Cheng Li-wun's Beijing Trip Isn't a Peace Mission — It's a Pre-Summit Stage Prop

KMT Chairwoman Cheng Li-wun landed in Shanghai today on what she calls a "journey for peace" — the first visit by a sitting KMT leader to China in nearly a decade. She'll travel to Nanjing and then to Beijing, where she may meet Xi Jinping himself¹. The itinerary is symbolically drenched: Sun Yat-sen's mausoleum, party-to-party meetings, photo ops projecting cross-strait comity. And all of it is happening, as the Associated Press noted², just weeks before Donald Trump travels to Beijing on May 14-15 for a summit where Taiwan will be, in Brookings' assessment, China's priority agenda item³.

I want to be precise about my argument here, because the lazy version of this take — "the KMT is a Chinese puppet" — is wrong and analytically useless. Cheng Li-wun is not a puppet. Her voters are not duped. The KMT represents a real constituency in Taiwanese politics that genuinely prefers managed engagement with Beijing over confrontation. In January 2024, candidates running on platforms friendlier to cross-strait dialogue than the DPP's collectively drew roughly 60% of the presidential vote, even as they split the ticket and lost. That's not a fringe position. It's a democratic mandate for a particular strategic orientation.

But here's the thing: the authenticity of Cheng's position is precisely what makes this visit so strategically useful to Beijing. And the timing is not an accident.

Start with what Beijing gets. The visit, as The Diplomat's analysis laid out⁴, allows China to frame Taiwan within a "Chinese political arena" in which it can bypass the DPP government and work with its preferred actors. By inviting the KMT chair now, Beijing sends a message not just to Taipei but to Washington: the Taiwan issue has multiple interlocutors, and some of them accept a one-China dialogue framework. Atlantic Council fellow Wen-Ti Sung told the AP² that the visit "may sideline the Taiwan Strait tension issue from the Xi-Trump summit, thus enabling the U.S.-China summit to focus on business areas of common interest." Think about what that means operationally: if Taiwan gets de-prioritized from the summit agenda because it looks like cross-strait dialogue is already underway, that is a concrete win for Beijing.

Chen Fang-yu, a political scientist at Soochow University in Taipei, is more direct⁵: Cheng may be playing into Beijing's "United Front" strategy, which includes welcoming Taiwanese politicians to "emphasize that Taiwan is a domestic or internal matter for China." The visit's choreography reinforces that framing whether or not Cheng intends it.

Now layer in the wider context, because the timing is doing a lot of work. This visit doesn't land in a vacuum. It lands in a specific information environment shaped by three converging developments. First, the Trump administration has delayed a $13-14 billion arms package to Taiwan⁶ to avoid upsetting Xi ahead of the summit, according to the New York Times. The White House told agencies not to move forward. Second, the KMT-dominated legislature has stalled President Lai's $40 billion special defense budget⁷, blocking it at least ten times. U.S. Senator Dan Sullivan directly called this out: "short-changing Taiwan's defense to kowtow to the CCP is playing with fire." Third, Trump himself told reporters in February⁸ that he was "talking to" Xi about arms sales to Taiwan — a remark that, as experts noted, could violate the spirit of the 1982 Six Assurances.

So the sequence looks like this: Washington is pausing arms sales. Taipei's legislature is blocking defense spending. And into that exact window walks the KMT chairwoman, arriving in Beijing to demonstrate that "peace" and "dialogue" are available alternatives to deterrence. You don't need a conspiracy theory to see the strategic utility. You just need a calendar.

The strongest counter-argument is the democratic legitimacy one, and I take it seriously. Taiwan is a democracy. The KMT has every right to pursue its foreign policy vision. Opposition parties worldwide maintain contacts with foreign governments. Labeling all of this as "influence operations" would narrow Taiwan's democratic space in ways that mirror the authoritarian logic critics claim to oppose. This is a genuinely important point, and anyone arguing for restricting KMT engagement needs to reckon with it.

But I think this counter-argument, while principled, misidentifies the problem. My concern is not that the KMT shouldn't be allowed to visit Beijing. It's that (1) Washington is not adequately tracking the narrative effects of this visit in the pre-summit environment, and (2) Beijing is using the visit to project a specific message — that cross-strait tensions are manageable without American involvement — at the precise moment that message most benefits its summit positioning.

The evidence for Beijing's intentionality here goes beyond timing inference. There are documented indicators that Beijing actively shaped the conditions for Cheng's rise to the KMT chairmanship. Taiwanese security officials found⁹ over 1,000 TikTok videos and 23 YouTube accounts, more than half based outside Taiwan, discussing the KMT chair election. According to reporting by the Sasakawa Peace Foundation¹⁰, the CCP allegedly supported Cheng's candidacy through "organization, public opinion, and funding," mobilizing Taiwanese business people in China (Taishang) and instructing certain media outlets to strengthen positive coverage. The Diplomat reported¹¹ that rival KMT candidates accused China of interfering to boost Cheng. She rejected these as "very cheap labels," and Beijing denied official involvement.

I am not saying these allegations are proven. But they paint a picture of a CCP that doesn't merely wait for convenient KMT positions to emerge. It actively cultivates the actors who hold them.

Here is what I think the Taiwanese public opinion data actually shows, and it's less flattering to the KMT than the raw election numbers suggest. A Taiwan Public Opinion Foundation poll from October 2025¹ found only 13.9% of respondents supported unification with China, versus 44.3% supporting independence and 24.6% supporting the status quo. A TPOF poll found¹² that half of respondents believed the KMT aligns itself more with China than with Taiwan. And 47% of voters believed Cheng's stance leaned toward unification¹² — a position that represents a marginal fraction of Taiwanese opinion. The KMT draws voters for many reasons — dissatisfaction with the DPP, economic concerns, opposition-party loyalty — that have nothing to do with endorsing Beijing's preferred framework.

So the KMT's cross-strait engagement mandate is thinner than the 60% election figure implies, and Cheng's specific positioning is further toward Beijing than most of her own voters actually want. That gap is the space in which Beijing's influence operates most effectively.

The practical question is what Washington should be doing. The Atlantic Council's Melanie Hart wrote after the Busan summit¹³ that "there is still no indication that the US side has the same strategic clarity" as Beijing heading into the 2026 meetings. The bipartisan Senate delegation that visited Taiwan last week¹⁴ — Shaheen, Curtis, Tillis, Rosen — was explicitly designed to signal that Congress's commitment "is unwavering and will endure well beyond any one administration." That's Congress doing counter-narrative work the executive branch isn't doing.

What to watch next: the content of whatever joint statement or readout emerges from a potential Cheng-Xi meeting later this week, compared against the Trump-Xi summit readout in May. If the Cheng-Xi framing — particularly any reaffirmation of the 1992 Consensus as the basis for cross-strait dialogue — shows up in Trump's public characterization of the Taiwan issue at or after the summit, that will be the strongest evidence that Beijing's pre-conditioning operation succeeded. If Trump's Taiwan language instead tracks with the Senate delegation's framing or the existing US One China Policy, the pre-conditioning failed. The test is specific and falsifiable, and it will be available within six weeks.