Crisis communication dimensions shape governance responses but remain understudied, particularly in transitional contexts like China, where provincial governments navigate multiple priorities amidst institutional changes. Using dynamic topic modeling on 245,022 COVID-19-related social media posts from Chinese provincial governments (2019-2022), we identify three interpretive dimensions: economy, livelihood, and administration. Results indicate administration received the greatest attention. Regression analysis reveals that emphasizing the administration dimension significantly reduced monthly confirmed COVID-19 cases, while economy and livelihood dimensions showed no significant effects. Findings suggest strategically highlighting administrative competence enhances governance outcomes by reinforcing legitimacy and accountability to citizens and central authorities. Theoretically, the study elucidates how institutional incentives shape crisis framing strategies that, in turn, influence governance performance. Practically, it shows that aligning communication dimensions with crisis phases and institutional priorities improves response performance and offers transferable insights for structured crisis governance. Methodologically, it demonstrates the value of dynamic topic modeling for capturing temporal shifts in crisis communication.
Li et al. (Sun,) studied this question.