Government responsive selectivity and public limited mediation role in air pollution governance: Evidence from large scale text data content mining

Social media has been widely discussed as an informal way for the public to present their environmental appeals. However, evidence on whether and how the government's decision-making will take the public in the non-Western electoral system under consideration remains limited. In this study, we extracted 9.25 million haze-related posts from Sina Weibo and analyzed the topics using the latent Dirichlet allocation topic model. Furthermore, we provide evidence of the Chinese governments’ responsiveness and the role of online public participation in environmental governance. The results show that the Chinese government responds to the public, both online and offline. Specifically, when online public appeal increases by 1%, the local government's posts on social media increase by 0.347%, and investment and regulations increase by 0.0676% and 0.074%, respectively, in the next phase. However, there are selective characteristics of government response: the local government is more concerned about its “political achievements” than undertaking “governance responsibility.” Furthermore, online public participation plays a minor role in the implementation of regulations.