On September 20, 1987, a Chinese computer expert sent China’s first email from Beijing with the message, “Across the Great Wall we can reach every corner in the world.”1 This email hailed the arrival of China’s Internet era with its statement that territorial obstacles between China and the world, as symbolized by the Great Wall, could be overcome by this new technology: the Internet. This message seems somewhat ironic in retrospect, because the Chinese state has since managed to construct the world’s most sophisticated virtual “Great Wall” to filter content exchange between the country and the outside world, a system nicknamed the “Great Firewall.” Why and how has the Party-state aimed to establish and enhance its control over the Internet? How has state control evolved over time? What are the dynamics shaping its evolution, particularly with regard to the state’s institutional adaptation?
This chapter takes a state-centric perspective to explore China’s Internet content control regime, defined as the complete set of organizational, institutional, administrative, and technical tools used by the state to limit the boundaries of online expression. The analysis shows that the Chinese state has developed a multi-agency, multi-level, multi-faceted approach to systematically and comprehensively censoring online content. However, this censorship system is plagued by both the Internet’s hard-to-control nature and the regime’s internal fragmentation, which together have rendered state censorship in China overly rigid but also arbitrary, and thus ineffective and even counterproductive to the state’s goals in some cases.
Admittedly, the Chinese state has attempted to shape online expression in more subtle and complicated ways than censorship, often using a combination of control, propaganda, and accommodation approaches.2 For instance, the state sometimes accommodates citizen participation online through channels such as e-government platforms or official accounts on popular social media platforms. To more directly shape online opinion, the state has adopted more innovative propaganda tactics, such as engaging the public via social media,3 embracing popular cyber culture,4 and fabricating pro-regime voices. These manipulative tactics are qualitatively different from censorship, as they ultimately do not depend on the state’s coercive capacity. Chapter 5 will explore how the state manipulates rather than censors online expression, with a special focus on the Internet commentator system.
INTERNET GOVERNANCE AND INTERNET CONTENT GOVERNANCE
Internet governance in China is driven by a complicated set of dynamics. In particular, the Chinese state faces a dictator’s dilemma. On one hand, it is highly concerned with the economic, technological, and industrial potential of the Internet and related information and communication technologies.5 On the other hand, it has to deal with the sociopolitical challenges brought about by these new technologies, particularly the freer flow of information. As the political scientist Yongnian Zheng argues, these contradictory tasks have pushed the Party-state to establish a regulatory regime that promotes the development of the Internet as an industry while at the same time attempting to establish a control regime to tame what it considers to be the Internet’s disturbing implications.6 Similarly, Min Jiang argues that China’s Internet policies “reflect an Internet development and regulatory model—authoritarian informationalism—that combines elements of capitalism, authoritarianism, and Confucianism.”7
As the Internet empowers social actors by promoting civil society, facilitating collective mobilization, and encouraging public expression and online activism,8 authoritarian states are increasingly recognizing a necessity to control it. According to Lawrence Lessig, controlling the Internet involves four mechanisms: the law, technical architecture (i.e., code), social norms, and the market.9 In China, all of these mechanisms are subject to the heavy influence of or direct intervention by the Party-state. Indeed, studies have found that the Chinese Party-state not only firmly controls the physical network infrastructure, but also makes serious efforts to censor the flow of online information.10 Scholars and human rights advocates have documented many censorship techniques employed by the state, ranging from automatic taboo word filtering to manual surveillance, and from limiting access to undesired websites using the Great Firewall to shutting down deviant websites and jailing dissenters.11 The state even completely shut down Internet and mobile services in Xinjiang after ethnic rioting in 2009.12 Moreover, the state has also attempted to take preemptive measures, such as promoting real-name registration, requiring the pre-installation of filtering software on all personal computers,13 and encouraging service providers to conduct self-censorship.14
These studies reveal how the state controls or attempts to control the Internet, but are insufficient in providing a historical and systematic understanding of Internet governance beyond specific control measures. In addition, they often treat the control regime as a single, undifferentiated entity, which makes it difficult to discern variations in state control over different targets or to recognize the dynamics involved in the state’s strategies and techniques. The following sections historicize and contextualize the evolution of China’s Internet control, especially the state’s institutional, organizational, technical, and administrative adaptations to the ever-increasing challenges of online expression over time.
