Contesting Cyberspace in China: Online Expression and Authoritarian Resilience

Many studies of Internet politics in authoritarian regimes—including those discussed in the first half of this book—choose to focus on the state–society struggle over content control; that is, what can and cannot be expressed online.2 These studies offer valuable insights into the dynamics of state control and social resistance in the digital era. As the previous three chapters have shown, the state, intermediary actors, and netizens have wrestled over the limits of online expression in China. Despite its legal, administrative, and technical efforts, the Chinese state has barely succeeded in controlling online expression, with discontented but compliant intermediary actors taking a fence-sitting stance while rebellious netizens regularly circumvent censorship in creative ways. However, focusing exclusively on censorship and counter-censorship is limiting. This perspective implies a cohesive group of netizens fighting unanimously against the state, which is inaccurate, and overlooks certain aspects of state adaptation. High-capacity authoritarian regimes such as those of China and Russia not only enjoy considerable popular support,³ but also demonstrate a strong adaptability to reform and cope with new challenges.⁴ To accurately gauge the resilience of authoritarianism in the digital era, we must look at how authoritarian states adapt beyond censorship, and how various actors interact with each other within the state-stipulated boundaries of expression.

This and the following two chapters focus on discourse competition in Chinese cyberspace by exploring the pluralization of online expression. They place special emphasis on how the state, its critics, and various netizen groups engage each other in ways other than through censorship and counter-censorship. Unlike the cat-and-mouse censorship game in which the state dominates with its coercive power, these chapters reveal a discourse competition in which involved actors rely more on expressive tactics and identity mobilization to shape online discourse. Seeing online expression as a discourse competition helps us understand Internet politics in China from the perspective of the public sphere, while also shedding light on how regime legitimacy is maintained or weakened through public debates in virtual space.

This chapter examines state adaptation beyond censorship. By demonstrating how the state has deployed paid Internet commentators—popularly known as the “fifty-cent army” (wumao dang, 五毛党)—to manufacture seemingly spontaneous pro-regime voices online, this chapter reveals authoritarian adaptability in the digital era. Given that old-fashioned propaganda has increasingly become ineffective, the introduction of Internet commentators represents an innovative adaptation of the state to promulgate its preferred messages and guide online opinion. However, this seemingly smart move has produced mixed results: Although the fifty-cent army may have managed to increase the state’s PR effectiveness on specific issues, it often backfires. This is because the adaptation is embedded in the apparatus of state propaganda and thought work, which though still powerful and resourceful, functions according to a logic that offers little incentive for Internet commentators to excel at their job and often leads to the exposure of their existence. The mixed results of the fifty-cent army strategy suggest that it is necessary to problematize “state capacity” and “state adaptability” in the digital age, as the Chinese state has been struggling to translate its despotic and infrastructural power of the pre–Internet era appropriately to shape online expression.

Relatively free expression online empowers social actors to challenge authoritarian regimes in many ways. In China, the Internet not only serves as an emerging public sphere, but has also promoted the development of civil society and facilitated collective action, online activism, and dissident mobilization. 5 Thanks to the Internet, Chinese citizens now enjoy some agenda-setting power,⁶ are able to question state agents more effectively,⁷ and can even influence state policies.⁸ In response, the Party-state has adapted to tame the Internet by constructing a comprehensive censorship system to control the flow of information online.⁹

However, the state has neither the capacity¹⁰ nor the intention to completely shut down the Internet or eliminate all political expression online.¹¹ In fact, perfect censorship may actually be undesirable for authoritarian regimes. On one hand, because much political expression takes place on popular web platforms, imposing overly strict censorship or shutting down such platforms may cause collateral damage, provoking public protests quite unnecessarily.¹² On the other hand, freer online expression may work to the advantage to the authoritarian regime as a form of policy feedback, at least by helping the state discipline its local agents.¹³ This logic is nothing new to China scholars, who have commented on the Party-state using village elections or popular unrest as a means of holding local officials accountable.¹⁴ Likewise, the state may selectively censor online expression and adjust the level of censorship in response to social tensions so it can harness the benefits of free information.¹⁵ The strategic censorship consideration, together with capacity constraints, may explain why the state tolerates critical online expression to a considerable degree except in the case of some well-patrolled “forbidden zones.”¹⁶

There is one problem with the strategic censorship logic. It assumes an omnipotent state that can differentiate online content according to the threat it poses and exercise its censorship perfectly. This is unrealistic, even given the strong capacity of the Chinese Party-state. As discussed in chapter 2, state censorship is at best semi-successful, with netizens and dissent groups still capable of penetrating its limits. Moreover, even limited freedom of expression can be risky for the Party-state. By breaking the state’s monopoly on mass media, the Internet, with its freer expression, enables citizens to contest the regime’s authority, propagate social discontent, and diffuse protest tactics. In this way, it calls the Party-state to account for discrepancies between its own views and those of the public.¹⁷ Freer expression may also expose the pervasiveness of social ills and official misconduct, thus promoting a shared awareness among currently scattered citizens about the need for fundamental change.¹⁸

