“Surveillance with Chinese Characteristics”: The Development & Global Export of Chinese Policing Technology

Sheena Chestnut Greitens1 Assistant Professor, University of Missouri

Paper prepared for Princeton University IR Colloquium Fall 2019

This version: 30 September 2019

DRAFT: Please do not circulate without author’s permission.

1 The author thanks the National Asia Research Program for financial support, the Igarape Institute for use of their Homicide Monitor data, and Adriana Boersner, Steven Yunpeng Bai, Edward Goldring, Yuan Yuan Hsu, Yu Bin Kim, and Emir Yazici for excellent research assistance. Thanks also for helpful feedback from attendees at NBR’s 2019 Asia Policy Assembly, CNAS/Bridging the Gap’s “Digital Authoritarianism” workshop, and APSA 2019, as well as Benjamin Della Rocca, Vanya Krieckhaus, Bryce Reeder, Christopher Sullivan, Laron Williams, and Joseph Wright.

1

Media and policy discussions of late have been chock-full of discussions about the use of

Chinese surveillance and policing technology for social control. Reports by policy analysts warn of

the dangers of high-tech surveillance and artificial-intelligence (AI)-based approaches to policing in

autocrats’ hands, 2 and journalists document the rise of a “dystopian surveillance state” inside China

– especially in the far western region of Xinjiang, where Turkic Muslim minorities, particularly

Uyghurs, have been subject to mass detention and involuntary ideological re-education in the name

of counterterrorism.3 Globally, China’s export of these surveillance and policing technologies has

also received increasing media and policy scrutiny. News organizations have provided detailed

accounts of the spread and use of surveillance technology from China to places like Venezuela,

Ecuador, Zimbabwe, and Uganda,4 and Huawei, the largest supplier of these kinds of platforms, is at

the center of debate over U.S.-China trade and technological competition.5 At a U.S. House of

Representatives hearing in May 2019 on “China’s Digital Authoritarianism,” both the chairman and

the ranking member warned that the export of these technologies would give “countries the

technological tools they need to emulate Beijing’s model of social and political control.”6

This project – which is still in early stages – examines the domestic development and global

export of Chinese surveillance and policing technologies. Domestically, it situates its findings in the

rapidly-evolving literature on information and authoritarian control, building on recent work that

shows that authoritarian regimes use information to make repression more precise and selective, and

that this repression is often preventive rather than reactive.7 Here, the project’s main contribution is

to specify theoretically why information is useful for preventive repression: it enables autocrats to

identify not just the overall distribution of concealed preferences among citizens, but to match

2 Fontaine and Frederick 2019.

3 Lucas and Feng 2018; Mozur 2018; Buckley et al 2019; Chin and Burge 2017; Wang 2019. 4 Joplin 2018; Shahbaz 2018; Cave et al 2019; Daw 2019; Feldstein 2019b; Parkinson et al 2019.

5 “U.S. Unlikely to Extend Waiver for U.S. Firms to Supply China’s Huawei,” Reuters, 26 September 2019. 6 U.S. House of Representatives 2019.

7 Pierskalla 2010; Nordas and Danveport 2013; Danneman and Ritter 2014; Sullivan 2015; Ritter and Conrad 2016; Greitens 2016; Blaydes 2017; Truex 2018; Dragu and Przeworski 2018; DeJaegher and Hoyer 2018. .

2

citizens to specific points on that distribution. Information also facilitates strategy selection: it

informs autocrats of the level but also the type of grievance, which allows regime agents to engage in

differentiated demobilization: a process of selecting and applying strategies that will convince

potential dissenters to stand down at the lowest possible risk to the regime. In many cases, what

information facilitates is not repression but coercion: the threat rather than the outright use of

punishment; it is an important factor shaping the extent to which an authoritarian regime mixes the

approaches of concession, coercion, and physical punishment, and is especially helpful for

understanding how authoritarian regimes think about the legibility of populations under conditions

of relative ethnic homogeneity.

Moreover, the rise of China’s surveillance state has had international as well as domestic

consequences. Evidence from a new dataset, discussed in the final section below, suggests that

policymakers have significantly understated the number of countries in which Chinese surveillance

and policing technologies are used, as well as the speed with which the use of this technology has

spread. It is now in use in at least 80 countries, both democratic and autocratic, on every continent

except Australia. China, therefore, can be thought of as the ‘primary’ or ‘index case’ from which a

new set of informational and technological conditions has spread across the world, impacting

political life in a wide range of political systems and geographic contexts.8

This paper proceeds in five main sections. In the first, I sketch a nascent theory of

authoritarian surveillance, identifying what problems surveillance can solve and how information can

be (but is not always) translated into differentiated demobilization that limits physical repression,

atomizes citizens, and avoid the potential for collective backlash. The second section outlines the

context in which China developed its surveillance and social control systems, drawing on new

sources to explain how the CCP sought to solve information problems at both the national and the

local level. The third section provides qualitative evidence of the theorized mechanisms, showing

8 In epidemiology, the “primary case” is the first case that introduces a condition to a population, while the “index case” is the first documented case; a case may be both primary and index. In political science, Finkel (2015) refers to the Holocaust as an index case for the study of genocide and repression; see also King 2012.

3

that local leaders in China explicitly seek information for the purposes of preventive repression and

use it to engage in differentiated demobilization. The fourth section turns to international effects,

documenting the spread of Chinese surveillance and policing technology and discussing potential

drivers of this diffusion. The fifth section concludes by reflecting on the project’s theoretical and

policy implications, and by highlighting necessary directions for future research.

From Information to Demobilization: Sketching a Theory of Authoritarian Surveillance

Recent scholarship has emphasized that information is a key component of coercive capacity

under authoritarianism.9 Although popular mobilization is not the only way that autocrats lose

power,10 protest and contentious collective action occur frequently under authoritarian rule, and can

spark authoritarian breakdown and democratization, or even revolution.11 In China, protests are a

common feature of political life; though the majority are not regime-threatening, scholars agree that

the Chinese political system takes them seriously and has put immense effort into developing

flexible tactics and strategies to channel, contain, and when necessary suppress, organized

contention and collective action. 12 This section theorizes the role that surveillance plays in China’s

efforts to demobilize collective action from within the population, with the hope of explaining why

9 In the pages that follow, therefore, coercive capacity specifically refers to the ability to deal with the threat of collective action or of mobilization (the latter of which is typically understood as a sustained and organized form of collective action that poses a more serious challenge to authorities or regime elites than individual episodes of collective action). Tarrow 1996; McAdam, Tarrow, and Tilly 2003; Tilly and Wood 2015. 10 Svolik 2012 finds that coup or insider removal is the most common way for autocrats to lose power.

11 Friedrich and Brzezinski 1965; Arendt 1973; Tilly 1978; Marwell and Oliver 1993; Bratton and Van de Walle 1997; Geddes 1999; Bunce 2003; Boix 2003; McAdam, Tarrow, and Tilly 2003; Acemoglu and Robinson 2001, 2006; Ulfelder 2005; Beissinger 2007; Bueno de Mesquita and Smith 2009, 2010; Chenoweth and Stephan 2011; Fu 2017; Geddes, Wright, and Frantz 2018. 12 Guo 2001; Perry 2002; O’Brien 2008; Elfstrom 2019; Mattingly forthcoming (2019). Wong 2012 provides an estimate of 180,000 mass incidents per year; Goebel and Ong 2012 believe it is somewhat higher. On accuracy, see Freeman 2010. See also O’Brien and Li 2006; Lee 2007; Perry 2009; Lianjiang Li 2008, 2013; O’Brien 2008; Wasserstrom 2009; Li, Liu, and O’Brien 2012; Florini, Lai, and Tan 2012; Heilmann and Perry 2011.

4

the Chinese Communist Party (CCP) has invested so much – financially, organizationally, politically

– into the informational components of its “social governance” regime.

The role of information in policing differs in autocracies versus in democratic political

systems.13 In either system, effective policing relies on information that comes from the citizenry,

but the relationship between information and political authority diverges. Under democracy, where

citizens are principals directing police agents, coercive agents take action if it is generated by and

guided by citizens. Police begin investigations based on citizen complaints (information), and, in

combination with other actors in a criminal justice system, use additional information from citizens

(provided in their roles as victims, witnesses, and sometimes suspects) to investigate, make arrests,

and prosecute crimes. 14 Democratic citizens, therefore, are simultaneously principals in control of

the police and subjects who are vulnerable to the authority they have granted police, a duality that

constrains the power of coercive agents.

Under autocracy, where the regime is the principal directing a coercive agent, and citizens are

primarily subject to police power, the dynamics of information provision change. Regime rather than

citizen demand initiates police action, and that action is directed against citizens who do not have a

dual role as principals. In this context, information must be obtained from citizens whether or not it

will benefit or protect them, and sometimes when it will harm them to advance the interests of the

regime. Citizens therefore do not always provide information voluntarily, especially types of

information that may trigger the use of coercive agents against them. Moreover, if citizens believe

that there are divergent interests within the regime (or between the regime and agents), they may

13 The following paragraph is, of course, an oversimplification of the principles of democratic policing. The model I describe in this paragraph should be thought of as an ideal type rather than an empirical reality. 14 Black 1972; Sklansky 2007.

5

offer more/less information, or different information, to different parts of the state (or party-state)

apparatus.15

The current consensus in comparative politics is that non-democracies face significant

information problems. Ordinary citizens falsify preferences to avoid being targets of state coercion,

and rumors proliferate in the absence of trust in official statements.16 Ethnically exclusive regimes

may have difficulty obtaining information from areas dominated by outside ethnic groups, as

happened both in Uganda and in Iraq under Saddam Hussein.17 Free media, which can be a source

of reliable and useful information for ruling elites, is often deemed too risky for autocrats to allow,

because it can generate shared knowledge of regime failures, facilitate anti-regime coordination, and

catalyze external support for opposition.18 In China, despite some commercialization of media and

some investigative or narrative journalism, media outlets remain under relatively tight CCP control.19

Misrepresentation and information problems occur regularly within authoritarian regimes as

well. Subordinates (including those closest to the autocrat) have incentives to misrepresent

information on citizen dissatisfaction, especially if doing so will have negative career consequences.20

Leaders who fear their own security forces may pursue coup-proofing strategies that handicap those

organizations’ ability to gather accurate intelligence on popular opposition, or to convey it to the

actors who can do something about those threats.21 Thus regimes often face difficulty collecting

accurate information, and even if they do, may have organizational handicaps that prevent that

15 As Dragu and Przeworski (2018) emphasize, among others, regime leaders and coercive agents may have divergent interests.

16 Kuran 1995; Chen and Xu 2017; Huang 2017; Wintrobe 1998; Charron and Lapuente 2011; Egorov, Guriev, and Sonin 2009.

17 Lewis 2017; Lewis and Larson 2018; Blaydes 2018. 18 Tilly 1978; Persson and Tabellini 2006; Egorov, Guriev, and Sonin 2009; Chwe 2013.

19 Lorentzen 2013; Stockmann 2013; Repnikova 2017; Jaros and Pan 2017; Qin, Stromberg, and Wu 2016. 20 Kapuscinski 1983; Rosen 2006; Woods et al 2006; Gao 2016; Pan and Chen 2018; Dragu and Przeworksi 2018. 21 Greitens 2016; Talmadge 2015.

6

information from travelling across the bureaucracy to inform action. This generates variation in the

ability to use preventive coercion: regimes that resolve more of these informational problems will

have a comparative advantage in identifying potential instability in time to address it effectively,

relative to those where information problems are frequent and endemic.

Faced with the constraint of limited information provision from citizens – which varies

across different temporal, national, and subnational contexts – autocracies have several options.

They can forego information (potentially costly to both agents and their principals); they can forcibly

extract information (which is costlier for agents and risks inaccuracy for principals); or they can

persuade citizens to supply information voluntarily (possible, though difficult). Alternatively,

coercive agents can attempt to obtain information from sources that can be collected without

human interaction and that human providers cannot falsify. The appeal of technologically-based

surveillance mechanisms is therefore partly in the fact that they do not require direct human

interaction – but as yet, we know relatively little about how effective these sources can be in meeting

regime informational needs absent context provided from human sources.

In what contexts, and for what exact reasons, do authoritarian coercive institutions seek

information on citizens? To begin, consider the process by which public collective action

materializes from individual dissent, represented in Figure 1, below:

Figure 1: From Private Individual Dissent to Public Collective Action

Individuals within a society under authoritarian rule have a range of opinions and preferences.

People’s true opinions, however, especially of the regime, are often private and concealed, a

Public Collective Action

Semi-Private Aggregation

Individual Dissent

7

phenomenon that Kuran calls “preference falsification.”22 If their preferences include some level of

dissent, they may engage in small-scale contention that remains individualized and private, whether

these are everyday acts of resistance that James Scott terms the “weapons of the weak,”23 or

oppositional rhetoric, humor, and cultural behavior that fall short of collective action.24 Individual

participants are drawn into collective action, therefore, by “mobilizing structures” – organizations,

networks, or associations that have the ability to gather and organize the people, resources, and

strategy necessary to turn a group into a movement or sustained challenge.25 These aggregating

structures may recruit participants, communicate knowledge to create shared understanding, shape

group identity around movement goals and objectives, teach tactics or some combination of all of

these functions. Mobilizing structures can be formal organizations, but in authoritarian political

settings, they are often informal networks – both in China26 and in other comparative contexts.27

If individual dissent is private and concealed, and collective action is public and visible, the

aggregation process that occurs via mobilizing structures lies somewhere between the two. The fact

that aggregating processes are collective means that they are not wholly private, but they are also not

formally visible, especially to the authorities, in the way that collective action must be to make claims

and achieve goals. This stage is also socially collective, in that people are communicating and

forming a shared identity and structure, but not necessarily physically collective in the way that a

protest or other contentious action (often) involves participants gathering in the same location. This

semi-private aggregation stage can result either in public collective action, or in “disguised collective

22 Kuran 1995; Wintrobe 1998.

23 Scott 1985, 1987. 24 Wedeen 1999; Thornton 2002; Johnston 2005.

25 Zald and McCarthy 1987; Snow and Benford 1988; McAdam et al 2001; Tarrow 2011; Tilly and Tarrow 2015.

26 Zhou 1993; Wright 2001; Lee 2007; Vala and O’Brien 2008; Hurst 2009; Yang 2009; Becker 2012; Fu 2018; Byler 2019.

27 Denouex 1993; DellaPorta 1995; Wiktorowicz 2003; Singerman 2004; Beinin and Vairel 2011; Bayat 2013; Fu 2018.

8

action” that has been organized collectively, but is carried out individually (in public) to increase

chances of success while lowering the risks of state repression.28

Responsive repression and preventive repression target different parts of the process shown

in Figure 1. The types of repressive activities that the field has (until recently) focused on, especially

those measured by military spending or personnel, are violent crackdowns on citizens who have

physically assembled for public collective claims-making, and is therefore reactive to contentious

action. Preventive repression, by contrast, targets the earlier stages of the process, when activity is

still private or semi-private.29 Figure 2 depicts this difference.

Figure 2: Preventive vs. Responsive Repression

The key difference between responsive and preventive repression is in the stage of mobilization that

they target, and the nature of the activities that take place during that part of the mobilization

process. If preventive repression seeks to stop individuals from proceeding to public collective

action, it must by definition intervene while beliefs and behavior are still private or semi-private.

28 Fu 2017, 2018. 29 Of course, the deterrent component to military manpower and deployments is partially preventive.

Public Collective Action

Semi-Private Aggregation

Private Individual Dissent

Preventive Repression

Responsive Repression

9

That earlier intervention places much higher informational requirements on the coercive

apparatus of an authoritarian regime to solve the problem of citizens’ preference falsification.30

Protests reveal information; 31 once citizens are shouting and waving placards in front of the

presidential palace, the capacity for violence required to remove them may be relatively high, but the

informational requirements are relatively low. The answer to the question asked by coercive agents

“who should we worry about?” is straightforward: “the people right in front of you.” By contrast,

identifying targets when behavior is still private and atomized requires more information, and that

information is harder to gather. Moreover, if citizens strategically mislead coercive agents – or offer

conflicting or divergent information to different parts of the coercive bureaucracy – then integration

and cross-checking of information becomes especially important to ensure accuracy. In sum,

information is more important to preventive repression than to reactive repression.32

Showing that information is important, however, is not a complete theory. How exactly do

authoritarian regimes use information to prevent unrest and maintain stability? I argue that having

good information is key to effective coercive capacity because it assists authoritarian regimes with

two key tasks. First, information helps officials identify potential threats and assess their severity. This

means that information must be accurate; inaccurate information leads either to false positives that

create a backlash or false negatives that can jeopardize the security of a regime. Because one piece

of information on its own is seldom enough to conduct a threat assessment, any given data point

must also be obtained early enough to be put together and integrated with other information that

makes it possible for intelligence and coercive agents to make judgments on threat type and severity.

30 Preference falsification also places informational constraints on citizens seeking to coordinate in authoritarian environments, as has been noted by Kuran 1989; Little 2015; Tyson and Smith 2018 (among others). 31 Lohmann 1993, 1994; Bueno de Mesquita 2010; Meirowitz and Tucker 2013; Casper and Tyson 2014.

32 There are also significant implications here for studies of coercion and authoritarian survival. As this figure illustrates, regimes select themselves into responsive repression through a failure in earlier demobilization or deterrence processes. Any study of how observed repression affects authoritarian survival must account for this selection effect. See Ritter and Conrad 2016; Greitens 2016.

10

Note that to engage in preventive repression, authoritarian coercive agents must solve specific

informational problems related to citizen preference falsification that go well beyond simply

discerning the overall “underlying distribution of preferences,” which is the aspect of the problem

most discussed by Kuran and others. Citizens are not homogenous or interchangeable; they exhibit

both a) a range of grievances with the regime, and b) varied thresholds for anti-regime

mobilization.33 A secret police organization must possess not just an understanding of how much

opposition is out there, but “the ability to identify enemies correctly.”34 Failure to do this risks both

Type I and Type II errors: repressing citizens who are not at risk of mobilizing in opposition, and

prompting a backlash;35 or failing to repress citizens who are planning to mobilize against the regime,

and therefore allowing private dissent to turn into collective action. For a security service, knowing

what the overall distribution of preferences is will not be sufficient; they must also know who falls

where on the particular distribution, and must find this out it at a time when they can still do

something about it. Thus, precise threat identification and assessment is the first contribution that

information makes to coercive capacity.

The second task that information facilitates, after threat identification and assessment, is

strategy selection: choosing how the coercive apparatus should deal with the threat it has identified.36

Because individuals living under authoritarian rule are not homogenous, identifying the potentially

threatening individuals earlier on provides the regime with more time and potential options for

demobilization. However, individualized information on a citizen’s preferences, grievances, and

33 Kuran 1991; Siegel 2011.

34 Gregory 2009; Gregory et al 2011.

35 On the risks of excessive violence producing backlash, see Condra and Shapiro 2012; Felter and Shapiro 2017; Francisco 2005; Downes 2007; Kalyvas 2006; Kocher, Pepinsky, and Kalyvas 2011. A subset of these studies argues that collective or indiscriminate repression can be effective under specific conditions; see, for example, Downes 2007; Lyall 2009; Daxecker and Hess 2013.

36 There is an extensive security studies literature on strategy selection in different contexts, from classic works on military proclivity for offensive strategy in conventional conflicts (Posen 1984; Snyder 1984; Van Evera 1984; Avant 1993; Rosen 1991) to nuclear strategy (Narang 2014) to counterinsurgency (see footnote 34 above).

11

thresholds must not only be acquired early enough to conduct a full threat assessment; it must be

acquired, integrated, and analyzed early enough that officials can then share it with the agents who

actually decide on and implement demobilization strategies. In short, to be useful for preventive and

differentiated demobilization, information must be accurate; obtained early, and distributed to the

right place in the coercive apparatus. This means that authoritarian regimes need an institutional

framework capable of collecting and integrating accurate information, deciding what to do with it,

and distributing it to the right actors in time for preventive action. This also reinforces the idea that

spending alone, though a common metric for coercive capacity, is insufficient: high levels of

spending on internal security are likely to be ineffective unless they are applied within the right

institutional framework.37

What kinds of strategies can authoritarian coercive agents employ if they identify a citizen as

a potential threat to the regime? I propose that authoritarian regimes apply one or more of three

primary demobilizational strategies: concessions, coercion, or pre-emption. Each of these strategies has

been documented in the general literature on comparative authoritarianism, but also appears

specifically in scholarship on Chinese politics. I argue that information, in essence, places individuals

at specific points on a distribution that tells the party-state and coercive apparatus what strategy to

apply to maximize the likelihood of successful demobilization – it is one of the key variables that

helps shape the balance between co-optation and repression in non-democracies.