The Evolution of China’s Internet Control Regime
Since the early 1990s, China’s Internet control regime has undergone a process of enhancement, expansion, and fine-tuning, with different focuses, characteristics, and implications involved at different stages. Taking a historical approach, Yongnian Zheng has argued that the Chinese state focused on constructing a regulatory regime in the late 1990s, before shifting emphasis to a control regime.15 Guobin Yang has divided regulatory evolution into three stages, with the first (from 1994 to 1999) focusing on network security, service provision, and institutional restructuring. The second stage (from 2000 to 2002) was characterized by the expansion and refinement of control with stronger content regulation, and the third stage (from 2003 onward) has centered on the further expansion of Internet regulation.16 Yonggang Li has also argued that Internet control has evolved through three periods. Adopting a policy-learning perspective, Li’s analysis provides a better understanding of the development of China’s Internet control regime, as it correctly emphasizes the agency of the state.17
Like Zheng’s, Yang’s, and Li’s work, my study also considers the state’s Internet control regime to have evolved through three phases (table 2.1). I examine the escalation and expansion of state control over time, as well as its shifting emphases at different stages: from regulation to control, and from network security to content control. But my periodization also highlights that the evolution of state control has in fact corresponded to the expansion of the Internet and its ever-increasing sociopolitical impact. After all, as Yongnian Zheng aptly puts it, “The state and social forces are mutually transformative via their interactions.”18
TABLE 2.1 The Evolution of Internet Content Control by the State
| |
Phase 1: Before 1999 |
Phase 2: 1999–2003 |
Phase 3: 2004–Present |
| Growth of Internet Use |
The Internet, with people having limited access, is viewed more as an economic and technological opportunity than a political challenge. |
Online expression becomes politically challenging; for example, the case of Sun Zhigang.a State adaptation lags behind the expansion of the Internet and online expression. |
The Internet becomes increasingly powerful and has more constraints imposed on it by the state. Online expression continues to boom, with new applications such as blogs, WeChat, and Weibo. |
| Features of State Control |
The focus is on network security and promoting the technology.b Control is unsystematic and reactive, often via technical and administrative means. |
Rapid institutional and organizational adaptation occurs, with a more specific and heavier focus on content control, but overall, state control remains in a trial-and-error process. There are frequent changes in control agencies and conflicts over jurisdiction.c |
Policy learning has come to a tentative conclusion. Control continues to expand and adjust, becoming increasingly assertive and adept, with bolder administrative and technical practices in use than ever before. |
| Policy Enactment (Table 2.2) |
Regulations focus on physical network security. Prescriptions regarding content control are often ambiguous and boilerplate. |
Content control regulations become more specific, and a systematic framework emerges for licensing, registration, and the monitoring and punishment of online activities. |
Regulations are enacted or redrafted to regulate new online applications, fine-tune content control, clarify division of labor, and promote coordination among state agencies. |
| Organizational Adaptation (Table 2.3) |
The state tries to accommodate Internet governance with its existing apparatus. |
The jurisdiction of existing state agencies expands, with specialized content control agencies established and the promotion of self-disciplining organizations. |
There is further organizational restructuring, an enhancement of division of labor and coordination, and the promotion of self-disciplining service providers. |
| Administrative and Technical Control (Table 2.4) |
Administrative and technical control are sporadic, but constant attention is paid to campus forums. |
Content control practices are intensified to promote official discourses and suppress unwanted information. |
Efforts to enforce control are made more rigid and bolder, notably via campaigns with multi-agency cooperation and coordination. |
a See Zixue Tai, The Internet in China: Cyberspace and Civil Society (London: Routledge, 2006), 259–68.
b Yongnian Zheng, Technological Empowerment: The Internet, State, and Society in China (Stanford, CA: Stanford University Press, 2008), 50–23; Guobin Yang, The Power of the Internet in China: Citizen Activism Online (New York: Columbia University Press, 2009), 48.
c Yonggang Li, Women de Fanghuoqiang: Wangluo Shidai de Biaoda Yu Jianguan (Our Great Firewall: Expression and Governance in the Era of the Internet) (Nanning: Guangxi Normal University Press, 2009), 75.
To highlight both the process and the nuances of state adaptation, the chapter groups the state’s Internet control efforts into three major categories: policy enactment (table 2.2), organizational adaptation (table 2.3), and administrative and technical enforcement (table 2.4). Consciously or unconsciously, the state has employed an uneven development strategy in building its control capacities. Its initial responses to online expression came out of its governance inertia and were mostly administrative and technical in nature, thus appearing to be sporadic and unsystematic. Institutional adaptation followed, with the creation of formal rules and norms, as well as organizational capacity building. At this stage, the state did not strictly enforce control regulations. The third stage has been one of capacity use, adjustment, and refinement. During this stage, the state has been trying out its control mechanisms more boldly and adjusting in response to social reaction.
TABLE 2.2 Policy Enactment
| Phase |
Major Regulations |
| Before 1999 |
■ In 1994, the State Council Information Office issues the Regulations on the Safety and Protection of Computer Systems of the People’s Republic of China, prohibiting the use of “computer information systems to conduct activities against national interests, public interests, or legitimate interests of citizens.”a |
| |
■ In 1997, the Ministry of Public Security releases the Computer Information Network and Internet Security Protection Management Regulations, prohibiting nine types of information.b |
| 1999–2003 |
■ In 2000, the State Council Information Office publishes the Telecommunications Regulations of the People’s Republic of China and the Administrative Measures on Internet Information Services, which both reiterate the nine prohibitions of the 1997 Ministry of Public Security regulations. The Administrative Measures on Internet Information Services also require the registration and licensing of Internet content providers.c |
| |
■ In 2000, the Ministry of Information Industry issues the Regulation on Internet News and Bulletin Boards, requiring the registration of forums.d |
| |
■ In 2000, the State Council Information Office and the Ministry of Information Industry issue the Interim Provisions for the Administration of News Publication by Internet Sites, prescribing qualifications for online news service providers.e |
| 2004–present |
■ In 2005, the State Council Information Office and the Ministry of Information Industry jointly issue the Administrative Provisions of Internet News Information Services, adding two more types of prohibited information: information that incites illegal gathering, association, demonstration, or disruption of social order, and information on the activities of illegal nongovernment organizations.f |
| |
■ In 2005, the Ministry of Information Industry issues the Administrative Measures of the Registration of Noncommercial Internet Information Services to enhance control over noncommercial content providers.g |
| |
■ In 2006, sixteen central party and state departments jointly issue the Work Program for the Coordination of Internet Website Management, dividing control responsibilities across departments.h |
| |
■ In 2007, the State Administration of Radio, Film, and Television and the Ministry of Information Industry jointly issue the Administrative Provisions on Internet Audiovisual Program Services, extending control to multimedia content.i |
| |
■ In 2012, the Cyberspace Administration of China and the Ministry of Industry and Information Technology jointly issue a revised draft of the Administrative Measures on Internet Information Services for public comment. These regulations explicitly tie online information control to national security, fine-tune the licensing and registration requirements of service providers, promote real-name registration, and further describe the responsibilities of service providers.j |
a State Council Information Office, Zhonghua Renmin Gongheguo Jisuanji Xinxi Xitong Anquan Baohu Tiaoli (Regulations on the Safety and Protection of Computer Systems of the People’s Republic of China), February 18, 1994.
