With their twin summits in May and September 2026, Presidents Donald Trump 
and Xi Jinping have now succeeded in reestablishing what can be metaphorically 
described as the floor and ceiling in the U.S.-China building. By a “ceiling,
” I mean establishing warm personal leader-to-leader relations. The history 
of U.S.-China relations since the Nixon-Mao rapprochement in 1972 has repeatedly 
demonstrated that leaders' personal engagement and investment in the relationship 
is critical to minimizing tensions and maximizing productive interactions. 
By establishing a “floor,” I mean that through their joint agreement to 
pursue a relationship of “constructive strategic stability based on mutual 
respect, fairness, and reciprocity,” the two leaders have put a foundation 
under the wildly fluctuant and steadily deteriorating relationship of the 
past decade.

Now, in some ways, comes the hard part: rebuilding the “walls” of the 
U.S.-China edifice. No structure can stand with only a floor and a ceiling. 
There must be walls to connect the ceiling and floor and hold the building 
up. This refers to the bilateral institutions and mechanisms that provide 
governmental channels for direct discussions, deconfliction of differences, 
and forging cooperation where possible＂across a wide range of bureaucratic 
domains and issue areas. Today, regrettably, there hardly exist any meaningful 
and institutionalized channels between the two governments and security establishments.
 Thus the relationship remains intrinsically unstable.

Inter-Governmental institutionalization was previously a key part of what 
became known as the U.S. “engagement” policy (the other components largely 
consisted of non-governmental exchanges). By the end of the Obama administration 
the two sides had erected nearly 100 such bilateral channels (the most notable 
of which was the Strategic & Economic Dialogue or SAED), before the first 
Trump administration slashed them to two in 2017, adding two more by 2020, 
while the Biden administration maintained six formal mechanisms (National 
Security Advisor Sullivan and State Councilor Wang Yi, a financial working 
group, economics working group, counter-narcotics working group, AI governance 
and safety dialogue mechanism, and some military-to-military communication 
channels). The current Trump administration has been highly reluctant to 
reestablish any such formal institutionalized mechanisms, although Presidents 
Trump and Xi did agree to establish a Board of Trade and Board of Investment 
at their May summit in Beijing, as well as a “Super Intelligence” (SI) 
dialogue mechanism at the recent Washington summit.

It is indeed doubtful that the Trump administration wishes to resurrect 
and rebuild some of the inter-bureaucratic mechanisms of earlier administrations,
 since the entire administration's foreign policy is highly transactional 
and centered on President Trump himself. Beyond commercial and perhaps student 
exchanges it is not at all evident that President Trump seeks a resurrection 
of the past comprehensive relationship of “engagement” with China (as 
I explain in my recent book Breaking the Engagement, those days are over). 
Both governments have now considerably “securitized” the relationship＂
establishing significant limits on exchanges with the other on the grounds 
of national security＂which is going to be nearly impossible to undo.

Relatedly, over the past decade, national and sub-national bureaucracies 
and legal establishments in both countries have increasingly based their 
budgets and institutional missions on countering one another. The financial 
resources that allow all of these bilateral governmental bureaucracies to 
operate are now baked into a competitive＂if not adversarial＂relationship.

In this regard, the recent Xi-Trump bonhomie offers a classic example of 
the classic “agent-structure” paradigm in political science. That is, 
the two leaders have exercised their agency not only in establishing a 
“friendly” atmosphere in their past two summits, but President Trump has 
also exercised his presidential agency by ordering his administration to 
take certain specific steps intended to not aggravate President Xi and to 
deconflict relations with Beijing＂specifically holding up the $14 billion 
arms package for Taiwan, warning Japanese Prime Minister Takaichi Sanae 
to provoke Beijing over the Taiwan issue, saying absolutely nothing about 
human rights abuses in China, and instructing his administration not to 
say anything publicly that might aggravate Beijing and that every issue 
of significance concerning China must cross his desk for approval.

Thus, rebuilding the government edifice of U.S.-China relations not only 
involves reconstructing (or launching new) intergovernmental bureaucratic 
dialogues and mechanisms between the two governments. It also includes redirecting 
the structural institutional missions and underlying financial resources 
within each government. Decades of political science research has proven 
that in some instances leaders can exercise their agency and impose their 
personal imprint on bureaucracies, usually temporarily＂much more often 
these institutional structures prove stronger, more resilient and patient, 
and resist the leader's initiatives while protecting their own core institutional 
missions and resources. Even if the two sides were to restart bilateral 
bureaucratic exchanges, the intrinsic comprehensively competitive nature 
of the relationship will persist indefinitely and thus there remains inherent 
instability and considerable fragility to the new detente. It would indeed 
behoove both sides to revisit, resurrect, and revise mechanisms that the 
United States and Soviet Union successfully forged during the detente phase 
of the late Cold War＂such as the Helsinki Process, the Conference on Security 
and Cooperation in Europe (CSCE), the SALT/START nuclear arms control agreements,
 and a variety of diplomatic and security confidence building measures (CBMs)
. That U.S.-Soviet era was different, but the intrinsic logic of these mechanisms 
may well have positive potential in the U.S.-China case today.

Thus, after the two recent summits between Presidents Trump and Xi Jinping, 
an improved and warmer atmosphere has been established at the leadership 
level and between the two governments. This is good and the relationship 
is now more stable as a result. But, as described above, the hard work of 
re-institutionalizing the relationship, deconflicting their very real differences 
and ameliorating the intrinsic competition will be much harder to accomplish.