First, authoritarian regimes can provide concessions to assuage dissent and limit contention.38

Some concessions reward loyalists,39 while others are aimed at co-opting opposition or converting

37 On how this applies to the China case prior to Xi Jinping’s ascent to power, see Greitens 2017.

38 Tullock 1987; Wintrobe 1998; Gandhi and Przeworski 2006; Haber 2007; Gallagher and Hanson 2009; Mares and Carnes 2009; Cammett and Issar 2010; Distelhorst and Hou 2017.

39 Slater 2003; Chang and Tsai 1985; Kwon 2003; Geddes 2006; Magaloni, Diaz Cayeros, and Estevez 2007; Blaydes 2011.

12

wavering/neutral citizens to supporters.40 These concessions can be policy concessions or material

ones, and the latter can be in the form of patronage, or targeted public goods and social welfare

provision (a form of redistributive politics). Bueno de Mesquita and Smith suggest that provision of

material benefits will improve citizens’ welfare and “diminish their desire for revolutionary

change”;41 in East Germany, for example, social unrest in 1953 produced redistributive responses on

the part of the GDR authorities.42 Similarly, in China, one study found that around 30% of large-

scale mass incidents in China resulted in the government providing some form of monetary

compensation to protestors.43 While post-unrest compensation is a responsive form of concession,

policy and material concessions can also be deployed preventively, to demobilize citizens in

anticipation of a protest or petition that then does not materialize. The GDR appeared to think of

redistribution this way in maintaining its program after the 1953 unrest had subsided, and the

contentious politics literature on China has documented cases where local officials simply pay

citizens not to protest: a direct exchange of material benefit for political demobilization.44

Second, information can enable coercion. Following standard usage in international relations, I

define coercion as a strategy that emphasizes the threat of punishment or violence, rather than its

actual use, to deter actors from mobilization or compel them to desist from it.45 These threats may

be communicated privately, so they are difficult to observe; they can either arrest mobilization in the

early stages of the process depicted in Figures 1-2, or over time, deter it entirely—leading to what

Judt referred to in post-1945 Eastern Europe as the “peace of the prison-yard.”46 In China, survey

40 O’Donnell 1973; Schwedler 2006; Smith 2008; Mares and Carnes 2009; Desai et al 2009; Bueno de Mesquita 2010; Conrad 2011; Wallace 2014; Cho, Lee, and Song n.d. 41 Bueno de Mesquita and Smith 2010; Roessler 2011; Reuter and Robertson 2015.

42 Thomson 2017. 43 Lee and Zhang 2013.

44 Ibid; see also Cai 2010; X. Chen 2012; Lee and Shen 2014; Shadmehr 2014. 45 For a classic definition of coercion, see Schelling 1960; see also George 1991; Byman and Waxman 2002.

46 Judt 2005; Sullivan 2015.

13

and ethnographic research has found evidence of coercive approaches in the phenomenon of

“relational repression” – where a colleague, friend, or family member is intimidated or threatened in

order to convince an individual to refrain from or stop challenging the party-state.47 As applied to

contentious politics, coercion can be used for either compellent purposes (to make an individual

stop contentious actions that they would otherwise continue to engage in, or convince them to

actively cooperate with the coercive apparatus) or deterrent purposes (to prevent them from

engaging in contentious or threatening action to begin with). Either way, the distinguishing feature

of coercion is the use of threats, rather than violent behavior by the state, to change individual

calculations and produce citizen compliance.

Finally, the coercive apparatus can use intelligence to target citizens for pre-emptive use of

force. Pre-emption is the actual use of physical punishment; it is an example of high-intensity

repression (though that term sometimes conflates the scope of repression with its intensity). Studies

of dissident repression in China have found clear evidence of the use pre-emption. In preparation

for sensitive anniversaries that could provide potential focal point for collective action, for example,

authorities proactively place protestors and petitioners under detention or house arrest.48 Here, in

particular, is where regimes face the problem of avoiding group-level punishment that can reinforce

communal identity and foster collective resistance. 49 Accurate intelligence in the hands of the right

coercive agents helps to mitigate this risk by enabling “precise, selective, lower-intensity violence,

against the right people and only when necessary.”50 This assists the coercive apparatus in two

distinct ways: first, it lowers the number of people who have been repressed, thereby diminishing the

probability that members of this group will find common ground and a shared basis for future

action. Second, selective pre-emption can also be conducted more privately, making violence less 47 Deng and O’Brien 2013; O’Brien and Deng 2015; Greitens and Truex 2019. 48 Truex 2018; Liao and Tsai 2018; Carter and Carter n.d.

49 Blaydes 2018. 50 Greitens 2016: 43; see also Kalyvas 2006: 173; Wintrobe 2000.

14

visible and reducing the risk of backlash in that way sparked by media coverage or exposure to an

increasingly pro-democratic international environment.51

One of the primary benefits of information, therefore, is that it generates capacity for

differentiated demobilization – the calibration of demobilizational strategies on an individual basis to

fit a particular citizen’s grievances and assessed threshold for action.52 Differentiation enables a

targeted approach that divides citizens based on whether or not their grievances are solvable

(governance/material concessions) and what their threshold is for mobilization (if low, they can be

coerced; if too high, they must simply be pre-empted).53 A targeted approach also makes coercive

threats, specifically, more credible – because threats issued to a specific individual demonstrate, at

the time the threat is made, the regime’s capacity to identify/detect and punish dissenters. And

because differentiated repression can be conducted privately, it lowers the potential for costly

backlash by disaggregating and atomizing citizens’ experiences of repression, rather than allowing

collective experience of it to solidify out-group identity and harden resistance.54 How a regime

obtains, organizes, and uses information on citizens, therefore, is a foundational attribute of

authoritarian control because information helps first to identify threats, and second to match them

to disaggregated de-mobilization strategies targeted to achieve the highest chances of success at the

lowest possible political cost.

Two words of methodological and theoretical caution emerge from this framework. Because

authoritarian regimes have a desire to deter individuals from mobilizing, they have incentives to

51 Overall, studies of global political violence suggest that mass killing and state violence against civilians are both on the decline in the post-Cold-War period, perhaps due to increasing awareness of the costs of visible violent repression. Ulfelder and Valentino 2008; Guriev and Triesman 2015; Levitsky and Way 2010; Conrad and Ritter 2019. 52 Roberts (2018) notes similar dynamics: porous censorship creates informational barriers between educated, affluent citizens (more likely to circumvent censorship), and average citizens (less likely). 53 These judgments are temporally specific; it may be possible to approach a citizen who requires pre-emption in the short-term with other strategies in the long-term. 54 Friedrich and Brzezinski 1965; Dallin et al 1970; Wintrobe 2000; Francisco 2005; Siegel 2011; Blaydes 2018.

15

exaggerate the effectiveness of the surveillance and control capabilities that they possess. Some of

the stories and glossy promotional videos on the efficacy of these systems should be interpreted as

exercises in domestic propaganda, aimed at convincing the population that political misbehavior will

be swiftly identified and caught by the authorities.55 Less likely to be emphasized in a regime’s public

rhetoric are failures of experimentation, bureaucratic and personnel hurdles, and gaps in the regime’s

ability to watch and coerce citizens.56 In general, then, scholars should use caution in assessing the

nature, source, and audience of regime statements, to check for the possibility that portrayals of

surveillance efficacy are intentionally upwardly biased by authorities to deter citizen mobilization.

Internal assessments and perceptions, though more difficult to access, are likely to provide a better

(though still imperfect) basis for judging efficacy.

This is especially applicable in China, where the idea of “striking hard” to intimidate and

deter criminals is deeply embedded in public security thinking, and where deterrent idioms like “kill

the chicken to scare the monkey” (shajijinghou, 杀鸡儆猴) or “kill one to warn a hundred” (杀一儆百,

shayijingbai) are frequently quoted by police officials.57 Given the clear desire for deterrent effects on

the part of the Chinese coercive apparatus, rhetoric about the awe-inducing power of technology to

regulate citizen behavior and deter unwanted mobilization should be regarded with skepticism. The

pages that follow, therefore, treat surveillance capability as a policy objective and a hypothesis to be

tested, rather than a reality to be assumed.

The second cautionary point is that surveillance technologies can be used this way, but do not

have to be. If a regime chooses to use ascriptive or group-based characteristics to assign citizens to

55 They could also be interpreted as international marketing to generate sales of these platforms abroad (more on that below). Carter and Carter (forthcoming); see also Bloomberg, “China’s Most Advanced Big Brother Experiment is a Bureaucratic Mess,” 18 June 2019, https://www.bloomberg.com/news/features/2019-06-18/china-social-credit-rating-flaws-seen-in-suzhou-osmanthus-program

56 Domestic security officials may also be unlikely to fully acknowledge these shortcomings in public or official assessments if doing so could damage their chances of promotion. On distortion of China’s crime statistics due to police career incentives, see Bakken 2005:17, 75; Dutton 2005: 282; Scoggins forthcoming. 57 Su 1994; originally quoted in Tanner in Bakken 2005: 173.

16

demobilization “treatments,” then fancy technology and a high volume of information will not

produce differentiated demobilization: the likely result is still collective repression, with all the

potential costs that accompany it. As I explore in a separate paper, this appears to be what has

occurred in Xinjiang, where in early 2017, the CCP shifted towards a strategy of collective repression

that relied on the mass detention and wide-scale re-education of Turkic Muslim minorities, on the

basis of a new judgment that their ethno-religious belief and identity itself posed a heightened

counterterrorism threat to regime security.58

Solving Information Problems: China’s Evolving Approach to Surveillance

Xi Jinping has made redefining the philosophy and management of domestic security a

hallmark of his tenure, a re-direction that was signaled early by his speeches about the importance of

“comprehensive” or “holistic” national security.59 In doing so, he has emphasized “prevention and

control” (防控, fangkong) rather than the regime’s previous catchphrase, “stability maintenance,” and

has emphasized the use of technology to collect information on citizens for preventive public

security purposes.60 This section traces the development of China’s surveillance state by describing

the coercive bureaucracy’s persistent information problems, and summarizing recent changes to

national doctrine that have attempted to address these challenges. It then outlines the key role of

local actors in implementing national directives, and explains two specific problems, information

blindness and information islands, that characterize information politics at the local level.

58 Greitens et al forthcoming (Winter 2019/20). Paper available upon request. 59 Xi 2015, 2018; see also Xinhua, “Book of Xi's discourses on national security published,” 15 April 2018, http://www.xinhuanet.com/english/2018-04/15/c_137112987.htm. 60 Hu 2015; for a summary of Xi’s changes to domestic security, see Greitens 2019.

17

Contemporary Dilemmas & Changing Doctrines

A non-democratic coercive apparatus is likely to find information-based, differentiated and

preventive repression especially desirable if it perceives visible mass repression as costly. Since the

mid-2000s, China’s public security forces have operated in such an environment: they are

responsible for maintaining stability, but told to achieve that goal with very limited use of force, and

in many cases, are told they will be disciplined for using force unnecessarily.61 Local cadres also have

career incentives that prompt them to focus on prevention of social conflict, which means that they

are (relatively) aligned on this score with local coercive agents. Collective action is considered

disruptive and costly to address ex post – but more practically for local party officials, simply having a

“mass incident” can be a negative/veto performance indicator for promotion (even if handled

successfully).62 Both public security doctrine and officials’ career incentives, therefore, incline

toward limiting the use of local state violence against citizens.63

Contemporary Chinese public security analysis emphasizes the importance of both

information quality and information timing in order to engage in effective prevention of social

conflict. Articles in public security journals highlight the importance of “foreknowledge” (xianzhi, 预

知) and timeliness (jishi, 及时) in public security intelligence, arguing that timing and purpose separate

intelligence from simple information.64 According to these studies, timing matters because defusing a

protest competently is not bad, but it is infinitely preferable if local officials can prevent the protest

61 Ministry of Public Security 2000; Meng 2008; Ministry of Supervision, Ministry of Human Resources/Social Security, and State Bureau for Letters and Calls, “Guanyu weifan xinfang gongzuo jilv chufen zanxing guiding [Tentative regulations on violations of work discipline for petition work],” 30 June 2008. 62 Whiting 2004; O’Brien and Li 2006; Cai 2008; Li et al 2012; Chen 2012.

63 A number of China scholars argue that this pressure has incentivized local authorities towards the use of informal coercion and third-party suppression using actors such as chengguan (城管) – and that this, in turn, has undermined the local state’s monopoly on violence and often visibly backfired by producing protest about the misbehavior and bullying of informal coercive actors. This may be one reason why CCP officials have emphasized intelligence: it provides an alternative to the use of informal actors whose behavior has become costly. Chen 2017; Ong 2018a, 2018b. 64 For a good English-language summary of some of these studies, see Schwarck 2018.

18

from coalescing in the first place—especially from the perspective of the local officials themselves,

since their prospects for promotion can be derailed by one major incident. To use a public safety

analogy invoked by Chinese officials, social management must be focused not just on effective fire-

fighting, but on fire prevention. In 2006, security chief Zhou Yongkang called for new mechanisms of

public opinion collection to provide “early warning” on potential threats to social stability;65 in 2009,

Minister of Public Security Meng Jianzhu instructed local leaders to seek accurate, actionable

information to forecast (yuzhi, 预知), forewarn (yujing, 预警), and prevent (yufang, 预防) threats to

political security and social stability;66 and thoroughout the 2000s, the Ministry of Public Security

worked to develop public security intelligence systems (gong’an qingbao tixi, 公安情报体系) focused on

“intelligence-led policing” (qingbao zhidao jingwu, 情报指导警务).67

The objective of obtaining complete, accurate, and timely information for preventive

purposes, however, faces significant obstacles that stem from the organizational makeup of the

Chinese political system. Contemporary scholarship on China has highlighted the role of

information in Chinese governance, contributing to and extending the findings from comparative

politics described above; on balance, its findings reinforce the view that obtaining accurate

information is a perennial problem for autocratic leaders. Accurate information on citizen

preferences and opinions is difficult for Chinese officials to obtain in the absence of traditional

democratic mechanisms such as free media, civil society, and electoral competition.68 Monitoring

government performance is also difficult; lower-level officials distort upward reporting of data from

economic statistics to online grievances and protest incidents, for reasons of both self-protection

and self-advancement,69 and fragmentation of social control efforts across “different agencies within

65 Zhou 2006; Schwarck 2018, pp. 8-9.

66 Meng 2009. 67 Schwarck 2018. See also CCP Central Committee 2004.

68 Liebman 2005; Lorentzen 2014, 2015; Stockmann 2012; Chen and Xu 2017; Roberts 2018. 69 Shirk 2011; Wallace 2014; Pan and Chen 2018.

19

a single local state” makes it difficult for the local party-state to gather the information necessary to

supervise civil society.70

To address its informational challenges, China’s party-state relies on various indirect

mechanisms—national/local people’s congresses and consultative bodies; petitions, protests, and

‘porous censorship’ of social media platforms; and government agencies like petition offices and

“mayor’s mailboxes”—to provide information on citizen preferences/grievances and government

performance.71 China’s “fragmented authoritarian” system, however, appears to have significantly

limited the extent to which this indirectly-collected information can actually be leveraged for stability

maintenance and social governance.72 In particular, the structure of overlapping vertical departments

(tiao, 条) and horizontal local authorities (kuai, 块) exacerbates challenges in local information-

sharing and preventive coercion: kuai are responsible for stability maintenance, but coercive power –

and the information needed to execute effective stability maintenance operations – is actually

controlled and operated vertically within departments/bureaus (tiao). Moreover, even single

ministries (such as MPS) lack institutionalized horizontal mechanisms for information-sharing across

their various departments.73 Indirect information mechanisms are of little help in solving this

fundamental organizational mismatch and achieving the process, outlined above, that is necessary

for preventive repression to work.

Under Xi Jinping, the Hu-Wen emphasis on “stability maintenance” has been replaced by

the byword “prevention and control” (fangkong, 防控). In 2015, for example, Zhou’s successor Meng

70 Fu 2017, 2018: 50-65; Deng and O’Brien 2013; Lee and Zhang 2013; Ong 2018a, 2018b; Chen 2017; Scoggins and O’Brien 2016. 71 Manion 2015; Truex 2016; Yan 2011; Lorentzen 2013, 2014; Dimitrov 2014, 2015; Chen and Xu 2017; King et al. 2013; Chen, Pan, and Xu 2016; Distelhorst and Hou 2017; Mattingly 2019. 72 Effective cooperation among China’s numerous stability maintenance organizations has long been a thorny problem. Lack of coordination across actors within a locality, among localities on an issue, or between central leaders and local agents commonly disrupts both policy implementation and Chinese leaders’ ability to receive accurate, timely information. Lampton 2015: 764; Guo 2012: 2-3; Fu 2017; Lieberthal and Lampton 1992. 73 Wang 2010; Schwarck 2018:7-8; Wuthnow 2017:895.

20

Jianzhu emphasized fangkong as the “correct direction” for political-legal work in a six-point address

that invoked the term “foresight” (不断增强工作预见性).74 In early 2019, at the national meeting of

Public Security Bureau directors, Minister of Public Security Zhao Kezhi urged his audience to

“Always insist on putting prevention of political risks as the first priority.”75 The use of information

technology in this effort is paramount: in mid-April 2015, the CCP Central Committee and PRC

State Council called for the creation of a “three-dimensional information-based prevention and

control system for public-social security” (创新立体化信息化社会治安防控体系) – another term

that appears repeatedly in CCP directives on “social governance” (社会治理, shehui zhili) – in order to

“comprehensively promote the construction of a peaceful China.”76 The directive calls for the

expansion of networked video surveillance and grid management, enhancement of predictive and

early warning capabilities in public security, and reorganization of local party and government work

in this area to limit information gaps and achieve smoother coordination of information and public

security intelligence.77 The CCP has also explored a number of ways to more directly connect

information-gathering with the infrastructure and management of public security at the ground level,

including major organizational and legal reforms to China’s political-legal apparatus (zhengfa xitong, 政

法系统).78

74 Meng 2015; see also Trevaskes 2013, pp. 51-77.

75 Emphasis added. His statement was “始终坚持把防范政治风险置于首位.” Li 2019.

76 CCP Central Committee/PRC State Council 2015. 77 This emphasis on informatization and technology also parallels a developing emphasis on networked intelligence and decision-making in China’s military strategy. 78 For more detail on these reforms, see Greitens 2019.

21

Understanding Local Information Problems

When Chinese leaders emphasize the importance of preventive capacity in social governance,

they consistently place primary responsibility on local communities and local leaders. The limited

amount of analysis that exists thus far, however, has focused on the CCP’s domestic security efforts

at the central/national level (which are still only incompletely understood) – and it is even less clear

how the initiatives that have been announced are playing out at the local level, where policies are

implemented, politicians’ careers are made or broken, and where – both in China and more broadly

in authoritarian political systems -- central and local coercive actors are not always in perfect

alignment.79

One of the most salient examples of this local emphasis is Xi Jinping’s repurposing of the

Mao-era “Fengqiao Experience” (Fengqiao jingyan 枫桥经验) as a model of the regime’s philosophy of

prevention and control. 80 Fengqiao, in Zhejiang province, became a model for removing

“reactionary elements” during the 1963 Four Cleanups Movement, and continued to be a post-Mao

benchmark for grassroots-level social control and “embodiment of the party’s mass line.” In

October 2013, Xi Jinping invoked the Fengqiao experience, bringing it back to the fore and defining

it as an approach in which the masses “resolve social conflicts or contradictions on the spot.” By

doing this, they do not transfer them to higher authorities (“small things don’t leave the village, big

things don’t leave the town, contradictions don’t escalate to higher authorities”81), which Xi depicted

79 On China, see Murray and Tanner 2007; Fu 2018. Generally, see Levitsky and Way 2010: 59-60; Earl and Soule 2006; Dragu and Przeworski 2018. 80 Wang and Chu 2018; Zhejiang PSB General Office 2018.