What, then, can the state do to offset these detrimental effects of online expression on its authoritarian rule? In addition to enhancing and refining the censorship system, the Chinese Party-state has attempted to expand its online presence and reinvigorate state propaganda online. To increase its online presence, the Chinese state has promoted e-government platforms, pushed state media online, and penetrated popular social media sites. First, to enhance its administrative capacity, improve public services, and expand citizen outreach, the state has pushed government agencies at all levels to set up e-government sites. This “Government Online Project,” launched in 1999, has resulted in the creation of 53,546 government websites as of December 2016.¹⁹ Though scholars debate the effectiveness of such platforms, they have indeed become a channel for the state to engage citizens.²⁰

Second, the Party-state has pushed state media, such as the Xinhua News Agency and the People’s Daily, to occupy cyberspace21 and has granted them exclusive rights to produce news reports that other online news services are only allowed to “reprint.”²² In fact, in addition to setting up their own online outposts, state media have also populated major social media platforms. For instance, as of October 2013, the two major microblogging service providers in China, Sina Weibo and Tencent, hosted at least 23,449 and 14,148 verified, official accounts respectively for traditional media outlets, including ones run by the state.²³ The increased online presence of state media has clearly allowed the state to respond to hot-button issues more quickly with its own voice.

Third, the state has attempted to engage the public directly on popular social media platforms.²⁴ For instance, as of December 2016, there were 164,522 verified, official accounts, representing 125,098 government agencies and 39,424 individual officials on Sina Weibo alone.²⁵ Some popular ones among the official Weibo accounts, such as that of Chinese Communist Youth League Central Committee, attract millions of followers and thus play a critical role in online agenda setting and issue framing.

Meanwhile, the Party-state has also updated its offline and online propaganda strategies to make state propaganda more appealing and effective. According to the political scientist Anne-Marie Brady, since the early 1990s, “China’s propaganda leadership has embraced new modern methods of social control and persuasion and introduced the use of new technology into propaganda work. In doing so, they have made a major contribution toward creating the conditions for ongoing CCP rule in China.’’²⁶ In the realm of traditional media, commercialization is a highly salient feature. Scholars such as Yuezhi Zhao have observed that by introducing market mechanisms into Party journalism, China’s media system has gradually evolved into a “propagandist/commercial model” that performs more subtle ideological work for the Party.²⁷ Mixed-method research by Daniela Stockmann has also found that media commercialization promotes regime stability, rather than destabilizing authoritarian rule, because it allows the state to tell more convincing and sophisticated stories while at the same time offering the leadership better information about what citizens think.²⁸ Alongside media commercialization and globalization, a recent study by Maria Repnikova has found that official influence still looms large in elite journalism training.²⁹ The state continues to place ideological teachings at the core of journalism curricula and employs control mechanisms, such as structural oversight, ad hoc surveillance, and coercion, to ensure journalism training does not deviate from official ideology. But since educators and students have been able to exploit limited openings to reinterpret some aspects of the Party’s ideology, the Party-state has merely achieved a “hegemony of form.”

Making state propaganda appealing online is a much more challenging task because the state can have no control over how particular topics are produced, distributed, or consumed without resorting to censorship, which reduces its success and often causes antipathy. Indeed, the state now finds itself in a new battlefield in which the power of beliefs, values, and identity, rather than coercive power, triumphs. It has to employ more innovative strategies to win the hearts and minds of netizens against competing discourses. According to the China specialist Johan Lagerkvist, one strategy the state employs to counter perceived enemies, to win over public opinion, and to shore up legitimacy is “ideotainment”—the use of popular expressional formats as vehicles to promulgate state ideological constructs.30 For instance, both President Xi Jinping and Premier Li Keqiang published official cartoon images of themselves in early 2014, making them look more approachable to the public.³¹ In fact, all seven of the new Politburo Standing Committee members had already been partially “cartoonized” as early as October 2013 for a video clip comparing how China, the United States, and the United Kingdom select top leaders.³² The five-minute video, using many popular cyber-cultural elements to convey the message that the Chinese Communist Party’s rule is justified, immediately went viral online.³³ According to Youku.com, the Chinese version of the video recorded more than 1.5 million views within two days.³⁴ Such efforts suggest that the Chinese Party-state may have shifted its focus from ideological indoctrination to a more subtle management of public attention in the norms competition with regime critics.³⁵

Even strategies such as ideotainment cannot prevent the regime from losing its ideational leadership, however. Netizens are so wary of state propaganda that any government connection to online content may be seen as an “instant negative.”36 Thus, hiding the “government connection” seems to be a natural strategy for the state and its agencies in order to exert influence over online opinion. In this sense, using “Internet commentators” to troll online expression represents a significant propaganda innovation. At least by deploying the fifty-cent army to fabricate pro-regime voices under the guise of ordinary netizens, the state can avoid a backlash over outright state propaganda and manufacture consent more effectively. The following sections examine this important yet understudied innovation with the purpose of examining and assessing this seemingly smart adaptation by the Chinese Party-state.

