B.J.Pol.S. 38, 411432 Copyright 2008 Cambridge University Press doi:10.

1017/S0007123408000215 Printed in the United Kingdom

Power Structure and Regime Resilience: Contentious Politics in China

YONGSHUN CAI*
Authoritarian governments may face serious uncertainties when dealing with popular resistance because of the unpredictable consequences of making concessions or repressing opposition. However, a political system with multiple levels of authority can help reduce the uncertainties by granting conditional autonomy to lower-level authorities. Such a power structure prevents excessive repression and unconditional concessions when the priorities of different levels of authority do not match. Under this political arrangement, the central authority can avoid blame when local authorities use repression. The divided power also helps reduce the uncertainties faced by the central authority because it will then have to deal with only a very limited number of instances of resistance. Using the case of China, this article shows that divided state power has allowed the party-state to maintain social stability amid numerous instances of social unrest during the reform era.

Authoritarian governments may face serious uncertainties in dealing with popular resistance. Making concessions tends to trigger more resistance or even the collapse of the regime, but reliance on repression damages the regimes legitimacy and makes it less sensitive to popular demands.1 How can authoritarian governments reduce these uncertainties? This article suggests that certain political arrangements can help address this dilemma. More specically, in a political system with multiple-level authorities, the regimes resilience will be enhanced if the top-level authority assigns the responsibility for dealing with popular resistance to lower-level authorities and grants them conditional autonomy. This arrangement places constraints on both resisters and lower-level authorities, and it therefore addresses the concessionrepression dilemma by discouraging both unconditional concessions and excessive repression. Moreover, the arrangement also protects the top-level authoritys image by allowing it to distance itself from lower-level authorities that use repressive measures. Authoritarian regimes are characterized by the lack of political opportunities for collective action, but this does not mean that popular resistance is absent in such regimes.2 In China, for example, there have been numerous instances of collective resistance in recent years.3 Reportedly, the number of such occurrences increased ten times from 8,700 in 1993 to 87,000 in 2005.4 Whilst not all of these instances were disruptive, many were. In 2003,
* Division of Social Science, Hong Kong University of Science and Technology. The author wishes to thank the anonymous reviewers and the Editor, Sarah Birch, for their suggestions. 1 Jack Goldstone and Charles Tilly, Threat (and Opportunity): Popular Action and State Response in the Dynamics of Contentious Action, in Ronald R. Aminzade, Jack A. Goldstone, Doug McAdam, Elizabeth J. Perry, William H. Sewell Jr, Sidney Tarrow and Charles Tilley, eds, Silence and Voice in the Study of Contentious Politics (New York: Cambridge University Press, 2001), pp. 17994. 2 See, for example, Kurt Schock, Unarmed Insurrections: People Power Movements in Non-democracies (Minneapolis: University of Minnesota Press, 2005). 3 These incidents include demonstrations, protests, strikes, disruptive collective petitions, sit-in demonstrations, fasting, posting slogans, attacking state agencies, surrounding leaders, blocking trafc, looting and damaging property (Chen Zhaxu, Actively Preventing and Appropriately Dealing with Instances of Collective Action, Research on Public Security, 5 (2001), 1619). 4 Jae Ho Chung, Hongyi Lai and Ming Xia, Mounting Challenges to Governance in China: Surveying Collective Protestors, Religious Sects and Criminal Organizations, China Journal, 56 (2006), 131.

412

CAI

there were 3,100 instances of highways, roads or railways being blocked and 3,900 attacks on state agencies.5 Citizens resistance has become a serious concern for the central party-state, and it is against this background that building a so-called harmonious society has recently become a top priority of the central party-state.6 In some other transitional economies, a large increase in collective actions, such as strikes, occurred only after the regime collapsed, which relaxed the political constraints on resistance. In Poland the number of strikes in 1989 was 894, whereas it was 7,443 in 1993 (or 8.3 times more); in the former Soviet Union, the number of strikes was 1,771 in 1990, and this number rose to 6,273 in 1992 (or 3.5 times more).7 In comparison, the large increase in popular resistance in China has occurred without regime change. The frequency and magnitude of collective actions in China appear puzzling: why are so many instances of resistance possible given the political system and state power? However, the many cases of resistance have not posed an insurmountable challenge to social and political order in a country that is undergoing tremendous socio-economic change. The coexistence of social disturbance and political stability in China reveals the states resilience.8 This article provides an explanation for this resilience by highlighting the importance of political arrangements. It shows that the Chinese party-state has been able to maintain social stability amid numerous social conicts and prevent the unpredictable consequences of making concessions or repressing resistance because the political arrangements grant conditional autonomy to local governments.
AUTHORITARIAN GOVERNMENTS AND THE CONCESSIONREPRESSION DILEMMA

