Contesting Cyberspace in China: Online Expression and Authoritarian Resilience

The role of intermediary actors in Chinese politics has long been recognized in the literature of Chinese politics. Scholars have critiqued the statist model that takes local cadres as the transmission belt of state policies,² and the Chinese Party-state has been depicted as “fragmented” by studies that have found increasingly pluralized policy implementation in the realms of both the political economy and contentious politics.³ In his conceptualization of “fragmented authoritarianism 2.0,” the political scientist Andrew Mertha further highlights how new policy entrepreneurs, such as marginalized officials, nongovernmental organizations, and policy activists, have started to influence the policy-making and policy-implementation process.⁴ Despite their diverse topics, these studies all emphasize the agency of intermediary actors in the exercise of state power.

Intermediary actors such as Internet service and content providers are indispensable in China’s Internet control system, as they are not only the primary information aggregators and distributors,5 but they have also been formally delegated the responsibility of censoring online expression. As scholars and human rights observers have documented, Internet service and content providers have facilitated censorship by accommodating state censorship actions. In fact, all major Internet content provider firms in China, such as Baidu, Sina, and Tencent, have a track record of complying or even collaborating with the Party-state. Even multinational IT giants have chosen “just doing business”—which means cooperating with the Party-state—over “doing just business” in China.⁶ For instance, Cisco Systems equipped China’s Golden Shield Project, which later became the backbone of the Great Firewall.⁷ Yahoo was heavily criticized for being complicit in the conviction of several Chinese dissidents by providing the state with account information.⁸ Skype has allowed its modified Chinese version, TOM-Skype, to censor users’ conversations with keywords.⁹ Similarly, Microsoft has been known to censor its blog services in China.¹⁰ Google, viewed by many as a positive example in the fight against censorship because of its withdrawal from China in 2010, has also cooperated with the regime, though perhaps unwillingly, by censoring search results before withdrawing from the country.¹¹

Indeed, work done by intermediary actors is an increasingly important component of censorship.¹² As the Chinese political scientist Yonggang Li puts it,

If the “core” of the “national firewall” is under the direct control of the state, its “periphery buffer zone” is then constructed by service providers and individual netizens. The self-censorship of these organizations and the self-discipline of netizens fence in or suppress information and opinion unapproved by the regime, thus constituting the first layer of preemption and filtering.¹³

Evidently, Li sees intermediary actors as the first “layer” of the Chinese censorship regime. However, the core-versus-periphery metaphor underemphasizes the role of intermediary actors who are pivotal to the censorship system in that, in most cases, they are the ones implementing state censorship directives, and their self-censorship efforts are much more pervasive than state censorship.¹⁴ Put simply, the state controls the Internet through intermediary actors. Much of state censorship is not directly carried out by the state apparatus, but by various actors who provide online platforms for public expression, mostly just before or shortly after “offensive” content is detected by the state.

However, depicting intermediary actors as loyal accomplices of the state is misleading, especially considering that both market incentives and the individual motivations of many intermediary actors often counsel against censorship. For instance, Rebecca MacKinnon, an advocate for Internet freedom and the director of the Ranking Digital Rights project at the New America Foundation, finds that Internet service providers enforce censorship differently owing to a number of factors such as a company’s features, a company’s owners, and the individual editors managing the portal.15 But MacKinnon does not elaborate on how these factors have affected the implementation of censorship by those service providers. Similarly, the China scholar Johan Lagerkvist argues that IT businesses only conditionally comply with the Party-state’s control policies owing to the principal–agent dilemma embedded in state regulation and social control.¹⁶ Lagerkvist argues that Chinese information communication technology entrepreneurs are pulled “in different directions by consumerism, social interest, and state control” and that the Party-state has managed to control them only through the threat of sanctions and policy rewards.¹⁷ Lagerkvist’s analysis is of theoretical significance because it highlights the principal–agent problem in censorship enforcement, particularly in terms of the emerging cracks in state–capitalist cooperation—such cooperation has been depicted in the literature as an explanatory factor of authoritarian resilience.¹⁸ However, his empirical evidence is somewhat limited because it relies primarily on one case, that of the Twitter-like microblog service Sina Weibo. Though the platform has served as a critical and popular arena for online expression since its introduction in 2009, it captures only one specific type of mechanism through which the state, intermediary actors, and netizens engage each other.