The Internet not only empowers average citizens to circumvent state control and engage public expression, but also potentially provides a way for any motivated actor to shape public opinion in more subtle and effective ways. Instead of directly promoting certain agendas, motivated actors can manipulate other users, who are vulnerable to such manipulation because they are not embedded in real-life accountability networks. For example, using multiple ghost accounts, a single manipulator can “stir up” (chaozuo, 炒作) a topic into a hot one. Public opinion is manipulated in this process not only because the initial momentum comes from a created “public,” but also because the “dominant” view of the “public” can be engineered through purposeful framing and information input, which in turn influences innocent users’ perceptions and their subsequent input. Such a trolling practice is essentially “astroturfing,” a widely used PR technique in politics and marketing in which incentivized actors (often paid) display apparent grassroots support for a product, policy, or event to shore up wider and more genuine support.³⁷ The tactic allows sponsors to hide their identity—to avoid associated negativity or maintain “plausible deniability”—and enhances the credibility of the message through the false impression of personal testimonies and widespread support.³⁸ The Internet serves as a natural platform for astroturfing given the anonymity of online expression. In the realm of politics, actors across the globe have used this PR tactic to their advantage. Authoritarian and illiberal regimes, such as those of China, Egypt, Kenya, Russia, Saudi Arabia, Syria, and Venezuela, have all recruited “electronic armies” to conduct political astroturfing.³⁹ Even in democracies, there are reports of political actors using astroturfing to gain an advantage over competitors.⁴⁰

Though political astroturfing is quite common, in-depth studies on the phenomenon are rare, largely owing to the secretiveness and sensitivity of the practice, especially in repressive regimes. Fortunately, a few existing studies may help us navigate the topic in the Chinese context. Chin-fu Hung, one of the first scholars in the field, has explored how Internet commentators were mobilized to pacify the Weng’an riot in Guizhou after the suspicious death of a 16-year-old girl.41 But Hung’s research does not investigate how Internet commentators actually operate on the ground. Based on the analysis of real fifty-cent army posts leaked from a district government, the political scientist Gary King and his colleagues have developed a sophisticated method to identify Internet commentators on social media.⁴² Finding that state astroturfing efforts often focus on cheerleading rather than confronting regime critics or controversial topics, they argue that the state’s strategy is not to engage but to distract the public. But this methodologically rigorous research is primarily based on data from one specific local government, making external validity a concern. To address this, Blake Miller, who studies the use of technology in state propaganda, proposes to detect Internet commentators using network and social metadata such as IP addresses and linkages to government accounts.⁴³ This approach overcomes the external validity problem and shows promise for better appreciating the nuances in online commentating. However, it is also limited, in that just like the work of Gary King’s group, it focuses exclusively on Internet commentators’ online activities without a holistic examination of how the whole system works, particularly how the Party-state mobilizes Internet commentators.

The existing studies are helpful, particularly in terms of how to identify state astroturfing online. But they are insufficient to assess the state’s intentions and capacity in this adaptive move. To fill the gap, it is necessary to situate the Internet commentator system into the broader context of state adaptation to the digital age and examine how it works systemically. The following section draws on many sources of data to show how Internet commentators were first introduced online; how they are recruited, trained, and rewarded; and how they operate in the online environment. Three primary sources of data are especially worth noting here. First, state propaganda directives and other official communication logs are direct evidence of how the system functions. These “internal” documents are not publicly available but may be exposed as a result of system glitches in the state propaganda platforms where they are stored or obtained from officials who leak them because of incompetence, carelessness, or disaffection. Though such incidences are rare, when available they provide rich data for researchers to peek into the “black box” of the state propaganda apparatus. Second, official media and government websites sometimes openly report on their Internet commentators because local authorities and propaganda agencies deem online commentating as a part of their routine work and thus boast rather than hide their achievements.44 These reports confirm the existence of Internet commentators and are particularly valuable for studying how they are introduced online, trained, and rewarded. Third, in-depth online ethnographic work produces additional data to analyze how the fifty-cent army functions in the online environment. Based on long-term observation, I have established a scheme to differentiate state-sponsored Internet commentators from regular users with quite some confidence. This scheme, which resembles that adopted in both the work of King and his colleagues and Miller, has been developed through learning from netizens, who are extremely sensitive to and thus cognizant of state astroturfing,⁴⁵ as well as by comparing expressional traits: Internet commentators tend to make only pro-regime comments on certain restricted political topics using official language codes and generally do not interact with other netizens. In-depth online ethnography also pays in that it exposes the researcher to many leaked official documents first discovered by netizens. I have even observed someone who used to work in the system talking about his or her experiences on Mitbbs.⁴⁶