A state can respond to popular resistance with repression, tolerance or concessions.9 Its choice has to do with, among other things,10 the type of political regime, which shapes political leaders calculations.11 In democracies, politicians face the pressure of elections and have to be cautious when using repression, and democratic systems are thus more likely to tolerate certain types of dissident behaviour and may use a mix of concessions and repression or more concessions.12 By comparison, political leaders in authoritarian regimes do not face the pressure of elections, but they are more sensitive to popular resistance.13 The occurrence of citizens resistance signals problems with social control or the weakness of the government because such resistance is not supposed to occur in a regime where
He Zuowen, Properly Handling Conicts in Our Country, Scientic Socialism, 2 (2005), 811. On 28 July 2005, The Peoples Daily, the Chinese Communist Partys mouthpiece, published a front-page commentary entitled Maintaining Stability, Promoting Development, urging citizens to obey the law and warning that any threats to social stability would not be tolerated by the authorities. 7 Zhang Sai, World Statistics 1995 (Beijing: China Statistics Press, 1996), pp. 3479. 8 Jean Oi, Realms of Freedom in Post-Mao China, in William Kirby, ed., Realms of Freedom in Modern China (Palo Alto, Calif.: Stanford University Press, 2003), pp. 26484. 9 Frances Fox Piven and Richard Cloward, Poor Peoples Movements: Why They Succeeded, How They Fail (New York: Vintage, 1977), p. 27. 10 One important factor that affects the states response is the power of the resisters. See, among others, William Gamson, The Strategy of Social Protest (Homewood, Ill.: Dorsey, 1990), pp. 812; McAdam, Political Process and the Development of Black Insurgency, p. 56. 11 Goldstone and Tilly, Threat (and Opportunity), pp. 1923. 12 Frances Piven and Richard Cloward, Regulating the Poor: The Functions of Public Welfare (New York: Vintage, 1993), p. 172. 13 Goldstone and Tilly, Threat (and Opportunity), p. 192.
6 5

Power Structure and Regime Resilience

413

citizens are denied the right to challenge the state.14 Repression is thus commonly used to show the states power and determination to protect the political system.15 However, authoritarian governments do not rely exclusively on repression; they may also make concessions. Excessively strong repression not only causes the opposition to gain allies: it may also leave the opposition little alternative but to revolt16 or it can produce a radicalization of collective action and a more effective organization of opponents.17 In addition, repression undermines the regimes legitimacy. Although authoritarian regimes are less reliant on legitimacy for survival than their democratic counterparts, they certainly have incentives to build and strengthen their legitimacy.18 Authoritarian states with a higher-level of legitimacy not only have a better chance of surviving crises but also save on the costs of imposing strict and constant control over the society.19 Nevertheless, authoritarian governments do face serious uncertainties in making concessions. Goldstone and Tilly suggest that modest concessions made by the government can be fatal to the regime; one reason for this is the de Tocqueville effect: minor changes made by the regime indicate its illegitimacy but do not fully address the causes of the illegitimacy, and thus lead to greater demands for elimination or transformation of the regime. Secondly, concessions may increase perceptions of state weaknesses, making others believe that they too can extract more from the regime.20 Therefore, the states failure to demonstrate its actual power or credible threats may trigger regime-threatening actions.21 Although authoritarian states do not always face such life-or-death crises from social unrest, small-scale resistance, if unchecked, chips away at state power. The emergence of Solidarity in Poland, for instance, originated from economic demands. The governments ineffective repression of the movement made it into a testing ground of various modes of resistance that contributed to the subsequent better-organized and forceful resistance
14 Susanne Lohmann, The Dynamics of Information Cascades: The Monday Demonstration in Leipzeig East Germany, 19891991, World Politics, 47 (1994), 42101; Charles Kurzman, Structural Opportunities and Perceived Opportunity in Social-Movement Theory: The Iranian Revolution of 1979, American Sociological Review, 61 (1996), 15370. 15 Christian Davenport, Multi-Dimensional Threat Perception and State Repression: An Inquiry into Why States Apply Negative Sanctions, American Journal of Political Science, 39 (1995), 683713; Goldstone and Tilly, Threat (and Opportunity), pp. 1923. 16 Scott Gartner and Patrick Regan, Threat and Repression: The Non-Liner Relationship between Government and Opposition Violence, Journal of Peace Research, 33 (1996), 27388, at p. 285; Paul Almeida, Opportunity Organizations and Threat-Induced Contention: Protest Waves in Authoritarian Settings, American Journal of Sociology, 109 (2003), 345400. 17 Sidney Tarrow, Power in Movement (New York: Cambridge University Press, 1998), p. 93; Will Moore, Repression and Dissent: Substitution, Context, and Timing, American Journal of Political Science, 42 (1998), 85173. 18 Saxonberg writes, even if the Soviet-type regimes never gained complete legitimacy in the sense of gaining popular support for the system, they certainly tried (Steven Saxonberg, The Fall: A Comparative Study of the End of Communism in Czechoslovakia, East German, Hungary and Poland (Amsterdam: Harwood Academic, 2001), p. 147). 19 Anthony Oberschall, Opportunities and Framing in the Eastern Europe Revolts of 1989, in Doug McAdam, John D. McCarthy and Mayer N. Zald, eds, Comparative Perspectives on Social Movements (Cambridge: Cambridge University Press, 1996), pp. 17299. 20 Goldstone and Tilly, Threat (and Opportunity), p. 188. 21 Lohmann, The Dynamics of Information Cascades; Kurzman, Structural Opportunities and Perceived Opportunity in Social-Movement Theory; Timur Kuran, Now Out of Never: The Element of Surprise in the East European Revolution of 1989, World Politics, 44 (1991), 748.

