Explore UCD

UCD Home >

Understanding authoritarian propaganda strategies with audio and text data from state-sponsored cable news

Seminar: Emotional Propaganda: Understanding Authoritarian Affective Manipulation with an Audio-as-Data Approach to Chinese State Media During the COVID-19 Outbreak Haohan Chen (University of Hong Kong)

14:00-15:00 (GMT) Wednesday, December 1.

Please register for this event (opens in a new window)here.

Abstract: Authoritarian regimes use propaganda for two purposes: changing minds and changing hearts. Existing works focus on the former, extensively analyzing how propaganda indoctrinates or threatens the population while overlooking the latter, propaganda for states’ strategic affective manipulation, namely, emotional propaganda. We theorize that authoritarian states employ emotional propaganda to preclude public grievances that undermine authoritarian control. We test the theory by applying an audio-as-data approach to China’s propaganda on state-run media during the COVID-19 outbreak in 2020. We collect original audio-visual recordings of China’s most-viewed daily evening news program, Xinwen Lianbo, during the period. We construct audio-based measures, including vocal pitch, speech rates, genders of speakers, and speaking modes, as indicators of the state’s strategic emotional propaganda. We pair them with daily reports of COVID-19 cases and social media sentiments about COVID-19 to examine how the state strategically deployed emotional propaganda as a response to the public health crisis and public grievances it caused. Our study provides the first systematic analysis of emotional propaganda, an important propaganda strategy of authoritarian regimes. Methodologically, we highlight the unique advantages of the audio-as-data approach in detecting emotions and apply this emerging computational tool to the study of comparative political communication.

About the speaker: Haohan Chen is an Assistant Professor of Politics and Public Administration at The University of Hong Kong. He develops computational tools for the collection and analysis of text and network data from social media. He applies these tools to study political communication under both authoritarian and democratic contexts from a comparative perspective.