The expansion of the Internet has clearly broken the state’s monopoly over the media and rendered outright propaganda increasingly ineffective.47 For instance, in a report on Libya’s response to the French-led airstrike in 2011, CCTV attempted to pass off an interpretation of local Libyans as Kaddafi supporters protesting French intervention. CCTV deliberately mistranslated their “Vive la France” banner as “French, go home.”⁴⁸ The deliberate misrepresentation was caught by critical netizens and attracted heavy criticism from many Internet forums. Facing such challenges, it is only natural for the state to resort to new propaganda tools and strategies. Introducing Internet commentators represents such an effort of state adaptation. In the political scientist Chin-fu Hung’s words, the state has been “revitalizing the propaganda apparatus through the utilization of these commentators.”⁴⁹ However, as my analysis will show, this seemingly smart adaptation is not as successful as the state may have expected, largely because the system is geared more toward pleasing upper-level state officials than engaging and persuading ordinary citizens.

There is little solid evidence showing when, why, or how the Internet commentator system was first created. It was in late 2004 when the term “Internet commentator” first appeared in an official report and online. Changsha, in the Hunan Province; the supervision department of the Chinese Communist Party Central Commission for Discipline Inspection; and Nanjing University are possibilities for the first local government, central state agency, and higher education institution, respectively, to deploy Internet commentators.⁵⁰ The first group of Internet commentators quickly earned the nickname of “fifty-cent army” because it is believed that they were paid 50 cents per post, plus a base monthly salary of 600 yuan (less than US$75 in 2005) in the Hunan case.⁵¹

Though some argue that Internet commentators were introduced by the state primarily to combat “Internet spies” sponsored by hostile foreign forces (wangte, 网特),52 the argument is not strongly supported. In fact, the adaptation was initiated by a few state agencies independently, suggesting that it was likely a local or departmental initiative rather than a centrally coordinated policy. That being said, the fact that state agencies in different sectors at different levels adapted this practice at around the same time implies that officials at all levels more or less simultaneously started to realize the potential of Internet commentators to shape public opinion.

Today, the deployment of Internet commentators is a systematic nationwide practice. On one hand, numerous state agencies, including all levels of authorities from counties to ministries, schools, and state-owned enterprises, sponsor their own fifty-cent armies.⁵³ On the other hand, the central government, through the propaganda system and the Communist Youth League, also controls a direct force of the trolling army. Unlike traditional news commentators who are often media professionals working at state-run media, Internet commentators undertaking astroturfing tasks are mostly employees of the sponsoring agency, whose trolling assignments are part of their regular duties.⁵⁴ Recruiting from within the system makes sense because it not only reduces the cost of recruitment, but also ensures a minimal level of accountability of Internet commentators, particularly in terms of their loyalty to the system. Only in rare cases are Internet commentators directly recruited from the general public. In such cases, recruiting agencies often stipulate certain criteria such as political loyalty and basic computer skills.⁵⁵ The vague and general nature of such criteria implies that they are mostly official clichés and thus unlikely to be strictly enforced. Indeed, universities often recruit employees from the student body as a form of financial aid, and recruitment is open to all.⁵⁶

In recent years, the state has turned college students into a “fifty-cent army reserve.” In 2014, the Chinese Communist Youth League Central Committee decreed universities across the country to set up student Internet propaganda troops to engage in online expression and monitor the “ideological trends” (sixiang dongtai, 思想动态) among students.⁵⁷ In this massive recruitment, student cadres in local Communist Youth League branches, student unions, campus media outlets, and student news media organizations, were given priority.⁵⁸ This again confirms the state’s emphasis on the political loyalty of fifty-cent army members.

Official reports indicate that Internet commentators often receive some training, but evidence suggests that such training is likely quite basic in most cases. Mostly, it would involve attending lectures by propaganda officials and media professionals, as well as exchanging experiences with each other.59 For instance, in October 2009, officials in the Jiangdong district of Ningbo, Zhejiang Province, held a training session for 102 newly appointed online commentators, which included lectures from the local Internet administrative center and Public Security Bureau Internet monitoring branch directors and peer experience exchange.⁶⁰ While there are no data on the specific techniques taught at these sessions, the training is likely quite basic and ultimately of no consequence, as will be discussed shortly. Leaked internal training documents suggest that Internet commentators are taught such basic skills as how to register an account and how to post or reply to threads. The only “technique” they are taught is how to create multiple ghost accounts to make their astroturfing activities less apparent.⁶¹ Such minimal requirements betray the low expectations of sponsoring state agencies, suggesting they may not care whether Internet commentators can effectively engage in public debates or sway the opinions of netizens.

Internet commentators tend to be poorly rewarded and are thus often unmotivated. Sponsoring agencies often provide them with monetary compensation, promotion opportunities, and recognition, but none of these constitutes a strong incentive. Monetary compensation is minimal in most cases—sometimes as little as 10 cents per post.⁶² Even when they are paid better—for example, a base salary of 600 yuan per month and 50 cents per post, as in the case of Hunan—the compensation is still far from the average for government employees. In some cases, Internet commentators are rewarded with opportunities. This is especially the case for politically ambitious college students, because working as Internet commentators helps open doors to state-sponsored organizations such as student unions or the Communist Youth League. But since there are many alternative options to get into those organizations, becoming an Internet commentator is not particularly attractive. Occasionally, top-performing Internet commentators are selected to receive honors and rewards, showing that their efforts have been recognized by the state.⁶³ Yet, such recognition is unlikely a strong motivator for most Internet commentators, because only a small percentage of them are eligible for such recognition, and, further, such recognition is not something to be proud of publicly.