Another example that had been highlighted by the Chinese leadership was that of Wukan, which became prominent after provincial authorities, including then-leader of Guangdong Wang Yang, successfully defused protests through dialogue and compensation. In 2013, Trevaskes called it “China’s most influential mass incident in years,” and noted that after the fall of Bo Xilai, Zhou Yongkang had proclaimed the “Guangdong model” as the new approach for China. The re-emergence of protest in Wukan several years later, however, appears to have damaged the credibility of this approach, and today the Fengqiao experience is a more prominent model than Wukan. See Trevaskes 2013, p. 71; Zhao 2012.

81 “小事不出村、大事不出镇、矛盾不上交.”

22

as a means of preventing the vertical escalation and geographic spread of contention. He revisited

Fengqiao in January 2019 remarks at the Central Party School, calling on cadres to “modernize social

governance, adhere to and develop the “Fengqiao experience,” improve social coordination

mechanisms to construct safety, and enhance social stability capabilities and levels from the source,”

which is a phrase that refers to preventive defusing of social conflict.82 Fengqiao has since been

termed a “banner for the country’s political-legal frontline governance” and “the compass and

weathervane of grassroots social governance,”83 and local leaders have explicitly linked it to Xi

Jinping’s charge to “comprehensively upgrade the ability to prevent and control social risks.”84 The

Fengqiao experience epitomizes the CCP’s emphasis on the importance of local actors in preventive

conflict management, as they are the ones whose early actions can avoid escalation and contagion.

Chinese analysis tends to focus on two major problems that occur at the local level:

information blindness, and information islands. The former focuses on obtaining/collecting

information and making society fully legible to the CCP by reducing “blind spots,” while the latter

focuses on information management and coordination within the local party-state bureaucracy.

Reducing information blindness received early attention under Xi Jinping’s leadership. In

2011, Xi warned Central Party School trainees that failing to effectively gather information via the

mass line would blind the Party and harm the party’s policy implementation and image; shortly after,

the Central Committee and State Council circulated updated guidance on how to use information

82 The sentence read: “要推进社会治理现代化，坚持和发展“枫桥经验”，健全平安建设社会协同机

制，从源头上提升维护社会稳定能力和水平.” See “习近平：提高防控能力, 着力防范化解重大风险 保持经济持续健康发展社会大局稳定 [Xi Jinping: Improve Prevention and Control Capabilities; Try to Prevent and Resolve Major Risks; Maintain Sustainable and Healthy Economic Development and Overall Social Stability],” 21 January 2019, http://www.qstheory.cn/yaowen/2019-01/21/c_1124021825.htm (hereafter Xi 2019b).

83 Emphasis added. “新时代“枫桥经验”是基层社会治理的指南针和风向标.”

84 “全面提升社会风险防控化解能力.”

23

for effective social management.85 The 18th Party Congress (2012) advocated a governance model

focused on understanding the origins of social disputes (yuantou zhili, 源头治理), using dynamic

management of social conflicts (dongtai guanli, 动态管理), and efficient crisis management (yingji chuzhi,

应急处置), reinforcing the preventive direction of public security work.86 Echoing themes from the

Fengqiao experience, the 19th Party Congress (2017) vowed to “improve mechanisms for preventing

and defusing social tensions… [and] shift social governance focus to the community level.”87 The

CCP emphasized the need for local leaders to increase civic involvement in information provision,

and required local governments to adopt comprehensive information platforms that would integrate

grassroots self-governance into “grid management” (described below).88

Local governments responded by experimenting with ways to involve citizens in new

information provision and stability maintenance programs. In Beijing’s Chaoyang District, police

recruited ~140,000 civilian informants, and built an app that they could use to send in reports; in

2017, these “Chaoyang masses” provided 8,300 pieces of information that police used to solve 370

cases, arrest 250 suspects, and eliminate 390 security threats.89 In Shanghai in 2014, then-Party chief

Han Zheng, who was subsequently elevated to the Standing Committee of the Politburo in October

2017, designated “grassroots social governance” as the “No. 1 research project.” Following his

direction, the municipality produced a grassroots social control policy that emphasized information-

based social management; it required streets, townships, residential communities, and villages to

establish comprehensive information platforms based on citizen-participation and sought to

85 Xi 2011; “Policy document on innovation in social management issued,” 20 September 2011, http://www.china.com.cn/policy/txt/2011-09/20/content_23451516.htm. 86 Hu Jintao, 18th Party Congress report, 8 November 2012.

87 Xi Jinping, 19th Party Congress report, 18 October 2017. 88 CCP Central Committee/PRC State Council 2017.

89 Xinhua, “The Chaoyang Masses’ Numbers Exceeded 140,000 Last Year, and Cracked Over 370 Cases,” 29 January 2018, http://www.xinhuanet.com/legal/2018-01/29/c_1122330010.htm

24

enhance big data analysis, and dissolve ‘information islands’ (see below).90 While public narratives of

successful flagship cases are useful, they offer relatively little insight into “typical” local experiences,

the range of variation therein, and the internal assessments and processes by which local officials

attempted to construct the requisite informational capacity.

Moreover, China’s fragmented authoritarian system has posed particularly thorny problems

for local officials tasked with maintaining social stability. Under Hu Jintao, fragmentation within the

local party-state meant that various actors, all sharing a general mandate to maintain social stability,

nevertheless followed distinctive bureaucratic mandates, possessed divergent informational

capacities, and employed different tactics to deal with contention—resulting in actions that were

sometimes complementary, but sometimes diametrically conflicting. 91 This fragmentation also

created “information islands,” in which threats to social stability were more often ignored, distorted,

or hidden by local bureaucratic structure than identified and illuminated by it. In 2016, for example,

Li Yu wrote that coordination was a major issue for the construction of information-based social

governance, and that local bottlenecks – either technical incompatibility or political unwillingness –

would have to be resolved in order to achieve effective information-sharing.92 Again, while Western

scholarship has focused on intelligence and organizational reform at the national level, relatively little

work examines the local dimensions of information coordination.93

What all of this suggests, in sum, is that under Xi Jinping, the CCP’s domestic security

doctrine has specifically emphasized the objective of “prevention and control.” National guidelines

place heavy emphasis on the role of information in making preventive demobilization possible, and

direct local actors to intensively implement information-technology platforms to achieve this goal.

90 CCP Shanghai Municipal Committee/Shanghai Municipal Government 2014.

91 Fu 2017, 2018: 10, 50-65; Deng and O’Brien 2013; Lee and Zhang 2013; Ong 2018a, 2018b; X. Chen 2017; Scoggins and O’Brien 2016. On fragmented authoritarianism, see Mertha 2009; Lieberthal and Lampton 2002.

92 Li Yu 2016; Wang 2010; Schwarck 2018:7-8; Wuthnow 2017:895. 93 Greitens 2019; Lampton 2015; Wuthnow 2017; You Ji 2016; NPC 2017; Zhao 2017.

25

As yet, however, though the information problems faced by the local party-state are fairly well-

documented, little is known about how major efforts to address those problems have been

implemented by local officials. The following section of the paper provides qualitative evidence of

this implementation process, illustrating the theoretical mechanisms described above.

Building the Surveillance State: Qualitative Evidence from M District94

This section provides qualitative evidence on the development and use of surveillance and

social control technology in M District, a relatively typical district in a large city in eastern China.

With a population of around two million, M District – like many Chinese cities – experienced rapid

urban expansion, contention over house demolition and development, and the influx of a migrant

population that comprised almost half the city, and that authorities saw as a significant source of

instability and crime. 95 Following a public security assessment in 2009, M District’s officials

concluded in 2009 that they needed to change their approach. They emphasized making information

on residents more timely, accurate, and integrated, so that it could be used for preventive

demobilization. District officials aggressively recruited local informants to avoid information

blindness and created new coordination mechanisms/platforms to integrate intelligence and reduce

information islands. They then drew on this system to engage in differentiated demobilization of

potential contention.

94 This section draws on research with Huirong Chen; separate paper available upon request. To protect interlocutor identities, the names of the District, city, and specific programs have been omitted, and the text avoids citing news coverage or listing the titles of internal reports in cases where doing so would identify the District. The MU IRB approval number for this project is 2013784. 95 M District Government, Report on Stability Maintenance, 2009. Hereafter M District 2009.

26

Prevention Through Information Collection and Coordination

In 2009, following a review by public security officials, M District initiated a “Great

Coordination Program” (GCP) aimed at unifying information-gathering, coordinating information

among fragmented coercive agencies, and integrating the organizations and individuals involved in

stability maintenance. The 2009 assessment of M District’s stability maintenance work emphasized

the need to develop information and act early to achieve effective urban management and social

stability, making the goal of the program explicitly preventative:

We shall find out and report issues of social management at the first time, advance public service, manage society by providing good service, and turn ex post law enforcement into ex ante public service. Addressing disputes at the grassroots level and in the bud is good for resolving people’s concerns and transforming our urban management from a reactive type into a proactive and adaptive one.96

This focus appeared again in the assessment’s urging to adapt frontline public security work to

“move the focus downwards, the front forward, and management to the beginning” (zhongxin xiayi,

guankou qianyi, guanli qianzhi, 重心下移，关口前移，管理前置).97 It was also maintained over time;

a 2014 review of the program noted growing consensus among local officials that “proactively

exploring problems” is better than “reactively waiting for problems,” that “ex ante prevention” is

better than “ex post response,” and that “systematic solution” is better than “ad hoc treatment.”98

Officials stated that “timely and sufficient public service can largely prevent emergence of social

unrest; social disputes in the bud are much easier to resolve.” 99 In other words, M District’s officials

seemed to arrive at a stable consensus that addressing disputes early and preventively was the key to

enhancing social stability and resolving public security challenges.

96 M District 2009.

97 M District 2009. Several years later, Minister of Public Security Zhao Kezhi urged PSB directors across the country to “shift their focus/center of gravity downward” (推动重心下移, 警力下沉, 保障下倾); Li 2019.

98 M District Government, Report on GCP and City Management, 2014. Hereafter M District 2014. 99 Interview, Public Security Bureau researcher, 24 June 2015.

27

Among M District’s major areas of focus for constructing preventive coercive capacity was

the creation of new mechanisms for timely, efficient information collection and distribution. A GCP

official described the importance of information collection and analysis for identifying sources of

social unrest as follows:

The GCP aims to coordinate information and forces that scatter in many departments. If we build a comprehensive system of information collection and analysis, we can know in a timely way what is going on and take corresponding measures to prevent eruption of massive incidents…We can identify who is the ‘trouble maker” and the leader. With this kind of information, we can more effectively address problems and keep society in order.100

District leaders discussed two different types of useful information. Ex ante or “infrastructural

information” (jichu xinxi, 基础信息) on M District’s population could be collected before the

emergence of social unrest; ex post or “emergent information” (yingji xinxi, 应急信息) had to be

collected during/after unrest. The GCP collected and integrated both types of information to

support both normal/daily social management (richang guanli, 日常管理) and crisis management

(yingji guanli, 应急管理 ), but officials clearly preferred and prioritized ex ante information

collection.101 District leaders believed that accurate, timely, and comprehensive “infrastructural

information” could help them, gauge public sentiment, assess governance quality, identify potential

problems, and prevent the emergence of disruptive contention.102

In pursuing information-guided preventive coercive capacity, however, M District officials

faced the problems of information blindness vis-à-vis their citizens and information islands within

the local party-state structure. District officials defined “information blindness” as an outside-system

(tizhiwai, 体制外) problem and “information islands” as a within-system (tizhinei, 体制内) one:

To build a comprehensive information platform for our political-legal work, the government system needs to reach every corner of grassroots society and keep eyes on what is happening.

100 Interview, GCP district-level center director, 21 September 2012.

101 M District 2009. 102 See Mann 1984 on infrastructural and despotic power.

28

Otherwise, government would be blind… Many tiaos have their own unique information and we lack a unified platform to coordinate scattered information.103

The first problem the GCP sought to resolve had to do with information collection. Officials in M

District used the metaphors of sight and blindness to describe information collection work; one

officer commented, “The masses’ eyes are bright; they can be the eyes of the blinded

government.”104 To decrease information blindness, the GCP relied on three main sources of

information: street patrols (jiemian xunluo, 街面巡逻), situation reports from neighborhoods/villages

(jucun huibao, 居村汇报), and online public opinion (wangluo yuqing, 网络舆情). Street patrols and

situation reports used civic organizations at the grassroots to generate useful information for public

security/social management: committees were tasked to investigate, identify, and report upward on

possible conflicts and threats, and directed to pay special attention to public security challenges

including illegal housing/renting, illegal business, and other ‘social disputes’.105 Frontline reports

were integrated with online information from numerous websites and hotlines via the GCP centers:

in 2016, 90% of grid information was collected on the ground, versus 10% from online intake.106

To cover every inch of territory, M District implemented “Community Grid Management”

(CGM, shequ wanggehua guanli, 社区网格化管理 ). M District was an early adopter of grid

management, a system now implemented across much of urban China. (Less than half of cities used

grid management in 2014-15; an estimated 60+% use it today.107) Neighborhoods were divided into

grids (each containing 800-1000 residents in several hundred households), and a team was assigned

to watch over each cell of the grid. Infrastructural information was uploaded into the grid; sensitive

103 Interview, Committee on Comprehensive Social Management, 22 September 2012.

104 Interview, officer in charge of GCP information synthesis, 12 October 2017. This is a reference to a Mao-era saying.

105 During the day, Urban Management/Law Enforcement personnel carry out street patrols; at night, public security forces conduct patrols. M District 2014.

106 M District, Grid Management Annual Data Report, 2016. 107 Economist 2018; Wu 2014.

29

places that required special attention and monitoring were identified in advance; and information

generated by frontline work and online activity was cross-laid onto the grid system as it came in.

The scope and scale of surveillance was impressive: by 2010, M District’s grid” had 13 medium grids,

further divided into a) 75 small grids based on urban blocks (further divided into 189 “responsibility

pieces”), and b) another 516 small grids based on residential communities/villages (1299

“responsibility pieces”). A total of 36,360 officials and other personnel – from police to road

cleaners – were involved in providing information to M District’s grid management system: at one

for every 50 or so residents, a relatively high ratio of surveillance.108

Adoption of the GCP information-gathering system significantly affected information

collection in M District. According to official statistics, by April 2012, village/neighborhood GCP

centers had collected 49,797 pieces of information on illegal behavior, enabling departments to open

49,309 cases and resolve 42,657 (86.5%). Of these, 19,981 pieces of information were related to

public security; 18,092 (90.5%) had been addressed.109 M District officials also believed that the

accuracy of their infrastructural information, a major concern in a district with a large floating

population and significant illegal housing, had improved substantially. 110 After 2012, grid

management continued its major role in information collection and analysis; the number of data

points and cases increased each year.

M District’s GCP and grid management systems also aimed to break up information islands

by creating an information-sharing platform that integrated previously-fragmented local

bureaucracies, and by building a “united big-data center” that employed that information to generate,

‘public security intelligence’ and analysis. (This platform bears notable resemblance to the Integrated

108 Greitens (2016, p. 9) finds that the most intensive authoritarian surveillance projects, such as those in North Korea or East Germany, involved a ratio of one official per 25-50 people. 109 M District, Report on GCP Achievements, 2012. The slightly higher rate of resolution for issues that are public-security-related appears to corroborate the prioritization of these issues. 110 Population and housing information is considered the foundation of “infrastructural information.”

30

Joint Operations Platform that has been described in Xinjiang, but was implemented almost a

decade earlier.111) Two major reforms made this possible. First, prior to 2009, each department in

M District managed its own functional grid, unconnected to others; the GCP replaced these

uncoordinated systems with a single integrated grid across the district’s geographic space. All 52

district-level functional departments (policing, health, sanitation, housing, etc), 13 township/street

governments, and 521 village/neighborhood GCP centers were equipped with shared information

collection/processing software, and required to upload their information into the shared platform.112

The platform also integrated data from online sources (citizen calls/emails; netizen communications

via the GCP website or local government mailboxes; and analyses of online public opinion) as well

as the information provided by regular local patrols. Then, in real time, it recorded, classified,

overlaid, and integrated incoming data with existing sources of information, so that the platform

could identify and analyze cases of concern and assign them to departments or grids for resolution.

The GCP platform also tracked and evaluated the rate, efficiency, and effectiveness of government

response, and provided some of this evaluative data to citizens on its website.

The second reform integrated disparate personnel systems. In 2009, M District had 18,470

social assistants acting as informants or supplementary coercive forces, but they were recruited and

controlled by 16 different departments and managed by no fewer than 194 labor organizations.113

The GCP reorganized social assistants and the funding that paid for them to align with the new

authority granted to township/streets by the revised grid system. Assistants who were previously

differentiated into 26 different types all received a standard title, “Grid Patrolman (wangge xunchayuan,

网格巡查员),” and were given consistent responsibilities, pay, training, and evaluation. Each was

111 See Greitens et al, forthcoming (Winter 2019-20). 112 Officials also hoped that the use of software would depersonalize social ties in the bureaucracy and avoid interference by personal connections in policymaking. 113 M District 2009.

31

assigned to a small grid to patrol communities/streets, collect infrastructural information, and assist

township/street governments in providing security and public services.

From Information to Action: Differentiated Demobilization

M District’s information-centered, preventive approach to social management and coercion

played a key role in enabling differentiated demobilization: it triaged ‘emergent issues’ that could

threaten social stability, decided what approach should be used; and assigned the issue to a particular

bureaucratic actor for resolution.114 M District’s 2009 blueprint outlined a preferred order for dealing

with potential threats, illustrated in Figure 3:

Figure 3: Options to Address Potential Sources of Instability115

These preference orderings roughly translate to the approaches outlined above of concession,

coercion, and pre-emptive punishment; the GCP’s role was to decide how many cases could be

addressed via the tools on the left or middle of the chain, reserving “law enforcement” (physical

repression/pre-emption) for the smallest number of cases possible.

Concession. First, authorities used GCP-provided information to dampen the emergence of

instability simply by resolving citizens’ complaints. By eliminating (some) grievances around which

groups mobilize, local officials can reduce citizens’ willingness to challenge the state. During the

outbreak of major protest in Wukan in 2011, for example, the provincial party leader called

protestors’ demands reasonable, acknowledged problems with government performance, and

promised not to retaliate against protestors after they stood down.116 Jennifer Pan has demonstrated

114 Dong 2015.

115 Concept outlined in M District 2009; figure created by author. 116 Wee 2011.

Service/Governance Social Management Law Enforcement

32

that officials often distribute China’s minimum livelihood payments (低保 , dibao) to targeted

populations (重点人, zhongdianren), including former prisoners, because of the fear that they will

engage in contentious behavior. 117 Many of these efforts, however, have been either reactive (Wukan)

or mistargeted (dibao), which has compromised their effectiveness.

GCP and grid management systems are being used, therefore, to more effectively target

concessions to prevent escalation of protest or social conflict. At a general level, M District’s

government research offices and leadership increasingly rely on the GCP database (and GCP-

enabled analysis) to generate policy recommendations for ‘meticulous urban management’ (jingxihua

guanli, 精细化管理).118 Grid management is then used to identify and solve urban management

problems before residents can complain; in Shuangpai County, for example, a grid manager out on

daily patrol registered a new business, checked a boarding-house for new tenants, and then used her

phone’s ‘social governance e-communication’ app (社治 E通, shezhi E-tong) to report a road problem

to the county grid management command center, which dispatched the correct department for

repairs.119 Information can therefore direct public goods provision to the areas that authorities deem

most useful for preventing complaints or collective action.

