Containing Unrest: New Strategies, New Dilemmas

The terrific rise in protest incidents China has witnessed over the past decade has forced China’s internal security forces – both the People’s Armed Police (PAP) and the regular public security forces – to seek new ways for dealing with unrest.1  Several years of internal debate over policing protest have yielded new strategies to contain protest: from a gradual shift from trying to deter or quickly squelch all protests to a strategy that is implicitly more permissive and focuses on containing and managing protest. While the transition toward containment is far more realistic in a society where protest is an increasingly normal tactic of social bargaining, it also presents police with major new risks and dilemmas.


Before the 1989 Tiananmen demonstrations, it was almost a misnomer to speak of a ‘strategy’ for policing protest. Scattered reports on how police dealt with protests suggest largely ad hoc responses that occasionally failed to contain unrest effectively, or so angered protestors that violence escalated before control was restored. In the late 1980s in Tibet, PAP forces typically fired modest amounts of tear gas at demonstrators, or resorted to beatings or sometimes firing live ammunition over the heads or at the feet of crowds. When police confronted student demonstrators in Tiananmen square in Winter 1986-87, they reportedly iced the square the night before using fire hoses, and the next day large numbers of unarmed officers in heavily padded coats surrounded single demonstrators, wrestling them until they lost their footing, and dragged them off. In January 1989, police officials criticized local Nanjing and Sichuan PAP and police for using excessive violence in the handling of student protests which subsequently escalated. Just three months later, perhaps as an overreaction to this ‘institutional lesson’ about handling protest, police and PAP in Beijing completely failed to contain the Tiananmen protestors effectively and with minimal violence. On occasion they enraged students into stiffening their demands, as in the famed April 18 incident of alleged belt-beating outside Beijing’s Xinhua Gate. Overall, however, PAP and police sent students the unspoken and tragically mistaken message that the regime would not – or worse, could not – repress their protests.


Ultimately, Deng Xiaoping used the regular PLA to prove otherwise.  1989 was, of course, an unmitigated disaster for the PAP and police, and it touched off pitched battles over the proper handling of protests. The Party leadership imposed an official ideological lesson that took a highly conspiratorial view, insisting that protest leaders had implacable counterrevolutionary designs, and the big mistake of 1989 had been to underestimate their perfidy. Protest followers were portrayed as either dupes of the Party’s enemies or ‘bourgeois liberals’. Echoing the thinking that motivated China’s ‘strike hard’ anti-crime campaigns (call it ‘realist’, ‘behaviourist’ or perhaps ‘Legalist’ 2), the masses were seen as a potentially wavering element in between the Party and its enemies. If they were not firmly deterred and overawed by police power, there was always the risk that they could turn toward the enemy. Consequently, early post-Tiananmen police texts on handling protest incidents argued the two ‘key links’ were ‘prevention’ (fang) and ‘putting down’ (zhi), where zhi was defined as ‘hunting down’, ‘quelling’, ‘suppressing’ or ‘attacking’.3 In other words, force prospective protestors to recognize that they were choosing sides between the state and its enemies, the protest leaders.



The PAP, accordingly, was reorganized to build a tough-minded corps that would not hesitate to draw that line in the sand – or in the street. Although the need for modern anti-riot training and equipment was stressed by some, the dominant critique (associated with Yang Shangkun and Yang Baibing) emphasized tighter military discipline, quick reaction, political education, unquestioned loyalty to the Party, and a willingness to use the level of force that the regular PLA sincerely hoped it would never have to use again. Beginning in the early 1990s, and culminating in a March 1995 joint State Council-Central Military Commission directive, the PAP was reorganized and placed under the direct leadership and personnel management of the State Council and Central Military Commission. The PAP Internal Guards Units, both at the Central and provincial levels, came under the direct leadership and personnel management of the PAP Main Headquarters and were no longer a ‘constituent part’ of the local public security apparatus. Provincial and local Party and government leaders, who provide the bulk of the PAP’s budget, retained their prerogatives to activate the units within their regions, however, and these units were still supposed to accept command from local Public Security during these operations.