Internet commentators are often mobilized for crisis management to defuse collective mobilization online and to maintain the image of their sponsoring agencies. Such cases include social protests such as those that occurred during the 2008 Weng’an riot and the removal of top-ranking officials such as Bo Xilai in 2012. The Weng’an riot was a massive one in which protesters in Weng’an County, Guizhou, thronged the streets and set fire to government buildings and vehicles as a demonstration of outrage over the suspicious death of a young girl and the allegedly official efforts to cover it up.⁶⁵ The Bo Xilai case is an example of the central mobilization of Internet commentators. The ousting of Bo—formerly the party chief of Chongqing and a Politburo member—triggered a legitimacy crisis, because it revealed the corruption of a high-ranking Party official, the power struggle among the top leadership, and the ideological conflicts present in Chinese society. As a result, it was unsurprising to see Internet commentators flooding all major online platforms to justify the trial of Bo and to praise the system for championing the rule of law.⁶⁶

The following intercept from an official report authored by the director of the Internet Administration Center of Wenzhou, Zhejiang Province, outlines the conditions that trigger the mobilization of Internet commentators:⁶⁷

The italicized section of this statement implies that local authorities are highly concerned with crises under their own jurisdiction and the subsequent impact on their image. This is confirmed by an internal report from the Communist Youth League’s Shanghai branch, which listed its several major “achievements” in 2009. According to the report, Internet commentators were mobilized to defuse the following online crises: building collapses,68 the pre-installation of Green Dam censorship software,⁶⁹ violent law enforcement for city management officials in Putuo District,⁷⁰ controlling the spread of the H1N1 influenza virus, the case of self-immolation caused by forced eviction in Minhang District,⁷¹ the entrapment of illegal cabs,⁷² and glitches in the subway system. All these events but two—the pre-installation of Green Dam and controlling the spread of H1N1—were local crises, indicating the priorities of local authorities. That being said, Internet commentators at all levels can be centrally mobilized by the state when it deems necessary, as in the cases of Green Dam, H1N1, and the Bo Xilai trial. In other words, the Internet commentator system is both fragmented and centrally coordinated. While local propaganda agencies and the Internet commentators they sponsor primarily serve local interests, they must also answer to calls from higher levels.

In addition to defusing online crises, Internet commentators are also mobilized to promote propaganda campaigns at both central and local levels. For instance, authorities in Hengyang, Hunan Province, mobilized Internet commentators in late 2008 to engage the “Liberate Thinking and Develop Hengyang” propaganda campaign. They were asked to participate in thematic discussions, post comments on local and national online platforms, and even fake online interviews with local officials.73 One specific assignment was to post one thousand comments to the “Hengyang Municipality Propaganda Branch ‘Liberate Thinking Big Discussion’ Special Thread.”⁷⁴ Hengyang is not an exception. For example, Internet commentators from Ganzhou, Jiangxi Province, were deployed to manipulate an online interview of the local Party chief in January 2014. They were asked to leave at least one comment “like average netizens” and report content and links to the Internet Propaganda Office.⁷⁵

Central state agencies also mobilize Internet commentators, often through their local branches, to facilitate propaganda initiatives. In May 2013, the entire propaganda system coordinated to promulgate President Xi Jinping’s idea of the “Chinese Dream.” In mid-2013, Internet commentators at all levels were mobilized to denigrate “public intellectuals” (gongzhi, 公知).⁷⁶ In December 2014, to promote the first “National Memorial Day for Nanjing Massacre Victims,” the Chinese Communist Youth League Central Committee mobilized college students across the country to comment on and repost designated Weibo entries. Leaked internal communications from the Communist Youth League’s Shanghai branch reveal how universities responded to the call. For instance,

By 12:00 PM, the Shanghai Business College Internet Propaganda Team had completed the following number of reposts and comments: The Oriental Wealth Media and Management School has 347 reposts and 96 comments; the School of Foreign Languages has 593 reposts and 254 comments; the School of Tourism and Food has 651 reposts and 529 comments; the School of Management has 508 reposts and 195 comments; the School of Information and Computer Science has 438 reposts and 147 comments; the School of Arts and Law has 289 reposts and 99 comments; the School of Arts and Design has 324 reposts and 139 comments; the School of Finance and Economics has 862 reposts and 397 comments. In total, our college has produced 4,012 reposts and 1,856 comments.⁷⁷