Beyond assigning public goods provision, GCP software can increase overall authoritarian

responsiveness by clearly delegating responsibility to one particular actor, and then reminding the

tasked department of impending or overdue case deadlines. M District’s GCP platform records and

shares late/absent responses with District leaders (and citizens); this makes it difficult for officials to

conceal poor performance or corruption. Because the GCP allows District leaders to compare

performance data provided by subordinates with feedback reported by residents, it serves a dual

function of responsive governance and agent monitoring. Overall, then, GCP data can be used to 117 Pan forthcoming. 118 Interview, Committee on Comprehensive Social Management, 10 August 2018.

119 “Shuangpai wanggeyuan pingfan de yitian [An ordinary day for Shuangpai grid patrolmen],” Sanxiang Management Network, 16 April 2018, http://www.sxzzw.gov.cn/news/201804/16_170732_4196.html

33

improve the responsiveness of local governance, reducing citizen grievances and dampening their

general willingness to challenge the party-state.

Grid management is also used to enable much more targeted distribution of concessions to

limit contentious activity. In one Nanjing district, the grid manager learned that the situation of a

medically disabled resident was angering residents; she then arranged for the resident to receive dibao,

housing, and childcare support. An online profile of her work called serving the people “the

foothold for grid governance…from resolving small neighborhood disputes to excluding major

security risks.”120 In other cases, grid managers have helped petitioners find employment or other

social services in exchange for ceasing petition activity.121 Concessions can also take the form of

simple payoffs. One local official described paying a resident 2,000RMB not to petition Beijing as

“very economical,”122 and a 2010 analysis found that 30% of mass incidents produce a response

involving economic compensation, though at that point most compensation was reactive rather than

preventive.123

Coercion. Second, the GCP’s early identification of potential or emergent threats to stability

enabled local authorities to demobilize threats via coercion and implicit threat rather than either

concession or overt punishment. For example, Nanjing grid manager Li Aihua found out from the

grid system that residents who had lost money in an illegal investment scheme, recently broken up

by the local public security bureau, planned to visit Beijing to complain. That information included

the names of the petition organizers, so grid manager Li visited the home of one organizer, Resident

Ding, with a mediator (tiaojieyuan, 调解员) and police official (minjing, 民警). During her visit to

120 Nanjing Municipal CCP, “Wo shenbian de wanggeyuan [Grid patrolman alongside me],” Politics and Law Committee, 25 June 2018, http://www.njzf.gov.cn/cxwghshzl/201806/t20180625_5469071.html 121 Liao and Tsai 2018; Nanjing Municipal CCP, “Wo shenbian de wanggeyuan [Grid patrolman alongside me],” Politics and Law Committee, 4 June 2018, http://www.njzf.gov.cn/cxwghshzl/201806/t20180604_5416204.html

122 Liao and Tsai 2018:6. 123 Tong and Lei 2010.

34

Resident Ding’s home, Li helped the resident “analyze his own responsibility for participating in

illegal fundraising” and consider the “severe consequences and risks” (yanzhong houguo he weihai, 严重

后果和危害) of seeking extreme solutions to his problem. Not only did she convince Resident Ding

to stand down from his attempt to organize and petition in Beijing, the resident subsequently

actively cooperated with authorities to help manage other frustrated investors, thereby protecting

against a group incident.124 Other articles tell similar stories about grid management’s success in

preventing petitions and other collective actions that could damage social stability (and the careers of

local officials). Again, the GCP’s role is both to identify problems early, and to guide officials in

both their selection of a coercive strategy, and its actual implementation in the privacy of a resident’s

home.

Pre-emption. Finally, platforms like the GCP can provide information for local officials to

pre-emptively suppress contention by force, allowing them to do so in a more targeted and

individualized fashion than they would be able to do absent GCP-generated information. As shown

in Figure 2 above, pre-emptive approaches tend to be employed when information indicates that

concession or coercion are unlikely to work, or when previous efforts to use the other two strategies

of demobilization have been tried and proved ineffective.

There are clear examples of GCP-like platforms generating the information that enables pre-

emption as a form of demobilization. In one township in M District, for example, more than 5,000

houses have been demolished since 2011, and hundreds of angry petitioners had planned to protest

and travel to Beijing to deliver a petition to higher authorities. Using information from the GCP

grids, township officials were able to identify the movement’s leaders, detect their plans, and

discover the specific details of their complaints and intended appeals to central authorities. Because

their travel to Beijing was imminent, local officials used that information to locate the

124 Nanjing Municipal CCP, “Wo shenbian de wanggehua [Grid patrolman alongside me],” Politics and Law Committee, 4 June 2018, http://www.njzf.gov.cn/cxwghshzl/201806/t20180604_5416204.html

35

“troublemakers” and their associates, block travel pathways to Beijing, and demobilize planned

demonstrations.125 In another instance, the New York Times recounted that a protest in Beijing over a

fraudulent financial scheme, with an estimated 8,000 intended participants, fizzled because majority

were intercepted and physically detained en route to the protest location.126

Profiles of grid managers often highlight their roles in providing information to public

security to arrest criminals,127 but the same process can be used to demobilize contention using pre-

emptive interception and arrest. Scholarly work on the petition system, for example, has shown that

during “sensitive periods” such as regime anniversaries, petitioners are targeted for surveillance,

blocked from entering bus/rail stations, and even physically detained in isolation (often temporary

house arrest) to pre-empt their pursuit of contentious action.128 Petition history is one part of the

GCP database, and can be merged with other information to determine someone’s history and

patterns of contentious behavior, showing authorities what an individual is likely to do, what

demobilizational strategies have been tried in the past, and their results. Thus if other approaches

have previously failed, or are judged unlikely to succeed, GCP-generated information can be used

operationally, to target individuals and pre-empt their ability to act.

Although this last option is what most people probably think of when they see the term

“preventive repression,” in fact, all three strategies are forms of preventive social control that

interfere with contentious behavior before it escalates or becomes publicly visible. The state does 125 Interview, township police officer, 20 December 2015.

126 Qian 2019; Liu 2018; see also “Police Lock Down Beijing’s Financial District to Thwart Protests,” Financial Times, 6 August 2018, https://www.ft.com/content/58728458-9943-11e8-9702-5946bae86e6d

127 “Weixian! Fuhai Jiedao yi nanzi zai loudao nei guyi dakai meiqi ping famen houli qu,” Shenzhen News Network, 23 March 2019, http://www.sznews.com/news/content/2019-03/23/content_21497558.htm; “Wanggeyuan yanzhong de hanjiang wanggehua zhili fuwu moshi,” Fujian Fazhi Baoshe (Fujian CCP Political-Legal Committee), 6 July 2016, http://www.pafj.net/html/2016/toutiao_0706/56954.html; “Xixiang wanggeyuan yi ri lianzhu jingfang zhuahao 2 ming diaoqie diandongche xianfan,” Nanfang Dushibao, 26 October 2018, http://epaper.oeeee.com/epaper/H/html/2018-10/26/content_55063.htm 128 Chen forthcoming; Liao and Tsai 2018; Li et al 2012; “Wanggeyuan yanzhong de hanjiang wanggehua zhili fuwu moshi,” Fujian Fazhi Baoshe (Fujian CCP Political-Legal Committee), 6 July 2016, http://www.pafj.net/html/2016/toutiao_0706/56954.html

36

not simply choose whether to engage in pre-emptive repression; it engages in a more complex

problem of identifying a course of action from a range of possible demobilizational strategies. This

targeted, sometimes-private, and differentiated demobilization atomizes citizens’ experiences of

repression, thereby limiting their potential to become the basis of collective identity formation and

group-based mobilization.

The approach described in M District has been applied widely within China, and has grown

in sophistication and intensity as technological capabilities have advanced. By 2015, 168 of China’s

332 prefectural-level cities were using grid management for social control, with plans for future

expansion.129 Under initiatives such as Skynet (天网, Tianwang) or the “Bright Snow Project” (雪亮工

程, Xueliang Gongcheng, also called “Sharp Eyes”), video surveillance and facial recognition are being

integrated into grid management platforms, and officials are learning to use the data-integration

platforms to identify threats via predictive analytics.130 The Xueliang Gongcheng is now operating in all

provinces, and China’s 13th Five Year Plan calls for implementation to be completed by 2020.131 In

March 2018, a graduate student working on surveillance in Hunan province tested the system run by

his local PSB; it took the police four minutes and fifteen seconds to locate him, and just over five

minutes to take him into custody.132

129 Lin 2015; NDRC 2015. 130 Tianwang comes from a saying “Tianwang huihui, shu er bulou (Heaven’s net has large mesh, but lets nothing through)”; while Xueliang comes from the saying, “Renmin qunzhong de yanjing shi xueliang de (the eyes of the people and the masses are as bright as snow).”

131 “农村雪亮工程首入一号文件 公共安全监控如何全覆盖 [Nongcun Xueliang Gongcheng shouru yi bao wenjian gongong anquan jiankong ruhe quan fugai],” 法制日报 [Legal Daily], 13 February 2018, http://www.xinhuanet.com/2018-02/13/c_1122411094.htm

132 “天网工程智能人脸监控系统再迎挑战者 5 分 22 秒被抓获,” Security knowledge network [安防知识网], 19 March 2018, http://security.asmag.com.cn/news/201803/93606.html

37

International Implications: China’s Export of Surveillance and Policing Technology

The development of these surveillance and policing technology is not just relevant within

China; it has already had global consequences. Currently, the extent to which these technologies

have been adopted worldwide is unknown, and there is considerable debate over why third countries

are pursuing surveillance and policing technologies provided by China. Policymakers and scholars in

the U.S. fear that China is exporting these technologies to “make the world safe for autocracy,”

while suppliers and adopters tend to emphasize the need to deal with crime and the importance of

public safety for attracting tourism and investment.133 A full analysis of what determines patterns in

the diffusion of these technologies is beyond the scope of this paper, but the following section

provides descriptive data to indicate the importance of the phenomena for contemporary

international politics and offer some initial thoughts on patterns in the data.134

The analysis draws on a new dataset, constructed from corporate, media, and government

reports in English, Chinese, Spanish, French, Arabic, and Russian, that codes the adoption of

Chinese surveillance and policing platforms worldwide from 2008 to 2019. Sometimes referred to as

“Safe City” (安全城市) solutions – the term used by these platforms’ largest global supplier,

Huawei135– these platforms are generally characterized by a big-data integration and predictive

analytics platform that supports one or more high-tech command-and-control centers, and analyzes,

in close to real time, information inputs from a wide range of sources (criminal records, other

government databases, networked surveillance cameras, facial and license plate recognition

applications, etc). Huawei, for example, describes its Safe City Solutions as providing:

…[W]orld-leading Collaborative-C4ISR [Command, Control, Communication, Cloud, Intelligence, Surveillance, and Reconnaissance] solutions that enable crucial visualization and convergence to maximize public safety….The safe city solutions jointly developed by Huawei and our partners enable advance prevention, precise

133 Myat Pyae Pho 2019. 134 Eventually, I hope to spin this analysis off into a separate paper focused on the drivers of diffusion of surveillance and policing technology. 135 Huawei is the world’s biggest provider of network telecommunications equipment; Vaswani 2019.

38

resource allocation, efficient analysis, visualized command, and efficient coordination among multiple departments. They help governments reduce crime rates and prevent and respond to crises more effectively, ensuring a safer environment for all.136

Elsewhere in Huawei’s materials, the central data-integration and analytics platform is referred to as

the “nerve center” and various collection devices are analogized to senses. In addition to Huawei,

China National Electronics Import and Export Corporation (CEIEC), ZTE, and other Chinese

companies are also commonly involved in the provision of PST platforms and technology. It is not

uncommon for multiple companies to be involved as vendors or contractors in different parts of a

project, and the companies involved also sometimes provide consulting or training in how to

employ the technology in addition to selling the tools themselves.

This new data demonstrates that policy analyses have significantly understated the extent of

use of Chinese surveillance and policing technologies around the world, as well as the speed with

which they have been adopted globally. For example, a 2018 Freedom House report discussed use

of Chinese surveillance technology in 18 countries; an April 2019 report by the Australian Strategic

Policy Institute covered 43 countries.137 By contrast, Huawei’s 2018 annual report stated that its Safe

City Solutions “now serve over 700 cities across more than 100 countries and regions,” a figure that

had tripled since 2015 (see Figure 4).138

136 Huawei 2017a: 31. Safe City platforms are a subcomponent of “Smart City” projects, which Huawei describes as providing a “nervous system” for the city; SafeCity platforms are specifically focused on public safety. Huawei 2017b.

137 Shahbaz 2018; Cave et al 2019. 138 In 2015, Huawei reported serving 400 million people in over 100 cities across over 30 countries; in 2016, over 800 million people in 200+ cities in 80+ countries; in 2017, over 1 billion people in 230 cities in 90 countries. Huawei 2015: 28; Huawei 2016: 4; Huawei 2017a: 31; Huawei 2018: 30.

39

Figure 4: Number of Safe City Projects Reported by Huawei (by Country & City)

Corporate self-reporting on the number of countries is subject to two possible and cross-cutting

forms of bias. On the one hand, suppliers may have incentives to exaggerate the numbers to

advertise the popularity of their products; on the other hand, as scrutiny of these companies

(especially Huawei) has increased, companies may downplay the extent of their involvement or seek

to hide it.139 The research team therefore sought to independently confirm specific cases of adoption

on a country-by-country basis, and has independently confirmed adoption in approximately 80

countries thus far.140 The number and location of these adopting countries are shown in Figures 5

and 6.

139 Huawei declined to provide a list of the 100+ countries mentioned in its annual report, citing non-disclosure agreements with clients. Huawei, personal communication with the author, June 2019. 140 The use of countries as the unit of analysis is somewhat misleading. In many cases, the original contract or memorandum of understanding is signed by a subnational government (region or municipality). Huawei’s term “Safe City” also suggests the original scale envisioned for the platform, though region- and nation-wide systems have been established. The heterogeneity of subnational units involved, combined with availability of independent variable data at the country-year level, make this the most feasible design at present.

0

200

400

600

800

Citie

s

0

20

40

60

80

100

120Co

untri

es

2013 2014 2015 2016 2017 2018

Countries Cities

Source: Huawei annual reports (English, 2013-2018).Figures include Safe City projects in China.

40

Figure 5: Chinese Public Security Technology Platform Adoptions (2008-18)

Figure 6: Locations of Chinese Public Security Technology Platform Adoptions (2008-2019)

41

What factors might be driving the spread of these technologies worldwide? Current literature, drawn

from both scholarly work and policy analysis, suggests four main possible explanations.

! First, analysts have suggested that China is supplying these technologies to countries as “a

key component of Chinese geopolitical strategy.” 141

! A second family of explanations argues that China is “making the world safe for autocracy”

by exporting these technologies to bolster authoritarian control. 142 This could be interpreted

in two ways: that these technologies are more likely to be adopted in less democratic regimes,

or that weak/unstable countries are more likely to adopt these technologies. (This could also

be a statement about the effects of adopting these technologies, rather than drivers, but

effects are beyond the scope of this paper’s analysis.)

! Third, adopting these technologies could be a straightforward response to high crime

rates.143

! Fourth, the literature on military diffusion suggests that adoption of innovative technologies

is facilitated by higher financial and organizational capacity. We are skeptical of the

applicability to China for several reasons: the financial costs of these platforms are low

compared to advanced military hardware; platform adoption is often subsidized in various

ways by China; and governments that are struggling to perform may have higher incentives

to pursue adoption, even if they subsequently struggle with implementation.

It would be methodologically problematic to attempt to make inferences about the drivers or causes

of the adoption of Chinese surveillance and policing technology based on the dataset as it stands

now, for several reasons. Not only are these observational data, but missing data is a significant 141 Feldstein 2019a; 2019c. 142 Weiss 2019; see also the quote from a White House official in Garrison 2019. On China’s role in promoting autocracy (or not), see Tansey et al 2017; Bader 2015; von Soest 2015; Mattes and Rodriguez 2014; Kleine-Ahlbrandt and Small, 2008; Kastner & Saunders 2012. Comparatively, see Way 2015; Tolstrup 2015; Vanderhill 2013.

143 Huawei, for example, advertises Safe City platforms as “public safety solutions,” and its “global chief public safety scientist” has claimed that violent crime costs Latin America $261 billion annually (3% of GDP). It advertises that adoption of the Safe City platform in Nairobi and Mombasa reduced crime by 46% in the first year of operation, and boosted tourism (a major revenue source for Kenya) by 14%, and its agreement with the Philippines sets a target of 15% reduction in the crime rate. Koh 2018; see also Huawei n.d.; O’Rourke and Choy 2019; Siu 2019; Elemia 2018.

42

problem on several independent variables (approaching 50% in some cases), and missingness is

likely non-random.144 Moreover, there is heterogeneity in the dependent variable observations that

are coded as zeroes: some of these observations are true non-adopters, while others are missing

adopters; publicly available sources do not (yet) allow us to distinguish between the two with

confidence.

Some preliminary descriptive analysis, looking for initial associations in the data, appears in

the Appendix.145 Cumulatively, this preliminary analysis suggests the following. Chinese strategic

partnership appears to be positively and significantly associated with the adoption of Chinese

surveillance and policing technology platforms. Violent crime also correlates positively with the

adoption of Chinese surveillance and policing platforms. Both of these correlations are in the

expected direction. By contrast, neither the level of democracy nor the level of political instability in

a country correlates significantly with the likelihood of adoption – suggesting that any ideas about

China exporting this technology to “make the world safe for autocracy” hinge on the post-adoption

effects of this technology, rather than explaining where it is exported and adopted. Finally, variables

generated from theories of military-technological diffusion return mixed results: consistent with

existing theories, financial capacity is positively correlated with adoption.146 Contrary to existing

theory, however, government effectiveness is negatively correlated with adoption, meaning that less

effective governments are more likely to agree to adopt surveillance and policing technology from

China (a result that makes some sense given the findings on crime, above). Together, this seems to

144 For debate on how to deal with missingness when it is MAR/MNAR, especially the pros and cons of multiple imputation, see King et al 2001; Lall 2016; Pepinsky 2018.

145 The Appendix includes variable descriptions, summary statistics, bivariate correlations, and a descriptive logit. There are valid theoretical arguments for using either a logit (explaining adoption) or a Cox model (explaining survival/time to adoption). Although statistical significance is slightly lower in the Cox model, the two models return very similar results, with the exception that when strategic importance is operationalized as ordinal rather than binary in the Cox model, GDP per capita (logged) loses statistical significance. I have included the logit results in the Appendix for ease of interpretation. 146 This is somewhat surprising given the anecdotal evidence of Chinese subsidization of these projects, but consistent with theories on military diffusion, such as Horowitz (2010)’s adoption-capacity theory.

43

suggest a divergence in patterns of technological adoption for military vs. domestic security

technology. Again, however, given the data quality issues and potential threats to causal inference,

these results should be interpreted as exploratory and descriptive.

Conclusion

Under Xi Jinping, China has pursued a surveillance state of immense scale and ambition.

This paper suggests that the CCP has pursued intensified surveillance in order to enhance its

capacity for preventive repression. Information facilitates preventive repression because it not only

reveals citizens’ underlying distribution of preferences, but matches individuals to specific locations

on that distribution, identifying both the type and level of grievance they possess. It then also

facilitates strategy selection, allowing the regime to assign individuals to one of several differentiated

demobilization strategies in a process that maximizes the likelihood of success and minimizes risks

of backlash. Qualitative evidence from within a local district’s coercive apparatus demonstrated that

leaders prioritize prevention, see information as critical to achieving preventive coercive capacity,

and seek to employ surveillance and policing technologies to solve specific information problems

and pursue differentiated demobilization of individual citizens.

Theoretically, the project finds that information shapes the balance between concession,

coercion, and pre-emption in authoritarian regimes, and builds on existing literature by showing that

there are individual-level variations in a regime’s intelligence capacity vis-à-vis citizens even under

conditions of (relative) ethnic homogeneity. Practically, the rise of the Chinese surveillance state has

had global consequences; over the course of the past decade, Chinese surveillance and policing

technologies have been adopted in more than 80 countries worldwide, both democratic and

autocratic, on every continent except Australia (and Antarctica). As noted above, therefore, China is

an ‘index case’ for a set of technological and informational conditions that have spread to such an

extent that they now characterize political life in much of the world.

