Data, Methods and Analysis Strategies

To explore how local governments deal with unruly protesters under the legal government campaign, I spent three months in 2014 conducting ethnographic fieldwork in W City, Guangdong. The fieldwork was conducted right after the official launch of the campaign in November 2013, which, as I discuss below, proved to be a crucial turning point as the rhetoric of legal governance promptly spread throughout the propaganda system from 2014 to 2015. I gained access to a county-level bureau for letters and visits (xinfangju 信访局, BLV hereafter) in W City through personal connections. I chose a county-level BLV to conduct fieldwork because it is where protest issues are commonly contested and resolved, whereas higher-level BLVs mainly return petition cases to county-level BLVs after receiving complaints. During the fieldwork, I was allowed to observe officials as they received complainants and handled petition cases. I observed approximately 60 cases of contention. The time-limited observation did not allow me to observe the entire trajectory of the government response in these cases, as the resolution of protest cases following bureaucratic procedures can often take years and sometimes even more than a decade. However, the different levels of bureaucratic processing in these cases gave me the opportunity to gain insights into how local officials understand the role of bureaucratic processing in protest control and how they manage contentious activities before and after some bureaucratic practices. I conducted unstructured, informal interviews with six officials working in the BLV and typically asked them to provide more details after the observations. To examine whether local agents behave differently in more legally developed regions, I also conducted semi-structured interviews with five local officials working in district and subdistrict governments in Guangzhou between 2014 and 2015.

To examine the generality of my observations and to gather more empirical evidence for the theoretical explanation, I use legal repression cases collected from China Judgments Online to conduct sequential and statistical analyses. The collection includes 5,474 cases from across 30 provinces that were tried between 2014 and 2019. I trace the sequences of various events that unfold over time to explore how and when local authorities resort to criminal prosecutions across different cases. I also use some comparable cases as examples to conduct a statistical analysis to further examine the effect of bureaucratic processing on the choice of legal repression.

Employing both qualitative and quantitative data to conduct analysis helps to unpack the causal link between bureaucratic processing and government reaction in a politically ambiguous context and also maximizes the validity and generalization of the observations and explanations. The analysis is based on cases which did not go through the judicial process (qualitative data) as well as cases that did enter the judicial process (published court judgments) in order to reduce selection bias.

The Rise of the Legal Governance Campaign

Since 2013, the Chinese state has increasingly turned to legal approaches to sustain its rule. Although previous studies have explored how the Chinese government uses legal measures to suppress civil society organizations and activists,Footnote ²² little research has examined how such an approach is used in routine contention governance practice. In November 2013, the Chinese Communist Party (CCP) suggested that the government should “resolve social conflicts in line with the thought and approaches of law-based governance”Footnote ²³ and launched a campaign to manage popular contention in accordance with law. In line with the party-state rhetoric, the petition system (xinfang 信访), which functions as the official channel through which citizens can make claims, initiated a new reform in 2013. This reform stressed “petition in accordance with law” (fazhi xinfang 法治信访) and required local authorities to use legal means to suppress petition activities that violated law and order.

Under the campaign for law-based governance, legal repression has increasingly been used to deal with social protests. Legal repression is a repressive strategy that uses the criminal justice system to suppress protesters who “violate the law.”Footnote ²⁴ As Yongshun Cai points out, it is easy for aggrieved citizens to breach regulations and laws when staging protest activities; moreover, whether or not they have crossed the line mainly depends on the local government's interpretation of that line. Protesters might receive criminal penalties even if they did not cause serious damage.Footnote ²⁵ Although local authorities suppress protest activities by claiming that protestors are breaking the law, legal repression is a collectively coordinated action by the local government rather than a single act carried out by a law enforcement agency. The decision to use criminal law to target protestors is usually made by local government leaders before the launch of criminal justice procedures. For example, in a legal repression case in Henan, leaders of the Committee of Political and Legal Affairs at the county government called a special meeting with various related agencies to discuss a petition case and decided to deploy legal repression. After the decision, the protester was arrested by the police, prosecuted by the people's procuratorate and was eventually convicted by a court.Footnote ²⁶ Another court decision in a legal repression case in Hubei follows the same line.Footnote ²⁷ Once protesters are prosecuted, it is rare, if not impossible, for them to successfully defend themselves and be declared not guilty in the court.Footnote ²⁸

