This study examines the intricate relationship between censorship mechanisms and the formation of public opinion on Sina Weibo, China's prominent microblogging platform, with particular attention to implications for Sustainable Development Goal 16 regarding public access to information. Through analysis of platform policies, censorship patterns, and user behaviors between 2009 and 2024, this research demonstrates how digital control shapes public discourse in China's online sphere. The findings reveal sophisticated censorship strategies that extend beyond mere content deletion to include algorithmic manipulation, strategic timing of interventions, and the cultivation of self-censorship among users. The research identifies multiple adaptive strategies employed by Weibo users to circumvent restrictions, including linguistic innovation, metaphorical expression, and temporal coordination. This manuscript contributes to scholarly understanding of how information control mechanisms impact sustainable development targets related to institutional transparency and information access. The implications extend to broader questions about the evolution of public discourse under digital authoritarianism and how restricted information flows affect citizens' ability to engage with sustainable development challenges. The objective of this study is to investigate the relationship between censorship and public opinion on Weibo, with the aim of understanding how censorship mechanisms influence online discussions and affect collective action. This study is grounded in key theories that illuminate how censorship shapes discourse and public opinion on Weibo. The digital public sphere, adapted from Habermas (1989) and explored in digital and authoritarian contexts (Dahlgren, 2005; Papacharissi, 2002; boyd, 2010), provides a framework for understanding how online discussions unfold within state-imposed constraints. Censorship as a dynamic system (Roberts, 2018; Creemers, 2017) moves beyond simple content deletion to include strategies like “information friction” and “strategic censorship” (Bandurski, 2018), which influence not only what is removed but also how users preemptively alter their behavior. Public opinion formation under censorship builds on Noelle-Neumann’s (1974) “spiral of silence,” explaining how individuals self-censor when facing political and social risks, while research on algorithmic curation (Tufekci, 2015; Gillespie, 2014) and “engineered publics” (Jiang & Fu, 2018) highlights how visibility and discourse are shaped by both political controls and platform mechanisms. Together, these perspectives provide a comprehensive lens for analyzing Weibo’s evolving censorship landscape and its impact on digital expression. This study employs a mixed-methods approach to analyze censorship and public opinion on Weibo from 2009 to 2024. Data collection includes (1) analysis of Weibo’s policies and Chinese government regulations, (2) monitoring of 200 trending topics (2020–2024) using Weibo’s API, web scraping, and external censorship-tracking initiatives, and (3) investigation of user adaptation through content analysis of 10,000 posts, and digital ethnography of key events like COVID-19 and the 20th Party Congress. Data analysis combines content analysis to track censorship mechanisms and adaptation strategies, supported by computational tools and human coding to ensure both large-scale pattern detection and contextual depth. The findings reveal that Weibo’s censorship operates through keyword filtering, account restrictions, algorithmic controls, temporal censorship, and selective enforcement, shaping narratives, delaying discourse, and adapting to political events. The findings support digital governance theories, showing censorship as an evolving process rather than a fixed system. It delays public discourse, fosters multiple publics, and disproportionately suppresses early-stage information, reinforcing theories on networked publics and information control. These results highlight key implications for SDG 16, emphasizing transparency and information access. The 89% removal rate for government criticism versus 24% for corporate criticism reflects institutional narrative control. Censorship of environmental and feminist discourse limits advocacy, while suppression of early COVID-19 discussions worsens information disparities, affecting governance and public awareness. Given censorship’s opacity, future research should use machine learning to track evolving patterns, compare censorship across Chinese platforms, and examine global trends. Further studies should assess its impact on sustainable development and long-term discourse shifts, expanding frameworks on adaptive preferences in censored media. This research enhances understanding of digital censorship by illustrating how visibility manipulation and algorithmic controls restrict information. It informs policymakers, digital rights organizations, and platform regulators about evolving censorship practices, emphasizing the need for transparency, algorithmic accountability, and governance reforms. Additionally, it offers insights for activists, journalists, and social movements on adaptive resistance strategies. This study provides an empirical analysis of Weibo’s censorship from 2009 to 2024, integrating keyword filtering, algorithmic control, and user adaptation strategies. Its mixed-methods approach—combining content analysis, digital ethnography, and computational tracking—offers a comprehensive perspective on censorship’s impact. By documenting user resistance, this research contributes to global discussions on digital authoritarianism and information control while enriching studies on digital resilience in restrictive online environments.
Han et al. (Mon,) studied this question.