • Kuomintang chair Cheng Li-wun met Xi on 10 April 2026 (gov.cn readout).
    • The Taiwan Affairs Office has welcomed academic essays on a "Taiwan plan" under "one country, two systems."
    • Taiwan's Mainland Affairs Council said it would reject every one of Hong Kong's "twelve-point consensus" proposals.

  • Even the most pro-Beijing figure sets no timetable. Asked on 7 August 2026 whether she would push for a cross-strait peace agreement, Cheng said there was "no timetable" and that any change to the constitutional order must reflect the wishes of people on both sides (TaiSounds). A sympathetic China Times op-ed (9 July 2026) treats a peace agreement as a goal more than five years away.

  • Taiwanese law blocks a quick deal. Article 5-3 of the Cross-Strait Act (2019) sets three gates:

    • Before talks may open, the Executive Yuan must submit a negotiation plan and win a three-quarters legislative vote.
    • The negotiated draft then needs another three-quarters vote, followed by a referendum in which more than half of all eligible voters approve. Only after that may the agreement be signed.
    • Changes to sovereign status may not be negotiated at all, and violating agreements are void (law.moj.gov.tw).

    A future government would have to amend the law first, and that fight would be very visible.

  • Public opinion is hostile. In My-Formosa's September 2026 poll (published 6 Oct), only 4.6% of respondents liked the Chinese Communist Party, a record low, against 80.1% who disliked it. Lai's trust rating was 48.3%. Rejection of unification reached 61.7% in 2026 (Time, 11 Aug 2026).

  • Washington's stance is wobbly but unchanged. Trump and Xi met in Beijing in May and in Washington on 23–25 September 2026 without changing US declaratory policy (CSIS, FPRI). Trump is holding a $14 billion arms package as a "negotiating chip."

  • US officials see an invasion before 2028 as unlikely. Reasons given are purges in the Chinese military and Beijing waiting for Taiwan's January 2028 election. They expect risk to rise if Lai wins again (Reuters via CTV, about 1 Oct 2026). The current drills across five seas do not encircle Taiwan; the closest zone is off Zhejiang (Japan Times, 12 Oct).

Reference class

  • This dyad: no signed status agreement in 77 years. The Ma Ying-jeou years (2008–2016) are the closest test. Ma had a party majority in the legislature and Beijing was pressing for political talks, yet he signed 23 functional agreements and no political one. He floated a peace accord in 2011 and dropped it within weeks.
  • Divided nations generally: such agreements do occur, but rarely and only after long preparation. Examples are the 1972 inter-German Basic Treaty and the 1991 inter-Korean Basic Agreement, which set out an explicit interim status of the kind this question would count. The 1951 Tibet agreement, signed under military duress, is the precedent for a coerced deal.

My base rate is well under 1% a year, and close to zero while the governing party in Taipei rejects the 1992 Consensus.

Paths and the model

Path A: a Kuomintang win in January 2028, then a signed peace or status agreement before the end of 2031.

  • I put a Kuomintang (or joint opposition) win at about 40%. The Democratic Progressive Party leads in party identification (34.7% vs 19.3% for the Kuomintang and 6.3% for the Taiwan People's Party in Zhen Media's August poll), Lai's ratings are steady, and the Kuomintang's field is split.
  • If the Kuomintang wins, I put the chance of a qualifying signature by the end of 2031 at about 2.8%. That requires amending Article 5-3, negotiating with a Beijing that wants faster progress than the Kuomintang does (Ryan Hass in Time), and then signing before the January 2032 re-election campaign.
  • The text also has to qualify. A Ma-style deal resting on the "1992 Consensus" and opposing independence would not. Explicit sovereignty language or a "special relations before unification" clause would. I assume roughly half of any such deals would contain a qualifying clause.
  • Unconditional contribution: about 1.1%.

Path B: a coerced agreement after a blockade or war, signed by Taiwan's existing government.

  • I put a blockade or invasion by 2031 at roughly 10–15%, concentrated after a Democratic Progressive Party re-election.
  • Given force, I put the chance of a qualifying instrument signed by the existing government at about 3–4%. More likely outcomes are failure, stalemate, a ceasefire (excluded), or an administration installed by Beijing (excluded).
  • Contribution: about 0.5%.