b These nine types of information include information that (1) incites people to resist or obstructs the implementation of the constitution, laws, or administrative regulations; (2) incites people to subvert the government or the socialist system; (3) incites separatism or harms national unification; (4) incites ethnic hatred or discrimination or undermines ethnic solidarity; (5) forges or distorts facts or spreads rumors that disturb social stability; (6) spreads superstition, obscenity, pornography, gambling, violence, homicide, terrorism, or instigates crime; (7) openly insults or slanders people with fabricated information; (8) damages the credibility of state apparatus; and (9) any other type of information that violates the constitution, laws, or administrative regulations.
c State Council Information Office, Zhonghua Renmin Gongheguo Dianxin Tiaoli (Telecommunications Regulations of the People’s Republic of China), September 20, 2000; State Council Information Office, Hulianwang Xinxi Fuwu Guanli Banfa (Administrative Measures on Internet Information Services), September 25, 2000.
d Ministry of Information Industry, Hulianwang Dianzi Gonggao Fuwu Guanli Guiding (Regulation on Internet News and Bulletin Boards), October 27, 2000.
e State Council Information Office and Ministry of Information Industry, Hulian Wangzhan Congshi Dengzai Xinwen Yewu Guanli Zanxing Guiding (Interim Provisions for the Administration of News Publication by Internet Sites), November 6, 2000.
f State Council Information Office and Ministry of Information Industry, Hulianwang Xinwen Xinxi Fuwu Guanli Guiding (Administrative Provisions of Internet News Information Services), September 25, 2005.
g Ministry of Information Industry, Fei Jingyingxing Hulianwang Xinxi Fuwu Beian Guanli Banfa (Administrative Measures of the Registration of Noncommercial Internet Information Services), January 28, 2005.
h Chinese Communist Party Central Propaganda Department, Ministry of Information Industry, State Council Information Office, Ministry of Education, Ministry of Culture, Ministry of Health, et al., Hulian Wangzhan Guanli Xietiao Gongzuo Fangan (Work Program for the Coordination of Internet Website Management), February 17, 2006.
i State Administration of Radio, Film, and Television and Ministry of Information Industry, Hulianwang Shiting Jiemu Fuwu Guanli Guiding (Administrative Provisions on Internet Audiovisual Program Services), December 29, 2007.
j Cyberspace Administration of China and Ministry of Industry and Information Technology, Hulianwang Xinxi Fuwu Guanli Banfa: Xiuding Caoan Zhengqiu Yijian Gao (Administrative Measures on Internet Information Services: Revised Version for Public Comment), June 7, 2012.
TABLE 2.3 Organizational Adaptation
| Phase |
Evolution of the State’s Internet Control Apparatus |
| Before 1999 |
■ The Ministry of Information Industry is established to better regulate and promote the development of information technologies. |
| |
■ The Ministry of Public Security is delegated the power of content control. |
| 1999–2003 |
■ Extended power is granted to the State Administration of Radio, Film, and Television over online video and audio programs, to the General Administration of Press and Publication over online publications, and to the Ministry of Culture over online cultural and artistic activities, online gaming, and Internet cafés.a |
| |
■ In April 2000, the State Council Information Office establishes the Internet News and Propaganda Administration Bureau. Local authorities quickly follow suit and establish related offices.b |
| |
■ Self-disciplining organizations, such as the Internet Society of China, the China Youth Internet Association, and the Wireless Internet Trust and Self-Discipline Alliance are established. |
| 2004–present |
■ In 2006, sixteen central party and state departments jointly issue the Coordination Plan of Internet Website Management, dividing control responsibilities across departments.c Local authorities also establish a joint-leadership mechanism to coordinate Internet control.d |
| |
■ In 2008, the Ministry of Industry and Information Technology is established, inheriting Internet regulatory and control powers from the Ministry of Information Industry and the State Council Informatization Work Office.e |
| |
■ In April 2010, the State Council Information Office establishes a Ninth Bureau to guide, coordinate, and supervise Internet cultural development and management.f |
| |
■ The state continues to promote self-discipline among service providers through efforts such as the establishment of an illegal information reporting center.g |
| |
■ In 2011, the State Council Information Office establishes the Cyberspace Administration of China as a parallel apparatus to the State Council to enhance control over online content.h |
| |
■ In 2014, the Cyberspace Administration of China is reorganized into the top Internet censorship, oversight, and control agency, which answers directly to Xi Jinping and the Central Leading Group for Internet Security and Informatization. |
| |
■ In November 2016, the National People’s Congress passes the Cybersecurity Law. As the latest effort to tighten jurisdictional control over online content, the law reaffirms the state’s sovereignty over the Internet. In particular, the law requires network operators to cooperate with the authorities in crime or security investigations, to store within China data gathered or produced in the country, and to verify users’ identities before providing services to them. |
a State Administration of Radio, Film, and Television, “Guanyu Jiaqiang Tongguo Xinxi Wangluo Xiang Gongzhong Chuanbo Guangbo Dianying Dianshi Jiemu Guanli de Tonggao” (“Circular on Strengthening the Management of Broadcasting Radio, Film and TV Programs via the Internet”), October 1999; State Administration of Radio, Film, and Television, Hulianwang Deng Xinxi Chuanbo Shiting Jiemu Guanli Banfa (Regulations on Broadcasting Video and Audio Programs Through the Internet), January 7, 2003; General Administration of Press and Publication and Ministry of Information Industry, Hulianwang Chuban Guanli Zanxing Guiding (Provisional Regulations on the Administration of Internet Publications), June 27, 2002; Ministry of Culture, Hulianwang Wenhua Guanli Zanxing Guiding (Provisional Regulations on Internet Culture Management), May 10, 2003.
b In June 2000, the Beijing Municipal Government Information Office established its Internet Propaganda Administration Office to supervise online information services. See the Beijing Association of Online Media website, www.baom.org.cn/biannian/bn2000.html.