How do intermediary actors of various natures, purposes, and scales situate themselves within the censorship regime, and how do they perceive their own situation? How do market considerations and individual attitudes toward censorship affect intermediary actors’ strategies when balancing the demands of the state and their users? This chapter explores these questions by examining the role of forum administrators in censorship, how faithfully they carry out censorship, and the considerations involved and tactics used to survive and develop their boards or forums. While my analysis generally aligns with the principal–agent model,¹⁹ it further contextualizes interactions between the principal (i.e., the state) and specific agents (i.e., service providers of different scales, affiliations, and purposes).

On the other hand, from the users’ perspective, intermediary actors such as Internet forums are both platform providers and either willing or unwilling accomplices of the state. As far as censorship is concerned, the Party-state is invisible to most users most of the time; forum administrators are the primary censors directly moderating online discussions, removing sensitive threads, and punishing deviants. However, though many netizens dislike censorship and complain about it, they also acknowledge that it is the service providers who make online expression and virtual life possible. As a result, many do not want to risk an entire platform’s existence for the sake of a politically sensitive discussion. For instance, after the forced closure of the Ytht BBS, a campus forum established and run primarily by Beijing University students, some of its users blamed politically provocative users, complaining that not only were the political boards gone, but so too were apolitical ones, such as joke, picture, and sex boards, which were more popular among average netizens. Many users also feel obligated to censor themselves, especially when they are institutionally or emotionally tied to a platform, as is the case with most campus forums. For instance, the takeover of Bdwm by Peking University in early 2003 was justified by many users with the rhetoric of “true love for the BBS and the university.”20

However, for intermediary actors, factors such as market incentives and administrators’ personal motivations often run against state censorship.²¹ Leading Internet entrepreneurs, including Tencent (QQ.com, QQ, and WeChat) founder and board chair, Ma Huateng, and the former CEO of Sina (Sina.com and Sina Weibo), Wang Zhidong, for example, have openly complained about excessive censorship.²² Even when personal political orientations are put aside, these commercial service providers compete for users’ attention in order to survive and prosper. However, the risk of state repression is real, with anecdotes of websites being shut down and daring forum managers being “invited to tea” serving as control parables to warn intermediary actors not to cross the red line (the boundary that separates the permissible and taboo zones).²³ Constant state pressure makes it sensible to comply, even for those who with liberal political orientations. As the founder of Kdnet Xiao Zengjian (alias Mu Mu) has explained,

Feeling trapped, most intermediary actors demonstrate a pattern of behavior that can be described as “discontented compliance.” On one hand, they have chosen to cooperate with the state because they can hardly afford open revolts. On the other hand, since compliance is involuntary, they often engage in low-profile and, at times, more radical resistance. The analysis that follows focuses on forum managers’ discontented compliance and explains why forums vary along the discontent–compliance spectrum.

The most basic acts of compliance include following censorship decrees and obeying laws and regulations that govern Internet services and online expression; adherence is especially strong when the state takes transgressions seriously. For instance, largely owing to the anti-pornography campaign of 2009, Internet content provider registration and licensing has become a big issue for small and medium-size forums, many of which simply ignored the state’s registration and licensing requirement before the campaign.²⁶ As a result of the campaign, they had three options: register, shut down, or move abroad.²⁷ Similarly, the Ministry of Education’s 2005 campaign to turn campus forums into internal communication platforms only forced many campus forums to limit registration to their students and to restrict off-campus access. Bdwm, the official BBS of Peking University, has since frequently restricted overseas or off-campus access. Even when the restriction is lifted occasionally, the popular and politically sensitive “Triangle” board has denied access to anonymous off-campus users and limited registered off-campus users (mostly alumni) to reading only, not allowing them to post. In response to the same campaign, the administrators of Tsinghua University’s official BBS, Smth, had no choice but to move off campus to flee the jurisdiction of the Ministry of Education, becoming more commercialized and renaming the forum NewSmth. Moreover, the forum had to prohibit anonymous access to its most popular public discussion board, “NewExpress,” in 2005, subsequently introducing a scoring system that would allow only veteran users to post beginning February 2006.²⁸