The use of legal repression to manage unruly protesters is not unique to the Xi era; however, compared with previous leaderships, such practices have fundamentally changed under Xi's rule. Whereas acts of legally suppressing protest activities were mainly sporadic under previous administrations, this legal governance campaign has been waged nationwide by the CCP, from top to bottom. Legal repression has increasingly become a crucial measure in the local government's toolkit for dealing with social protests. Figure 1 shows that the number of newspapers publishing articles containing discourse “attacking illegal petitions” significantly increased in 2014 and 2015.Footnote ²⁹ These changes demonstrate that legal control has become an abiding principle by which state agents handle state–society contention in the Xi era; this is in stark contrast to the practice of “constructing [a] harmonious society” under the Hu–Wen administration.

Figure 1. Number of Newspapers Containing Discourse “Attacking Illegal Petitions”

Political Ambiguity in Contention Governance and Cautious Government Reactions

While the existing literature suggests that legal approaches enable state officials to deploy a more consolidated form of repression, this study reveals the political constraints faced by state authorities when legally suppressing popular contention and the consequent cautious reactions of local governments. Ambiguous principles, norms and signals regarding contention governance can impose considerable pressure on local authorities and prevent them from employing arbitrary repressive methods, even though the laws and regulations and the prevailing rhetoric of law-based governance provide a convenient basis upon which authorities can legitimize their coercive actions. The literature on contentious politics has underscored that ambivalent signals sent by the authoritarian state can contain activists’ resistance by encouraging self-censorship.Footnote ³⁰ However, this study argues that ambivalent norms and signals exist not only in the protester's struggle with the state but also in the state's management of popular contention.

Political ambiguity in contention governance under authoritarian legality exists for several reasons. Partly, it is caused by the tension between populist rule and law and order. Although President Xi has launched legal governance campaigns since he took office, he has not detracted from populist rule, and he significantly highlighted the importance of the “mass line” when he provided instructions on handling petitions:

Populist rule emphasizes that authorities should be responsive to public opinion, grievances and resistance, whereas law and order suggests that officials need to regulate protest activities through formal rationality and punish unlawful actions to protect the legal authority.Footnote ³² This inherent tension created ambiguity around the enforcement criteria used by local officials to handle protest activities. For instance, at the central conference on political and legal work held on 7 January 2014, President Xi underlined the importance of the “three satisfactions and one disposal” (san daowei yi chuli 三到位一处理) in dealing with petition activities: satisfactorily solve problems in legal and reasonable cases, satisfactorily carry out thought work in unreasonable cases, satisfactorily offer assistance in cases related to life difficulties, and punish lawbreaking cases in line with the law. This guidance was promptly adopted within the petition system.Footnote ³³

Political ambiguity is also a result of the ruling elites’ dilemma of how to balance political control and regime stability. In a regime with multiple hierarchies, one of the most important tasks for top leaders is to achieve effective political control over lower-level agents.Footnote ³⁴ Thus, the Chinese central government has a very ambivalent attitude towards petitioning activities. On the one hand, the central government expects local authorities to control petitioning to Beijing to some extent and requires them to legally suppress unruly petitioners. This is because the centre does not directly handle those petition cases and fears that a wave of petitions to Beijing would affect political stability.Footnote ³⁵ On the other hand, the central government hopes to learn about local conditions and achieve political control over local agents by receiving aggrieved citizens’ complaints.Footnote ³⁶ Thus, the central government also emphasizes “resolutely putting an end to all wrong practices of interpreting normal petition activities and of restricting petitioners’ personal freedom.”Footnote ³⁷