c See Chinese Communist Party Central Propaganda Department, et al., Hulian Wangzhan Guanli Xietiao Gongzuo Fangan. The sixteen agencies include the Central Propaganda Department; the China Food and Drug Administration; the Chinese Academy of Sciences; the General Staff Department Communication Department; the General Administration of Press and Publication; the Ministry of Commerce; the Ministry of Culture; the Ministry of Education; the Ministry of Health; the Ministry of Information Industry; the Ministry of Public Security; the Ministry of State Security; the State Administration for Industry and Commerce; the State Administration of Radio, Film, and Television; the State Council Information Office; and the State Secrets Bureau.
d The Beijing Municipal Government formed the Internet Propaganda Administration Leadership Group. See Wang Hao, “Beijingshi Hulianwang Xuanchuan Guanli Lingdao Xiaozu Huiyi Zhaokai” (“Beijing Municipal Internet Propaganda Administration Leadership Group Conference Convened”), Beijing Ribao (Beijing Daily), July 13, 2007.
e Though primarily a regulatory agency, the Ministry of Industry and Information Technology plays an indispensable role in censorship because all Internet data centres (IDCs), Internet service providers (ISPs), and Internet content providers (ICPs) must register with it before entering the market.
f Su Yongtong, “Guoxinban ‘Kuobian,’ Wangluo Guanli Jusi Yi Bian Er” (“State Council Information Office Expansion: One Internet Administration Bureau Becomes Two”), Nanfang Zhoumo (Southern Weekend), May 20, 2010.
TABLE 2.4 Technical and Administrative Control
| Phase |
Selected Technical and Administrative Control Efforts |
| Before 1999 |
■ Overseas websites considered hostile undergo an ad hoc blockade, and violators are jailed.a |
| |
■ Service providers are encouraged to conduct self-censorship.b |
| |
■ The state is highly attentive to campus forums: All major campus forums are carefully monitored and strictly controlled, with sites shut down owing to worries about collective mobilization or during sensitive periods, such as the event of Deng Xiaoping’s death. |
| 1999–2003 |
■ State media are encouraged to conquer “commanding heights of thought and public opinion” online.c |
| |
■ The Great Firewall starts to function, monitoring information flow online and blocking “hostile” sites through techniques such as IP and URL filtering.d |
| |
■ There is a crackdown on unregistered Internet cafés; those that remain open are required to record their customers’ identification information.e |
| |
■ More severe measures are taken to punish deviants, including shutting down websites and jailing more citizens for “subverting the state” or “leaking state secrets.”f |
| 2004–present |
■ Enforced registration and licensing are taken more seriously with the establishment of a database of Internet content providers, IP addresses, and domain names.g |
| |
■ Real-name registration is required for terminal devices (e.g., mobile phones and computers used in Internet cafés) and applications (e.g., blogs, microblogs, and forums).h |
| |
■ Control is tightened over campus forums: The Ytht BBS is shut down in 2004, and the Ministry of Education mandates campus forums to restrict off-campus access in the spring of 2005. |
| |
■ Internet commentators, a.k.a. the fifty-cent army, are introduced to guide public opinion anonymously (see chapter 5). |
| |
■ More efforts are made and state campaigns run to control pornographic, political, and other illegal information. For instance, the Ministry of Industry and Information Technology introduces the filtering Green Dam software (which is revoked following criticism in 2009).i In 2013, the state launches a massive anti-rumor-mongering campaign primarily targeting Weibo and other social media platforms.j |
a See Seth Faison, “E-Mail to U.S. Lands Chinese Internet Entrepreneur in Jail,” New York Times, January 21, 1999.
b Jack Linchuan Qiu, “Virtual Censorship in China: Keeping the Gate Between the Cyberspaces,” International Journal of Communications Law and Policy, no. 4 (2000): 12.
c CCTV Online, China.com.cn, China Daily Online, China-Youth Online, CRI Online, People’s Daily Online, and Xinhua Net are among the first designated as key Internet news service providers by the State Council Information Office.
e Li, Women de Fanghuoqiang, 90–93. As Li points out, such campaigns have regulatory purposes.
f For a detailed list, see Zheng, Technological Empowerment, 70–78.
g Enacted as early as 2000, the regulation was not seriously enforced until 2004, when fourteen ministries jointly launched a campaign to enforce it. The pressure intensified in the later anti-pornography and illegal information campaigns.
h See Jiao Likun, “Shouji Shimingzhi Zuiwan Xiayue Tuichu, Zhongguo Yi Fengsha Wanbu Shouji” (“Real-Name Registration of Cellphones Will Be Implemented Next Month Onward; China Has Banned More Than Ten Thousand Cellphones), China News, December 22, 2005, www.chinanews.com/news/2005/2005-12-22/8/668715.shtml; Ying Ni, “ ‘Wangyou Shimingzhi’ Jin Qi Shixing, Xinzheng Zaoyu Zhixingnan” (“Real-Name Registration of Online Gaming Effective Today; the New Policy Encounters Difficulty in Enforcement”), China News, August 1, 2010, www.chinanews.com/it/2010/08-01/2438659.shtml; “Tencent QQ Jiang Shixing Wangluo Shimingzhi, QQ Qun Chuangjianzhe Xu Shiming Dengji” (“Tecent QQ Intends to Introduce Real-Name Registration; QQ Group Owners Must Register with Real Names”), Sohu, July 21, 2005, http://it.sohu.com/20050721/n240175776.shtml; “Boke Shimingzhi Anran Tuichang, Wangluo Guanzhi Yiyou Fansi” (“Real-Name Registration of Blog Services Abandoned; We Need Reflections on Internet Control), Nanfang Dushibao (Southern Metropolis Daily), May 25, 2007.
i Wang Qihua, “ ‘Lüba’ Huangong” (“Green Dam Implementation Postponed), Caijing Magazine, no. 14 (July 6, 2009).
j For instance, see Chris Buckley, “Crackdown on Bloggers Is Mounted by China,” New York Times, September 11, 2013.
UNDERSTANDING INTERNET CONTROL IN CHINA
According to the Internet researcher Yonggang Li, state governance of the Internet in China has proceeded through the “garbage can model,” which features problematic preferences, unclear technology, and fluid participation19 in a “categorized governance” model that applies different governing strategies in different realms.20 By now, the state has established a comprehensive content control regime endowed with a full set of organizational, institutional, technical, and administrative tools.