But from the late 1990s to today, many PAP and public security analysts and officials have conceded rather frankly that unrest is getting worse, with protests increasing in number, average size, organization and tactical sophistication. Moreover, the post-1989 conspiratorialism, though still embraced in some cases (especially the Falun Gong and Muslim separatists), is more and more giving way to a blunt admission by many police that most protestors are members of ‘the masses’ who are motivated by essentially legitimate economic and political complaints such as unemployment, layoffs, withheld wages and pensions, and abusive, corrupt officials.4 The overwhelming majority of protest incidents are now officially declared to be ‘contradictions among the people’ and these social order incidents are now formally renamed ‘incidents of groups of the masses’ (quntixing shijian). Many police analysts now point out that police coercion, no matter how professional, will not be sufficient to contain unrest, and effective strategies must also include restrategizing of economic reform, strengthened social safety nets, and serious political and legal reform. A far more serious implication, quietly hinted at by social-psychologically trained scholars at the PAP  command School, concerns loyalty – because of an almost certain rise in police sympathy for the very  protestors whom they must contain, disperse, or suppress.


The result is that China’s police and its new senior leadership are quietly deemphasizingdeterrence reinforced by harsh, repressive responses. Police concede that they probably cannot deter most protests from starting, moderate levels of protest are inevitable, crowds are growing larger, and protestors enjoy considerable public (and even police) sympathy. They are also embracing a more permissive ‘safety valve’ view of protest, and are groping their way toward a more sophisticated strategy of ‘containment’, ‘management’, and ‘defusing’. Increasingly, the outcome police fear the most is shifting – away from a fear of failing to deter protests from starting, to a fear that if they misuse force, they will exacerbate popular rage and turn a small-scale peaceful demonstration into a large-scale, sustained, protest or riot. Recent internal police texts now cite ‘preventing the exacerbation of contradictions’ as second only to Communist Party leadership among five basic principles for police handling unrest.5 But a more permissive ‘containment’ strategy presents PAP and public security forces with complex dilemmas and profound organizational challenges. Just three are noted here: 1) the dangerous ambiguities of experimenting with less repressive strategies; 2) the risky choice of whether or not to make economic concessions to protestors; 3) the far higher organizational demands of professionalism and discipline that a more restrained strategy entails.


How and when to use force and coercive measures is among the thorniest dilemmas of more permissive, ‘containment’ protest policing. All security leaders understand that violent tactics may help deter future protests. But once a protest has started, the use of crude, hamfisted, or less-than overwhelming brutality risks enraging crowds and losing control. Recent police documents have therefore called on police to observe ‘three cautions’ in protest policing:


1 Cautiously use police power.

2 Cautiously use coercive measures.

3 Cautiously use weaponry and police armaments.6


Many leading public security officials and analysts call for limiting violent and coercive tactics to incidents of imminent mob violence, arson, looting, or attacks on key government buildings. Police leaders increasingly discourage officers from plunging into crowds or making mass arrests – urging them instead to focus on maintaining order at the scene, maintaining normal traffic flow and work patterns, etc. As a precaution, some police leaders urge commanders to keep weaponry in reserve away from the front lines. A key element of containment is ‘isolation’ – making protestors feel physically and psychologically separate from onlookers by establishing a temporary police line and preventing outsiders from joining in. Police are instead urged to do ‘propaganda and education’ work persuading demonstrators to take their demands to officially sanctioned institutions. Police are also urged to carefully gather ‘evidence’ to be used to detain and punish protest leaders later – after demonstrators have dispersed. While secret infiltration is an important aspect of this, and Anti-Riot Police and PAP are urged to co-ordinate with clandestine Public Security Domestic Security Protection (guonei anquan baowei) and State Security units, they are also urged not to be too secret about this evidence-gathering. It is hoped that the noticeable presence of police carrying video recorders and cameras will have a chilling effect on both protestors and onlookers.7



But more moderate tactics can leave local forces in dangerously paralyzing dilemmas. More restrained tactics may minimize the risk that any given protest will boil over, but they also make protesting appear far less risky and dangerous to the average citizen. Moreover, while police are ordered to do everything possible to avoid ‘sharpening contradictions’ with protestors, they are also strictly ordered to obey local Party and government leaders – in effect, forcing them to serve two masters. The police are well aware of this passive position, and many fear finding themselves trapped uncomfortably among an angry populace and local Party officials who demand that they ‘decisively restore order.’ Beijing, moreover, as is its wont, ultimately gives local police dangerously vague, contradictory directives, insisting that using power ‘cautiously’ does not  mean not using it at all. So police are ordered to overcome their contradictory fears that if they use too much force, the masses will seek revenge, and if they use too little, the leadership will punish them. One can almost hear the echoes of the impossibly contradictory Cultural Revolution directive that officials must simultaneously ‘grasp revolution and promote economic production’. The history of Chinese Communism is littered with disasters born of local officials’ efforts to resolve such connundra.