Not all universities are equally active, however. For instance, Shanghai Normal University produced 2,178 comments and reposts in total; the Shanghai Institute of Health Sciences contributed 816 reposts and 321 comments; the Shanghai University of Finance and Economics reported only 216 reposts and 58 comments; and East China Normal University produced 269 reposts and 56 comments.78 This variation suggests that the call was not treated as a mandatory task and thus each university felt it had considerable leeway. Nonetheless, contributions from student Internet commentators have made a huge splash online. As figure 5.1 shows, one Weibo entry by the Communist Youth League accumulated over 210,000 reposts and 31,000 comments within a matter of days. Table 5.1 provides an English translation.⁷⁹

Since Xi Jinping came to power, the state has made more aggressive efforts to regain ideational leadership. In particular, the propaganda machinery has worked hard to project and maintain a positive image of Xi as the new icon of the regime. Numerous Internet commentators have been mobilized to glorify Xi online. Such blatant praise of Xi by the fifty-cent army is quite easy to recognize: They “smell” official, are highly repetitive, and look similar to each other (figure 5.2, with translation provided in table 5.2). Moreover, the comments come in waves, with a large number of consecutive ones posted from the same city within a very short period of time. A check of user profiles reveals that Internet commentators making such comments tend to comment only on Xi Jinping, use official language codes, and never interact with other netizens (figure 5.3, with translation provided in table 5.3).

An analysis of comments on randomly selected news reports on Xi Jinping, Li Keqiang, Hu Jintao, and Wen Jiabao shows that Internet commentators are indeed much more active in the Xi Jinping era (table 5.4). The results are telling in that the coding scheme is likely to underestimate the pervasiveness of the fifty-cent army, particularly in the case of Xi Jinping, as many anonymous comments regarding him are likely made by Internet commentators. With only location information available (in the following format: Netease User from X Province Y City ip: xxx.xxx.*.*), these anonymous comments are admittedly quite difficult to code. But as figure 5.2 shows, these comments came in waves, with several or even dozens of consecutive comments from the same locality. This discernable pattern, together with the language traces, suggests that these comments are likely not genuine.

In addition to manipulating online opinion, Internet commentators often also function as a bridge connecting the state and the public by helping proliferate and clarify state policies and, more importantly, by reporting public concerns to the authorities. This function, though hardly appreciated by most netizens, is deemed beneficial by some to counterbalance unfounded rumors and to improve governance.80 In particular, monitoring online discussion provides the state with references for policy-making and implementation. For instance, online commentators are said to have contributed to the Changsha Public Opinion Express (Changsha Yuqing Kuaibao, 长沙舆情快报), which is edited by the local External Propaganda Office and delivered to municipal leaders daily.⁸¹ Li Guanghua, a former deputy director of Hengyang Propaganda Department’s Information Office, claimed that he used to organize online commentators to compile netizens’ complaints and directly submit them to the local Party chief, who would then push for solutions. Student Internet commentators also assume similar responsibilities, such as collecting and reporting students’ suggestions and criticisms.⁸²

While “linking the government and the people” sounds like empty self-praise, Internet commentators have indeed helped citizens in some cases. For example, in 2014, in Hukou County, Jiangxi Province, local Internet commenters detected a Weibo entry alleging that a company owed 3 million yuan to more than three hundred migrant workers. Local authorities then looked into it and solved the problem by remunerating the migrant workers.⁸³ Similar cases are actually quite common, according to leaked internal emails from Zhanggong District, Ganzhou City, Jiangxi Province.

The rise and spread of Internet commentators demonstrate the state’s strong capacity to learn and mobilize. Initially a local innovation, the practice has not only diffused horizontally across the country, but also become a “national policy,” in that it is now an integral part of the state’s apparatus of propaganda and thought work. Through existing institutions, such as the propaganda system and the Communist Youth League, the Party-state can mobilize tens of thousands of Internet commentators at any given time. For instance, in March 2014, the Chinese Communist Youth League Central Committee urged colleges and universities to set up a nationwide Internet propaganda troop (wangluo xuanchuanyuan, 网络宣传员) of at least 350,000 members, with quotas of the troop assigned to each provincial unit (table 5.5).⁸⁴ In February 2015, the Central Committee called for the establishment of an even larger army of “youth Internet civilized volunteers” (qingnian wangluo wenming zhiyuanzhe, 青年网络文明志愿者), with a minimum quota of 10.5 million members.⁸⁵

There is no doubt that the Party-state possesses formidable power and resources to institutionally and organizationally restructure itself, to mobilize its agents, and to adapt. But such strong capacity does not automatically guarantee successfully adaptation. Internet commentators have been mobilized for many purposes, including defusing online crises, facilitating state propaganda initiatives, cheerleading for the leaders, and connecting the state and the citizenry. But the state has generally failed in its mission of guiding public opinion online. Like other scholars who have studied this phenomenon, I find that Internet commentators communicate almost exclusively among themselves and very rarely interact with netizens.⁸⁶ One plausible explanation for this, of course, is that the state intends only to distract rather than engage netizens in online debates.⁸⁷ It is also possible that the state’s goal is to signal its strength in maintaining social control and political order rather than persuade netizens.⁸⁸ But my analysis shows that Internet commentators do not engage in critical online discussion because they are unmotivated to excel at their job. In particular, the state’s reliance on the propaganda system, the Communist Youth League, and colleges and universities to mobilize the fifty-cent army betrays how embedded the innovation is in the existing apparatus of propaganda and thought work. Such embeddedness, though able to facilitate the state to more effectively mobilize and control Internet commentators, has captured and bureaucratized the innovative adaptation, thus preventing the fifty-cent army from functioning effectively and, in many cases, rendering its work counterproductive.