Hazard shape. I computed quarterly hazards and accumulated them in Python:

  • About 0.01–0.015% a quarter before May 2028, when only the coercion path is open.
  • Rising to about 0.15% a quarter in 2030–31.
  • Cumulative: 0.1% by mid-2028, 1.1% by the end of 2030, 1.66% by the end of 2031.

Strongest case against my estimate

US commitment is eroding under a transactional Trump: arms sales are being used as bargaining chips, and Chinese commentary describes his Taiwan stance as "Four Noes." Xi is ageing and wants progress, and Beijing has far more leverage than it had in Ma's era. If the opposition wins in 2028 and amends Article 5-3, a peace deal by 2030–31 that says "both sides belong to one China" is conceivable, and a resolver might count that wording.

That could justify 3–4%. I stay lower for four reasons:

  • The likely Kuomintang nominees (Chiang Wan-an, Han Kuo-yu, Lu Shiow-yen) are moderates.
  • Taiwanese hostility to the Communist Party is at a record high.
  • The 2032 re-election falls right after the window closes.
  • The question explicitly excludes bare references to "one China" or the "1992 Consensus."

Interpretation

I read "both sides belong to one China" without any further definition as borderline, and as not qualifying on its own. An explicit interim-status or sovereignty clause would count.

Sources

Resolution criteria

Opened 16 Sep 2026Not yet resolved

Question. By the last day of each quarter, have the governments in Beijing and Taipei signed a formal agreement on Taiwan’s political status?

Definitions.

  • Governments in Beijing and Taipei: the central governments of the People’s Republic of China and the Republic of China (Taiwan), respectively, including successor central governments regardless of name or relocation. Municipal governments do not count. A Taiwan administration installed by Beijing and subordinate to it does not constitute a separate contracting government.
  • Sign: authorized representatives of both governments execute the same written agreement, including by signing separate counterparts. Signatures by the Straits Exchange Foundation (SEF) and the Association for Relations Across the Taiwan Straits (ARATS), or other delegates, count only if each government officially confirms that its representative was authorized to conclude that agreement on its behalf. Political-party agreements alone do not count.
  • Formal agreement on political status: a signed instrument containing an express, mutually accepted substantive provision defining Taiwan’s sovereignty, statehood, constitutional relationship with mainland China, or allocation of ultimate governing authority. This includes recognition of separate statehood, unification, federation or confederation, autonomy under an agreed sovereign, or an express interim status arrangement. The provision may preserve an existing status rather than change it. The instrument need not be called a treaty or be legally binding under international law.

Resolves YES if qualifying official publications establish both governments’ signatures and the agreement’s qualifying substantive provision. Ratification, a referendum, entry into force, or implementation is not required. An agreement signed under military, economic, or political pressure counts. Subsequent rejection, repudiation, or nonimplementation does not reverse a YES.

Exclusions (resolve NO).

  • Negotiations, agreed drafts, announcements of an intention to sign, unsigned joint statements, unilateral declarations, or a signature by only one side.
  • Agreements solely about trade, travel, technical cooperation, security, a ceasefire, or military surrender, unless they also contain the qualifying political-status provision.
  • An agreement merely to discuss or determine status later, to refrain from changing it without defining it, or to acknowledge that the parties hold different positions. References to the “1992 Consensus” or “one China” without a mutually accepted substantive definition of Taiwan’s status do not suffice.
  • Annexation or conquest without an agreement signed by both qualifying governments.
  • If no qualifying confirmation has been published by the horizon, resolve NO for that horizon.

Resolving source. A signed agreement published in an official government record, or official statements from both governments identifying the same signed agreement and confirming enough of its substantive terms to apply these criteria. Eligible issuing bodies are the PRC State Council, Taiwan Affairs Office, Ministry of Foreign Affairs, or Presidency, and Taiwan’s Office of the President, Executive Yuan, Mainland Affairs Council, or Ministry of Foreign Affairs, including their successor bodies and official gazettes. Publication by SEF or ARATS requires confirmation of authorization from the corresponding government as specified above. These are narrower resolving sources than Public confirmation (A2); news reports or third-party assessments alone do not suffice. Apply the Resolution window and Resolution date (A3); if multiple publications are necessary, use the date of the last publication needed to establish qualification.

Download all forecasts (JSON)