These ambiguous norms and signals have placed local authorities under considerable legitimacy and political pressures and so they tend to respond cautiously to avoid potential risks, even though protest activities may break the law and affect stability. During my fieldwork in January 2014, for example, approximately 20 ex-soldiers were unhappy about their resettlement arrangements and so staged a collective sit-in and petitioned in front of the county government for several days. They unrolled a huge banner in front of the government gate declaring, “Once we were the most respectable people, but now we are the most miserable people.” Rather than dispatching police to dispel these petitioners, local government officials in the BLV attempted to persuade them to go down the regular petition routes: “If you have any complaints or grievances, you can select not more than five representatives and come to the BLV to make claims. If your problems are reasonable, the government will try its best to solve your problems,” an official in the BLV told them. The veterans responded by demanding to meet the county leaders: “let the leaders come out and meet us!” Instead, the petition officials attempted to persuade the veterans to follow normal petition procedures: “Of course, you can meet the leaders. You just need to elect five representatives and then go in and meet the leaders.” However, the veterans did not accept this response and insisted that the county leaders come to meet them. After three days of sit-in protests, the local authority eventually compromised. A local government leader in charge of veteran affairs came to persuade the veterans face to face. This leader first emphasized that the veterans’ actions were indeed against the law, but he also promised to address their problems. After the face-to-face communication, the veterans agreed to end their sit-in and follow the normal petition procedures to resolve their grievances.

A similar episode also occurred with a small-scale petition case over a land dispute. During a meeting of the county people's congress, four villagers unfurled a banner in front of the meeting hall to draw attention to the land dispute. After being informed of this accident, the BLV leader and some officials quickly went to the site and took the villagers back to the office. The leader did not accuse these complainants of violating the regulations and rules; instead, he politely explained to them that they should follow the normal procedures. At the same time, the leader promised to try to resolve their problem and called the village cadres and relevant parties in for negotiations.

The prevailing populist rule makes the local government believe that the publicness of troublemaking activities may stain its political image and cause public distrust and discontent, damaging its legitimacy. As an official in the BLV in W City explained:

When I pressed this official on why the local government did not use the law to punish and prevent petitioners’ activities, he responded:

A local official in Guangzhou also acknowledged the pressures on the local government's legitimacy: “It is not easy for us [to handle popular resistance]. We need to be cautious. Ordinary citizens will not consider these [protest activities] as protesters’ responsibilities or faults. They do not think in that way. They would point the finger at the government and blame the government.”Footnote ⁴⁰ Similarly, in a Yingshan 英山 county (Hubei) petition working report, the deputy director of the BLV noted “the contradiction between the pressure of public opinion and the lack of legal punishment.” He suggested that the local authority had substantial concerns and fears that punishing troublemakers would draw the attention of the public and stir up new social conflicts.Footnote ⁴¹

Given the higher-level government's ambivalent attitude towards petitioning activities, the political pressure on local governments is also a crucial factor that shapes their calculations when dealing with popular resistance. Under the “delegated” model, local states have considerable autonomy in how they deal with popular contention yet still come under political pressure from higher-level governments.Footnote ⁴² Once protesters take their claims directly to the higher authorities, or threaten to do so through petition channels, local authorities will worry that the protest events will attract attention from above and invite political intervention. The divisions and differentiations within the political system as well as the higher-level power holders’ need to monitor lower-level agents create a space for aggrieved citizens to gain support from higher-level governments.Footnote ⁴³ Additional political pressure stems from the assessment requirements imposed on local officials under the cadre responsibility system. Local state officials’ performance in stability maintenance, especially in controlling the amount of “non-normal petitioning” to higher-level governments, functions as a “hard target” in cadre evaluations.Footnote ⁴⁴

In my ethnographic research in W City, I witnessed a case that illustrates how political pressure leads local governments to buy stability and how such compromises encourage further troublemaking behaviour. In 2010, some farmland in a village was flooded to make way for a dam project. Three villagers made claims for damages to the local authority. Initially, the county government gave them more than 10,000 yuan in compensation, and the villagers signed a “letter of assurance” (xifang basu baozhengshu 息访罢诉保证书) promising not to lodge any further petitions on this issue. However, the villagers lodged a second round of petitions on the same issue in 2012. The local government was furious with these villagers and refused their demands. During the standoff, the villagers filed a complaint with the provincial government, which transferred the case back to the local government and demanded that it deal with the problem. Finally, the local authority gave the villagers another 7,000 yuan in compensation. A petition official acknowledged that the pressure from the higher-level government forced the local authority to compromise with the villagers and offer them more compensation:

