Disaggregating Public Security Propaganda on Chinese Social Media

This paper aims to evaluate the effectiveness of public security propaganda on the Chinese social media network Sina Weibo. A piece of propaganda is disaggregated into two constituent parts: its style and its content. Using a dataset of roughly 2.8 million Weibo posts from a total of 330 public security accounts at the prefectural, provincial, and central levels, I examine whether the style and content of public security posts change over time and how these individually affect the posts’ propagation, as operationalized by the number of retweets. Preliminary results suggest that public security accounts have increasingly adopted a softer posting style as time progresses – though this adoption is not uniform – and that this softer style has facilitated the propagation of their posts. These preliminary results imply that a number of public security accounts are adapting to new forms of communication when spreading their message.