Internet forums and other service providers are responsible for implementing state censorship directives. Large commercial portals, such as Sina.com, Sohu.com, and Tianya.cn, often have special divisions that connect with state agencies that communicate censorship requests to on-duty editors or board managers. Since the introduction of the Internet content provider registration and licensing system, the managers of small forums must provide their contact information to authorities so that they can respond to state requests at all hours. For campus forums, censorship orders are often routed through their home universities. For politically sensitive campus forums such as Bdwm and Smth, managers are often scheduled around the clock. To ensure a prompt response to censorship requests, university authorities sometimes demand administrative rights so that they can directly step in when managers are unavailable.

Beyond responding to censorship directives, service providers such as forums are expected to provide user data to the authorities upon request.29 User agreements often explicitly stipulate that forum administrators may be required to release user data to state agencies. For instance, Tianya’s user agreement states that the forum will release users’ account information under any of the following conditions: (1) with prior authorization from the user; (2) according to relevant laws and regulations; (3) at the request of relevant government agencies; and (4) for the public interest.³⁰ If the second condition reflects the forum’s legal obligation, the third demonstrates that it has no intention of protecting its users’ privacy against any state agencies or agents. Similarly, NewSmth claims that it will deny any requests to check users’ data except those from the government.³¹

As well as cooperating with state censorship, forums are delegated the responsibility of monitoring online expression on a routine basis to prevent “deviant” expression.³² For most forums, this means strict self-censorship to avoid trouble with the state. This goal is achieved through a series of measures that deter, detect, and punish users who violate the rules.

Most forums have enacted specific regulations reiterating state prohibitions. Some forums have also drafted board-specific user guidelines that specify taboo topics and the subsequent punishments if users fail to follow the regulations. Such regulations often disclaim responsibility for user behavior and grant forum administrators the power to monitor user activity and take action they deem proper. Moreover, regulations encourage self-censorship among users by serving as constant reminders of what can and cannot be discussed.

Many forums pre-screen postings through an automatic system that identifies and denies any posts containing taboo words.33 All large commercial forums have such a filtering system in place, though it is not clear whether they receive a keyword list from the state—existing evidence suggests they do not.³⁴ Campus forums and small individual forums have also gone to significant effort to establish their own automatic filtering systems. Indeed, at the 2010 Beijing National Campus Bulletin Board System Managers Conference, sponsored by the Chinese Communist Youth League Central Committee, campus BBS managers asked for a keyword list from the state so that they would not have to devise their own, but Committee officials shied away from the request. I have also witnessed administrators of small-scale forums exchanging keyword lists on forums and QQ groups. The very fact that service providers have to devise their own keyword lists not only confirms their indispensable role in the censorship system, but also provides a technical explanation for the variation in censorship enforcement.³⁵

Forum managers also surveil published expression closely. Large commercial sites often hire special board or channel editors to weed out posts that are potentially in violation of state censorship regulations.³⁶ On campus forums, where board managers are usually selected from among a forum’s users, popular political boards are frequently directly staffed by administrators close to supervisory bodies, such as university Communist Party committees or Chinese Communist Youth League branches. Small-scale, particularly individual-run, forums often lack sufficient technical and human resources to maintain constant surveillance on their users, and thus they sometimes simply avoid political topics altogether. Many small forums have decided to host only apolitical discussion boards to avoid potential trouble. This appears to be a rational choice because even for forums willing to take risks to attract traffic, which in turn generates revenue, political topics are much less cost-effective than other content such as soft porn, violence, and popular entertainment.