However, the story does not end there. The villagers lodged a petition for a third time in late 2013. The local government was fed up with these villagers and refused to compensate them again. Seeing this, the villagers repeated their previous tactics. They continued to make claims to the BLV and threatened to lodge a petition with the provincial government again and even to take their petition to the central government in Beijing. An official in the BLV recognized that such persistent petition activities were incentivized by the local authority's strategy of buying stability, as he told me: “They promised that they would stop petitioning after compensation and signed the letter of assurance, but they are coming again. Now that they know they can get money from the government, they treat the government as a cash cow and come to petition once they have no money left to spend.”Footnote ⁴⁶ Even if these petitioners continued to break the rules, this local official acknowledged that there was nothing the local government could do: “There is nothing we can do if they come to petition; we just ignore them and let them sit here.”Footnote ⁴⁷

Bureaucratic Processing, Insulation from Pressure and Progressive Legal Repression

Although the prevailing rhetoric of law-based governance provides a basis from which local governments can legitimize their coercive actions, owing to the political ambiguity in the realm of contention governance, they are regularly cautious in their management of and response to contention and engage in progressive legal repression. This section further illustrates the elements of progressive legal repression and analyses how and why bureaucratic processing allows local officials to overcome political uncertainty and reduces the risks of formally using the criminal law to prosecute protestors in a vague political environment.

Generally, there are three main elements to progressive legal repression: inducing unruly protesters to engage with available bureaucratic procedures, utilizing bureaucratic procedures to reconceptualize protesters’ claims and behaviour as unreasonable and signal the fulfilment of government responsibilities, and applying criminal sanctions to suppress protesters after legitimacy and political concerns are partly dealt with. Local governments do not tend to initially apply criminal proceedings against protestors when illegal petition events occur unless the protesters’ behaviour is very disruptive and violent. Local officials are more likely to use soft tactics to placate protesters and encourage them to engage with the available legal-bureaucratic procedures to resolve their issues. As an official in the BLV in W City mentioned:

Previous literature has noted that the Chinese state uses various legal-bureaucratic institutions and procedures to absorb popular resistance.Footnote ⁴⁹ There are official procedures within the xinfang system that enable local governments to deal with citizen complaints, including investigation, resolution, mediation, public hearings, re-examination and review, and termination, although there are various different types of petitions and protests, and specific practices might vary. Once complainants decide to go down the official route, the legal-bureaucratic game begins. As Ching Kwan Lee and Yonghong Zhang find, aggrieved citizens are willing to participate in the game not only because of the legitimacy implied by bureaucratic rules and procedures but also because there is a real chance of winning the game.Footnote ⁵⁰ If complainants were unable to follow the official procedures, petition officials would attempt to guide them through the entire bureaucratic process. During my fieldwork, I became familiar with a case regarding a pension problem, which gave me an insight into the guidance offered by local officials. The petitioner was the employee of a construction company in W City. He felt unhappy about the local government's response to his complaint and so lodged petitions with the BLV several times. The head of the BLV advised the officials in the reception area: “you can ask him to apply for a re-examination and review and try to guide him go through the whole process. The procedures of re-examination and review are prescribed in the petition regulations. If you guide them to complete all these procedures, our responsibility is fulfilled. When he comes here next time, it will not be our responsibility.”Footnote ⁵¹

While the extant literature focuses on how bureaucratic institutions, rules and procedures demobilize popular resistance by channelling contentious activities and consuming protesters’ time, money and energy,Footnote ⁵² progressive legal repression emphasizes the role of bureaucratic processing in overcoming the constraints faced by state officials and shaping government reactions. Local governments are regularly required to satisfactorily address petitioners’ problems in legal and reasonable cases. However, as discussed above, owing to the political ambiguity in contention management, the tension between populist rule and law and order, and the state's balancing act between political control and regime stability, the lines between the legal and the illegal and between the reasonable and the unreasonable become blurred. Bureaucratic processing helps state officials to overcome such political uncertainty and lowers the potential risks of formally applying criminal proceedings through two mechanisms. First, once the petition cases have been through the legal-bureaucratic route, the procedure serves as a crucial institutional platform which helps local governments reinforce their discursive power by allowing officials to reconceptualize protesters’ claims and behaviour as unreasonable. Local governments can use the knowledge they have gained from passing the details of the complaint through the bureaucratic procedures to frame convincing and reasonable depictions, explanations and arguments for or against the complaint. As a result, local authorities have a dominant role in interpreting the meanings of protesters’ demands and behaviour and the roots of social conflicts. Complainants’ persistent contentious activities are often described as “illegal petitioning,” “unreasonable petitioning” or “disruptive petitioning,” and the causes of the contention may be reframed to show the protesters as at fault. By following this course, local authorities not only delegitimize petitioners’ protest activities but also legitimize further disciplinary steps taken against protestors. Second, by performing these bureaucratic practices, local governments can demonstrate that they are fulfilling their responsibilities by correctly following official procedures. All bureaucratic procedures are regulated and authorized by the state. When faced with inspections by the public and higher authorities, the performance of bureaucratic practices and the ensuing paperwork also can demonstrate the government's responsiveness to complainants – “we have made a strong effort to resolve petitioners’ demands” – no matter whether local authorities have substantively satisfied protesters’ demands or have merely paid lip service to procedures for evaluation purposes or to present an image of good management.

