Skip to main navigation Skip to search Skip to main content

How Citizens Respond to Infoganda: Evidence from Shanghai COVID Lockdown in 2022

Research output: Other Conference ContributionsPresentation

Abstract

Government-authored media content that blends propaganda with information has seen substantial growth over the past decade. How do citizens react differently to these types of content? The lack of variation in treatment status and the small number of observations—owing to the ubiquity and monopoly of authoritarian propaganda apparatus—make citizen-state interactions difficult to identify in observational data. Using daily street-level data during Shanghai’s spring 2022 COVID lockdown, we collected 90,000 street-level government WeChat posts and over 5,000 citizen petitions to precisely match propaganda senders with their attentive receivers. Advances in large language models allow us to discern information delivery from propaganda messages and classify the latter by the main actors in the content: the central state, local state, or ordinary citizens. Our results show that citizens can discern information from propaganda and pay much more attention to the former during a crisis. Among propaganda pieces, citizens focused more on those covering local government performances rather than central state ideologies or narratives highlighting ordinary citizen heroism.
Original languageEnglish
Publication statusPublished - 11 Apr 2025
EventChinese Politics Research in Progress (CPRP): A Virtual Workshop on Chinese Politics -
Duration: 11 Apr 202611 Apr 2026

Workshop

WorkshopChinese Politics Research in Progress (CPRP): A Virtual Workshop on Chinese Politics
Period11/04/2611/04/26

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

  1. SDG 3 - Good Health and Well-being
    SDG 3 Good Health and Well-being

Fingerprint

Dive into the research topics of 'How Citizens Respond to Infoganda: Evidence from Shanghai COVID Lockdown in 2022'. Together they form a unique fingerprint.

Cite this