Organizationally, before the Cyberspace Administration of China (also known as the Office of the Central Leading Group for Cyberspace Affairs) was founded in 2011, the two most important sets of state apparatus responsible for Internet control were the party propaganda system and the government’s information office system, headed by the Central Propaganda Department and the State Council Information Office, respectively.21 The propaganda system, nicknamed the “Department of Truth,”22 is in charge of ideological work and has ultimate control over the media. The State Council Information Office is in charge of the development and control of online news services and is the de facto top government agency for Internet control. Both the Central Propaganda Department and the State Council Information Office have specific sub-offices for Internet control. The Internet Division of the Bureau of Information and Public Opinion, the Internet News Office under the News Bureau, and the Internet Commentary and Criticism Group are sub-offices of the Central Propaganda Department, and the Fifth and Ninth Bureaus fall under the State Council Information Office.23 In May 2011, the State Council set up the Cyberspace Administration of China as a parallel agency to take on its Internet regulation and control functions.24 In 2014, the Cyberspace Administration of China became the central Internet censorship, oversight, and control agency, reporting directly to President Xi Jinping and the then newly formed Central Leading Group for Internet Security and Informatization. With the state’s constant organizational adaptation, a redivision of labor and redistribution of power have occurred from time to time among state agencies. For example, since the 1990s, the propaganda system has gradually increased its influence on Internet censorship.25 The role of the Ministry of Public Security in content control has become relatively less significant as the propaganda and state information systems have gradually built up their organizational capacities.26
Institutionally, the two most important regulations that serve as the major legal basis for online content censorship are the Administrative Measures on Internet Information Services, issued by the State Council in 2000, and the Administrative Provisions of Internet News Information Services, issued by the State Council and the Ministry of Information Industry in 2005. The former set out conditions for websites to operate legally, including registration and licensing. The latter established the online news publication qualification system, dividing online news service providers into three categories: those run by news entities, those run by non-news entities, and those established by news entities to carry already-published content. Only websites in the first category are allowed to report on news events. Those in the second and third categories may only reprint news from news sources approved by the state.
For both netizens and service providers, censorship rules become tangible only when they are enforced, because state regulations and laws are often ambiguous in terms of implementation. For instance, individuals were not allowed to register “.cn” domain names according to both the 2002 and the 2009 versions of the China Internet Network Information Center Implementation Rules for Domain Name Registration. But the China Internet Network Information Center did not enforce the regulation until 2009, when China Central Television criticized its loose supervision of domain names in an anti-pornography campaign.27 Apparently, state regulation in this case served more as a disclaimer to avoid imputation than actually as a means of controlling online content.
According to the former State Council director Wang Chen, state control can be implemented through various means, including the following:
1. Regulating domain names, IP addresses, registration and licensing, and service access.
2. Establishing an entry-and-exit mechanism for online information services; i.e., lawfully implementing registration and licensing of information services related to ideological security and public interests; establishing and improving management such as routine supervision, annual review, and administrative penalty; and forming a coordination mechanism among relevant agencies to dispose harmful information and prevent its infiltration from abroad.
3. Actively exploring real-name registration.28
Such control measures are implemented at multiple levels and target individual users and all levels of service providers. According to their functions, these measures can be categorized into four groups serving preventive, surveillance, crisis management, and public opinion manipulation goals. Preventive measures are designed to filter the flow of online information and prevent unwanted information from being accessed or published. Filtering can be done at various levels. At the national level, the Great Firewall employs a collection of filtering technologies to disrupt connections that the state deems harmful. Local authorities often then add taboo terms to filter content specific to their areas. Most content providers, such as blogs, Internet forums, and microblogs, also automatically and/or manually screen posts before letting them through.29 Measures such as the registration and licensing of Internet service providers and the real-name registration of users are also preventive measures in that they enable the state to easily identify violators and promote self-censorship. Similarly, in some cases, official takeovers of campus forums by universities could be considered a preemptive action to avoid trouble for university authorities.
Online surveillance is carried out through automatic and manual screening of the web to check for violators who have managed to elude preventive measures. Most observers suspect that the state uses search engine technologies to oversee online expression. Moreover, Internet police forces, the propaganda system, and many other control agencies employ inspectors to manually watch the web. Service providers and universities also keep close watch on online expression and are ready to remove any “inappropriate” content from sites under their supervision upon state notification or to preempt state intervention. State surveillance is also supplemented by the illegal information reporting system, which encourages peer monitoring.
Online crisis management has become routine for both central and local authorities.30 Online crises are defined broadly here and include the discussion of any topics considered taboo, official scandals, mobilization for online or offline collective action, and any viral cyber-events that the state deems to be going out of control. To defuse crises, monitoring agencies send out directives via phone, email, or instant messenger to service providers, instructing them to take specific actions, such as deleting messages within a time limit, banning users, warning or fining violating websites, and shutting down parts of or entire websites either temporarily or permanently. Individual violators may be “invited to tea” (bei hecha, 被喝茶, which refers to the situations where netizens are informally detained and questioned by authorities as if it is a causal meeting for a cup of tea) or even jailed. Sometimes, massive campaigns are launched as part of the state’s crisis management effort. For instance, after the Chongqing party chief, Bo Xilai, was removed from his post in early 2012, the Party-state shut down forty-two websites and detained six citizens for “fabricating and disseminating rumors” about Bo’s allies in Beijing staging a coup.31 Moreover, the two Chinese IT giants Sina and Tencent were punished by the state and forced to suspend the comment function of their microblog services for three days, from March 31 to April 3, to get rid of “harmful information.”32 In 2013, the state launched another massive anti-rumor campaign, primarily targeting social media platforms, including Sina Weibo. Many Internet opinion leaders were silenced or even jailed, including Charles Xue, an outspoken American-Chinese venture capitalist who had over twelve million Weibo followers. The campaign reportedly reduced the activity of Weibo opinion leaders by 40 percent compared to the previous year.33
Public opinion manipulation represents the state’s more aggressive and smarter adaptation to the challenges of online expression. Instead of trying to prevent or remove undesired information, such measures allow the state to engage citizens, promulgate its preferred voices, and directly shape online discourse. For instance, the state regularly issues decrees to content providers such as news portals and forums about its propaganda initiatives.34 It has also urged government agencies to set up e-government platforms,35 promoted state media outlets online,36 set up official accounts on popular social media sites,37 and embraced creative propaganda tactics such as “ideotainment” (by juxtaposing official ideological constructs and popular cyber culture) and “astroturfing” (by using the fifty-cent army to produce pro-government content under the guise of ordinary netizens).38 Through these measures, the state turns online expression into a new frontline for state propaganda and actively promotes its preferred content in cyberspace.