44

Many important questions remain unanswered at present. One of the central questions to be

pursued is simply: does this technology work – and if so, what results does it produce, under what

conditions? Future research could examine the efficacy of this technology for both political control

and crime reduction, and explore how different social, political, and bureaucratic factors that vary

across implementation contexts might generate variations in effectiveness, inside China and abroad.

Another line of research could analyze citizen reactions to technologically-based surveillance and

differentiated demobilization using survey or experimental approaches. Finally, the preliminary data

on international diffusion of Chinese surveillance and policing technology provided above would

benefit greatly from further empirical refinement, as well as comparative analysis and theorization:

how did great powers in previous eras disseminate models of public security and police technology,

and in what ways are these patterns similar or different to those of China today? What does a

comparison of these patterns tell us about the drivers of this diffusionary process?

The project’s findings also have important implications for policy. For American foreign

and security policy, the data provided here suggest that rhetoric expressing concern over about

China’s “digital authoritarianism” 147 should begin to incorporate a much more nuanced

understanding of the incentives and interests of adopting countries, and an acknowledgment that

these technologies are now widespread in democratic as well as autocratic political systems. Given

the virtual absence of de-adoption from the international system thus far, it may make sense for the

U.S. or the international community to focus more on establishing strong norms of legal constraint

around these technologies’ use than to try (belatedly) to prevent their adoption altogether. Indeed,

UN Special Rapporteur David Kaye recently called for the development of global standards to

regulate the domestic use and international export of surveillance technology, by both government

147 China’s expansion into the overseas “safe city” market has already raised concerns in the Pentagon that the move will increase these companies’ “access to foreign talent and data” in ways that are not beneficial to the U.S. or its partners, and other work expresses concern that the growing use of Chinese surveillance and policing technology in third countries could bifurcate the world into adopters and non-adopters. U.S. Department of Defense 2019: 101; Hammond 2019.

45

and private actors.148 The findings contained here about the rapid proliferation and contemporary

prevalence of this technology both confirm the importance of that discussion, and also provide

some indication of how difficult it is likely to be.

148 UN 2019.

46

APPENDIX:

VARIABLES & EXPLORATORY DATA ANALYSIS

Explanation of Variables:

Chinese Surveillance Technology Adoption (dependent variable): Binary variable coded 1 for the year in which the initial contract/agreement to adopt was signed, and every year thereafter. A separate Appendix with sources used to justify coding decisions will be published online.

Strategic Importance: Measured using an ordinal variable (0-4) created by the author to categorize variations in the level of “strategic partnership” listed on the PRC Ministry of Foreign Affairs’ webpage.149 Robustness checks include a binary version of this variable, and a binary variable measuring membership in the Belt and Road Initiative.

Level of Democracy: We test whether less democratic countries are more likely to adopt Chinese surveillance and policing technology using Polity (-10 to 10 scale).

Political Stability: Measured using V-Dem’s Political Stability indicator (-3 to 2), which “combines several indicators [that] measure perceptions of the likelihood that the government in power will be destabilized or overthrown by possible unconstitutional and/or violent means, including domestic violence and terrorism.” 150

Violent Crime: Measured using homicide statistics from the Igarape Institute’s Homicide Monitor, which uses and expands on World Bank and United Nations Office of Drugs and Crime data.151

Financial Capacity: Measured using GDP per capita. Robustness checks employ military spending to proxy for willingness/capacity to spend on security.152

Organizational/Bureaucratic Capacity: Operationalized using the World Bank’s Government Effectiveness indicator, which assesses the capacity of government bureaucracies (defined as “Perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies.”153

149 PRC MOFA 2019; on interpreting these partnerships, see Feng and Huang 2014. Note that China does not pursue formal alliances, with the partial exception of North Korea, which we code at the highest level of strategic partnership. On the strategic logic of arms transfers, see Spindel 2018; Yarhi-Milo et al 2016; Johnson and Willardson 2018.

150 See https://www.v-dem.net/. 151 Data provided to the author upon request.

152 Conventional wisdom is to use military spending to proxy internal coercive capacity. Albertus and Menaldo 2012; Bellin 2004; Wang 2014; for a China-specific critique of these approaches, see Greitens 2017.

153https://govdata360.worldbank.org/indicators/h1c9d2797?country=BRA&indicator=388&viz=line_chart&years=1996,2017

Table 1: Summary Statistics

Obs. Mean SD Min MaxChina surveillance technology 2153 0.108 0.310 0 1China partnership (binary) 2153 0.262 0.440 0 1China partnership (ordinal) 2153 0.756 1.404 0 4BRI membership 2153 0.105 0.306 0 1Polity 1592 4.162 6.131 -10 10Political Stability 1727 -0.189 0.966 -3 2Homicide number (Logged) 1284 5.519 2.252 0 11GDPPC (Logged) 2072 8.613 1.471 5 12Military Spending (Logged) 1639 6.727 2.312 -3 13Government Effectiveness 1917 -0.072 0.991 -2 2

Table 2: Bivariate Correlations with Chinese Surveillance TechnologyExpected Direction Indicator Correlation

H1 + Strategic Partnership (binary) 0.30Strategic Partnership (ordinal) 0.35BRI Membership 0.31

H2 − Polity -0.09

H3 − V-dem: Political Stability -0.01

H4 + Homicide (World Bank) 0.01Homicide Number (Homicide Monitor) 0.15Homicide Number (Homicide Monitor) (Logged) 0.20Homicide Rate (Homicide Monitor) 0.04

H5 + GDP per capita -0.05GDP per capita (logged) 0.03Military Spending -0.004Military Spending (logged) 0.14

H6 + World Bank: Gov’t Effectiveness -0.05

1

Table 3: Factors Associated with the Likelihood of Chinese Surveillance Tech Adoption(Main Model: Logistic Regression with Duration Polynomials)

Chinese surveillance technologyChina partnership (ordinal) 0.454∗∗

(1.998)

Polity -0.026(-0.317)

Political stability 1.310(1.523)

Homicide number (Logged) 0.549∗∗

(3.258)

GDPPC (Logged) 1.428∗∗

(2.409)

Government effectiveness -2.396∗∗∗

(-3.875)

Duration (demeaned) -7.768∗∗

(-2.893)

Duration2 (demeaned) 1.049∗∗

(2.455)

Duration3 (demeaned) -0.040∗∗

(-2.009)

Constant -20.265∗∗

(-2.904)Pseudo-R2 0.7322Observations 1055

t statistics in parentheses∗ p < 0.10, ∗∗ p < 0.05, ∗∗∗ p < 0.001

2

1

References

Albertus, Michael, and Victor Menaldo. 2012. “Coercive Capacity and Prospects for Democratization.” Comparative Politics 44(2).

Thomas Ambrosio and Jakob Tolstrup, “How do we tell authoritarian diffusion from illusion?” Quality and Quantity (2019).

Alvaro Artigas, “Surveillance, Smart technologies, and the development of Safe City solutions: the case of Chinese ICT firms and their international expansion to emerging markets,”Institut Barcelona Estudis Internacionals (IBEI), 2017.

Julia Bader, “Propping up dictators? Economic cooperation from China and its impact on authoritarian persistence in party and non-party regimes,” European Journal of Political Research 54, no.4 (2015)

Bakken, Børge, ed. 2005. Crime, Punishment, and Policing in China. Rowman & Littlefield.

Bayat, A. 2013. Life as Politics: How Ordinary People Change the Middle East. Stanford University Press.

Becker, J. 2012. “The Knowledge to Act: Chinese Migrant Labor Protests in Comparative Perspective.” Comparative Political Studies, 45(11): 1379-1404.

Beinin, J. and F. Vairel, eds. 2011. Social Movements, Mobilization, and Contestation in the Middle East and North Africa. Stanford University Press.

Beissinger, Mark. 2002. Nationalist Mobilization and the Collapse of the Soviet State. Cambridge University Press.

Beissinger, Mark. 2007. “Structure and Example in Modular Political Phenomena: The Diffusion of the Bulldozer/Rose/Orange/Tulip Revolutions.” Perspectives on Politics 5, pp. 259-76.

Bellin, Eva. 2004. “The Robustness of Authoritarianism in the Middle East.” Comparative Politics 36(2):139-158.

----------. 2012. "Reconsidering the Robustness of Authoritarianism in the Middle East: Lessons from the Arab Spring." Comparative Politics 44(2): 127-149.

Benford, R., and D. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment.” Annual Review of Sociology, 26, pp. 611-39.

Benney, Jonathan. 2016. “Weiwen at the Grassroots: China’s stability maintenance apparatus as a means of conflict resolution.” Journal of Contemporary China 25(99), pp. 389–405.

Biddulph, Sarah. 2007. Legal Reform and Administrative Detention in China. Cambridge University Press.

Blaydes, Lisa. 2011. Elections and Distributive Politics in Mubarak’s Egypt. Cambridge University Press.

Blaydes, Lisa. 2018. State of Repression: Iraq under Saddam Hussein. Princeton University Press.

Blanchard, Ben. “China passes controversial counter-terrorism law,” Reuters, 27 December 2015, at https://www.reuters.com/article/us-china-security-idUSKBN0UA07220151228.

2

Blanchard, Ben. “China expels former justice minister from party for graft,” Reuters, 14 October 2017, at https://www.reuters.com/article/us-china-congress-corruption/china-expels-former-justice-minister-from-party-for-graft-idUSKBN1CK027

Boix, Carles (2003). Democracy and Redistribution. Cambridge University Press.

Brehm, John, and Emerson Niou. 1997. “Police Patrol Versus Self-Policing: A Comparative Analysis of the Control Systems Used in the Former Soviet Union and Communist China.” Journal of Theoretical Politics, Vol. 9, No. 1: 107-30;

Brook, Timothy. 1998. Quelling the people: The military suppression of the Beijing democracy movement. Stanford University Press.

Brownlee, Jason, Tarek Masoud, and Andrew Reynolds. 2015. The Arab Spring: Pathways of Repression and Reform. New York: Oxford University Press.

Bruce, Gary S. 2003. “The prelude to nationwide surveillance in East Germany: Stasi operations and threat perceptions, 1945–1953.” Journal of Cold War Studies 5(2):3–31.

Buckley, Chris, “China is Detaining Muslims in Vast Numbers. The Goal: Transformation.” New York Times, 8 September 2018, https://www.nytimes.com/2018/09/08/world/asia/china-uighur-muslim-detention-camp.html

Buckley, Chris, Paul Mozur, and Austin Ramzy, “How China Turned a City Into a Prison,” New York Times, 4 April 2019, https://www.nytimes.com/interactive/2019/04/04/world/asia/xinjiang-china-surveillance-prison.html?action=click&module=Top%20Stories&pgtype=Homepage

Chinese Academy of Social Sciences [CASS], “Guonei Shequ Wanggehua Guanli Yanjiu Zongshu [Summary of Research on Internal Community Grid Management],” 28 October 2013, http://www.cssn.cn/sf/bwsf_gl/201312/t20131205_895684.shtml;

Cai, Yongshun. 2008. “Local governments and the suppression of popular resistance,” China Quarterly 193: 24–42.

Cai, Yongshun. 2010. Collective Resistance in China: Why Popular Protests Succeed or Fail. Stanford University Press.

Carey, Sabine. “The Use of Repression as a Response to Domestic Dissent,” Political Studies Vol. 58 (2010): 167–86.

Carter, Erin Baggott, and Brett Carter. 2019. “When Autocracies Threaten Citizens with Violence.” Working Paper.

Danielle Cave et al, Mapping China’s Tech Giants, Australian Strategic Policy Institute, 18 April 2019, https://www.aspi.org.au/report/mapping-chinas-tech-giants;

Chin, Josh, and Clement Burge. 2017. “Twelve Days in Xinjiang: How China’s Surveillance State Overwhelms Daily Life.” Wall Street Journal. 9 December. https://www.wsj.com/articles/twelve-days-in-xinjiang-how-chinas-surveillance-state-overwhelms-daily-life-1513700355

Charron, Nicholas, and Lapuente. 2011. “Which Dictators Produce Quality of Government?” Studies in Comparative International Development 46 (4), 397-423.

3

Chen, Huirong. Forthcoming. “The Myth of Authoritarian Institutions: Conversion of China’s Consultative Petition System.” China: an international journal.

Chen, Xiaoyan, and Ang, Peng Hwa. 2011. “Internet Police in China: Regulation, Scope and Myths.” In Online Society in China: Creating, Celebrating, and Instrumentalising the Online Carnival. David Herold and Peter Marolt, eds. New York: Routledge, pp. 40–52.

Chen, Baifeng. 2012a. “Teding zhie qunti shangfang de fa sheng jizhi [The Mechanisms of Professional Group Petitions].” Shehui Kexue [Social Science] .

Chen, Baifeng. 2012b. “Teding zhiye qunti shangfang de fasheng jizhi [Causes of Professional Petitioning].” Shehui Kexue [Social Science] 8:59–68.

Chen, Jidong, and Yiqing Xu. 2017a. "Why Do Authoritarian Regimes Allow Citizens to Voice Opinions Publicly?" Journal of Politics 79(3): 792-803.

Chen, Jidong and Yiqing Xu. 2017b. “Information Manipulation and Reform in Authoritarian Regimes.” Political Science Research and Methods 5(1):163—-178.

Chen, Jidong, Jennifer Pan, and Yiqing Xu. 2016. “Sources of Authoritarian Responsiveness: A Field Experiment in China.” American Journal of Political Science 60(2): 383-400.

Chen Quanguo, “陈全国：筑起反恐维稳的铜墙铁壁 确保新疆社会大局和谐稳定 [Build an Impenetrable Defensive Counterterrorism Wall to Ensure Xinjiang’s Overall Harmony and Social Stability],” China National Radio, 19 August 2017, http://news.cnr.cn/native/city/20170819/t20170819_523908751.shtml.

Chen, Xi. 2009. “Power of Troublemaking: Chinese Petitioners’ Tactics and Their Efficacy.” Comparative Politics 41(4):451–471.

----------. 2012. Social Protest and Contentious Authoritarianism in China. Cambridge University Press.

----------. 2017. “Origins of Informal Coercion in China.” Politics and Society 45(1): 67-89.

Chen Xiangyang, “Why China Needs the CNSC: Three Reasons/Requirements,” Zhongguo Guofan Bao [China National Defense], 19 November 2013;

Chen, Yung-fa. 1986. Making revolution: the communist movement in eastern and central China, 1937-1945. Berkeley.

Chenoweth, Erica and Maria Stephan (2011). Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press.

CCP 16th Central Committee, 4th plenary session, “Zhonggong Zhongyang gunyu jiaqiang dang de zhizheng nengli jianshe de jueding [Decisions on Enhancing Ruling Capacity],” 19 September 2004.

CCP Central Committee, “Decision of the Central Committee of the Communist Party of China on Some Major Issues Concerning Comprehensively Deepening the Reform,” China.org.cn, 16 January 2014, http://www.china.org.cn/china/third_plenary_session/2014-01/16/content_31212602.htm

“CCP Central Committee Publishes Plan on Deepening the Reform of Party and State Organizations,” Xinhua, 21 March 2018, www.gov.cn/zhengce/2018-03-21/content_5276191.htm#1.

4

“CCP Central Committee Decision Adjusts the Leadership and Command System of the People’s Armed Police,” Xinhua, 27 December 2017, www.xinhuanet.com/2017-12/27/c_1122175909.htm;

CCP Central Committee/PRC State Council, “Guanyu jiaqiang he wanshan chengxiang shequ zhili de yijian [Suggestions to Enhance/Improve Social Management in Urban and Rural Communities],” 12 June 2017.

CCP Central Committee/PRC State Council, “关于加强社会治安防控体系建设的意见 [Opinion Regarding Strengthening the Construction of a Societal Security Prevention and Control System],” 13 April 2015, http://www.gov.cn/xinwen/2015-04/13/content_2846013.htm

CCP Central Committee/PRC State Council, “Zhonggong zhongyang guowuyuan fachu ‘guanyu kaizhan sao hei chu e zhuanxiang douzheng de tongzhi’ [The Central Committee of the Communist Party of China and the State Council issue a "Notice on the Launch of the Special Struggle for Eliminating Evil and Eliminating Evil”],” Xinhua, 24 January 2018, at http://www.mps.gov.cn/n2255079/n5967516/n6203051/n6203341/c6204686/content.html

CCP Shanghai Municipal Committee/Shanghai Municipal Government, “Guanyu jinyibu chuangxin shehui zhili jiaqiang jiceng jianshe de yijian [Suggestions on Improving Innovations in Social Management and Enhancing Grassroots Construction],” 31 December 2014.

Central Military Commission [CMC], “Central Military Commission Opinion on Deepening National Defense and Military Reforms,” Xinhua, 1 January 2016, www.xinhuanet.com/2016-01/01/c_1117646695.htm.

Chen Fang, “Xinjiang qu jiduanhua diaocha [Investigating Extremism in Xinjiang],” Fenghuang zixun, October 12, 2015, at http://news.ifeng.com/mainland/special/xjqjdh/

Chen, Shaowei, and Chao Zhang. 2019. “From Suppressive to Proactive: Chinese Governments’ Media Control Strategies in Popular Protest.” China: an International Journal 17(2).

Josh Chin and Clement Burge, “Twelve Days in Xinjiang,” Wall Street Journal, 19 December 2017, https://www.wsj.com/articles/twelve-days-in-xinjiang-how-chinas-surveillance-state-overwhelms-daily-life-1513700355;

Cody, Edward. “Chines Police Bring Villagers to Heel After Latest Uprising,” Washington Post, 21 December 2005.

Condra, Luke, and Jacob Shapiro. 2012. “Who Takes the Blame? The Strategic Effects of Collateral Damage,” American Journal of Political Science, Vol. 56, No. 1 (January), 167-187.

Conrad, Courtenay. “Constrained Concessions: Beneficent Dictatorial Responses to Domestic Political Opposition.” International Studies Quarterly 55:1167-1187.

Conrad, Courtenay, and Emily Hencken Ritter. 2019. Contentious Compliance: Dissent and Repression under International Human Rights Law. Cambridge University Press.

Dallin, Alexander , GeorgeWBreslauer, et al. (1970). Political Terror in Communist Systems. Stanford University Press.

Danneman, Nathan, and Emily Hencken Ritter. 2014. “Contagious Rebellion and Pre-Emptive Repression,” Journal of Conflict Resolution 58(2): 254-79.

Davenport, Christian. “Multi-Dimensional Threat Perception and State Repression,” American Journal of

5

Political Science, Vol. 39, No. 3 (1995), pp. 683-713;

Davenport, Christian, Hank Johnston, and Carol Mueller, eds. 2004. Repression and Mobilization. University of Minnesota Press.

Davenport, Christian. 2007. “State Repression and Political Order.” Annual Review of Political Science 10: 1-23.

Chris Daw, “China’s Surveillance Technology on Show in London,” Spectator, 6 July 2019, https://www.spectator.co.uk/2019/07/chinas-surveillance-technology-is-terrifying-and-on-show-in-london/amp/

De Jaegher, Kris, and Britta Hoyer. 2018. “Preemptive Repression: Deterrence, Backfiring, Iron Fists, and Velvet Gloves.” Journal of Conflict Resolution: 1-26.

DellaPorta, Donna. 1995. Social Movements, Political Violence, and the State. Cambridge University Press.

Deng, Yanhua, and Kevin O'Brien. 2013. “Relational Repression in China: Using Social Ties to Demobilize Protesters.” China Quarterly 215: 533-552.

Dimitrov, Martin. 2015. “Internal Government Assessments of the Quality of Governance in China.” Studies in Comparative International Development 50(1): 50-72.

----------. 2014a. “Internal Government Assessments of the Quality of Governance inChina.” Studies in Comparative International Development pp. 1–23.

----------. 2014b. “What the Party wanted to Know: Citizen Complaints as a “Barometer of Public Opinion” in Communist Bulgaria.” East European Politics & Societies and Cultures 2: 271-295.

Distelhorst, Greg. 2017. “The Power of Empty Promises: Quasi-Democratic Institutions and Activism in China,” Comparative Political Studies 50(4): 464-98.

Distelhorst, Greg, and Yue Hou. 2017. “Constituency Service under Non-Democratic Rule: Evidence from China.” Journal of Politics 79(3): 1024-40.

