Fighting COVID-19 in China's Greater Bay Area: how do different policy styles affect policy effectiveness?

Background: The COVID-19 pandemic has highlighted how political systems influence public health policy effectiveness. This study examines how different governance styles within China's Greater Bay Area shaped pandemic responses, comparing Guangdong Province, Hong Kong Special Administrative Region (SAR), and Macau SAR as representatives of society-mobilizing, market-facilitated, and targeted-control styles, respectively. This quasi-natural experiment setting controls for cultural and geographic variables while allowing for meaningful institutional comparison. Methods: Using a mixed-methods approach, the study analyzed COVID-19 infection statistics and 75,870 social media posts from Weibo and X (December 2019-December 2022). The analysis employed statistical methods, sentiment analysis via the Baidu Application Programming Interface (API), and Latent Dirichlet Allocation (LDA) topic modeling to examine policy styles, effects, and public feedback. Interrupted time series (ITS) analysis was also applied to assess policy impacts across three distinct pandemic phases. Results: Guangdong's society-mobilizing approach maintained stable case numbers (mean = 31.03 daily cases in Guangzhou) with the highest sentiment scores (0.54) and 51.64% positive reactions. Hong Kong's market-facilitated approach showed the highest infection rates (mean = 442.95) and lowest sentiment scores (0.46). Macau's targeted-control approach achieved the lowest infection rates (mean = 0.79) with moderate sentiment scores (0.47). Interrupted time series (ITS) analysis revealed distinct transmission trends in each region, with significant changes observed during the Omicron phase in Guangzhou and Hong Kong, and sustained low transmission in Macau. Topic modeling identified region-specific concerns: overseas case imports (Guangdong), vaccine and local case monitoring (Hong Kong), and casino impacts (Macau). Conclusion: The study demonstrates that effective pandemic response depends on governance-society alignment, particularly during early outbreaks. While all three systems achieved relative success, their effectiveness varied based on institutional capacity, suggesting that successful crisis management requires consideration of political social structures while maintaining adaptability in transitioning from containment to endemic management. These findings offer a transferable framework for evaluating governance effectiveness in public health crises beyond the Greater Bay Area.