Such reconceptualization and signalling practices can help to insulate state officials, at least to some extent, from the legitimacy and political pressures that they face and so encourage them to use legal repression. During my fieldwork, I observed a laid-off state worker who continued to come to the BLV even after his case had been re-examined and reviewed. The following dialogue between the petitioner and the BLV official highlights the effects of bureaucratic processing:

Bureaucratic practices can effectively help to overcome the political constraints faced by local authorities when information asymmetry occurs within government hierarchies and bureaucratic practices matter for the lower-level government's discourse domination and the higher-level government's scrutiny. In China, given the principle of “territorial jurisdiction” (shudi guanli 属地管理), higher-level governments shift the main responsibility of contention management to lower-level authorities and rely heavily on local officials’ reports for knowledge of local conditions. However, local state officials tend to protect their positions and, thus, do not tell the whole truth to their superior authorities, leaving higher-level power holders ignorant of the problems and realities at the lower level.Footnote ⁵⁴ Although higher-level governments sometimes conduct field investigations into petition cases at the local level, these investigations are highly selective and usually conducted within a campaign. According to public announcements on the website of the State Bureau for Letters and Visits (guojia xinfangju 国家信访局, SBLV hereafter), 414 petition cases, which is just a fraction of the total number of national cases, were investigated by the central agencies from December 2014 to November 2019.Footnote ⁵⁵ Therefore, local authorities play a dominant role when interpreting the meanings of protesters’ demands and behaviour after bureaucratic processing.

If protesters can obtain support from the upper authorities, local governments might still be faced with the possibility of disciplinary action, despite framing complainants’ claims and behaviour as unreasonable and signalling that they have done their duty. However, the higher-level governments’ selective scrutiny strategy provides an opening for local authorities to decrease political pressure after bureaucratic processing. When information asymmetry occurs and upper authorities have limited capacity to fill this information gap, screening protest cases through bureaucratic processing and then adopting a differentiated attitude become a low-cost strategy for higher-level rulers to balance political control and regime stability. More specifically, higher-level governments still present opportunities for protest events that have yet to go through legal-bureaucratic procedures but are less likely to support veteran cases that have been processed and terminated by some bureaucratic practices. This selective and differentiated attitude and strategy can be traced to the central government's significant effort to emphasize the termination of petition cases. The Xi administration has further highlighted closing the window for veteran petitioners. In late 2013, the SBLV further clarified the petition termination procedures and underlined that petition bureaus should no longer accept petition cases that have been concluded.Footnote ⁵⁶

Further Analysis Based on Legal Repression Cases

To examine the generality of the theory derived from the fieldwork and to better consider other competing explanations of the government's response, this section uses legal repression cases collected from China Judgments Online to conduct a sequential comparison and statistical analysis.

After collecting 5,474 cases from China Judgments Online, I used 3,102 cases to conduct sequential analysis.Footnote ⁵⁷ This study distinguishes two patterns of legal repression: direct legal repression (“protest event → legal repression”) and progressive legal repression (“protest event → bureaucratic processing→ protest event →legal repression” or “protest event → bureaucratic processing →legal repression”). Table 1 examines how the attributes of protest events (before the bureaucratic processing) affect whether a case enters the bureaucratic process or experiences direct legal repression. As Table 1 demonstrates, the pattern of progressive legal repression is more suitable when local authorities deal with boundary-pushing protest activities. Progressive legal repression was used in almost all of the 2,569 cases of troublemaking protests targeted at lower-level and higher-level governments. In contrast, among the 83 cases with violence or collective violence, only 9 cases were legally repressed after bureaucratic processing. Interestingly, even if protesters stage collective resistance against higher-level governments, they might not be directly repressed by local authorities through criminal proceedings. Among 78 cases of collective petitions at high-level governments, 55 cases fit the mode of progressive legal repression. Of the 372 cases of collective actions at the local level, progressive legal repression was employed in 149 cases.