中美关系重塑是否迎来转折点

沈大伟教授 美国乔治华盛顿大学

特朗普总统与习近平主席在 2026 年 5 月和 9 月举行的两场峰会，成功重塑了中
美关系大厦中可喻为 “天花板” 与 “地板” 的架构。所谓 “天花板”，指两国
领导人之间融洽的个人关系。自 1972 年尼克松与毛泽东实现关系破冰以来，中美
关系发展史反覆证明，领导人的亲自参与和对双边关系的投入，对于最大限度减少
紧张局势，最大限度拓展良性互动至关重要。而所谓 “地板”，是指两国领导人共
同努力，构建 “基于相互尊重、公平和互惠的建设性战略稳定关系，为过去十年剧
烈波动、持续恶化的中美关系筑牢根基。

现在，某种程度上说，真正的困难已摆在眼前，那就是重筑中美关系这座大厦的 
“墙体”。任何建筑仅有地基与顶层都无法屹立，必须依靠墙体连接上下，来支撑
整座建筑。这里的墙体，指的就是各类双边制度与机制，它们覆盖众多政府部门与
议题范畴，为两国政府提供渠道，用以直接对话、管控分歧，并在力所能及的领域
开展合作。遗憾的是，当前两国政府及安全机构之间几乎不存在具备实质意义的制
度化沟通渠道，这也使得双边关系本质上依旧不稳定。

政府间的制度化安排，曾是美国所谓 “接触政策” 的核心组成部分（该政策其余
内容大多由民间交流构成）。奥巴马政府末期，两国搭建了近百个这类双边沟通渠
道，其中最知名的便是中美战略与经济对话（S&ED）。然而特朗普第一届政府在 2017 
年将这些机制压缩至仅剩两个，到 2020 年也只增加了两个。拜登政府时期则保留
了六个正式机制，包括国家安全顾问沙利文与国务委员王毅的沟通渠道、金融工作
组、经济工作组、禁毒工作组、人工智能治理与安全对话机制，以及若干两军沟通
渠道。本届特朗普政府对于重建这类正式的机制化渠道态度十分勉强，不过，特朗
普总统与习近平主席仍在今年 5 月北京峰会上同意设立贸易委员会和投资委员会，
并在最近的华盛顿峰会上设立了一个 “超级智能” （SI）对话机制。

特朗普政府是否有意恢复并重建往届政府时期的跨部门机制，这一点确实存疑。毕
竟，本届政府的整体外交政策带有浓厚的交易属性，并且高度以特朗普总统个人为
核心。除经贸往来，或许再加上学生交流之外，尚无明显迹象表明特朗普总统希望
恢复过去那种全面的对华 “接触” 关系（正如我在新作《终结接触》中所述，那
样的时代已经一去不返）。如今，两国政府都已在很大程度上将双边关系 “安全化”，
以国家安全为由，对彼此间的各类交流设置重重限制，而这种局面几乎难以逆转。

与之相关的是，过去十年间，两国中央及地方政府部门、司法体系，越来越多地将
自身预算与机构使命建立在对抗对方的基础之上。支撑所有双边政府机构运转的财
政资源，如今已被锁定在竞争性的──甚至可以说是敌对性的──关系格局之中。

就此而言，习特二人近期的融洽互动，正是政治学经典“施动者-结构”范式的典型
案例。也就是说，两位领导人不仅通过发挥自身能动性，在两次峰会中营造出友好
氛围，特朗普总统还通过行使总统权力，下令其政府采取若干特定措施，来避免激
怒习主席，缓和与北京的关系。具体措施包括暂缓推进价值 140 亿美元的对台军售
方案、告诫日本首相高市早苗不要在台湾问题上挑衅北京、完全不提所谓中国人权
问题，并指示其政府不得公开发表可能激怒北京的言论，所有涉及中国的重大事项
均须经他本人审批。

因此，重建中美关系的政府间架构，不只是要恢复（或新设）两国政府间的跨部门
对话与机制，还需要调整两国各自内部结构性的机构使命，以及背后的财政资源。
数十年政治学研究表明，在某些情形下，领导人能够发挥其能动性，将个人意志施
加于行政体系，但这种影响通常是暂时性的。更多时候，这类制度性架构被证明韧
性更强、存续更久，在守护自身核心机构使命与资源的同时，抵制领导人的政策尝
试。即便双方重启两国政府部门间交流，双边关系根深蒂固的全面竞争属性仍会长
期存在，因此这次新的缓和局面天生带有内在不稳定性，是相当脆弱的。双方理应
重新审视、恢复并修订那些美苏在冷战后期缓和阶段成功搭建的各类机制，例如赫
尔辛基进程、欧洲安全与合作会议、战略武器限制谈判/削减战略武器条约等核军控
协议，以及各类外交与安全领域的建立信任措施。美苏时代虽与当下不同，但这套
机制的内在逻辑，对今天的中美关系仍有积极的借鉴价值。

综上，在特朗普总统与习近平主席近两场峰会之后，领导人层面及两国政府之间已
经建立了更为良好融洽的氛围。这是好事，双边关系也因此变得更加稳定。但正如
上文所述，重建两国关系制度框架、管控双方切实存在的分歧、缓解两国固有的竞
争状态，这些艰巨工作要完成起来则困难得多。

(The source of the article comes from  chinausfocus.com)