First, the existing apparatus of propaganda and thought work is incapable of evaluating and improving the performance of Internet commentators. In fact, as most Internet commentators have been recruited from within the system, with selection criteria that give priority to political loyalty, they often lack proper skills to communicate appropriately with other netizens. As a result, it is not surprising at all that they are not functioning as effectively as the state might have anticipated. According to a former propaganda official in charge of Internet commentators in Hengyang, Hunan Province, Internet commentators vary in terms of their capabilities, and many of their comments “fail to guide public opinion online and even backfire sometimes.”⁸⁹

This capacity deficit is made worse by the improper incentive structure, because many Internet commentators simply are unmotivated to excel at their job. Evidently, most Internet commentators are part-time workers trolling for the state and poorly compensated. To many of them, commenting online for their sponsoring agencies is simply extra work. To others, the fact that they are paid makes the job seems to be a cheap “sale of souls” (chumai linghun, 出卖灵魂).⁹⁰ This is precisely why “fifty-cent army” has become a disgraceful label. As a campus forum commentator confessed,

Second, the embeddedness of the fifty-cent army has made it vulnerable to the problems that plague state propaganda work. While agencies such as the propaganda system and the Communist Youth League can mobilize Internet commentators en masse, their behavior and mindset are not adjusted to fulfill the new task of opinion manipulation.⁹² In particular, these agencies have mostly turned political astroturfing into a mundane bureaucratic process of propaganda work, which is often mobilized in a top-down manner, with pleasing higher-ranking officials being the ultimate goal. Subsequently, Internet commentators care more about recognition from above than actually persuading netizens. This explains why Internet commentators have made little effort to hide their distinguishing characteristics or engage other netizens in online discussion. My analysis shows that in addition to being constantly pro-regime and using official language codes, Internet commentators tend to work only during office time, often publicize a large number of consecutive comments in a short period of time, and focus only on tasks assigned to them. Even online commenting on President Xi Jinping—clearly a task with paramount political significance—is implemented in such a crude fashion that any savvy netizen will be able to see through these comments. Apparently, either the entire propaganda system is so incompetent that it does not realize it is doing a bad job, or it could not care less about the actual impact of its work. The latter is more likely considering the incentive structure: For Internet commentators (and propaganda officials), the primary audience is not netizens, but their superiors. In fact, a Hengyang Information Office official admitted that “pleading for achievements” (yaogong, 邀功) was one of his goals when mobilizing Internet commentators to interact with the local Party chief in an online interview.⁹³

The same rationale explains why the fifty-cent army has also been frequently exposed by state propaganda. In fact, the Party-state, which has repeatedly stressed the importance of strengthening its ideational leadership and maintaining social stability, is unabashed about its intention to control and manage public opinion. For local authorities and propaganda agencies, the fifty-cent army is a means of demonstrating their efforts to upper levels of government. This is why reports on recruiting, training, and rewarding Internet commentators often make their way to online news portals, as well as traditional media such as newspapers and TV channels, which are much more closely watched by the state. For instance, in December 2006, Shanxi Province sponsored its first training session for Internet editors and commentators. The local media outlet, Jincheng News Online, which is under the direct management of the Jincheng Propaganda Department, not only reported on this event proudly, but also included in its report links to other influential media platforms that covered the news.94 Apparently, for local propaganda officials, it is a significant accomplishment to get attention from central media such as People’s Daily Online (people.cn) and national commercial portals such as QQ.com and 163.com. There are also many other cases of mainstream media reporting on the work of Internet commentators. For example, Southern Metropolis Daily reported on Gansu Province’s initiative to hire 650 Internet commentators.⁹⁵ Similarly, when Guangdong Province planned to set up a new “Internet public opinion guidance troop” of ten thousand members, the news was covered by both Guangzhou Daily and the online news portal 163.com (Netease).⁹⁶ These examples demonstrate that the propaganda machinery does not intend to keep Internet commentators in the dark, which is clearly a legacy of past approaches to propaganda and thought work.

