Quantifying government guidance in online policy communication: A hypernetwork approach to China's delayed retirement reform

Social media has become a central arena in which public policies are interpreted, circulated, and emotionally negotiated. Existing studies of online policy communication often rely on single-layer indicators such as repost frequency, comment volume, topic proportion, or sentiment polarity, which are insufficient for diagnosing whether structurally visible official communication is semantically aligned with public concerns. Taking China's delayed retirement reform as an empirical case, this study develops an agent–content–topic hypernetwork framework and proposes an exploratory Guidance Capacity indicator integrating actor position, textual influence, and topic–sentiment alignment. Based on 482 Weibo posts and 1527 noise-reduced comments collected from September 11 to November 18, 2024, the results show that official accounts retained institutional visibility and guidance potential, but their textual influence was weaker than that of several non-official and media-oriented actors. Public discussion concentrated on pension expectations, employment pressure, cost-of-living concerns, market conditions, and career adaptation. Although positive sentiment accounted for the largest share of the observed corpus, this finding is interpreted as platform-specific observed polarity rather than as nationally representative policy support. Topic trajectories further reveal a shift from national policy interpretation to household and occupational concerns. The study contributes to digital policy communication research by showing how higher-order relations among actors, content, topics, and sentiment can reveal structural–content mismatches that are not visible in conventional social network analysis or text-only analytics.