In addition to implementing state censorship initiatives and conducting self-censorship, intermediary actors sometimes seek to lower the risk of state sanctions by cultivating good relations with state agencies. Netizens believe that forums such as Kdnet and Tianya dare to exploit politically sensitive topics from time to time because of their close relationship with local propaganda departments. NewSmth provides another telling example. In early 2012, the forum featured notices and news from the local police branch on its welcome page, which were clearly posted in an effort to please the government and signal its connections with the state.37 These examples echo what Lagerkvist has found with Century China—one of the most influential websites on which intellectuals congregate—which cooperated closely with the Party-state for financial resources and political connections.³⁸

It is essential for every campus BBS to maintain good relations with its home university, not only to obtain financial and technical resources,³⁹ but also to earn some protection from state intervention, which is often routed through the university.⁴⁰ However, to gain trust, a campus BBS often must enforce strict self-censorship to assure the university that it will not cause trouble. Moreover, its managers often try to convince the university that a campus BBS will bring benefits such as enriching campus culture, facilitating dialogue with students, and enhancing the image of the institution.⁴¹

To demonstrate compliance and win trust from the state, forums sometimes even attempt to embed themselves further in the censorship regime. At both National Campus Bulletin Board System Managers Conferences I attended, a group of forum managers sought incorporation into the Communist Youth League system. They even circulated a proposal asking for specific censorship guidelines. They also promised strict self-regulation. Apparently, these forum managers were attempting to exchange autonomy for safety. Their hope, as one interviewee explained, was to “get some guidance from the state agency to avoid unintentional trespassing.”⁴²

Almost all forum administrators I met expressed discontent toward the censorship system. Some criticized state censorship because of their pro-liberal stance. For instance, a Qiangguo Luntan editor I interviewed expressed his dissatisfaction toward censorship in a subtle way. Though he refused to talk about his work, he started criticizing Peking University for failing its motto of “freedom of thought, all-embracing attitude” (jianrong bingbao 兼容并包) once he learned that I had graduated from the institution. He argued that China’s best time was between 1976 (the year of Mao’s death) and 1989 (the year of the Tiananmen Square democratic movement) when “ideas interacted with the reality naturally and with passion,” and today Chinese intellectuals lack independent personality, and thus are “either parasites or servants.”⁴⁵

Many forum managers have complained about being trapped between the state and netizens and blamed by both sides for things beyond their control: the state holds them responsible for netizens’ deviance, while netizens blame them for censoring their expression. Even those who deem regulation somewhat necessary grumble about the arbitrariness, rigidity, and ambiguity of censorship measures. “We have to be extremely careful, or someday we will lose the platform.”⁴⁶ Grumbling alone can hardly undermine the censorship regime. But it helps build solidarity among forum administrators and serves as a signal to both the state and netizens about the discontent of intermediary actors. Indeed, some forum administrators have even complained semi-openly online or to the state. For instance, at the 2009 National Campus Bulletin Board System Managers Conference in Suzhou, a forum manager complained that real-name registration was unnecessary because “truth is never afraid of debate,” and “censorship leads only to distrust and facilitates the spread of rumors.”⁴⁷

Beyond grumbling, forum administrators have also demonstrated their discontent by enforcing censorship haphazardly; for example, by allowing boundary-spanning expression and delaying the implementation of censorship directives. Daring forum managers may turn a blind eye to boundary-spanning expression when the forum is not closely watched by the state. For example, when Xu Zhiyong, a civil rights activist and law professor at the Beijing University of Posts and Telecommunications, was jailed in July 2009, the topic was prohibited on Xu’s home BBS and boards such as NewSmth’s “NewExpress,” which were supervised closely by the state. Yet, NewSmth’s “Beijing University of Posts and Telecommunications” board allowed such discussion, largely because the apolitical alumni board was not a major surveillance target. Similarly, a few discussion threads on the 1989 Tiananmen Square democratic movement, a highly censored topic, appeared on NewSmth’s “NetNovel” board in April 2016, and board managers simply ignored them for weeks.48

