Narrative dynamics on social media platforms are shaped not only by content but by the governance regimes that structure their circulation. This study applies the $SEI_AI_DZ$ epidemiological model to examine how competing tariff narratives spread differentially across X, a Western open-access platform, and Weibo, a Chinese government-moderated platform. Analyzing 90,150 Weibo posts and 16,979 X posts collected from February through August 2025, we operationalize the transmission rate ($beta$) and the basic reproduction number ($R_0$) for narratives expressing agreement or disagreement with U.S.-China tariff positions. Results reveal stark platform-specific contagion patterns. On X, both agreeing and disagreeing narratives exhibit suppressed contagion ($R_0 < 1.0$), with near-parity between stances ($R_0 = 0.28$ vs. $0.35$), indicating that open platforms suppress viral spread regardless of narrative alignment. On Weibo, agreeing narratives achieve $R_0 = 1.8$ ($beta = 0.706$) while disagreeing narratives collapse to $R_0 = 0.2$ ($beta = 0.219$), a 190-fold transmission-rate difference. These findings demonstrate that platform governance operates as an epidemiological intervention, structurally enabling or suppressing narrative reproduction at scale, with direct implications for understanding information ecosystems in polarized geopolitical contexts.