Policing Social Media in China | rheacock | December 10, 2013


Policing Social Media in China

Robert Faris

China is renowned for its lukewarm embrace of the open Internet and its willingness to go to great lengths to curtail online speech. Its longstanding Internet filtering apparatus, the so-called Great Firewall, is intended to prevent users from accessing thousands of websites hosted outside of China. This centrally coordinated system is based on maintaining a running list of keywords and web addresses to be blocked. In technical terms, it is quite sophisticated. In terms of content control, it is crude; thousands of innocuous sites are caught up by the keyword-based logic, while much controversial content continues to leak through. Amid the thousands of keywords and web addresses on the block list, the blocking of a handful of social media sites—Facebook, Twitter, YouTube, and various blog hosting platforms—has arguably had the biggest impact on the Internet in China by ensuring that domestic firms have come to dominate social media markets in China. This means that control of social media content in China is a domestic affair.

Over the past several years, microblogging has emerged as the heart and soul of a remarkably vibrant networked public sphere in China. Social media in China is staggering in scale, both in the number of participants (registered accounts are currently estimated at about half a billion) and in the breadth of topics that are discussed. For government censors, this represents a very different and far more challenging task.

A number of studies over the past year have shed a great deal of light on the mechanisms that are employed in China. The first step is holding the intermediaries responsible for the content that passes through their platforms. This in turn has prompted software companies to produce tools for social media sites to support a hybrid approach in which technical filters flag content for subsequent human review. This approach offers a more fine-grained approach to blocking content that incorporates human judgment, but at a cost. Back of the envelop calculations suggest that social media companies employ tens of thousands of people to manually review individual posts.

Among the estimated one hundred million posts each day, a substantial number never make it through the review process for public viewing. And for those that survive the initial technical screen, studies estimate that another 10-15 percent of posts are subsequently taken down. These activities leave a digital record that allows researchers to study the targeting of social media censorship. The evidence supports the view of a system that allows discussion of many controversial topics, but responds quickly and decisively to prevent selected topics from catching fire in digital media. The surprising twist is that criticism of the government is apparently not a factor in social media censorship. Many posts that are highly critical of the government are allowed online as long as they are not related to hot button topics, while posts that are supportive of government positions are taken down if related to the most sensitive issues.

This is partial vindication for those who believed that preventing ideas from being spread via the Internet would prove to be impossible: technology is triumphing over the political will of repressive governments. However, it supports the notion that authoritarian regimes fear collective action above all else. The most recent wave of blogger arrests, which includes many high profile bloggers, is a sign that the government is wary of the power of social media in China, despite the massive monitoring and take-down regime in place. It also points to the inherent fragility of civil society action online.

Additional Reading

Gary King, Jennifer Pan, and Margaret Roberts, “How Censorship in China Allows Government Criticism but Silences Collective Expression,” American Political Science Review 107, no. 2 (2013): 1-18, http://gking.harvard.edu/publications/how-censorship-china-allows-government-criticism-silences-collective-expression.

Tao Zhu, et al., “The Velocity of Censorship: High-Fidelity Detection of Microblog Post Deletions,” July 10, 2013, http://arxiv.org/abs/1303.0597.

David Bamman, Brendan O’Connor, and Noah Smith, “Censorship and deletion practices in Chinese social media,” First Monday 17, no. 3 (2012), doi:10.5210/fm.v17i3.3943.


  1. Gary, Jennifer Pan, and Margaret Roberts, “ A Randomized Experimental Study of Censorship in China,” October 6, 2013, http://gking.harvard.edu/publications/randomized-experimental-study-censorship-china.

Text Information

December 10, 2013


Author Stats


Leitura Garamond Futura Verdana Proxima Nova Dagny Web
small medium large extra-large