You choose, we deliver
If you are interested in this story, you might be interested in others from The Journal Gazette. Go to and pick the subjects you care most about. We'll deliver your customized daily news report at 3 a.m. Fort Wayne time, right to your email.

Editorial columns


New censorship study reveals what Beijing fears

While living for more than a decade in China, and using its thriving social media, no question came to mind quite so often as: “Who is the idiot who just censored that online post, and what on Earth was so dangerous about it?”

I was hardly alone in my frustration. While online social media has transformed civil society in China, most users have come up against the limits of free expression. The one impediment to Chinese people connecting as a whole was, and is, their own government.

As a group of Harvard University researchers show in a study, however, responsibility for this state of affairs rests with social-media platforms as much as the Chinese regime. Censorship in China has evolved into a kind of private-public partnership, with the government setting the parameters, and Internet companies free to “innovate” in finding ways to meet them – at their own expense. The system is a perverse form of blackmail: If the companies don’t play ball, they risk attracting users who defy the state’s edicts on information.

The Harvard group used subterfuge, setting up its own Internet bulletin board in China where users could foster and engage in online discussions. The software didn’t include censoring tools (necessary if you don’t want to be shut down). So the researchers reached out to their Web hosting company and found them “forthcoming when we asked for recommendations as to which technologies have been most useful to their other clients.”

The researchers found that most Internet companies censor posts by curating sets of sensitive keywords (provided by government agencies, officials and common sense), which are fed into software that matches them to actual use. Posts that don’t trip the keyword search are approved automatically; those that do are held for review (users are sometimes informed), or simply disappear. Similar filtering can be done by other criteria.

Not all of what the study discovered is new. In 2009, guidelines for Internet monitoring and censorship that paralleled much of what’s in the Science paper leaked from Baidu, China’s top search engine. More significant, in 2013, Chinese state media reported that about 2 million people were employed in China as “Internet opinion analysts.” Though these analysts supposedly weren’t involved in censoring posts, it’s unlikely that such a large army would be recruited for the purpose of passive observation alone.

What is actually censored? The authors devised a means to view posts before they were eliminated. What they found: Posts critical of the government were no more likely to be censored than those supportive of it. Of course, this would be obvious to anyone who spends time on Chinese social media, where rants against the government are rife.

However, posts that hint at “collective action” were 20 percent to 40 percent more likely to be censored than any others. The result is a digitally connected China where communications are both hyper-modern and co-opted by the government. For a Communist Party interested primarily in the perpetuation of its own power, the situation is ideal. But it means that Chinese social media is doomed to remain compromised, suspect and, outside of China’s borders, all too irrelevant.

Adam Minter is an American writer based in Asia, where he covers politics, culture and business. He wrote this for Bloomberg News.