A further debate is simmering over whether or not it makes sense for police to encourage local Party, government and economic officials to ‘buy off ’ protestors by paying lump sums of back wages or pensions. Some police see no problem in this tactic, since it spares them the chore of suppressing the protestors, who probably deserve the money anyway. But some other police officials are frankly fearful of ‘contagion’, and warn that such payoffs will simply prove to demonstrators in nearby towns and factories that ‘protest pays’.


Finally, a shift toward more permissive ‘containment’ strategies places enormous  additional organizational challenges on the PAP and public security units – in particular challenges of personnel levels, training, and especially discipline in the face of angry crowds. If these strategies are to succeed in defusing rather than exacerbating unrest, they require the PAP and public security police to overcome some major organizational shortcomings and become several things that, at present, they are not.  Containment, and especially prevention, requires consistently effective sources of social intelligence. But recent police discussions clearly suggest that China’s vaunted social monitoring network is not nearly what it was, and police leaders are not at all clear on how they will revive these networks.


Anti-protest forces must be increasingly capable of nimble, effective, and unified command, control, communication, and information-sharing, and especially co-ordinating activities among many branches of police and security services – patrol police, traffic police, anti-riot units, special police, social order police, PAP, domestic security protection, State Security, fire-fighting, security protection companies, etc. – in addition to working with the government departments or enterprises where the protest occurred. But recent police analyses suggest that problems with such co-ordination are widespread. Moreover, the 1995 PAP regulations are not detailed or explicit on how the PAP is to be activated, and implicitly rely upon effective informal co-operation between local Party and government leaders, local police, and PAP officials at multiple levels. Containment requires police who are carefully recruited, extensively trained and tightly disciplined in the face of angry protestors. This, in turn, requires that local Party officials actively support this strategy by allowing the PAP and police to recruit as they see fit, and to refrain from plunging in to repress protestors. But numerous open source police writings contend that this is often not the case at local levels, and local officials frequently treat the police as their own private security units (‘dog beating brigades’) or as ‘unemployment programmes’ for unqualified friends and relatives. Since provincial and local Party and government offices supply the vast majority of police and PAP budgets in China, this sense of local entitlement will be hard to overcome.


Murray Scot Tanner

The author is a Senior Political Scientist at the RAND Corporation



1 Many of the views in the article are discussed at much greater length in the author’s forthcoming article ‘China Rethinks Unrest’, The Washington Quarterly (Summer 2004).

2 For an apt comparison of classical Chinese Legalism with simple ‘reward-punishment’ behaviourism, see Benjamin Schwartz’s classic The World of Thought in Ancient China (Cambride Massachusetts: Belknap Press, 1985).

3 Hu Guanwu, ed., Jinji Zhi’an Shijian yu Duice (Emergency Social Order Incidents and their Countermeasures), (Beijing, Zhongguo Renmin Gongan Daxue Chubanshe, April 1996), p. 101.

4 Much of the material for this discussion draw on a major set of PAP and police studies on unrest edited by the Ministry of Public Security’s Fourth Research Institute. Li Zhongxin, ed., Quntixing Shijian Yanjiu Lunwen Ji (Collected Research Essays on Mass Incidents), (MPS Social Order Research Institute, Beijing, Zhongguo Renmin Gongan Daxue Chubanshe, June 2001).

5 Zhang Shengqian, Shehui Zhian Shijian Chuzhi (Handling Social Order Incidents), (Beijing: People’s Public Security University Press, 2001), p. 63. These five principles are cited so widely in this and other police volumes that it is almost certain they are enshrined in a major recent internal police policy document.

6 Zhang Shengqian, p. 63.

7 Police handling the 2002 Liaoyang worker protests skillfully deployed many of these tactics. See Phillip P. Pan’s excellent The Washington Post report, ‘Three Chinese Workers: Jail, Betrayal and Fear’, December 28, 2002, p. A-1.

Explore our related content