The Chinese Communist Party ceases to govern China

Current status (as of 2026-10-14)

Nothing points to a near-term threat to Communist Party rule. Since the 2026-09-16 opening date:

• Xi completed a state visit to Washington (Sept 23–25).
• The Politburo set the Fifth Plenum for Oct 26–29, the main meeting before the autumn-2027 21st Party Congress, where most observers expect Xi to take a fourth term.
• On Sept 21 the party expelled Zhang Youxia and Liu Zhenli, the military's two most senior generals (Reuters, AP, 2026-09-21). Anadolu reports this leaves only two members on the Central Military Commission, the party's top military body, Xi among them. Analysts read this as Xi tightening control, not losing it (Foreign Affairs, "China's Purges Are Working as Designed", Sept 2026).
• Minxin Pei, a long-time critic of the party, rated its grip on power "nine, 9.5" out of 10 (Brookings Beijing Brief, 2026-09-15). His reasons: no organized opposition, divisions far milder than in the Cultural Revolution, and the party in better shape than in 1989 or 1976.
• The real new risk is Xi's health. He is 73. The WSJ (via Times of India, Oct 2026) reported unusual rest breaks at the Washington summit, and he skipped a gala dinner at the BRICS summit in India. He has no designated successor.
• Polymarket "Xi Jinping out before 2027?" trades at 2.8% (gamma API, 2026-10-14). Losing Xi is a far weaker event than the party losing power.

Base rates (computed)

• Surviving communist party-states. China, Vietnam, Laos, Cuba and North Korea have had 0 failures in about 174 regime-years since 1992. That gives 0.57%/yr (Laplace), about 3.0% over 5.25 years; or 0.29%/yr (Jeffreys), about 1.5%.
• Broader class: party-based autocracies. Source: Geddes–Wright–Frantz dataset, gwf_autocratic_extended.rda from github.com/xmarquez/democracyData, 1946–2010.
  • Regimes aged 30 or more: 20 failures in 643 regime-years (3.1%/yr).
  • Excluding the 1989–92 Soviet-bloc cluster: 10 in 572 (1.75%/yr), about 8.8% over 5.25 years.
  • That class is dominated by electoral dominant-party regimes (Mexico, Taiwan, Kenya, Senegal) and coup-prone African regimes. The party has no electoral exit and fuses party and army, so I give it about 20% weight.
• Expert survey. Atlantic Council Global Foresight (Jan 2024): 86% of about 290 strategists expected the party to still be in power in 2034, 3% did not, 11% were unsure. Read as probabilities, that is roughly 6–7% over 10 years, or about 3.5% over 5.25 years.
• Prediction markets. Thin, long-dated Manifold markets are higher: "China collapse before 2030" at 11% (broad definition, 24 traders) and "party in power till 2033" at 81%. Markets like these are biased away from extremes, so I give them little weight.

Paths to YES

1. Disorderly succession if Xi dies or is incapacitated (perhaps 25% by 2031). Leninist regimes almost always survive this (USSR 1953, China 1976, North Korea 1994 and 2011), so I put the chance of collapse at about 1%.
2. Lost war over Taiwan. About 10% chance of war, times about 5% chance of collapse given war: about 0.5%.
3. Economic or social breakdown, or an elite or military split that sidelines the party's institutions: about 0.8%.
4. AI-driven upheaval. This world is in an apparent AI takeoff. Amodei's "We Must Pace the Frontier" (Sept 2026) and OpenAI's "An Alien Mind" both describe AI starting to drive its own improvement. Routes to YES include a decisive US advantage, an openly reported AI takeover, or AI-driven domestic collapse. I give these about 1.3%, concentrated in 2029–31. Even the AI Futures Project's best-case "Plan A" (ACX, July 2026) assumes a US–China deal with the party intact, and AI also strengthens party surveillance.

The resolution bar is high. It needs actual loss of control over more than half the population, confirmed by two of Reuters, AP and BBC, or by an official announcement. A gradual Soviet-style transition could also finish after a given quarter-end. Both trim the odds slightly.

Forecast shape

Quarterly hazards rise as Xi ages and AI effects build:

• Q4 2026: about 0.03%
• 2027: about 0.36% per year
• 2028: about 0.56%
• 2029: about 0.8%
• 2030: about 0.96%
• 2031: about 1.08%

Cumulative: 0.4% by end-2027, 1.7% by end-2029, 3.7% by end-2031 (about 2.4% non-AI and 1.3% AI).

Strongest case for a higher number: if AI reaches superintelligence by 2029, the world could change in ways no base rate covers. Personalist decay and a hollowed-out military command could also make the 2027–32 succession unusually fragile. 1989 showed that regimes that look stable can fall within months. Case for lower: the narrow base rate (1.5–3%) and Pei's assessment. I settle near the top of the narrow-class range.

Question. By the last day of each quarter, has the Chinese Communist Party ceased to be the governing party of mainland China?

Definitions.

• Chinese Communist Party (CCP): the Communist Party of China (CPC), including a renamed or reorganized continuation of the same party. A change of name, ideology, general secretary, or leadership faction alone does not constitute the end of its rule.
• Mainland China: the territory administered by the People's Republic of China on the opening date, excluding Hong Kong and Macao; Taiwan is excluded. This includes Hainan and other PRC-administered islands.
• Governing party: the party exercising ultimate political control over the national executive and its armed forces, whether through constitutional authority or effective command. The CCP remains the governing party if it leads a coalition or retains binding authority over executive appointments or decisions. Participation as a subordinate coalition partner without those powers does not by itself make it the governing party.
• Ceased: an actual loss or relinquishment of that authority, not merely an announced intention, proposed constitutional amendment, election result awaiting a transfer of power, or prediction. A nonparty, military, foreign, or AI-led successor authority can qualify; democratic government is not required. No causal connection to AI is required.

Resolves YES if qualifying sources establish that the CCP has ceased to exercise the governing authority defined above and either:

1. A successor national government has assumed effective executive and security authority over territory containing more than half of mainland China's population; or
2. National government has collapsed or the country has fragmented, and the CCP no longer exercises effective executive and security authority over territory containing more than half of mainland China's population.

For territorial tests, use the latest pre-transition National Bureau of Statistics of China census population counts for the affected areas; if those areas do not follow census boundaries, use estimates expressly supplied by a qualifying source. A successor government need not control every province or the former capital. No minimum duration is required once an actual transfer or loss of authority is confirmed.

Exclusions (resolve NO). Changes of leader, internal party coups, leadership purges, temporary incapacity of individual officials, or delegation to AI while the CCP retains final authority do not count. Neither do local revolts or secessions that fail the territorial test, repeal of the constitutional reference to CCP leadership without an actual end to its rule, or recognition of a rival government that lacks effective control. A successor staffed by former CCP members counts only if they govern independently of the party's governing institutions. Resolve NO at a horizon if no qualifying confirmation has been published by then.

Resolving source. Either (a) an official public announcement by the CCP Central Committee, National People's Congress, or State Council acknowledging the completed end of CCP rule and the qualifying transfer or loss of authority; or (b) concordant factual reporting from at least two of Reuters, Associated Press, and BBC News establishing the required facts. Statements by a self-proclaimed successor, diplomatic recognition, and changes to China's UN representation are not sufficient alone. The existing constitutional baseline is Article 1 of the Constitution of the People's Republic of China, as amended March 11, 2018. These are question-specific resolving sources, rather than Public confirmation (A2). Apply the Resolution window and Resolution date (A3); where two reports are required, use the publication date of the later report. Once YES, subsequent restoration of CCP rule does not reverse the resolution.