Shirley Hung on control of the Chinese Internet

I spent part of Wednesday at MIT’s Communications Futures Program, where the topic at hand was internet openness. Political scientist Shirley Hung opened the afternoon session with a discussion of “a different vision of openness”, looking at the Chinese internet.

Hung suggests that the authorities who control the Chinese Internet believe their internet is open, just not the way we think of it in the west. To explain how this “walled garden” works, she asks us to think through questions of freedom of speech. We think of freedom of speech as a universal right… and it is, as enshrined in the Universal Declaration of Human Rights. And freedom of speech is explicitly addressed in the Chinese constitution, in article 35. It’s in the body of the document, not in an amendment, as in the US constitution… and she suggests that it’s worth remembering that the Chinese constitution is based on the US constitution.

The apparent disparity comes from a disagreement over what freedom of speech means. A 2010 Chinese government white paper declared, “Chinese citizens fully enjoy freedom of speech on the Internet”. A few months later, a party official in charge of China’s internet censorship apparatus, Wang Chen, celebrated the fact that 350 million pieces of “harmful information”, including text, pictures and video, had been removed from the Chinese internet in the previous year. That’s 1 million pieces of content a day. And this doesn’t count content like Facebook or Twitter that’s simply blocked in China – that’s content manually removed from websites.

Why do Chinese authorities believe they have an open internet, despite this heavy, pervasive control of content? When the West – and America, in particular – talks about internet freedom, we talk about a single, open internet that works the same way everywhere. That’s a value-laden conception of the internet, one associated with political ideas about expression and participation. China values the internet in terms of economic development and advancement. When China talks about the internet, they talk about national sovereignty and cultural differences, and express the idea that internets might be different in different countries. From their perspective, the US is trying to export their view of the internet, while China is asking for each country to determine its own priorities and future.

To understand the Chinese perspective, it’s helpful to review China’s history. Hung offers us an extremely abbreviated view:
– The top priority for the current government is stability, which means preservation of Communist party rule
– The party’s legitimacy is based on two pillars – economic development and territorial integrity
– China has a history of social unrest leading to revolutions. Almost every shift between dynasties has come through social unrest.

It’s worth remembering that this is the first regime to have solved the warmth and hunger problems – people aren’t generally starving, Hung reminds us, and they have money to buy televisions and mobile phones. This success in meeting people’s basic needs may give the government more latitude to control the internet.

The Chinese approach to control apparently rests on three principles:
– Push responsibility and implementation of control downwards, through the network
– Create multiple levers to enable fine-grained control, so you don’t need to shut off the internet, ala Egypt
– Rely on the panopticon and on deterrence to force users to self regulate.

She shows us an organizational chart of the various bureaucracies that control the Chinese internet. Her chart – an oversimplification she suggests – includes more than 20 entities, divided into three general groups. One set of groups are party organs. Another are government agencies. And a third set are “quangos”, quasi-NGOs, which are funded and approved of by the government but are not technically part of the government. This complex system exists at national and local levels – much of the control over the Chinese internet is delegated to the Beijing government because companies like Baidu are based in the capital city.

A first line of regulation is red tape – websites require licenses, registrations and permits. “You need a stamp and a seal to do almost anything.” These regulations look silly and inefficient, but they’ve got a purpose – they allow for multiple, different checks against certain behaviors. If you don’t like what a site is doing, you can deny its owners a stamp or seal and shut it down.

Almost all Chinese sites feature a button that can summon the Cyber110 police. That’s the Beijing police reporting center – click the button and you get a screen showing animated policemen, each of whom are willing to take your report of various different online behaviors where a web publisher might be crossing a redline. Those reports can cause publishers to lose points – each publisher has 100 points when they begin, and they can gain more by publishing pro-government stories and lose points for failing to remove content in a timely fashion. Sites that publish content are required to run third-party monitoring systems, separate bureaus that report to the editor in chief and which monitor whether the content the site publishes is appropriate.

Control is delegated down to these quasi-NGOs, and from them to individual citizens. The Beijing Association of Online Media recruits a team of citizens who monitor the internet, each of whom is required to report 50 pieces of “harmful information” each month. The 50 cent party (it’s more like 7 US cents when converted from RMB to US dollars) compensates individuals for posting pro-government information on bulletin board systems and fora. In this sense, the structure of monitoring reflects older structures on societies. Since the emperors, neighbors have been spying on neighbors, and this pattern continues in a digital age.