414

CAI

which was fatal to the regime.22 This process of testing the boundary of state tolerance is, to a great extent, what Scott describes as the weak resisting the powerful through the hidden transcript (i.e., discourse taking place offstage): The hidden transcript is continually pressing against the limit of what is permitted on stage, much as a body of water might press against a dam. If the act of insubordination is not rebuked or punished, then others will exploit that breach and expand the territory: A small success is likely to encourage others to venture further, and the process can escalate rapidly.23 The balance between concessions and repression is thus a serious issue faced by authoritarian governments. In China, unprecedented socio-economic changes have given rise to numerous conicts in the society.24 Tolerating certain modes of resistance (such as disruptive action) will certainly lead to more resistance, which threatens economic reforms and social stability. However, reliance on repression will only fuel the resentment, seriously damage the regimes legitimacy, and produce backlashes that threaten the regimes survival in crises, as reected in the 1989 Tiananmen incident.25 While a regime with strong resilience is better able to address the concessionrepression dilemma, the question is how the state can balance the two options. Divided State Power and Regime Resilience It is a reasonable assumption that the states response to popular resistance depends on its perception of the costs and benets associated with the choice of a particular response. However, as Goldstone and Tilly show, the states calculations, regardless of how sophisticated they are, may still fail to reduce the uncertainties.26 This article suggests that the states response is shaped by the political arrangements, and certain arrangements can help the state address the concessionrepression dilemma and enhance the regimes resilience. In dealing with a large number of instances of resistance, it is essentially impossible for authoritarian governments to give up repression. The issue becomes what type of repression can be used and, more importantly, who carries out the repression and takes the blame. A political system will help maintain social stability if the political arrangements allow the state to adopt repression whilst reducing the damage to the regimes legitimacy and to prevent concessions from being interpreted as signs of weakness. This is possible when there is more than one authority which deals with popular resistance. The existence
22 Roman Laba, The Roots of Solidarity: A Political Sociology of Polands Working-Class Democratization (Princeton, N.J.: Princeton University Press, 1991). 23 James Scott, Domination and the Arts of Resistance: Hidden Transcript (New Haven, Conn.: Yale University Press, 1990), p. 196. 24 Dali Yang, Economic Transformation and Its Political Discontents in China: Authoritarianism, Unequal Growth, and the Dilemmas of Political Development, Annual Review of Political Science, 9 (2006), 14364. 25 The 1989 Tiananmen incident not only showed the consequence of a divided ruling elite but also revealed the serious impact of the lack of legitimacy on regime survival. Central leaders acknowledged that many people protested in the incident because of their resentment against corruption. What was more astonishing was the participation of state agents. Between 15 and 19 May 1989, about 700 work units had employees who held demonstrations in Beijing, including about seventy party and government agencies. Fifty of the seventy agencies, including the State Council, belonged to the central party committee and the central government (State Education Commission, The Soul-Stirring 56 Days (Beijing: Land Press, 1989), p. 138). 26 Goldstone and Tilly describe how there can be a variety of undesirable scenarios that the state encounters in using repression or concessions because of the unknown consequences of a response (Goldstone and Tilly, Threat (and Opportunity), pp. 18892).