Lax forum administrators sometimes also delay the implementation of state censorship regulations. Many small forums and websites had ignored the Internet content provider registration and licensing policy until the state became serious about it. The policy was stipulated as early as 2000 in the State Council Information Office’s Administrative Measures on Internet Information Services, but was only gradually implemented through a series of state campaigns. Its enforcement was first emphasized in 2004 when fourteen ministries and commissions launched a joint anti-pornography campaign.⁴⁹ Then, in 2007, another anti-pornography campaign forced the Ministry of Information Industry to enforce the policy more seriously.⁵⁰ The pressure was further intensified in the 2009 anti-pornography and anti–illegal information campaign.⁵¹ Similarly, for a time, forum administrators pursued real-name registration only half-heartedly. The policy was dutifully implemented by some universities after the Ministry of Education’s 2005 mandate to transform campus forums into internal communication platforms, but a manager of an unofficial campus BBS told me that some forum administrators never took the mandate seriously. Here is his clever justification: “We ask for a valid email address to register. That should be considered real-name registration since people are supposed to register their real names when signing up for an email account.”⁵²

Forum administrators sometimes even dare to delay enforcing specific censorship directives that come with deadlines. According to a large-scale commercial forum editor, upon receiving state directives to delete postings, her colleagues often would not act immediately but instead hold off until the last second. “For every additional second [the post exists], thousands more netizens can read it,” she explained.53

Sometimes forums establish technical barriers to make state censorship less efficient or effective. As they have been delegated responsibility for routine surveillance, forums are under constant state attention to ensure their compliance: State monitoring agencies screen forum activities manually and use keyword filtering technology similar to that of search engines. Manual monitoring is hard to avoid, but forums can fool scanning software with simple technical barriers. For instance, one reason NewSmth prevents anonymous access to its “NewExpress” may be to prevent censoring software from scanning posts. Some forums recruit new users on an invitation-only basis, thus fending off state monitoring more effectively. A good example is 1984BBS (1984bbs.com). Known for its pro-liberal and anti-censorship stance, this BBS recruited new users only through invitations distributed by existing users and had more than ten thousand registered users before its forced closure in 2010.⁵⁴

Forum administrators may even promote discussion on boundary-spanning topics such as local scandals. For instance, on June 11, 2012, a netizen started a thread on Huashang Forum accusing local family planning officials in Ankang, Shanxi Province, of coercing a woman into an abortion in her seventh month of pregnancy.⁵⁵ The topic soon gained momentum online, leading to an investigation and punishment of the involved cadres. Forum administrators played an important yet under-recognized role: Instead of deleting the thread, they highlighted it nine minutes after its appearance and pinned it to the top of the forum webpage five days later (figure 3.1). Surely not all forum managers would be willing to take such risks, nor is the strategy feasible for all topics. But this case shows how forum managers can facilitate boundary-spanning protest by Chinese netizens.

Notes: “Canjue Renhuan! 7 Ge Yue de Taier bei Qiangzhi Yinchan, Haizi Shiti jiu Fangzai Mama Shenbian” (Such a Tragedy! A 7-Month Baby Aborted by Force and the Dead Body Placed Right Next to Her Mom) http://bbs.hsw.cn/read-htm-tid-3697258.html, retrieved Sept, 25, 2012. The text box above the post content shows the operations of forum administrators. One administrator highlighted the posting on June 11, 2012, and another administrator pinned the thread to the top of the forum on June 16, 2012.

In addition to tolerating or promoting boundary-spanning online activism, intermediary actors may themselves stage protests against state censorship and authoritarian rule in general. For instance, though forum administrators generally avoid open confrontation with the state, they may choose to engage in online activism when state repression jeopardizes the survival of the platform. For instance, many forum and board managers of Bdwm resigned when Peking University officially took over the forum in early 2003. Many board managers also protested by banning the president’s official account, PKU, on their boards.56 The struggle between administrators of Smth BBS (which later moved off campus and became NewSmth) and Tsinghua University authorities during the 2005 Ministry of Education campaign to constrain campus forums was no less dramatic. Many managers resigned, and hundreds of them posted protest messages or sabotaged the forum by deleting all threads on their boards or posting random content such as lists of prime numbers. More importantly, a few top administrators fought hard to “steal” user data, which were stored on university-owned servers, and they succeeded, allowing them to establish NewSmth.⁵⁷