Dong, Youhong. 2015. “Dachengshi jiceng zonghe zhili jizhi chuangxin de lujing xuanze” [Path selection in innovation of grassroots comprehensive governance mechanisms in big cities), Shanghai xingzheng xueyuan xuebao [Journal of Shanghai Administration Institute] 6:31–37.

Dooley, Ben. “‘Eradicate the tumors’: Chinese civilians drive Xinjiang crackdown on separatism,” Japan Times/AFP, 26 April 2018, https://www.japantimes.co.jp/news/2018/04/26/asia-pacific/eradicate-tumors-chinese-civilians-drive-xinjiang-crackdown-separatism/#.W8FlylK1vOQ

Dragu, Tiberiu, and Adam Przeworski. 2018. “Preventive Repression: Two Types of Moral Hazard.” American Political Science Review 113(1):77-87.

Du Zhongming[杜忠明], 毛泽东名言故事 [Mao Zedong's Quotes and Stories] (Beijing: Central Party Literature Press/中央文献出版社, 2006),

Dutton, Michael. 2005. Policing Chinese Politics: A History. Duke University Press.

Earl, Jenifer, Sarah Soule, and John McCarthy. 2003. “Protest Under Fire: Explaining the Policing of Protest.” American Sociological Review 68(4): 581-606.

6

Earl, Jennifer. 2003. “Tanks, tear gas, and taxes: Toward a theory of movement repression.” Sociological Theory 21(1), pp. 44–68.

Economist, “China wants eyes and ears on every street,” 30 June 2018, https://www.economist.com/china/2018/06/30/china-wants-eyes-and-ears-on-every-street;

Egorov, Georgy, Sergei Guriev and Konstantin Sonin. 2009. “Why Resource-poor Dictators Allow Freer Media: A Theory and Evidence from Panel Data.” American Political Science Review 103(4):645–668.

Elfstrom, Manfred. 2019. “Two Steps Forward, One Step Back: China’s Response to Labour Unrest.” China Quarterly (Online First).

Steven Feldstein. 2019a. “How Artificial Intelligence is Reshaping Repression,” Journal of Democracy, Vol. 30, No. 1 (January 2019), https://carnegieendowment.org/2019/01/09/how-artificial-intelligence-is-reshaping-repression-pub-78093

Steven Feldstein. 2019b. “How AI Systems Could Threaten Democracy,” Carnegie, 24 April 2019, https://carnegieendowment.org/2019/04/24/how-artificial-intelligence-systems-could-threaten-democracy-pub-78984;

Steve Feldstein. 2019c. “Artificial Intelligence and Digital Repression: Challenges to Global Governance,” working paper, 9 May 2019, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3374575

Felter, Joseph, and Jacob Shapiro. 2017. “Limiting Civilian Causalities as Part of a Winning Strategy: The Case of Courageous Restraint,” Daedalus, Vol. 146, Issue 1 (January), pp. 44-58.

Feng Zhongping and Huang Jing, “China’s Strategic Partnerships: Engaging With a Changing World,” European Strategic Partnership Observatory, No. 8 (Jun 2014).

“Fengqiao jingyan weihe lijiu mixin [The Fengqiao experience is still relevant],” People’s Daily, 12 November 2018, http://paper.people.com.cn/rmrb/html/2018-11/12/nw.D110000renmrb_20181112_4-01.htm.

Fewsmith, Joseph. China Since Tiananmen (Cambridge: Cambridge University Press, 2001);

Finkel, Evgeny. 2015. “The Phoenix Effect of State Repression: Jewish Resistance During the Holocaust.” American Political Science Review, 109(2): 339-353.

Florini, Ann M., Lai Hairong, and Yelin Tan. 2012. China Experiments: From Local Innovations to National Reform. Brookings Institution Press.

Richard Fontaine and Kara Frederick, “The Autocrat’s New Tookit,” Wall Street Journal, 15 March 2019, https://www.wsj.com/articles/the-autocrats-new-tool-kit-11552662637

Francisco, Ronald. 2005. “The Dictator’s Dilemma.” In Davenport and Mueller, eds., Repression and Mobilization (University of Minnesota Press).

Frantz, Erica and Andrea Kendall-Taylor. 2014. “A dictator’s toolkit: Understanding how cooptation affects repression in autocracies.” Journal of Peace Research 51(3), pp. 332–346.

Freeman, Will. 2010. “The accuracy of China’s ‘mass incidents’.” Financial Times. March 1.

7

Friedrich, Carl J and Zbigniew K Brzezinski (1965). Totalitarian dictatorship. Cambridge, MA: Harvard University Press.

Fu, Diana. 2017. “Disguised collective action in China.” Comparative Political Studies 50(4):499–527.

Fu, Diana. 2018. Mobilizing without the Masses: Control and Contention in China. Cambridge University Press.

Fu, Diana, and Greg Distelhorst. 2018. "Grassroots Participation and Repression under Hu Jintao and Xi Jinping." China Journal 79:100-122.

Fu Hualing, “Re-education through Labor in Historical Perspective,” China Quarterly, 185 (2005): 811-803;

Gallagher, Mary, and Jonathan K. Hanson. 2009. “Coalitions, Carrots, and Sticks: Economic Inequality and Authoritarian States.” PS: Political Science & Politics, 42(2), pp. 667-672.

Gandhi, Jennifer (2008). Political Institutions under Dictatorship. Cambridge: Cambridge University Press.

Gandhi, Jennifer and Adam Przeworski. 2006. “Cooperation, Cooptation, and Rebellion under Dictatorships.” Economics and Politics 18(1), pp. 1–26.

Gandhi, Jennifer, and Adam Przeworski. 2008. “Authoritarian Institutions and the Survival of Autocrats.” Comparative Political Studies.

Gao, Jie. 2016. ““Bypass the Lying Mouths”: How Does the CCP Tackle Information Distortion at Local Levels?” The China Quarterly 228: 950–969.

Cassandra Garrison, “Safe Like China: in Argentina, ZTE finds eager buyer for surveillance tech,” Reuters, 5 July 2019, https://www.reuters.com/article/us-argentina-china-zte-insight-idUSKCN1U00ZG.

Geddes, Barbara. 1999. "What Do We Know about Democratization after Twenty Years?" Annual Review of Political Science, 115-44.

Geddes, Barbara, Joseph Wright, and Erica Frantz. 2018. How Dictatorships Work. Cambridge University Press.

Gieseke, Jens. 2014. The History of the Stasi: East Germany's Secret Police, 1945-1990. Berghahn Books.

Goebel, Christian, and Lynette Ong. 2012. “Social Unrest in China: a birds-eye perspective.” Working Paper.

Gohdes, Anita. 2019. “Repression Technology: Internet Accessibility and State Violence.” American Journal of Political Science. Forthcoming.

Emily Goldman and Leslie Eliason, The Diffusion of Military Technology and Ideas (Stanford, 2003);

Edward Goldring and Sheena Chestnut Greitens, “Rethinking Democratic Diffusion: Bringing Regime Type Back In,” Comparative Political Studies, 2019.

Goldstone, Jack, and Charles Tilly. 2001. “Threat (and Opportunity): Popular Action and State Response in the Dynamics of Contentious Action.” In Aminzade et al, eds. Silence and Voice in the Study of Contentious Politics. Cambridge University Press.

8

Rosie Gray and Adrian Carrasquillo, “Documents Illuminate Ecuador’s Spying Practices,” Buzzfeed, 25 June 2013, https://www.buzzfeednews.com/article/rosiegray/exclusive-documents-illuminate-ecuadors-spying-practices

Gregory, Paul (2009). Terror by Quota: State Security from Lenin to Stalin (an Archival Study). New Haven: Yale University Press.

Gregory, Paul, Phillip Schroeder, and Konstantin Sonin. 2011. “Rational dictators and the killing of innocents: data from Stalin’s archives.” Journal of Comparative Economics, vol. 39(1), pp. 34-42.

Greitens, Sheena Chestnut. 2016. Dictators and Their Secret Police: Coercive Institutions and State Violence. Cambridge University Press.

--------------. 2017. “Rethinking China's Coercive Capacity: An Examination of PRC Domestic Security Spending, 1992–2012.” China Quarterly 232: 1002-1025.

--------------. 2019. “Domestic Security in China under Xi Jinping.” China Leadership Monitor (March). Online at www.prcleader.org/greitens.

Greitens, Sheena Chestnut, and Rory Truex. 2019. “Repressive Experiences Among China Scholars: new evidence from survey data.” China Quarterly (OnlineFirst).

Gunitsky, Seva. 2015. “Corrupting the Cyber Commons: Social Media as a Tool of Authoritarian Stability.” Perspectives on Politics 13(1): 42-54.

Guo, Xuezhi. 2012. China's security state: philosophy, evolution, politics. Cambridge University Press.

Guriev, Sergei and Daniel Treisman. 2015. “How modern dictators survive: An informational theory of the new authoritarianism.” National Bureau of Economic Research.

Haber, Stephen. 2007. “Authoritarian Government.” In Oxford Handbook of Political Economy, Barry Weingast and Donald Wittman, eds. New York: Oxford University Press p. 688-711.

Hagenloh, Paul. 2009. Stalin’s Police: Public Order and Mass Repression in the USSR. Baltimore: Johns Hopkins.

Kelly Hammond, “Reconfiguring Geopolitics in the Era of the Surveillance State: Uyghurs, the Chinese Party-State, and the Reshaping of Middle Eastern Politics,’ The Caravan (Stanford: Hoover Institution, 2019), https://www.hoover.org/research/reconfiguring-geopolitics-era-surveillance-state-uyghurs-chinese-party-state-and-reshaping

Hao Zaijin, China’s Secret Wars: A Record of Chinese Communist Intelligence and Defense Work (Beijing: Zuojia Chubanshe, 2005) / 郝在今, 中国秘密战: 中共情报、保卫工作纪实 (北京: 作家出版社, 2005).

Hassan, Mai. 2016. “The Strategic Shuffle: Ethnic Geography, the Internal Security Apparatus, and Elections in Kenya.” American Journal of Political Science 61(2): 382-395.

Heilmann, Sebastian, and Elizabeth Perry, eds. 2011. Mao's invisible hand: the political foundations of adaptive governance in China. Harvard University Press.

Michael Horowitz, The Diffusion of Military Power (Princeton, 2010).

9

Hou, Meng. 2003. “Sifa Gaige Beijing Xiaa de Zhengfa Zhili Fangshi [A Model of Political-Legal Governance in the Context of Legal Reform],” Huadong Zhengfa Xueyuan Xuebao [Eastern China Law & Politics University], No. 6, pp. 99-106.

Houle, C., Kayser, M. A., & Xiang, J. (2016). Diffusion or confusion? Clustered shocks and the conditional diffusion of democracy. International Organization, 70, 687-725.

Human Rights Watch [HRW], “’Eradicating Ideological Viruses’: China’s Campaign of Repression against Xinjiang’s Muslims,” 9 September 2018, p. 35.

Human Rights Watch [HRW], China’s Algorithms of Repression: Reverse Engineering a Xinjiang Police Mass Surveillance App (2019).

Hu Jintao, 18th Party Congress report, 8 November 2012.

Huawei. n.d. “Safe Cities: a Revolution Driven by New ICT.” https://e.huawei.com/us/publications/global/ict_insights/201608271037/ecosystem/201608271557. Accessed 15 August 2019.

Huawei. 2015. Annual Report. https://www-file.huawei.com/-/media/corporate/pdf/annual-report/annualreport2015_en.pdf

Huawei. 2016. Annual Report. https://www-file.huawei.com/-/media/CORPORATE/PDF/annual-report/AnnualReport2016_en.pdf?la=en.

Huawei. 2017a. Annual Report. https://www-file.huawei.com/-/media/corporate/pdf/annual-report/annual_report2017_en.pdf

Huawei. 2017b. “Sharing Global Practices at the SCEWC 2017 in Barcelona,” 14 November 2017, https://www.huawei.com/en/press-events/news/2017/11/Huawei-Smart-City-Nervous-System-SCEWC2017

Huawei. 2018. Annual Report. https://www-file.huawei.com/-/media/corporate/pdf/annual-report/annual_report2018_en_v2.pdf?la=zh

Huang, Guangzhi. 2017. “Policing Blacks in Guangzhou: How Public Security Constructs Blacks as SanFei,” Modern China, 45(2): 171-200.

Hurst, William. 2009. The Chinese Worker after Socialism. Cambridge University Press.

Igarape Institute, “Homicide Monitor,” accessed 17 June 2019, https://igarape.org.br/en/apps/homicide-monitor/.

“In Your Face: China’s All-Seeing State,” BBC, 10 December 2017, https://www.bbc.com/news/av/world-asia-china-42248056/in-your-face-china-s-all-seeing-state.

Johnston, Hank. 2006. “Let’s Get Small: the Dynamics of Small Contention in Repressive States.” Mobilization 11, pp. 195-212.

Richard Johnson and Spencer Willardson. 2018. “Human Rights and Democratic Arms Transfers: Rhetoric Versus Reality with Different Types of Major Weapon Systems.” International Studies Quarterly, Volume 62, Issue 2 (June): 453–464.

10

Ty Joplin, “China’s Newest Export? Policing Dissidents,” Albawaba, 31 May 2018, https://www.albawaba.com/news/china%E2%80%99s-newest-global-export-policing-dissidents-1139230;

Judt, Tony. 2005. Postwar: a history of Europe since 1945. Penguin.

Kalyvas, Stathis. “The Logic of Terrorism in Civil War,” Journal of Ethics Vol. 8, No. 1 (2004), pp. 98-137.

Kalyvas, Stathis. The Logic of Violence in Civil War (Cambridge, 2006).

Kang, X., and H. Han. 2008. “Graduated Controls: The State-Society Relationship in Contemporary China.” Modern China, 34, pp. 36-55.

Khan, Sulmaan Wasif. 2018. Haunted by Chaos: Chinese Grand Strategy from Mao Zedong to Xi Jinping. Harvard University Press.

King, Charles. 2012. “Can There Be a Political Science of the Holocaust?” Perspectives on Politics, 10(2): 323-41.

Gary King, James Honaker, Anne Joseph, and Kenneth Scheve, “Analyzing incomplete political science data: An alternative algorithm for multiple imputation,” American Political Science Review 95/1 (2001).

King, Gary, Jennifer Pan, and Molly Roberts. 2013. “How Censorship in China Allows Government Criticism but Silences Collective Expression.” American Political Science Review 102(2): 1-18.

King, Gary, Jennifer Pan and Margaret E Roberts. 2014. “Reverse-engineering censorship in China: Randomized experimentation and participant observation.” Science 345(6199):1–10.

Stephanie Kleine-Ahlbrandt and Andrew Small, “China’s New Dictatorship Diplomacy,” Foreign Affairs, January/February 2008, https://www.foreignaffairs.com/articles/asia/2008-06-01/chinas-new-dictatorship-diplomacy.

Kocher, Matthew, Thomas Pepinsky, and Stathis Kalyvas, “Aerial Bombing and Counterinsurgency in Vietnam,” American Journal of Political Science 55(2): 2011, pp. 201-18.

Koesel, Karrie. 2014. Religion Under Authoritarianism. Cambridge University Press.

Koesel, Karrie, & Valerie Bunce, “Diffusion-proofing: Russian and Chinese responses to waves of popular mobilizations against authoritarian rulers,” Perspectives on Politics, Vol. 11, No. 3 (2013), pp. 753-768.

Koh Hong-Eng, “How video cameras can make cities safer and contribute to economic growth,” South China Morning Post, 3 June 2018, https://www.scmp.com/comment/insight-opinion/article/2148860/big-brother-surveillance-how-video-cameras-can-make-cities;

Koss, Daniel. 2018. Where the Party Rules: Party-Based Authoritarianism and the Reach of the Chinese State. Cambridge University Press.

Kuran, Timur (1989). Sparks and prairie fires: A theory of unanticipated political revolution. Public Choice 61(1), pp. 41–74.

Kuran, Timur. 1991. “Now out of never: The element of surprise in the East European revolution of 1989.” World Politics 44(1), pp. 7–48.

11

Kuran, Timur. 1995. Private Truths, Public Lies: The Social Consequences of Preference Falsification. Harvard University Press.

Kurzman, Charles. 1996. “Structural opportunity and perceived opportunity in social-movement theory: the Iranian revolution of 1979.” American Sociological Review pp. 153–170.

Jeremy Kuzmarov, Modernizing Repression: Police Training and Nation-Building in the American Century (Univ. of Massachusetts Press, 2012).

LaFree, Gary, Min Xie, and Aila Matanock, “Contagious Diffusion of Worldwide Terrorism: Is it Less Common Than We Think?” Studies in Conflict and Terrorism, Vol. 16, No. 4 (2018), pp. 261-280.

Ranjit Lall, “How Multiple Imputation Makes a Difference,” Political Analysis, vol. 24 (2016), pp. 414-433.

Lampton, David. 2015. “Xi Jinping and the National Security Commission: policy coordination and political power.” Journal of Contemporary China 24(95): 759-777.

Landry, Pierre. 2008. Decentralized authoritarianism in China. Cambridge University Press.

Lee, Ching Kwan and Yonghong Zhang. 2013. “Unraveling the Microfoundations of Bargained Authoritarianism in China.” American Journal of Sociology 118: 475-508.

Lee, Ching Kwan. 2007. Against the Law: Labor Protests in China’s Rustbelt and Sunbelt. University of California Press.

Lee, Ching Kwan and Yuan Shen. 2014. “Seeing like the Grassroots State: Using Market and Mass Line for Stability Maintenance in China’s Bargained Authoritarianism.”

Leibold, James. 2019b. “Surveillance in China’s Xinjiang Region: Ethnic Sorting, Coercion, and Inducement,” Journal of Contemporary China, DOI: 10.1080/10670564.2019.1621529

Leventoglu, Bahar and Nils W. Metternich (2018). “Born weak, growing strong: Anti-government protests as a signal of rebel strength in the context of civil wars.” American Journal of Political Science 62(3), 581–596.

Levitsky, Steven, and Lucan Way. 2010. Competitive Authoritarianism. Cambridge University Press.

Lewis, Janet. 2017. “How Does Ethnic Rebellion Start?” Comparative Political Studies 50(10): 1420-1450.

Lewis, Janet, and Jennifer Larson. 2018. “Rumors, Kinship Networks, and Rebel Group Formation.” International Organization 72(Fall): 871–903.

Li, Lianjiang. 2008. “Political Trust and Petitioning in the Chinese Countryside.” Comparative Politics 40(2): 209–226.

Li, Lianjiang, Mingxing Liu and Kevin J. O'Brien, “Petitioning Beijing: the high tide of 2003–2006,” China Quarterly 210 (2012): 313-334.

Li Yukun, “赵克志: 防范抵御“颜色革命”, 打好政治安全保卫仗 [Zhao Kezhi: Prevent Color Revolutions and Defend Political Security,” Beijing News, 18 January 2019, http://www.bjnews.com.cn/news/2019/01/18/540767.html

12

Li Yu, ‘Hulianwang + shehui zhili yingyong tansuo’ [Probing the uses of the Internet for social governance], Renminwang, September 1, 2016, accessed May 9, 2018, http://theory.people.com.cn/n1/2016/0901/c40531-28682384.html.

Li, Pei, and Cate Cadell, “At a Beijing Security Fair, An Arms Race for Surveillance Tech,” Reuters, 30 May 2018, https://www.reuters.com/article/us-china-monitoring-tech-insight/at-beijing-security-fair-an-arms-race-for-surveillance-tech-idUSKCN1IV0OY;

Lichbach, Mark Irving. 1989. “An evaluation of “does economic inequality breed political conflict?" World Politics 41(4), pp. 431–470.

Lieberthal, Kenneth, and David M. Lampton. 1992. Bureaucracy, Politics, and Decision-Making in Post-Mao China. University of California Press.

Liao, Xingmiu, and Wen-Hsuan Tsai. 2018. “Managing Irregular Petitions in China.” Journal of East Asian Studies 10: 1-18.

Lin Zhongliang, Geiji Danwei Zhengfawei de Zhineng Ji Hongguan Zhengfa Gongzuo [The Political-Legal Affairs Work of Political-Legal Committees at All Levels of Government] (Beijing: Chang’an Chubanshe, 2004).