CHALLENGES FROM OUTSIDE AND FROM WITHIN
The Chinese Party-state, as a powerful, well-structured, and quick-learning machine, has adapted its control regime to “harmonize” the Internet.39 Some argue that it has thus far only reaped economic benefits without having a major political impact on authoritarian rule.40 However, the difficulty of controlling online expression and the state’s internal fragmentation have undermined the state censorship, making it at best quasi-successful. This section situates the Party-state as an internally divided organization within a complicated environment of high-volume anonymous online information exchange and explains how both external and internal challenges have resulted in the rigidity and arbitrariness of state censorship.
External Challenges to the Censorship Regime
The capacity of the state to calibrate “acceptable” expression is out of the question. Online expression challenges the Party-state in several additional ways. For example, the state must reduce an overwhelming amount of information to a manageable level, address the anonymity of online expression to track and punish violators, and operationalize its propaganda initiatives. These external challenges have conditioned the strategies of the state, forcing it to (1) “solidify” virtual space by embedding the Internet in existing institutions to make it less fluid and more accountable;41 and (2) rely on automated tactics, blacklisting, and, more recently, whitelisting to reduce the volume of information it has to deal with.
The regime has attempted to embed virtual space into the physical world by making the web less anonymous and more accountable. It has pursued the real-name registration of Internet users while establishing a complex system of registration and licensing of service providers. Real-name registration directly targets anonymity. The state has more or less succeeded in controlling Internet cafés insofar as most now require customers to show their identification cards.42 The real-name registration of cellphones has been partially successful, though SIM cards can be purchased from vendors who seldom check their customers’ identities. The state’s attempts to force netizens to register on forums, blogs, and microblogs have been met with greater resistance. Except for some campus forums in which student email addresses or identification numbers are required, most forums ask only for a valid email address to register an account. Given that campus forums have become a much smaller force in online expression since the Ministry of Education mandated them to re-form into campus-bound platforms in 2005, the impact of real-name registration has been more symbolic than significant.43
The state has also contracted censorship responsibilities to intermediary actors, including Internet service providers, Internet content providers, and research and educational institutions. Through measures such as registration and licensing, service providers and other relevant intermediary actors are held responsible for preventing unwanted information from appearing online. Large service providers, such as Sina and Sohu, have special offices for monitoring content.44 For smaller websites, webmasters are directly held responsible, with local Internet monitoring agencies often requiring them to be accessible via cellphone around the clock.45 Most campus forums are now under the direct supervision of university student affairs offices, Communist Youth League committees, or Party committees. Generally, these intermediary actors have considerable discretion to monitor routine online activities, but the state will occasionally intervene to patrol, issue directives, or punish offenders.
As intermediary actors usually have higher stakes than individual netizens, outsourcing censorship responsibilities to them helps reduce the state’s workload and encourages self-censorship. For most netizens, the risk of engaging in taboo topics is having their posts deleted or accounts suspended. Such punishments are inconsequential. But for intermediary actors, allowing politically sensitive expression may result in large fines, administrative punishments, or even the loss of their businesses. However, the system that delegates responsibility is often ineffective or even counterproductive. It is ineffective largely because intermediary actors can often evade state control measures. Many small and medium-size forums have managed to bypass licensing and registration by moving overseas or by going through agents who do not authenticate information.46 The system is also counterproductive at times because it introduces uncertainty and arbitrariness into the censorship system: Though intermediary actors may strictly follow state restrictions, they may also choose not to enforce them faithfully, either for commercial considerations or owing to personal inclinations. In particular, measures to hold intermediary actors accountable sometimes politicize them. This is the main reason that many small-scale website owners sided with Google when the company decided to withdraw from China in early 2010.
In addition to embedding cyberspace in existing institutions, China’s Internet control also relies on blacklisting taboo topics and actors. The cybernetic model of organizational choice suggests that when facing great uncertainty, large-scale organizations tend to engage in a servo-mechanistic pattern of decision making; that is, they base their decisions on key indicators and react only when these indicators reach certain thresholds.47 The high volume and extreme creativity of online expression have made the Internet an uncertain and complex environment for the Party-state. By monitoring a list of taboo words (mingan ci, 敏感词) and groups or individuals considered dangerous, the state simplifies its censorship task: It must react only when a taboo topic is detected or when a discussion of a dangerous topic reaches a certain level of intensity. This explains why state censorship often follows a significant burst of participation in a discussion of certain topics.48
The blacklisting strategy not only circumscribes the number of indicators the state must watch, but also facilitates automatic filtering. In addition to pornography, most taboos are political; that is, anything the regime deems threatening to its stability or damaging to its image, particularly events such as the 1989 Tiananmen Square democratic movement49 and any online expression deemed to have the potential for collective mobilization.50 For instance, the primary function of the Great Firewall is to prevent netizens from accessing blacklisted hostile websites or any webpages that contain forbidden keywords. If any of these sensitive words is detected, the system will not only interrupt the connection, but also reject any sequential attempts to access the server for at least a few minutes. Similarly, the Green Dam software, which the Ministry of Industry and Information Technology attempted, but failed, to have pre-installed on all personal computers in 2009, works by blacklisting keywords and web addresses, as netizens who hacked the software discovered.51 Internet content providers use the same method to detect politically sensitive expression. Major forums, blog services, and microblog platforms all automatically filter postings that their users try to post. When sensitive words are found, the filtering system either directly rejects the posting, asks the user to modify it, or sends the post to a nonpublic board for manual review by moderators.