Exit and resurrection can also be viewed as a way to counter state censorship and demonstrate discontent. Forums may take the exit option for business purposes. Many small apolitical forums exit to avoid arbitrary censorship, which can disrupt their services and destroy their hard-earned user bases, as users tend never to return if a forum goes offline, even for just a few days. Outbursts of such exodus occurred when the state tightened up licensing and registration, as well as following Google’s withdrawal. This is similar to Albert Hirschman’s conception of exit; that is, withdrawal from a relationship.58 Forums may also exit involuntarily when forced to shut down by the state. In these cases, their user data and reputation may sometimes be carried forward by former administrators and loyal users. For instance, during the 2005 Ministry of Education campaign, a few administrators of the Nanjing University BBS, Lily, the second-largest campus BBS at the time, refused to cooperate with the university, instead re-establishing itself under the name “Wild Lily.”⁵⁹ In fact, the largest overseas Chinese forum, Mitbbs, is a successor of earlier Peking University and Chinese Academy of Sciences BBS sites that were shut down by the state. Ytht, then the largest campus BBS provides an even better example. After it was shut down in 2004, at least three forums claimed to be its successors.⁶⁰ A Peking University graduate inspired by Ytht’s free spirit also attempted to re-create a new Ytht from scratch. His insistence on including “ytht” in the domain name and forum name caused him a lot of trouble, as it was considered a challenge to the regime. He was frequently “invited to tea” and forced to shut down the site during sensitive periods even though it had attracted only a few dozen users.

For forums run by state media outlets, the primary concern is political correctness, which results in strict monitoring measures. Take Qiangguo Luntan as an example. The forum attracts thousands of users, and its topics are mostly political. Postings go through software filtering and manual scrutiny by editors before being published. And, unlike most forums that run twenty-four hours a day, the forum used to close down between 10:00 PM and 10:00 AM before its upgrade in July 2012.61 When it is open, there is at least one editor on duty (zhiban, 值班) monitoring the discussions. My interaction with one of Qiangguo Luntan’s editors demonstrated that the editors are politically alert. Once I had briefed him about the purpose of my research project, he refused to talk about his job, even though his opinions on other topics betrayed a pro-liberal stance.⁶²

Though strictly monitored, Qiangguo Luntan still allows various and sometimes critical voices. There are two possible reasons. On one hand, the forum’s affiliation with the state might have provided its managers knowledge about where boundaries lie, thus reducing the uncertainty of state censorship faced by other types of forums. On the other hand, as Gary King, Jennifer Pan, and Molly Roberts have suggested, the censorship regime may aim more at curtailing collective action than silencing general criticism of state leaders or policies.⁶³ Discussion on Qiangguo Luntan is indeed more in line with abstract ideological debate than conversations that might spur collective mobilization.

Large commercial sites rely on users to generate profit. Though the stakes may seem to be higher than for smaller forums if they are forced to shut down, the risk is actually smaller because large commercial websites also have stronger bargaining power vis-à-vis the state: They represent the high-tech industry, which the state supports, and often enjoy better connections with the state.64 Moreover, the popularity of these sites also forces the Party-state to think twice when considering whether to shut them down, because doing so would affect millions of apolitical users who access the platforms for mundane purposes.⁶⁵ As a result, for the largest platforms, such as Sina (Sina.com and Weibo), Tencent (QQ.com and WeChat), and Tianya, the possibility of being shut down is hardly imaginable; fines, reprimands, and punishments for responsible personnel are more likely repercussions. For instance, according to an internal official propaganda circulation, Tencent’s QQ.com, one of the largest portals, was severely reprimanded in early 2009 because it produced several critical reports under its own brand and allowed some harsh news comments to be posted.⁶⁶ Yet, nothing more serious happened to the online portal. In short, though large commercial sites still keep away from forbidden zones, their business interests—and the pro-liberal values of their leaders and administrators—have often incentivized them to tolerate, or even encourage boundary-spanning expression. According to Yawei Liu of the Carter Center, Tencent deployed a “promotion-and-protection strategy” toward politically sensitive celebrities like himself on its microblog platform. Though he is a verified user, Tencent would temporarily remove his verification whenever he had posted something sensitive. This is an indirect way of protecting Liu because Tencent can plausibly claim to have no knowledge of his identity if the state were to attempt to trace him.⁶⁷