Lin, Xuefei. 2015. “Zhengfujian zuzhi xuexi yu zhengce zaishengchan [Organizational Learning among Governments and Policy Reproduction]”, Gonggong guanli xuebao [Journal of Public Management] 12 (1): 11-23.

Liu, Yujing, “Police Thwart Protest Rallies by Victims of China’s Underground Financial System,” South China Morning Post, 6 August 2018, https://www.scmp.com/business/banking-finance/article/2158533/police-thwart-protest-rallies-victims-chinas-underground

Lohmann, Susanne (1993). “A signaling model of informative and manipulative political action.” American Political Science Review 87(2), 319-333.

Lohmann, Susanne (1994a). “The dynamics of informational cascades: Monday Demonstrations in Leipzig, East Germany, 1989–91.” World Politics 47(1), 42–101.

Lohmann, Susanne (1994b). Information aggregation through costly political action. The American Economic Review, pp. 518–530.

Lorentzen, Peter L. 2013. “Regularized Rioting: Permitting Protest in an Authoritarian Regime.” Quarterly Journal of Political Science 8: 127-158.

----------. 2014. “China’s Strategic Censorship.” American Journal of Political Science 58(2): 402-414.

----------. 2017. “Designing Contentious Politics in Post-1989 China.” Modern China 43 (5): 459-493.

Loveman, M. “High Risk Collective Action: Defending Human Rights in Chile, Uruguay, and Argentina.” American Journal of Sociology 104, pp. 477-525.

Lucas, Louise, and Emily Feng, “Inside China’s Surveillance State,” Financial Times. 19 July 2018, https://www.ft.com/content/2182eebe-8a17-11e8-bf9e-8771d5404543;

Lyall, Jason. “Does Indiscriminate Violence Invite Insurgent Attacks? Evidence from Chechnya,” Journal of Conflict Resolution, Vol. 53, No. 3 (June 2009), pp. 331-62.

13

Ma Haoliang, “解放军高层调动大幕拉开 五正大军区级将领履新,” Ta Kung Pao, 25 December 2014, http://news.takungpao.com/mainland/topnews/2014-12/2869552.html.

Magaloni Beatriz. 2008. “Credible Power-Sharing and the Longevity of Authoritarian Rule.” Comparative Political Studies 41(4/5):715–741.

Magaloni, Beatriz, Alberto Diaz-Cayeros and Federico Estévez. 2007. “Clientelism and portfolio diversification: a model of electoral investment with applications to Mexico.” In Patrons, clients, and policies: Patterns of democratic accountability and political competition, pp. 182–205.

Niharika Mandhana, “Huawei’s Video Surveillance Business Hits a Snag in the Philippines,” Wall Street Journal, 20 February 2019, https://www.wsj.com/articles/huaweis-video-surveillance-business-hits-snag-in-philippines-11550683135.

Manion, Melanie. 2004. Corruption by Design: Building Clean Government in Mainland China and Hong Kong. Harvard University Press.

Manion, Melanie. 2015. Information for Autocrats: Representation in Chinese Local Congresses. Cambridge University Press.

Mann, Michael. 1984. “The autonomous power of the state.” European Journal of Sociology 25(2): 185-213.

Mares, Isabela and Matthew Carnes. 2009. “Social Policy in Developing Countries.” Annual Review of Political Science 12:93–113.

Marwell, Gerald and Pamela Oliver. 1993. The Critical Mass in Collective Action. Cambridge University Press.

Michaela Mattes and Mariana Rodriguez, “Autocracies and International Cooperation,” International Studies Quarterly Vol. 58, No. 3 (2014): 527-538.

Mattingly, Daniel. Forthcoming 2019. The Art of Political Control in China. Cambridge University Press.

McAdam, Sidney Tarrow, and Charles Tilly. 2001. Dynamics of Contention. Cambridge University Press.

McAdam, Doug, Sidney Tarrow and Charles Tilly. 2003. “Dynamics of contention.” Social Movement Studies 2(1), pp. 99–102.

Meirowitz, Adam and Joshua A Tucker (2013). “People power or a one-shot deal? A dynamic model of protest.” American Journal of Political Science 57(2), pp. 478–490.

Meng Jianzhu. 2008. “Shenru xuexi shijian kexue fazhanguan, zuo dangde zhongcheng weishi he renmin qunzhong de tiexinren [Learning and Practicing the Scientific Outlook of Development; Being a Loyal Defender of the Party and Intimate of the Masses],” Qiushi 21: 15-18.

Meng Jianzhu, “Tuijin gong’an xinxihua jianshe, goujian wu fengxi fangkong wangluo” [Promote public security informatization construction, create seamless prevention/control networks].” China News. 12 January 2009, http://www.chinanews.com/gn/news/2009/12-01/1992812.shtml.

“Meng Jianzhu: Qieshi tigao zhengfa jiguan fuwu daju de nengli he shuiping” [Effectively improve the ability and level of political and legal organs to serve the overall situation], Renmin fayuan bao, 18 March 2015, at http://www.court.gov.cn/fabu-xiangqing-13840.html

14

Meserve, Stephen A., and Daniel Pemstein. 2018. “Google Politics: The Political Determinants of Internet Censorship in Democracies.” Political Science Research and Methods 6(2): 245-63.

Metternich, Nils, Shahryar Minhas, and Michael D. Ward, “Firewall or Wall on Fire? A unified framework of the role of conflict contagion and the role of ethnic exclusion,” Journal of Conflict Resolution, Vol. 61, No. 6 (2015), pp. 1151-1173;

Millward, James. “’Re-educating’ Xinjiang’s Muslims,” New York Review of Books, 7 February 2019, https://www.nybooks.com/articles/2019/02/07/reeducating-xinjiangs-muslims/

Ministry of Defense, People’s Republic of China, “Full Text: China’s Military Strategy,” 26 May 2015, http://eng.mod.gov.cn/Press/2015-05/26/content_4586805.htm;

Ministry of Public Security, “Gong’an jiguan chuzhi quntixing zhi’an shijian guiding [Public security regulations on handling mass security incidents],” 5 April 2000.

Minzner, Carl. 2011. “China’s Turn Against Law.” American Journal of Comparative Law 59: 935-84.

Moore, Will H. 1998. “Repression and dissent: Substitution, context, and timing.” American Journal of Political Science 42(3):851–873.

Moore, Will. 2000. “The Repression of Dissent: A Substitution Model of Government Coercion,” Journal of Conflict Resolution, Vol. 44, No. 1, pp. 107–27.

Mozur, Paul. “Inside China’s Dystopian Dreams,” New York Times, 8 July 2018, https://www.nytimes.com/2018/07/08/business/china-surveillance-technology.html

Paul Mozur, “Made in China, Exported to the World: The Surveillance State,” New York Times, 24April 2019, https://www.nytimes.com/2019/04/24/technology/ecuador-surveillance-cameras-police-government.html.

Myat Pyae Pho. 2019. “Huawei to Supply Mandalay’s Safe City Project with Security Cameras, Equipment.” The Irrawaddy. May 9. https://www.irrawaddy.com/news/burma/huawei-supply-mandalays-safe-city-project-cameras-security-equipment.html

Ellen Nakashima, Gerry Shih, and John Hudson. 2019. “Leaked Documents Reveal Huawei’s Secret Operations to Build North Korea’s Wireless Network.” Washington Post. July 22.

Nanjing Municipal CCP, “Wo shenbian de wanggeyuan [Grid patrolman alongside me],” Politics and Law Committee, 25 June 2018, http://www.njzf.gov.cn/cxwghshzl/201806/t20180625_5469071.html

Nanjing Municipal CCP, “Wo shenbian de wanggeyuan [Grid patrolman alongside me],” Politics and Law Committee, 4 June 2018, http://www.njzf.gov.cn/cxwghshzl/201806/t20180604_5416204.html

National Development and Reform Commission [NDRC], “Guanyu jiaqiang gonggong anquan shipin jiankong jianshe lianwang [Strengthening Public Security Video Surveillance Network Construction],” 2015, http://www.ndrc.gov.cn/zcfb/zcfbtz/201505/t20150513_691578.html

National People’s Congress [NPC], 2017a. “Zhonghua Renmin Gongheguo Qingbaofa” [PRC Intelligence Law], 26 June 2017, http://www.npc.gov.cn/npc/xinwen/2017-06/27/content_2024529.htm.

NPC 2017b. “Explanation of the Draft Decision on Temporary Adjustments to Relevant Laws and Regulations of the People’s Armed Police During the Reform Period,” China National People’s Congress

15

(online), 8 November 2017, www.npc.gov.cn/npc/xinwen/2017-11/08/content_2032260.htm.

Nordas, Ragnhild, and Christian Davenport. 2013. “Fight the Youth: Demographic Bulges and State Repression.” American Journal of Political Science 57(4): 926-40.

Novaretti, Simona. ‘Social Governance vs. Social Management’, Comparative and Transnational Law 11, (2014) http://www.cdct.it/wp-content/uploads/2014/10/Master-Wp-Novaretti.pdf.

O’Brien, Kevin. 2008. Popular Protest in China. Cambridge, MA: Harvard University Press.

O’Brien, Kevin, and Yanhua Deng. 2015. “Repression Backfires: Tactical Radicalization and Protest Spectacle in Rural China.” Journal of Contemporary China 24(93): 457-70.

O’Brien, Kevin, and Yanhua Deng. 2017. “Preventing Protest One Person at a Time: Psychological Coercion and Relational Repression in China,” China Review, 17:2 (June): 179-201.

O’Brien, Kevin, and Lianjiang Li. 2006. Rightful resistance in rural China. Cambridge University Press.

O’Donnell, Guillermo, and Schmitter, Philippe. 1986. Transitions from Authoritarian Rule. Johns Hopkins University Press.

One Belt One Road Portal [中國一帶一路網], “已同中國簽訂共建一帶一路合作文件的國家一覽 [List of countries that have signed cooperation documents with China for construction of the Belt and Road],” 中國一帶一路網 [China One Belt One Road portal], 12 April 2019, https://www.yidaiyilu.gov.cn/info/iList.jsp?tm_id=126&cat_id=10122&info_id=77298

Ong, Lynette H. 2018 a. “Subcontracting of State Coercion and State Capacity in China.” Perspectives on Politics 16(3): 1-16.

-----------. 2018 b. "Thugs and Outsourcing of State Repression in China." China Journal 80 (1): 94-110.

Opp, K., and Roehl, W. 1990. “Repression, Mobilization, and Political Protest.” Social Forces 69, pp. 521-47.

Ben O’Rourke and Gigi Choy, “Big Brother Huawei Kitted Out this Philippine City,” South China Morning Post, 28 January 2019, https://www.scmp.com/week-asia/economics/article/2183540/big-brother-huawei-watches-philippine-city-does-china-too

Oster, Shai. “China Tries its Hand at Pre-Crime,” Bloomberg, 3 March 2016, https://www.bloomberg.com/news/articles/2016-03-03/china-tries-its-hand-at-pre-crime

Pan, Jennifer. 2019. Hush Money: How China Uses Surveillance and Social Welfare. Unpublished book Manuscript.

Pan, Jennifer, and Kaiping Chen. 2018. "Concealing Corruption: How Chinese Officials Distort Upward Reporting of Online Grievances." American Political Science Review 112 (3): 602-620.

Parkinson, Joe, Nicholas Bariyo, and Josh Chin. 2019. “Huawei Technicians Helped African Governments Spy on Political Opponents.” Wall Street Journal. 14 August. https://www.wsj.com/articles/huawei-technicians-helped-african-governments-spy-on-political-opponents-11565793017?mod=e2tw

Patane, Christopher. 2019. “Regime Rhetoric and Protest: How Government Statements Can Repress.” Terrorism and Political Violence (online first), pp. 1-42.

16

Pei, Minxin. 2012. “Is CCP Rule Fragile or Resilient?” Journal of Democracy 23(1): pp. 27-41.

People’s Republic of China (PRC), Ministry of Foreign Affairs. 2019. "Guojia he Zuzhi [Countries and international organizations].” https://www.fmprc.gov.cn/web/gjhdq_676201/gj_676203/yz_676205/

Thomas Pepinsky, “A Note on Listwise Deletion versus Multiple Imputation,” Political Analysis, vol. 26 (2018), pp. 480-488.

Elizabeth J. Perry, Shanghai on Strike: the politics of Chinese labor (Stanford: Stanford University Press, 1993).

Perry, Elizabeth. 2002. Challenging the Mandate of Heaven: Social Protest and State Power in China. Armork, NY: M. E. Sharpe.

Perry, Elizabeth J. 2006. Patrolling the Revolution: Worker Militias, Citizenship, and the Modern Chinese State. Lanham: Rowman & Littlefield.

Perry, Elizabeth. 2008. “Permanent Revolution? Continuities and Discontinuities in Chinese Protest.” In Popular Protest in China, Kevin O’Brien, ed. Cambridge, MA: Harvard University Press, pp. 205–216.

Perry, Elizabeth J. 2009. “A new rights consciousness?” Journal of Democracy 20(3), pp. 17–20.

Peter, Woelert. 2015. “Political rationality vs. technical rationality in China’s target-based performance measurement system: The case of social stability maintenance.” Policy and Society 34(1):37–48.

Pierskalla, Jan Henryk. 2010. “Protest, Deterrence, and Escalation: the strategic calculus of government repression.” Journal of Conflict Resolution 54(1), pp. 117-145.

“Policy document on innovation in social management issued,” 20 September 2011, online at http://www.china.com.cn/policy/txt/2011-09/20/content_23451516.htm.

Posen, Barry. 1984. The Sources of Military Doctrine: France, Britain, and Germany Between the World Wars. Cornell University Press.

Qian, Jianan. “Feeling Safe in the Surveillance State,” New York Times, 10 April 2019, https://www.nytimes.com/2019/04/10/opinion/china-internet-surveillance.html;

Qin, Bei, David Stromberg and Yanhui Wu. 2017. “Why Does China Allow Freer Social Media? Protests versus Surveillance and Propaganda.” Working Paper.

Rajagopalan, Megha. “They Thought They’d Left the Surveillance State Behind. They Were Wrong,” Buzzfeed, 9 July 2018, https://www.buzzfeednews.com/article/meghara/china-uighur-spies-surveillance

Repnikova, Maria. 2017. Media Politics in China: Improvising Power Under Authoritarianism. Cambridge University Press.

Reuter, Ora John and Graeme B Robertson (2015). “Legislatures, cooptation, and social protest in contemporary authoritarian regimes.” Journal of Politics 77(1), 235–248.

Reuter, Ora John, and David Szakonyi. 2013. “Online Social Media and Political Awareness in Authoritarian Regimes.” British Journal of Political Science 45(1): 29-51.

Ringen, Stein. 2016. The Perfect Dictatorship: China in the 21st Century. Hong Kong University Press.

17

Ritter, Emily Hencken, and Courtenay Conrad. 2016. “Preventing and Responding to Dissent: The Observational Challenges of Strategic Repression.” American Political Science Review 110(1): 85-99.

Roberts, Margaret. 2018. Censored: Distraction and Diversion Behind China’s Great Firewall. Princeton University Press.

Rød, Espen Geelmuyden, and Nils B. Weidmann. 2015. “Empowering Activists or Autocrats? The Internet in Authoritarian Regimes.” Journal of Peace Research 52(3): 338-51.

Roessler, Philip (2011). “The enemy within: Personal rule, coups, and civil war in Africa.” World Politics 63(2), pp. 300–346.

Sadowski, Jathan. “Potemkin AI,” 6 August 2018, https://reallifemag.com/potemkin-ai/

Saich, Tony, and Hans Van de Ven, eds. 1995. New Perspectives on the Chinese Communist Revolution. Armonk: M.E. Sharpe.

Sarotte, M.E. “China’s Fear of Contagion: Tiananmen Square and the Power of the European Example,” International Security, Vol. 37, No. 2 (2012), pp.156-182;

Saunders, Phillip C., ed., Chairman Xi Remakes the PLA (NDU Press, 2019);

Saunders, Phillip C. and Joel Wuthnow, “Large and in Charge: Civil-Military Relations under Xi Jinping,” in Saunders, ed., Chairman Xi Remakes the PLA (NDU Press, 2019); Joel

Schatz, Edward. 2009. “The Soft Authoritarian Toolkit: Agenda-Setting Power in Kazakhstan and Kyrgyzstan.” Comparative Politics, 41, pp. 203-22.

Schoenhals, Michael. 2013. Spying for the people: Mao's secret agents, 1949-1967. Cambridge University Press.

Schwarck, Edward. 2018. “Intelligence and Informatization: The Rise of the Ministry of Public Security in Intelligence Work in China,” China Journal 80: March.

Scoggins, Suzanne. 2019. Policing in the Shadow of Protest. Unpublished book manuscript.

Scoggins, Suzanne, and Kevin O'Brien. 2016. “China's Unhappy Police.” Asian Survey 56(2): 225-242.

Scott, James. 1985. Weapons of the Weak: Everyday Forms of Peasant Resistance. Yale University Press.

Scott, James. 1990. Domination and the Arts of Resistance: Hidden Transcripts. Yale University Press.

Adrian Shahbaz, Freedom on the Net 2018: The Rise of Digital Authoritarianism, (Washington, DC: Freedom House, October 2018), https://freedomhouse.org/report/freedom-net/freedom-net-2018/rise-digital-authoritarianism

Shadmehr, Mehdi. 2014. “Mobilization, repression, and revolution: grievances and opportunities in contentious politics.” Journal of Politics 76(3): 621–635.

Shambaugh, David. China’s Communist Party (Berkeley: UC Press, 2008).

Shirk, Susan. China: Fragile Superpower (Oxford University Press, 2007).

18

“Shuangpai wanggeyuan pingfan de yitian [An ordinary day for Shuangpai grid patrolmen],” Sanxiang Management Network, 16 April 2018, http://www.sxzzw.gov.cn/news/201804/16_170732_4196.html

Siegel, David. 2011. “When Does Repression Work?” Journal of Politics: 993-1010.

Skocpol, Theda. 1979. States and Social Revolutions. Cambridge University Press.

Slater, Dan. 2010. Ordering Power: Contentious Politics and Authoritarian Leviathans. Cambridge University Press.

Smith, Amanda. Thought Reform and China’s Dangerous Classes (Rowman and Littlefield, 2013);

Snow, D., and R. Benford. 1988. “Ideology, Frame Resonance, and Participant Mobilization.” International Social Movement Research, vol. 1(1), pp. 197-217.

Song Yongge, ‘Qianxi Wulumuqi shi shequ “sihua” guanli moshi’ [Review of Urumqi’s ‘four transformations’ model of social management], Zhonggong wulumuqi shiwei dangxiao xuebao 3, (2016).

Stern, Rachel, and Kevin O’Brien. 2012. “Politics at the Boundary: Mixed Signals and the Chinese State,” Modern China 38(2): 174-98.

Stockmann, Daniela. 2013. Media Commercialization and Authoritarian Rule in China. Cambridge University Press.

Strauss, Julia. 2002. “Paternalist Terror: The Campaign to Suppress Counterrevolutionaries and Regime Consolidation in the People's Republic of China, 1950-1953.” Comparative Studies in Society and History, (44)1, pp 80-105.

Strauss, Julia. 2014. “Communist Revolution and Political Terror.” Oxford Handbook of the History of Communism.

Su Yingjie. 1994. “Dui Danqian ‘Yanda’ Tongyi Xingdong Wenti de Jidian Sikao [Reflections on ‘Strike Hard’ Unified Operations’ Current Problems],” Gong’an Yanjiu 6: 22-35.

Sullivan, Christopher M. 2015. “Undermining Resistance: Mobilization, Repression, and the Enforcement of Political Order.” Journal of Conflict Resolution.

Sullivan, Christopher M. 2016. “Political Repression and Destruction of Dissident Organizations.” World Politics 68(4): 645-676.

Svolik, Milan. 2012. The Politics of Authoritarian Rule. Cambridge: Cambridge University Press.

Talmadge, Caitlin, and Varun Piplani. 2015. “When War Helps Civil-Military Relations: Prolonged Interstate Conflict and the Reduced Risk of Coup Attempts.” Journal of Conflict Resolution (February): 1-27.