The blacklisting mechanism also applies to “dangerous groups or individuals,” such as dissidents and online opinion leaders. State agents watch these groups and individuals closely to prevent them from publishing disturbing information online. For instance, Liu Xiaobo, a democratic activist and the 2010 Nobel Peace Prize laureate, was regularly put under house arrest and allowed only restricted telephone and Internet access during politically sensitive periods before he was jailed again in 2009 (after having been jailed several times from the late 1980s through the 1990s).52 However, actions against high-profile dissidents often prove counterproductive by turning these individuals into iconic heroes internationally, as was the case with Ai Weiwei, the artist who was jailed in 2011 for his dissenting activities, including the investigation of school building collapses in the 2008 Sichuan earthquake (see chapter 4 for more discussion). The state’s image is harmed even more among average netizens when the target is not quite political, as was the case when Gao Yaojie, a renowned and respected physician and AIDS activist, was placed under house arrest for weeks to prevent her from receiving an international award for her work in the United States.53
Blacklisting may effectively reduce the workload of state censorship but lacks the intelligence required to make contextualized judgments. At times, blacklisting also reveals the hand of the state and even backfires. The blacklisting system is triggered when and only when keywords are detected, which means any variants of the keyword can potentially circumvent the system. For instance, instead of typing “89” (referring to the 1989 Tiananmen Square democratic movement), netizens can insert an asterisk between the two characters (“8*9”) to circumvent automatic filtering. (Similarly creative tactics employed by netizens to explore loopholes of state censorship will be discussed more systematically in chapter 4.) Worse than being ineffective, indiscriminate filtering often enrages and politicizes otherwise indifferent netizens. When text such as “in eight or nine cases out of ten” is denied because it contains “eight” and “nine,” netizens previously unaware of or indifferent to censorship may start to complain or become curious about the blacklisted topic. Additionally, indiscriminate filtering also stifles pro-regime voices, as netizens supporting the regime also get censored. As a result, they may feel frustrated and start to see the irony of defending a regime that inhibits them from promoting their support.
The Party-state continually updates and refines its blacklist. But its adaptation often comes too late and fails to take into account nuanced and specific situations. For example, new taboo words are often added only once a threshold is reached, by which time the topic may have already been disseminated across the web, potentially reaching the mainstream media. The state’s “harmonizing” efforts often only invite criticism and, in some cases, energize the topic by reframing it as an anti-repression story. More than that, state censorship directives typically disregard specific situations, often asking service providers to delete all postings on certain topics regardless of whether they are for the regime or not, or to post a unified official statement (tonggao, 通稿) from Xinhua News Agency, the official press agency of the People’s Republic of China.54 Such “one-size-fits-all” approaches have also been used in a series of regulation campaigns since 2005. In such campaigns, many small-scale websites were shut down as collateral casualties because the entire Internet data center that hosts their servers was taken offline due to improper content found on only a few websites.55
Internal Challenges to the Censorship Regime
As China scholars have long recognized, the Chinese authoritarian regime is not monolithic, but rather fragmented both vertically and horizontally.56 If the uncertain and complex environment of the Internet poses external crises for the Party-state, fragmentation within the state constitutes internal challenges to the censorship system. Conflicting functions and interests, along with ideological discrepancies within the state, provide opportunities for online expression, and fragmentation allows for arbitrary censorship actions.
Above all, discrepancies of function and interest between state Internet governing bodies have led to competition and differing agendas. Aside from attempting to contain the Internet’s disturbing effects, the Party-state has also been promoting information technology as an industry. Pursuing these sometimes-incompatible goals requires not only a control regime, but also a regulatory regime that regulates the IT industry according to market principles.57 Compared to the ideology-driven propaganda system, the Ministry of Industry and Information Technology (formerly the Ministry of Information Industry) emphasizes the development of the sector and often does not favor restrictive policies. For instance, the China Internet Network Information Center, taking orders from the Ministry of Industry and Information Technology, was relatively unconcerned with domain name registration until China Central Television criticized it for “passively providing convenience for porn websites” during the 2009 anti-pornography campaign:58
Prior to 2007, nobody cared about domain name or website registration. But now without registration, government agencies will shut down the website. It is undoubtedly a severe consequence. Since the online rectification campaign started, China Telecom59 has shut down over one hundred thirty thousand unregistered websites, and there might be tens of thousands more accidental injuries.60
The pressure not only intensified the Ministry of Industry and Information Technology’s efforts, but also forced it to adopt ham-fisted tactics: All small websites were temporarily shut down for close scrutiny, which the Ministry of Industry and Information Technology Minister Li Yizhong claimed was a necessary overcorrection.61
State censorship also provides rent-seeking opportunities for involved state agencies and officials. That is, censorship can be used for business purposes. For instance, a PR manager for a multinational corporation boasted the corporation’s good relationship with the propaganda system, which he claimed was a more effective PR strategy than any other typical PR technique. Amidst a PR crisis following the voiding of promotional coupons, the company did not care what netizens said, because “whatever they complain about will be eliminated.”62 It is unclear how the propaganda system benefited from this process, but it is unlikely that it did everything for free. In fact, official reports have exposed several corruption cases associated with censorship. At the local level, an Internet police officer in Hainan Province allegedly took bribes from counterparts from other localities for removing negative publicity about them.63 At a higher level, the former deputy director of the Fifth Bureau of the Cyberspace Administration of China and the International Communication Office, Gao Jianyun, was charged for deleting online postings in exchange for bribes.64