Small for-profit forums, many of which are run by individuals, not only lack scale or strong affiliations with institutions or big businesses to bargain with the state, but also have smaller and weaker user bases than large forums or even campus forums. As a result, they are much more vulnerable to both policy and market turbulence. To avoid risk, small forums discourage political discussion and sometimes host only apolitical boards. They also tend to impose stricter self-censorship to play it safe. For instance, when the state attempted to crack down on rumors centering on the ousted Chongqing Party secretary, Bo Xilai, in early 2012, small forums increased their self-censorship. Consenz, the company that developed the popular platform software Discuz! used by many small-scale forums, added a special search function allowing administrators to more effectively target sensitive information and monitor user activities, which managers welcomed. The following comment reveals the rationale of such small-forum managers: “It is an unusual time, and we need to take it seriously! Where there is life, there is hope!” (Liude qingshan zai, bupa meichai shao; 留得青山在, 不怕没柴烧).⁶⁸

Nonetheless, administrators of small-scale forums are not always apolitical. An online survey conducted by a forum popular among forum administrators found that the overwhelming majority of them sympathized with Google when the company chose to withdraw from China.69 Further, many small forums moved their sites abroad to avoid state intervention, which can easily ruin a fragile business. After all, small-scale forums have little bargaining power with the state and can hardly afford more radical or open actions. In fact, even when they take risks to attract an audience, they often do so with violent or sexual content rather than political topics. The state shows more tolerance toward violent and soft erotic content as long as administrators keep a low profile and clean up their sites when campaigns come along.

Unlike early BBSs run by students, most official campus forums today are supported and controlled by their home universities. The official status has a twofold implication: On one hand, official status means that the forums will receive technical, financial, and even administrative support from their universities;⁷⁰ on the other hand, official status also comes with the cost of bringing the forums under the close supervision of university Party committees, Communist Youth League branches, and student affairs offices. Moreover, students serving as administrators of the forum and sensitive discussion boards are often co-opted by university authorities with stipends, promotion opportunities as student cadres, or even scholarships.⁷¹ As a result, the degree of freedom campus forums enjoy hinges on their relationship with the university and the tacit understanding that it is best to avoid political risk, especially following the 2005 Ministry of Education campaign to restrict off-campus access. This struggle against the Ministry of Education was the last major wave of activism against censorship among most campus forums.

Unofficial campus forums often serve mixed purposes. Many such forums are established by interested students to serve their classmates, and profit is not their priority. Yet, these forums still have to raise funds to sustain themselves and expand.72 Many unofficial forums have become increasingly similar to small for-profit forums, as their founders treat their projects as a business endeavor. Without official affiliation, these forums enjoy no protection from universities. Their small user bases and limited resources also mean that they have little bargaining power vis-à-vis the state. As a result, these forums cannot afford politically sensitive discussion that may lead to repercussions. However, since their users are mostly students, unofficial campus forums sometimes allow freer and more daring expression when idealistic students dominate. This was especially the case in the earlier years of the BBS era when it was not a business model, as the cases of Ytht and NewYtht demonstrate.⁷³

Evidently, affiliation, scale, and the primary purpose of an online platform influence its will and capacity to negotiate with the state, which in turn shape its strategies. Affiliating with the state, large businesses and universities often impose constraints on forums, but also provide advantages such as protection from state censorship or market turbulence.⁷⁴ Forums attached to state media outlets are monitored most closely and do not have to appeal to users as much as commercial forums do. They are unlikely to push the boundaries of online expression. However, since they have better ideas about where the red line lies, they are also likely to tolerate certain types of political expression. Large commercial forums are subject to market incentives, rendering them naturally averse to state control that hinders them to meet the market demand for timely and interesting information.⁷⁵ Their strong user bases—particularly compared to those of small-scale forums—and links to big businesses also add to their leverage when bargaining with the state. All these factors make them more likely to tolerate boundary-spanning activities even though they are closely watched by the state. The state uses both “carrots” and “sticks” to ensure their compliance.⁷⁶ Official campus forums are often controlled by a university supervisory apparatus, but enjoy stable user bases as well as financial, technical, and administrative support from their home institutions. As a result, they can allow boundary-spanning expression as long as it does not upset their supervisors.