Table 1. Protest Attributes and the Pattern of Legal Repression

Notes: Cases involving more than five people were considered collective resistance, and cases mentioning property damage and injury were coded as violent resistance. Higher-level governments refer to provincial governments and the central government.

To further examine the effect of bureaucratic processing on local governments’ responses, I selected some cases of troublemaking petitions lodged in Beijing as examples to conduct statistical analysis. The reasons for this selection and the criteria for including a case can be found in Appendix B (online). I used 351 cases for analysis. The immediate employment of legal repression after some bureaucratic practices can indicate the effect of bureaucratic processing, given that the theory suggests that bureaucratic practices can help local authorities to reduce potential risks and encourage them to bring criminal charges against the protestors. I used logistic regression to examine whether the level of bureaucratic processing and other factors affect the likelihood of local governments employing immediate legal repression following the bureaucratic process. If local governments are more likely to deploy prompt legal repression after the bureaucratic process has been sufficiently exhausted, relative to limited processing, this can provide empirical evidence for the theory.

The dependent variable is a dummy indicating whether local governments deploy immediate legal repression after bureaucratic procedures. It is to be noted that not all of the cases experienced prompt legal repression after some bureaucratic procedures, and variations exist in how immediately the local government employed legal repression after bureaucratic processing.Footnote ⁵⁸ In model 1, a protest case was coded “1” if (1) protesters staged non-normal petitions in Beijing several times before the final bureaucratic procedure, and (2) protesters only staged non-normal or normal petitioning in Beijing 0-1 time after the last bureaucratic procedure and then experienced immediate legal repression. Other cases were coded as “0.” In all, 129 cases (36.8 per cent) were recorded as “1.”Footnote ⁵⁹ This measurement of the dependent variable might run the risk of overlooking the effect of a protestor verbally threatening to lodge a petition in Beijing and those actual but unsuccessful petitions to Beijing which are intercepted by the local authorities. To reduce the impact of this problem, an alternative measurement of dependent variable is used in model 2. In addition to meeting the two criteria mentioned above, only those cases in which local governments deployed legal repression within 95 days after the final bureaucratic process were coded as “1” (the mean time interval of all immediate legal repression cases in model 1 is 190 days). The independent variable is the level of bureaucratic processing, which was coded into three categories: limited processing, sufficient processing and terminated processing. The control variables include the type of demand, period of time, government concession, the nature of the crime, protest intensity and region. Please refer to Appendix C (online) for the detailed measurement of the independent variable and control variables.

Table 2 presents the results of the logistic regression analysis on the local government's choice of legal repression. In model 1, the findings show the positive effect of the level of bureaucratic processing on the use of legal repression, which provides empirical support for the main argument in this study. More specifically, after controlling for other variables, for cases with sufficient processing, the odds of immediately deploying legal repression following bureaucratic procedures are 288 per cent higher than for cases with limited processing. For cases with terminated processing, the likelihood of immediately using legal repression after bureaucratic procedures is 328 per cent higher than for cases with limited processing. The results also show that the decision on legal repression is shaped by the type of demand. All other factors being equal, local officials are more likely to immediately carry out legal repression in citizen–government conflict cases, relative to civil conflict cases. This finding is consistent with the cost-benefit explanation in previous studies.Footnote ⁶⁰ In addition, local governments increasingly deploy immediate legal repression after bureaucratic processing over time. For example, the odds of immediately using legal repression after bureaucratic processing in cases in which the last bureaucratic procedure was performed in 2018 is 1,776 per cent higher than for cases in which the last bureaucratic procedure was performed in 2014. In model 2, even if an alternative measurement of the dependent variable is used, the positive effect of the level of bureaucratic processing is still statistically significant.

Table 2. Effects of Bureaucratic Processing on Immediate Legal Repression

Notes: *** p<0.001, ** p<0.01, * p<0.05