The political and economic benefits involved in censorship may drive state agencies to compete for regulatory and control power over the Internet.65 For instance, the Ministry of Culture and the General Administration of Press and Publication have openly fought for the authority to pre-approve online games.66 Similarly, some observers suspect that the onset of the anti-pornography campaign in 2009 was a result of the long rivalry between the Ministry of Industry and Information Technology and the State Administration of Radio, Film, and Television.67 Such competition takes place not only among governmental agencies, but also between party and government bodies. Sources claim that the real-name registration of blogs was an effort by the Central Propaganda Department to wrest control over online information services from the State Council Information Office.68
State agencies may have different perceptions of risks and benefits. For instance, the Ministry of Education pushed for the real-name registration of campus forums and attempted to turn them into intra-university platforms in 2005. By isolating campus forums and making them less likely to take up hot social topics, the Ministry of Education attempted to limit its own responsibility in this area. But the Chinese Communist Youth League Central Committee showed interest in campus forums because it saw opportunities to expanding its regulatory reach. At the National Campus Bulletin Board System Managers Conference, the sponsoring Chinese Communist Youth League Central Committee subsidiary showed particular interest in establishing a self-disciplinary association and in promoting online ideological indoctrination, job searches, and entrepreneurship through campus forums. These goals are clearly safe for the Chinese Communist Youth League Central Committee since none threatens control over online expression—and achieving these goals would improve the performance evaluations of its officials, as doing so would demonstrate their competence.69
Ideological conflicts within the Party-state constitute an even more fundamental and deep-rooted threat to the coherence of the censorship system. Constant struggles, occurring most prominently among different factions during the 1989 democratic movement, illuminate the ideological divisions within the Chinese Communist Party. Many former Party-state officials have defected overseas70 or departed greatly from Party discipline. They have started to call for constitutional rights, such as freedom of speech, the press, assembly, association, and demonstration.71 Such ideological and factional struggles have now spilled over into cyberspace. For instance, when former Party chief of Chongqing Bo Xilai was removed in 2012, online debates erupted, with many netizens and several left-wing sites, such as Utopia (www.wyzxsx.com), boldly supporting Bo against the central leadership.72 To quell such voices, the propaganda machine took a number of measures, including shutting down those online platforms. Such “targeted censorship” (dingxiang shencha, 定向审查) reveals that online expression and the censorship regime have become emblematic of political struggle within the Party-state.
Increasing ideological pluralization and increasing media commercialization have started to shake the base of the propaganda state. As Johan Lagerkvist has rightly noted, state agents and political leaders may “slowly begin to doubt the legitimacy and sustainability of the control and censorship regime,”73 and media elites, reformist officials, and propaganda cadres may influence each other in ways that allow for a freer discursive space.74 Some mainstream media outlets, such as Southern Weekend (Nanfang Zhoumo, 南方周末), are known as liberal strongholds, despite heavy-handed state control.75 When supposed ideological strongholds of the propaganda system and mouthpieces of the Party-state are flawed, the control regime can hardly be as effective as the regime might hope.
Fragmentation exists not only across sectors but also across levels of the Party-state. As power to control the Internet is delegated, principal–agent problems emerge.76 Being faithful in implementing censorship policies may not be a sensible choice for subordinates when it conflicts with other priorities. For instance, local legislators in Guangzhou worried about the impact of the excessively harsh registration of Internet cafés on public and small businesses.77 This example helps explain the variation in content control across regions and even across websites.78 Such a vertical divide provides opportunities for online expression but sometimes also induces more rigid and arbitrary censorship measures, as local agencies tend to showcase their competence or cover up negative publicity under the pressure of the cadre responsibility system.79 This explains why local authorities have aggressively sought to eliminate “harmful” online information and suppress netizens daring to disclose local scandals heavy-handedly. For instance, a young man named Wang Shuai was detained in Shanghai simply for posting an online complaint about improper land grabs in his hometown of Lingbao County, Henan Province.80 Such interprovincial pursuits of netizens (kuasheng zhuibu, 跨省追捕) demonstrate how far local governments can go.81 The paramount concern of “stability maintenance” (weiwen, 维稳) efforts by local authorities is clearly not just to preserve the legitimacy of the regime, but to reduce trouble for local governments.82
CONCLUSION
According to the Chinese media expert David Bandurski, the state’s primary means of control is the “fuzzy line” that does not clearly define acceptable and unacceptable content, and its control apparatus is built on “uncertainty and self-censorship, on creating this atmosphere of fear.”83 The analysis in this chapter suggests that the fuzziness of state censorship can be both intentional and unintentional. It can be intentional in that the strategy is indeed cost-effective and thus an optimal option for the state, and it can be unintentional because “fuzziness” can result from structural factors and the state’s limited capacity to control. Indeed, China’s Internet control regime has undergone a process of policy learning through which the state has gradually built up its institutional, organizational, administrative, and technical capacities to tame the Internet. However, the seemingly formidable content control regime still faces severe external and internal challenges, which create tensions in Internet censorship. On one hand, anonymous and creative online expression has forced the state to rely on servo-mechanistic surveillance techniques and to delegate censorship power to intermediary actors. Though these strategies may help reduce the workload of censoring agencies, they have also made the censorship regime overly rigid, arbitrary, and fuzzy, thus providing room for both intermediary actors and netizens to maneuver. On the other hand, internal fragmentation of the authoritarian state, particularly in terms of interest and value conflicts among state agencies, has introduced heterogeneity into the state censorship apparatus, producing opportunities for online expression in some cases and inducing harsher censorship in others.84
Internet content control is not only a process of state adaptation to digital challenges, but also a process of nonstate actors exploring the opportunities and limits of online expression. In particular, nonstate actors, such as various service providers and netizen groups, are continually reacting to the Party-state’s censorship efforts. In this chapter, the role of these nonstate actors has been acknowledged, but not detailed. The following two chapters will examine how the governed, particularly forum managers and users, respond to state censorship.