The size of an online platform affects its bargaining power with both the state and users in a more straightforward way than affiliations. In general, larger online platforms are better able to bargain with the state, because (1) there will be a bigger impact if the state attempts to shut them down, and (2) they have more financial and social resources that can be mobilized either to cultivate stronger ties with the state or to stage protests. For instance, the state is unlikely to completely shut down giant commercial sites such as Sina Weibo and Tianya.77 But it can be bold enough to terminate popular forums such as Ytht, which has a smaller user base compared to the commercial sites mentioned earlier but is quite political. Further, for all the criticism that followed the closure of Ytht, tens of thousands of smaller forums have died almost silently. For instance, in the 2009 anti-pornography campaign alone, more than one hundred thirty thousand small-scale websites were shut down, many of which were simply “collateral casualties.” In the eyes of the Ministry of Industry and Information Technology minister, Li Yizhong, the loss of these websites was merely a result of a “necessary over-correction.”⁷⁸

If scale and affiliation are the primary factors affecting an online platform’s capacity to bargain with the state, an online platform’s mission shapes the strategies available to administrators. Forums run by state mouthpieces care less about profit than political risk, which explains why Qiangguo Luntan used to close down between 10:00 PM and 10:00 AM. Campus forums, both official and unofficial, are often geared toward serving the students; thus, their purposes are shaped by their users as well as by university authorities. Large commercial forums are in a market that competes for user traffic and user attention; therefore, they are willing to allow limited boundary-spanning expression. Small for-profit forums that are vulnerable to state and market turbulence try hard to attract audiences while avoiding political expression. Even overseas forums, which have not yet been discussed, may choose to compromise with the state if they attempt to attract an audience in China. For instance, Mitbbs established a “cleaner” mirror site—mitbbs.cn—hoping to gain access to the domestic Chinese market. Even the pro-regime Ccthere, a United States–based forum, worked to depoliticize itself by directing political topics, particularly ones related to domestic Chinese politics, to a new site to avoid being blocked by the Great Firewall.⁷⁹

As Daniela Stockmann has argued, the Party-state’s control over traditional media is built on an institutional framework through which the state exerts “control over the organization, personnel decisions, and the editorial process.”⁸¹ By delegating censorship responsibility to intermediary actors, the Party-state has established an institutional framework that is similar to but also distinct from that governing the traditional media, one that enhances its control over the Internet. In this way, the Party-state attempts to bring Internet control back into a realm in which it has rich experience, strong capacity, and abundant resources. However, as this chapter has shown, state control over intermediary actors has been implemented inconsistently, because the intermediary actors situate themselves differently vis-à-vis the Party-state. These intermediary actors thus have opted for a variety of strategies to balance the need for state censorship with the demands of netizens.

The Party-state and intermediary actors are not the only players in the struggle over control of online expression, however. After all, netizens are the ultimate targets of state censorship, both as consumers and producers. With the stage set by the state and forum management, how do Chinese netizens pursue their virtual experience? In particular, how do they react to the censorship regime, given their pluralized interests and motivations? Chapter 4 completes the picture of the censorship game in China by focusing on popular activism by Chinese netizens.

	State-Run	Commercial/For-Profit		Campus Forum
		Large	Small	Official	Unofficial
Examples	Qiangguo Luntan	Kdnet, Sina, Sohu, Tianya	Many	Bdwm, Lily, Smth	NewSmth, Ytht
Affiliations	State media	Large businesses	None	Colleges and universities	None
Primary Purpose	Political	Profit	Profit	Non-profit	Non-profit or Profit
Risk of Shutdown	Low	Low	High	Medium	High
Bargaining Power	Strong	Strong	Weak	Strong	Weak
Options and Strategies	Avoid taboos	Boundary-spanning	Avoid politics	Boundary-spanning	Avoid politics or boundary-spanning