Oisín Tansey, Kevin Koehler, and Alexander Schmotz, “Ties to the Rest: Autocratic Linkages and Regime Survival,” Comparative Political Studies Vol. 50, No. 9 (2017): 1221-1254;

Tanner, Harold. 1999. Strike Hard! Anti-Crime Campaigns and Chinese Criminal Justice 1979-89. New York: Cornell University.

Tanner, Murray Scot. 2000. “State Coercion and the Balance of Awe: The 1983-86 ‘Stern Blows’ Anti-Crime Campaign.” China Journal 44: 93-125.

19

Tanner, Murray Scot. 2016. China’s Response to Terrorism. CNA (Arlington, VA). https://www.uscc.gov/sites/default/files/Research/Chinas%20Response%20to%20Terrorism_CNA061616.pdf

Tanner, Murray Scot. 2017. “Beijing’s new national intelligence law: From defense to offense,” Lawfare, 20 July. https://www.lawfareblog.com/beijings-new-national-intelligence-law-defense-offense;

Tanner, Murray Scot and Eric Green. 2007. “Principals and secret agents: central versus local control over policing and obstacles to ‘rule of law’ in China.” China Quarterly 191: 644–70.

Tang, Wenfang. 2016. Populist authoritarianism: Chinese political culture and regime sustainability. Oxford University Press.

Tarrow, Sidney. 1996. “Social Movements in Contentious Politics: A Review Article.” American Political Science Review, 90(4), pp. 874-883.

Tarrow, Sidney. 2011. Power in Movement: Social Movements and Contentious Politics. 3rd ed. Cambridge University Press.

Thomson, Henry (2017). “Repression, redistribution and the problem of authoritarian control: Responses to the 17 June uprising in socialist East Germany.” East European Politics and Societies and Cultures 31(1), pp. 34–42.

Thornton, Patricia. 2002. “Framing Dissent in Contemporary China: Irony, Ambiguity, and Metonymy.” China Quarterly 171, pp. 661-681.

Tilly, Charles. 1978. From Mobilization to Revolution. New York: McGraw-Hill Publishers.

Tilly, Charles and Lesley Wood. 2015. Social Movements 1768-2012. Routledge.

Tilly, Charles. 2006. Regimes and Repertoires. University of Chicago Press.

Tilly, Charles, and Sidney Tarrow. Contentious Politics. 2nd ed. Oxford University Press.

Jacob Tolstrup, “Black knights and elections in authoritarian regimes: Why and how Russia supports authoritarian incumbents in post-Soviet states,” European Journal of Political Research 54, No. 4(2015): 673–690/

Tong, Yanqi, and Shaohua Lei, “Large Scale Mass Incidents and Government Responses in China,” International Journal of China Studies, Vol. 1, No. 2 (2010), pp. 487-508.

Tong, Yao, and Sui Yunyan. “Xinjiang fankong biaozhang ji dong’yuan bushu hui’yi zhokai Zhang Chunxian, Meng Jianzhu chuxi bing jianghua” [Zhang Chunxian and Meng Jianzhu speak at the XUAR counter-terrorism meeting], CPC News, 13 December 2015, http://cpc.people.com.cn/n1/2015/1213/c64094-27922213.html.

Trevaskes, Susan. 2007. Courts and Criminal Justice in Contemporary China. Lanham: Lexington.

Trevaskes, Susan. 2010. Policing Serious Crime in China: From “Strike Hard” to “Kill Fewer.” New York: Routledge.

Trevaskes, Susan. 2013. “Rationalizing Stability Preservation Through Mao’s Not-So-Invisible Hand,” Journal of Current Chinese Affairs, Vol. 42, No. 2, pp. 51-77.

Truex, Rory. 2016. Making Autocracy Work: Representation and Responsiveness in Modern China. Cambridge

20

University Press.

Truex, Rory. 2018. “Focal Points, Dissident Calendars, and Pre-Emptive Repression.” Journal of Conflict Resolution. https://doi.org/10.1177/0022002718770520

Tyson, Scott A. and Alastair Smith (2018). Dual-layered coordination and political instability: Repression,co-optation, and the role of information. Journal of Politics 80(1), 44-58.

Ulfelder, Jay. 2005. “Contentious collective action and the breakdown of authoritarian regimes.” International Political Science Review 26(3):311–334.

Ulfelder, Jay and Benjamin Valentino. 2008. “Assessing risks of state-sponsored mass killing.” Working Paper. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1703426

United Nations. 2019. “Moratorium call on surveillance technology to end ‘free-for-all’ abuses: UN Expert.” 25 June. https://news.un.org/en/story/2019/06/1041231

U.S. Department of Defense, Military and Security Developments Involving the People’s Republic of China 2019, Annual Report to Congress, 2 May 2019.

U.S. House of Representatives, Permanent Select Committee on Intelligence, “China’s Digital Authoritarianism: Surveillance, Influence, and Social Control,” 16 May 2019, https://docs.house.gov/Committee/Calendar/ByEvent.aspx?EventID=109462

U.S. SFRC Senate Foreign Relations Committee, “The China Challenge, Part 3: Democracy, Development, and the Rule of Law,” 4 December 2018, https://www.foreign.senate.gov/hearings/the-china-challenge-part-3-democracy-human-rights-and-the-rule-of-law-12042018;

U.S. Senate Foreign Relations Committee, Subcommittee on East Asia, “Testimony of Deputy Assistant Secretary of State Scott Busby,” 4 December 2018, https://www.foreign.senate.gov/imo/media/doc/120418_Busby_Testimony.pdf

Vaswani, Karishman. 2019. “Huawei: The Story of A Controversial Company.” BBC. 6 March. https://www.bbc.co.uk/news/resources/idt-sh/Huawei

Christian von Soest, “Democracy prevention: The international collaboration of authoritarian regimes,” European Journal of Political Research 54, no.4 (2015).

Walder, Andrew. 2009. Fractured Rebellion: the Beijing Red Guard Movement. Harvard University Press.

Walder, Andrew. 2009b. “Political Sociology and Social Movements.” Annual Review of Sociology, pp. 393-412.

Wallace, Jeremy. 2014. “Juking the Stats: Information Problems in China.” British Journal of Political Science, June: 1-19.

Walton, Jonathan. “China Plans for Internal Unrest,” in Andrew Scobell et al, eds., The People’s Liberation Army and Contingency Planning in China (NDU Press, 2015), pp. 55-84;

Wang Meng, “At this High-level Seminar, Xi Jinping Delivered What Important Signal to Protect Our Country’s Political Security,” [习近平在这个高层专题研讨班上提出确保我国政治安全，传递出什么重要信号？], Shanghai Observer, 22 January 2019, https://www.jfdaily.com/news/detail?id=128522

21

Wang Mingshan, “Kaichuang weihu shehui wending he changzhijiu'an xin jumian [Create a New Situation of Social Stability Maintenance and Long-Term Peace and Stability],” Renmin gong’an, 7 November 2017, http://www.mps.gov.cn/n2255079/n5590589/n5822616/n5822768/c5860294/content.html

Wang Ning, “Resolutely Defend and Implement the CMC Chairman Responsibility System, Ensure the Party’s Absolute Leadership Over the People’s Armed Police” [坚决维护和贯彻军委主席负责制 确保党对武警部队绝对领导], Study Times [学习时报], 11 October 2017, http://theory.people.com.cn/n1/2017/1011/c40531-29580014.html

Wang, Shirong, and Chenge Chu. 2018. Fengqiao jingyan: jiceng shehui zhili tixi he nengli xiandaihua shizheng yanjiu [Fengqiao Experience: Empirical research on modernization of grassroots social management systems and capacity]. Beijing: Falü chubanshe [Law Press].

Wang, Xiaohai. 2015. Empowerment of Chinese Police Force’s Role in Social Service. New York: Springer

Wang, Yuhua, and Carl Minzner. 2015. "The rise of the Chinese security state." China Quarterly 222.

Wang, Yuhua. 2014a. "Coercive capacity and the durability of the Chinese communist state." Communist and Post-Communist Studies 47(1):13-25.

--------------. 2014b. "Empowering the Police: How the CCP Manages Coercive Leaders." China Quarterly 219: 625-648.

Wasserstrom, Jeffrey, and Elizabeth J. Perry. 1994. Popular Protest and Political Culture in Modern China. 2nd ed. Westview Press.

Wedeen, Lisa. Ambiguities of Domination: Politics, Rhetoric, and Symbols in Contemporary Syria. University of Chicago Press.

Wee, Sui-Lee, “China Seeks Rebel Village Concessions as New Protest Flares,” Reuters, 20 December 2011, https://www.reuters.com/article/us-china-unrest-village/china-seeks-rebel-village-concessions-as-new-protest-flares-idUSTRE7BJ0I320111220

Jessica Chen Weiss. 2019. Testimony to the House Permanent Select Committee on Intelligence, “China’s Digital Authoritarianism: Surveillance, Influence, and Social Control,” 16 May 2019, https://docs.house.gov/meetings/IG/IG00/20190516/109462/HHRG-116-IG00-Wstate-ChenWeissJ-20190516.pdf

White, Lynn T. 1989. Policies of Chaos: Organizational Causes of Violence in China’s Cultural Revolution. Princeton University Press.

Whiting, Susan. 2004. “The Cadre Evaluation System at the Grassroots.” In Holding China Together (Cambridge University Press).

Woods, Kevin, James Lacey, and William Murray. 2006. “Saddam’s Delusions: The View from Inside.” Foreign Affairs (May/June).

World Bank, “Intentional Homicides,” https://data.worldbank.org/indicator/vc.ihr.psrc.p5, accessed 17 August 2019.

Wright, Teresa. 2001. The Perils of Protest: State Repression and Student Activism in China and Taiwan. University of Hawaii Press.

22

Wu Aiying, “Chongfen fahui sifa xingzheng zhineng zuoyong weihu shehui wending” [Give full play to the role of the judicial administrative functions and maintain social stability], Chinese Academy of Social Sciences, Institute of Law (2017), http://www.iolaw.org.cn/showNews.asp?id=19749

Wu Qiang, “Urban Grid Management and Police State in China,” 12 August 2014, https://chinachange.org/2013/08/08/the-urban-grid-management-and-police-state-in-china-a-brief-overview/

Wuthnow, Joel. 2017. “China's New ‘Black Box’: Problems and Prospects for the Central National Security Commission.” China Quarterly 232: 886-903.

Wuthnow, Joel China’s Other Army: The People’s Armed Police in an Era of Reform (NDU: INSS), April 2019,

Wuthnow, Joel, and Phillip Saunders, Chinese Military Reforms in the Age of Xi Jinping: Drivers, Challenges, and Implications (Washington: NDU Press, 2017), pp. 32-35.

Xi Jinping, “Tantan diaocha yanjiu [Discussing field research],” 16 November 2011; CCP Central Committee/PRC State Council, “Guanyu jiaqiang shehui chuangxin guanli de yijian [Suggestions on Enhancing Innovations in Social Management],” July 2011.

Xi Jinping, “Safeguard National Security and Social Stability,” 25 April 2014, in Xi Jinping, Governance of China (Beijing: Foreign Languages Press, 2014).

Xi Jinping. “Xi Jinping jiu zhengfa gongzuo zuochu zhongyao zhishi” [Xi Jinping issues important instructions on political-legal work], Xinhua, 20 January 2015, at http://www.xinhuanet.com/politics/2015-01/20/c_1114065786.htm; http://www.court.gov.cn/fabu-xiangqing-13840.html

Xi Jinping, 19th Party Congress report, 18 October 2017.

Xi Jinping, On the Holistic Approach to National Security (Central Party Literature Press, 2018).

Xi Jinping 2019a. “Xi Jinping zai zhongyang zhengfa gongzuo huiyi shang qiangdiao [Xi Jinping’s emphasis at the Central Political & Legal Work Conference],” Xinhua, 16 January 2019, https://www.chinacourt.org/article/detail/2019/01/id/3707480.shtml

Xi Jinping 2019 b. “习近平：提高防控能力, 着力防范化解重大风险 保持经济持续健康发展社会大局稳定 [Xi Jinping: Improve Prevention and Control Capabilities; Try to Prevent and Resolve Major Risks; Maintain Sustainable and Healthy Economic Development and Overall Social Stability],” 21 January 2019, http://www.qstheory.cn/yaowen/2019-01/21/c_1124021825.htm

Xie, Yue. 2012. “The political logic of weiwen in contemporary China.” Issues and Studies 48(3):1–41.

Xinhua, “The Chaoyang Masses’ Numbers Exceeded 140,000 Last Year, and Cracked Over 370 Cases,” 29 January 2018, http://www.xinhuanet.com/legal/2018-01/29/c_1122330010.htm

Xinhua, “Zhe liang nian Xi Jinping dai huo de 12 ge re ci” [Xi Jinping’s hot phrases during the past two years], 6 February 2015, at http://www.xinhuanet.com//politics/2015-02/06/c_127467007.htm

Xinhua, “ “Xi calls for efforts to break new ground in national security,” Xinhua, 17 April 2018, at http://www.xinhuanet.com/english/2018-04/17/c_137118101.htm

23

Xinhua, “Book of Xi's discourses on national security published,” 15 April 2018, http://www.xinhuanet.com/english/2018-04/15/c_137112987.htm;

Xinhua, “ “Dang he guojia jigou gaige fang’an quanwen gongbu” [The full text of the party and state institutional reform plan is announced], Xinhua, 21 March 2018, at http://news.sina.com.cn/c/nd/2018-03-21/doc-ifysnhat1097446.shtml

Xinhua, “ “关于国务院机构改革和职能转变方案的说明 [Description of the State Council’s Institutional Reform and Transformation Plan],” Xinhua, 10 March 2013, http://www.xinhuanet.com/2013lh/2013-03/10/c_114969788_2.htm;

Xinhua. 2018. “Chinese technology brings falling crime rate to Ecuador.” 19 January. http://www.xinhuanet.com/english/2018-01/19/c_136908255.htm

Xu, Xu. Forthcoming. “To Repress or To Co-opt? Authoritarian Control in the Age of Digital Surveillance." American Journal of Political Science.

Yan Xiaojun. 2011. “Regime Inclusion and the Resilience of Authoritarianism: The Local People’s Consultative Conference in Post-Mao Chinese Politics.” China Journal 66: 53-75.

Yang, Dali L. 2017. “China’s troubled quest for order: leadership, organization and the contradictions of the stability maintenance regime.” Journal of Contemporary China 26(103): 35–53.

Yang, Guobin. 2003. “The Internet and Civil Society in China: A Preliminary Assessment.” Journal of Contemporary China. 12(36): 453-475.

Yang, Guobin. 2009. The Power of the Internet in China: Citizen Activism Online. Columbia University Press.

Yang Kuisong. 2008. “Reconsidering the Campaign to Suppress Counterrevolutionaries.” The China Quarterly, Vol. 193 (March): 102-21.

Yang Jingjie, “Xinjiang to see ‘major strategy shift,’” Global Times, January 9, 2014, at http://www.globaltimes.cn/content/836495.shtml#.UtS1ivaFZ0Q

Yarhi-Milo, Keren, Alexander Lanoszka, and Zack Cooper. 2016. “To Arm or Ally? The Patron's Dilemma and the Strategic Logic of Arms Transfers and Alliances.” International Security, Vol. 41, No. 2 (Fall): 90-139.

You, Ji. “China’s National Security Commission: theory, evolution, and operations,” Journal of Contemporary China 25(98): 178-196.

Yuan Shenggao, “Formerly Remote Area Gears Up to Drive into an Exciting New Future,” Guiyang City website, 20 October 2017, http://www.eguiyang.com.cn/2017-10/20/content_33482298.htm

Yue Huairang,“军委国防动员部部长盛斌、政委朱生岭双双晋升中将军衔,”The Paper, 12 August 2016, https://www.thepaper.cn/newsDetail_forward_1512738

Zald, M., and J. McCarthy. 1987. Social Movements in an Organizational Society: Collected Essays. Transaction Books.

Zarzecki, Thomas. 2002. Arms Diffusion: The Spread of Military Innovations in the International System (Routledge).

24

Zeng, Jun, and Liangliang Shi. 2012. “Difang Zhengfawei Xiaotiao Chuli Xingshi Anjian De Zhidu Kaocha Ji Fenxi [An Institutional Analysis of Local Political-Legal Committtees’ Criminal Case Coordination Functions],” Xinan Zhengfa Daxue Xubao [Southwest University of Politics and Law Journal] 2: 65-72.

Zenz, Adrian. Corralling the People’s Armed Police: Centralizing Control to Reflect Centralized Budgets,” Jamestown China Brief, Vol. 18, No. 7 (24 April 2018), at https://jamestown.org/program/corralling-the-peoples-armed-police-centralizing-control-to-reflect-centralized-budgets/

Zenz, Adrian. 2018b. “China’s Domestic Security Spending: An Analysis of Available Data,” China Brief, Vol. 18, No. 4 (March 2018).

Zenz, Adrian, and James Leibold. 2017a. “Xinjiang’s Rapidly Evolving Security State.” China Brief 17(4). https://jamestown.org/program/xinjiangs-rapidly-evolving-security-state/

Zhang Hui, “Facial Recognition System Leaves Criminals With Nowhere to Run,” Global Times, 18 March 2018, http://www.globaltimes.cn/content/1093864.shtml

Zhang Liping, “Legal Considerations of the People’s Armed Police Management of Mass Incidents,” Police Practical Combat Training, No. 2 (2013), pp. 92-95;

Zhang Yunbi, “Xi outlines new strategy to protect nation,” China Daily, 16 April 2014, at http://www.chinadaily.com.cn/china/2014-04/16/content_17436806.htm;

Zhang, Xiaoling, and Yongnian Zheng. 2009. China’s Information Communications Technology Revolution: Social Changes and State Response. Routledge.

Zhao Lei, “Zhongguo sifa kaiqi shaosha shidai: congzhong congkuai chengwei lishi henji” [China’s justice begins a new era of fewer executions: Swift and Severe punishment becomes history],” Nanfang Ribao, 4 January 2007.

Zhao, Yang, “Zhou Yongkang: xiwang Guangdong wei quanguo xin weiwen chuangzao fangxiang jingyan [Zhou Yongkang: I Hope Guangdong Will Create a Guiding Experience in Innovating Concepts of Stability Maintenance], Nanfang Ribao [Southern Daily], 4 July 2012, http://china.nfdaily.cn/content/2012-07/04/content_49940818.htm

Zhejiang Public Security Department General Office [厅办公室], “Xinshidai Fengqiao Jingyan shi zenyang liancheng de [“Fengqiao experience” in the new era],” Ping’an Times, 2 January 2018, http://www.zjsgat.gov.cn/jwzx/jsyw/201801/t20180102_1441822.htm

Zhongguo Gongchandang Shanghaishi zuzhi shi ziliao (Shanghai: Renmin Chubanshe, 1991) / 中国共产党上海城组织史資料 (上海: 人民出版社, 1991).

Zhongguo Renmin Gong’an Shigao Bianxie Xiaozu, Zhongguo Renmin Gong’an Shigao [Draft history of Chinese people’s public security] (Beijing: Jingguan Jiaoyu Chubanshe, 1997, neibu faxing)/中国人民公安史稿编写小组, 中国人民公安史稿 (北京: 警官敎育出版社, 1997, 内部发行).

Zhou, X. 1993. “Unorganized Interests and Collective Action in Communist China.” American Sociological Review, 58, pp. 54-73.

Zhou Yongkang, “Jiaqing he gaijin shehui guanli, cujin shehui wending hexie,” Renmin Ribao, 26 October 2006.

25

Zhou, Baogang. 2008. Shehui zhuanxingqi qunti xing shijian yufang, chuzhi gongzuo fanglue [Strategies for Preventing and Managing Mass Incidents during the Reform Period]. People’s Public Security University Press.

Zwirko, Colin. 2019. “Chinese Surveillance Software in Use Across Range of North Korean Facilities.” NKNews. 12 July. https://www.nknews.org/2019/07/chinese-surveillance-software-in-use-across-range-of-n-korean-facilities-photos/