Hung explains that she’s stronger on policy than on specifics of technology, but offers a brief outline of China’s technological arms race. The “great firewall” – known locally as the “golden shield” – uses IP blocking, port blocking, keyword and URL filtering, packet filtering and other techniques to block content. When users access banned content, they often experience TCP resets, and sometimes longer bans from the internet. Recently, there’s evidence that commercial, subscription VPNs are being blocked. China has also signalled a willingness to filter on the client side – the failed Greendam project sought to put filtering on individual PCs, a further push towards decentralization. And we’re now seeing a rise of sophisticated attacks, like malware targeted at dissidents and independent media organizations.

She sees this as an indication of a growing vertical integration of control. China has a great deal of influence over equipment providers like Huwawei. State ownership of fiber allows another level of control, as does influence over and ownership of telephone companies. By blocking access to non-Chinese Web2.0 companies, the government has opened a market for domestic companies that compete in the social media space. Some of these systems are pretty amazing, like QQ, which has 500 million accounts, a pretty impressive metric in a country with 400 million internet users. When Chinese companies can’t build their own, they partner – we may see a Facebook/Baidu partnership in the near future. And control extends to devices, through registration of handsets and SIM cards, and the need to use a national ID to log in at a cybercafe.

The control isn’t just technical – it’s often focused on content. Hung tells us that there’s a set of approved media sources one should quote from, suggesting that you can’t write or report your own news. And given the presence of monitors within these platforms, it’s increasingly unlikely that you’d see original news reporting.

Questions for Hung focused on whether this overview of regulations accurately reflects the reality of the Chinese internet. If these regulations are all enforced, why is cinema and software piracy so rampant within the country? Clearly, there’s some disparity between the legal mechanisms that enable control and the actual practice.


While I thought this was a great overview – one of the very best I’ve heard – of the systems currently deployed to control conversations on the Chinese internet, I worry that looking at these systems may blind us to the richness of the content that gets created in China. My colleague Dong Hao offered a great discussion of some of China’s top sites, and the ways in which they often creative, participatory behaviors not seen on the English-language internet. I think many US observers of the Chinese internet hear the complexities of censorship and assume everyone posting on the Chinese internet is reciting party propaganda. It’s a shame we don’t have a site like Yeeyan.org helping translate large swaths of the Chinese internet into English. But even a quick visit to sites like EastSouthWestNorth or chinaSMACK should give a sense of the richness of content that’s available.

This isn’t to say that Chinese censorship isn’t onerous, or doesn’t profoundly shape online conversation. But it would be a mistake to limit an understanding of the Chinese internet to what isn’t permitted at the expense of what is.

This entry was posted in Human Rights, Media. Bookmark the permalink.

8 Responses to Shirley Hung on control of the Chinese Internet

  1. bill says:

    you might fin this essay, originally published in the China Economic Quarterly, interesting too:
    China’s Internet: The Invisible Birdcage | Sinocism http://bit.ly/gfr5wK

  2. ltlee says:

    “Almost all Chinese sites feature a button that can summon the Cyber110 police. That?s the Beijing police reporting center ? click the button and you get a screen showing animated policemen, each of whom are willing to take your report of various different online behaviors where a web publisher might be crossing a redline. Those reports can cause publishers to lose points ? each publisher has 100 points when they begin, and they can gain more by publishing pro-government stories and lose points for failing to remove content in a timely fashion. Sites that publish content are required to run third-party monitoring systems, separate bureaus that report to the editor in chief and which monitor whether the content the site publishes is appropriate.”

    Please show an example or two to validate your claim. Please tell where can one find evidence on the “point system.”

  3. - says:

    I feel unsatisfied.
    The reading was really great, which made me expect a lot, but missed far too important questions, especially what makes them feel they are open?
    I felt I was about half of it when suddenly it ended.

  4. kailing says:

    Haha, and the Dong Hao link is… out of reach inside China w/o a VPN…

  5. kailing says:

    Well, feelings are just that, feelings. They feel they are open enough, and as openness is relative (no country’s internet is 100% open [pedofilia, bestialism, nazism, terrorist talk, bullying, gender abuse…]), they are as everyone, that is relatively open, so being “relatively” a common denominator [and indeed quite a “relative” term] can be eliminated from the equation.

    It is a bit of sophism argumentation but this is their logic… and it serves their purpose: to show that no one is absolutely open, be it due to moral or political pre-conceptions. They are open enough (as stability is the utmost value to preserve), and as everyone they are relatively open.

  6. Pingback: (Asien)-Linktipps | azia.fm

  7. Pingback: Too much information: week ending 13 May | The Barefoot Technologist

  8. Pingback: Recensione: Felici e sfruttati

Comments are closed.