All authoritarian regimes seek to manage the flow of information to create a “mainstream consensus” (Geddes & Zaller, 1989) that shapes public opinion. While traditional authoritarian leaders like Stalin or Hitler instilled fear and relied on violence, modern, tech-savvy authoritarian rulers adopt “third-generation” strategies. These methods employ soft, subtle tactics to influence public opinion in ways favourable to the regime (Deibert et al., 2010), often proving more persuasive than traditional hard propaganda (Huang, 2018; Mattingly & Yao, 2022). The advancement of ICTs has not only empowered citizens and become a liberation technology (Diamond, 2010), but also expanded the toolbox available to authoritarian rulers, resulting in “spin dictators” who manipulate information to engineer support (Guriev & Treisman, 2023). The subtle management of public opinion involves innovative strategies, such as hiring online “troops” to fabricate support for the regime or garner consent for policies through what appear to be unsolicited citizen comments (Han, 2015; King, Pan, & Roberts, 2017; Sobolev, 2019; Bradshaw & Howard, 2017), and appropriating popular cyberculture and digital populism to reinvent official culture and make propaganda appealing (Guo, 2018). However, the effectiveness of these strategies often hinges on the representation of meaning into storylines – narratives, which are the actual realization of materials/elements into stories, with the instrumental aim of influencing the opinions and behaviour of others (Miskimmon, O’Loughlin, & Roselle, 2014). This paper focuses on narratives and investigate the questions: How do authoritarian regimes construct and disseminate mainstream consensus on political and social events of different natures?

Building on the “Narrative Policy Framework” (Jones et al., 2014), we propose the theoretical framework of “event-based narratives”, which refers to the representation of political and social reality through narratives composes of elements, strategies, and beliefs. Empirically, we randomly sampled verified Weibo users and collected approximately three million posts from government, media, and celebrity users between January and May 2022. Using the subject-verb-object (SVO) extraction, we reverse-engineer these messages to uncover the stories of “who does what to whom” and infer underlying strategies and beliefs. Furthermore, we trace the spread of specific narratives over time by comparing the timestamps of their dissemination by different opinion leaders online. We find that the construction and dissemination of mainstream consensus in China’s cyberspace largely depend on the nature of the events. For domestic events, the mainstream consensus centres on the image of benevolent leadership, while for international events, it focuses on collective national identity. In state-initiated events, the state can craft and coordinate a response that highlight positive aspects, thereby boosting popular support. In contract, during unexpected events, finding a scapegoat and shifting blame becomes the easiest solution. Our work offers new insights into the routine and operational consensus-building and its role in propaganda and political campaigns within and beyond authoritarian regimes. The quantitative narrative analysis of text data enables us to extract the SVO network of the dominant narratives and the tracing of diffusion patterns, providing rich insight into the construction and dissemination of the mainstream consensus.

Speaker: Dr Yan Wang (University of Manchester)

Chair: Professor Bingchun Meng (Director of the LSE-Fudan Global Policy Hub)

Venue: OLD 2.21 (2nd floor, Old Building, LSE)

This seminar is open to LSE staff and students only.

Lunch and refreshments will be served before the seminar.

Hosted by the LSE-Fudan Global Public Policy Hub and the Department of Social Policy, LSE