The fact that the state has failed to keep Internet commentators covert has made what had at first appeared to be a smart state adaptation ineffective. The exposure of the fifty-cent army has only contributed to a growing antipathy toward state propaganda and authoritarian rule, leading to the erosion, rather than consolidation, of the state’s ideational leadership. In particular, as netizens learn about the fifty-cent army, it is only natural for them to be doubtful about any pro-regime expression. This explains why voices in favor of the government have become “politically incorrect” among netizens and why any government connection in online expression is seen as an “instant negative.”97 As peer pressure distances netizens from pro-regime stances, potential government supporters are demoralized and silenced. As the Beijing Normal University professor Zhang Shengjun puts it, “Now, the fifty-cent army has become a baton waved toward all Chinese patriots.”⁹⁸ A flood of criticism has intensified the mutual distrust felt among netizens, exacerbating the labeling wars, as will be further discussed in chapters 6 and 7.

The evolution of the Internet commentator system also confirms the fragmented authoritarianism thesis that depicts the Chinese Party-state as divided across different levels and sectors.99 Begun as a local innovation and spread through imitation and mutual learning, the Internet commentator system was clearly not a policy push from the central government. Central coordination, which exists in some cases today, is still mostly absent, as Internet commentators are sponsored and mobilized by local authorities or specific state agencies for their own purposes. That being said, there are signs of increasing activity among Internet commentators in central-coordinated propaganda campaigns, especially since Xi Jinping came into power. The increased reliance on manufactured pro-regime voices online betrays the new administration’s profound sense of insecurity.

The state is not the only actor with the will and capacity to manipulate popular opinion online for political purposes. Other sociopolitical actors such as aggrieved netizens and dissident groups have also been found to employ PR tactics like astroturfing to spread their own voices. To better understand the discourse competition in Chinese cyberspace, nonstate actors’ attempts to manufacture discontent and the subsequent implications also deserve close examination. The next chapter explores how online opinion engineering efforts by regime critics have intensified netizens’ identity anxiety. Such anxiety fuels discourse wars among netizens, contributing to the rise of a nationalistic narrative that depicts regime critics, pro-liberal media, and foreign countries as national enemies trying to sabotage the rise of China. This explains why the erosion of state discourse has yet to be translated into support for regime-challenging discourse.

CCYLCC School Department				+Follow
2014-12-12 14:43 from 360 Safe Browser
#National Memorial Day Student Awareness and Action# December 13 is the first National Memorial Day for the Nanjing Massacre Victims. We hereby launch the Weibo Relay theme of “remembering the national calamities through a national memorial with three lines of oration to express heartfelt emotions.” Pals, please forward this Weibo post and write three lines of oration with your most sincere feelings to commemorate our deceased compatriots and memorialize the national history. For example: Tens of thousands of martyrs died on the battlefield, whose patriotic mission will never fade, as millions of compatriots commemorate their great deed.
Bookmark	Forwards: 210218	Comments: 31130	Likes: 3377

Topic: Implications for Xi Jinping’s first talk on the Four Comprehensives
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:48:38
The general secretary sets a good example for cadres across the country, making us ordinary citizens feel assured.
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:47:17
General Secretary Xi bears the people in mind and is a close friend of the people.
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:45:36
I believe President Xi will lead the people to make our motherland stronger and more prosperous!
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:44:13
General Secretary Xi’s resolution to fight corruption is the sweet spring that we the masses have been longing for. Firmly support General Secretary Xi.
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:42:55
The people’s yearning for a good and beautiful life is the goal for us to strive for.
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:41:53
General Secretary Xi Jinping is absolutely an excellent chairman. Every sentence of his speech is close to our hearts, and his actions also warm our hearts.
Netease user from Heilongjiang Shuangyashan ip：61.180..	2014-12-18 09:40:42
Our people love life. They expect better education, more stable jobs, better income, and more reliable social security.

	Comments Coded	Anonymous	Likely Internet Commentator	Likely Genuine	Unclear
Xi Jinping	1,133	193	786	140	14
	(100%)	(17.03%)	(69.37%)	(12.36%)	(1.24%)
Li Keqiang	446	238	58	138	12
	(100%)	(53.36%)	(13%)	(30.94%)	(2.69%)
Hu Jintao	292	103	9	145	35
	(100%)	(35.27%)	(3.08%)	(49.66%)	(11.99%)
Wen Jiabao	399	140	5	177	77
	(100%)	(35.09%)	(1.25%)	(44.36%)	(19.3%)

Province	Quota	Province	Quota
Beijing	20,000	Hubei	20,000
Tianjin	10,000	Hunan	15,000
Hebei	15,000	Guangdong	20,000
Shanxi	10,000	Guangxi	10,000
Inner Mongolia	7,000	Hainan	2,000
Liaoning	20,000	Sichuan	12,000
Jilin	10,000	Chongqing	10,000
Heilongjiang	10,000	Guizhou	6,000
Shanghai	10,000	Yunnan	8,000
Jiangsu	20,000	Tibet	1,500
Zhejiang	15,000	Shaanxi	15,000
Anhui	15,000	Gansu	6,000
Fujian	10,000	Qinghai	2,000
Jiangxi	10,000	Ningxia	2,000
Shandong	20,000	Xinjiang	3,000
Henan	15,000	Bingtuan^a	500
Total：350,000