The United States and China entered the 2015–2025 period at rough bilateral soft power parity. By its close, that parity was gone, and the deterioration on the American side had less to do with declining cultural appeal than with a deliberate, institutionally organized Chinese effort to foreclose the conditions under which American soft power could function. Joseph Nye's original formulation defined soft power as the capacity to attract through culture, values, and diplomacy rather than compel through force. What Joseph Nye labeled "sharp power" in 2018 describes something different: the manipulation of information environments, suppression of rival narratives, and coercive use of institutional levers to achieve influence outcomes that cultural attraction alone could not produce.
This brief measures that shift across three bilateral channels — educational exchange, digital media, and cultural tourism — using a Balance of Exchange ratio that tracks directional influence flow between the two nations from 2015 to 2025. The aggregate finding: American outflow soft power scores fell from 15.7 to 4.4, while China's scores rose from 15.0 to 20.2. A net shift of 15.8 normalized points over a single decade constitutes a structural transformation, not a cultural trend.
The education channel is where the distance between surface appearance and structural reality is most pronounced. Chinese enrollment at U.S. universities remained the largest foreign national cohort in 2024/25, at 265,919 students and 22.6% of total international enrollment, per IIE Open Doors 2025. The decline from the 2019/20 peak of 372,532 is notable but leaves Chinese educational influence in the United States essentially intact.
The reverse flow has effectively ended. American enrollment at Chinese universities fell from roughly 15,000 in 2012 to approximately 1,200 by 2025, a 92% reduction that represents, for practical purposes, the elimination of meaningful American academic presence in China. Three institutional mechanisms drove this asymmetry. Optional Practical Training (OPT) allows international graduates of U.S. institutions to remain for up to three years of post-graduation work experience, extending and deepening Chinese students' exposure to American society. No reciprocal program exists for Americans in China. Visa reciprocity friction, documented by Lau (2019), created substantially higher administrative burdens for Americans seeking Chinese study visas during the trade dispute period. Third, the security climate surrounding Chinese student enrollment — concerns about surveillance, intellectual property exposure, and state-directed influence operations — corroded the institutional trust on which genuine educational exchange depends, further suppressing American interest in China-based study.
The net educational exchange score registers China at +6.4 (normalized), driven by institutional asymmetry rather than relative academic quality or cultural attractiveness.
The digital domain provides the clearest illustration of sharp power in operation and produces the decade's largest bilateral shift: a net swing of 8.4 normalized points in China's favor.
The Great Firewall, the suite of DNS filtering, IP blocking, deep packet inspection, and keyword monitoring that controls Chinese citizens' access to the external internet, is not incidental infrastructure. Faris and Villeneuve (2008) and Gao (2014) both document it as a deliberate political choice, initiated in the late 1990s and systematically intensified under Xi Jinping after 2012. Google, Meta, YouTube, Wikipedia, Netflix, and the major Western news organizations are all blocked. American digital soft power in China registered a normalized score of 2.2 in 2015; continued enforcement tightening reduced that figure to 0.8 by 2025.
"The Great Firewall does not merely filter — it forecloses. It eliminates, at the architectural level, the conditions under which American soft power could generate the attractions that Nye's theory describes."
At the same time, the American digital market remained structurally open to Chinese platform presence. TikTok, owned by ByteDance and subject to China's National Security Law (2017), grew to 170 million monthly U.S. active users by early 2025, per ByteDance's own Supreme Court filing. This represents the largest single penetration of a foreign state-linked platform in American digital history.
A precision worth establishing here: TikTok is Chinese-owned infrastructure predominantly carrying American-generated content. The normalized score of 9.2 reflects platform dominance, data architecture reach, and algorithmic influence capacity, not Chinese cultural content dominance. These are meaningfully different claims. A CSIS brief from October 2024 documented specific concern that the Chinese Communist Party could influence user feeds, suppress dissent, or spread disinformation through platform-level controls that are difficult to detect externally. Whether or not that capacity has been actively exercised, its structural existence constitutes a form of sharp power in its own right.
The 2025 legislative episode underlines the point. The Protecting Americans from Foreign Adversary Controlled Applications Act required ByteDance divestiture or a ban. The Supreme Court upheld the law in January 2025. TikTok went briefly dark before the requirement was suspended through executive action. A platform whose removal required simultaneous engagement from the legislative, judicial, and executive branches has achieved structural embeddedness that places it well outside the category of peripheral commercial presence.
The BRI does not appear in the bilateral Balance of Exchange table because its primary effects are not bilateral. It operates in the Global South, reshaping political alignments in regions where American soft power once operated without Chinese competition. Its inclusion here reflects its function as the macroeconomic scaffold beneath the bilateral findings: the mechanism by which the sharp power transition documented domestically is operationalized globally.
Shen and Chan (2018) draw the explicit comparison to the Marshall Plan: both programs win political alignment through material benefit rather than cultural attraction. The Marshall Plan was accompanied by democratic institution-building and ideological content. The BRI generates dependency without an equivalent ideological prerequisite, achieving political outcomes through structural leverage. Over 140 nations have entered BRI partnership arrangements (UNCTAD, 2019), receiving infrastructure investment, loan disbursements, and technical assistance. The cumulative effect is the construction of durable political alignments in regions where American soft power previously operated largely uncontested, and those alignments reduce the effective global audience for American influence narratives without requiring a single Chinese cultural export to have succeeded on its own merits.
The Great Firewall, TikTok, and the BRI share a common strategic orientation. The Firewall forecloses American influence within China's information environment. TikTok expands Chinese platform infrastructure within America's open one. The BRI reshapes the global terrain through material dependency in regions that cannot be reached by cultural appeal alone. Together they describe a coherent strategy, not a set of parallel policy instruments operating by coincidence.
The American structural constraint is equally coherent. The United States has not built a firewall against Chinese digital influence, and its constitutional culture makes such a measure politically implausible. It has not deployed state-linked platforms into the Chinese market; the GFW would prevent them from operating. The openness that constitutes American soft power's fundamental appeal is precisely what a sharp power strategy targets. The TikTok legislative episode demonstrated both the recognition of this problem at the highest levels of American government and the difficulty of resolving it within a liberal constitutional framework designed for a different threat environment.
One qualification that matters for the analytical record: nothing in this analysis implies that Chinese culture, political institutions, or values have become more globally attractive than their American counterparts. American universities, cultural exports, technological capacity, and democratic institutions retain genuine international appeal. The bilateral metrics document what happens when one party deploys sharp power instruments against a counterpart committed to an open information environment, not what would result from a straightforward contest of cultural merit.
The decade from 2015 to 2025 did not produce a Chinese cultural victory. It produced a structural advantage built on institutional asymmetry: a firewall that sealed China's information environment against American influence, a platform that penetrated America's open digital market to an unrivaled depth, and an infrastructure program that reshaped the political geometry of regions where American soft power had previously operated without meaningful competition. Whether the United States develops systematic responses to this challenge, without compromising the openness on which its own soft power fundamentally depends, is the question the decade ahead will need to answer.