Power Structure and Regime Resilience

415

of multiple authorities integrated into a political hierarchy provides signicant space for the top-level authority to deal with popular resistance when it grants conditional autonomy to a lower-level authority. Conditional autonomy means that the lower-level authority or the local government has considerable autonomy in dealing with popular resistance while facing constraints imposed by the top-level authority or the central government. Therefore, while it is commonly suggested that a government with strong power is better able to deal with resistance,27 this article shows that in an authoritarian state with a number of lower-level authorities, the central government can be better positioned by giving up part of its power to local governments.28 This structure of divided state power assists the central government in dealing with resistance in several important ways. First, the divided power creates space for the central government to avoid blame. Authoritarian states are said to be less able to avoid blame because the concentration of power in the hands of the government also means the concentration of responsibility and blame.29 By delegating the power to deal with popular resistance to lower-level authorities, the central government shifts the responsibility to those lower-level authorities, thereby avoiding being in blame-generating situations itself. Secondly, while tolerating resistance which does not threaten the regime, this structure places constraints on both resisters and local governments. Unlike the situation where there is only a single authority, the existence of multiple authorities implies that the states policies towards protesters are inconsistent when the interests of those state authorities differ. These inconsistencies not only create political space for resistance but also prevent local authorities from adopting certain modes of response (for example, violent repression). However, resisters face more serious constraints. The local authorities autonomy, though conditional, still allows them to adopt a range of responses to deal with most of the instances of resistance in the light of their own interests, which implies that concessions to resisters are conditional. Therefore, the divided state power prevents both excessive repression that damages the regimes legitimacy and unconditional concessions that lead to the lack of discipline.30 Thirdly, this structure reduces the uncertainties for the central government. Goldstone and Tilly suggest that the state faces considerable hazards in dealing with popular resistance largely because the state, in general, will have a hard time picking the right level of concessions and repression to respond to a group.31 In other words, concessions or repression at the wrong time may prove costly or fatal.32 A power structure with multiple authorities reduces the hazards for the central government because only a very limited number of instances of social discord will need to be addressed by the central government itself. The uncertainty is further reduced because those limited instances have already been
27 Doug McAdam, Political Process and the Development of Black Insurgency, 19301970, 2nd edn (Chicago: University of Chicago Press, 1999), pp. 447; Charles Tilly, From Mobilization to Revolution (New York: Random House, 1978), pp. 11415. 28 Authoritarian regimes can be fragmented. The fragmented authoritarianism model, for example, has been used to describe policy making in China. However, the fragmented authoritarianism model focuses more on the horizontal divide among state agencies, whereas this article focuses on the vertical divide between the central and local governments. See Kenneth Lieberthal and David Lampton, eds, Bureaucracy, Politics, and Decision Making in Post-Mao China (Berkeley: University of California Press, 1992). 29 Kent Weaver, The Politics of Blame Avoidance, Journal of Public Policy, 6 (1986), 37198. 30 Federick Barghoon, Politics in Russia (Boston, Mass.: Little, Brown, 1972), p. 308. 31 Goldstone and Tilly, Threat (and Opportunity), p. 187. 32 Lohmann, The Dynamics of Informational Cascades.

416

CAI

(mis)handled by the local government, which gives the central government more time and information to work on solutions. As discussed below, the divided state power is crucial to the regimes resilience in dealing with popular resistance in China today.
THE POWER STRUCTURE IN CHINA AND ITS IMPLICATIONS

Chinas political hierarchy has multiple levels of government.33 For analytical convenience, the political structure can be seen as consisting of the central government and local governments. Decentralization in China has granted considerable power and autonomy to local governments. Indeed, they enjoy more autonomy than their counterparts in some former socialist countries (for example, the Soviet Union) or even democracies (for instance, India).34 The autonomy assumed by local governments in China has been consolidated or strengthened because of their role in economic development, which has led some scholars to conclude that there is a Chinese style of federalism.35 Given their power, local governments in China are doomed to play a crucial role in managing local affairs, which is also the reason for their abuse of power.36 The delegation of power, however, also means the delegation of responsibility (and blame). Decentralization shifts most of the responsibility of dealing with citizens resistance to local governments, and the central government holds local governments accountable by assigning the responsibility directly to local leaders. The Chinese political system is different from the federal system in a democracy because local ofcials in China are appointed by the upper-level authorities and have to full assigned responsibilities.37 During the reform period, maintaining social stability has been one of the most important responsibilities assigned to local governments. In the Chinese political system, top leaders (i.e., the party secretary and the head of the government) at each local level assume the most power. Also, because dealing with citizens resistance is often an urgent matter, local leaders, especially the party secretary, assume the most decision-making power and thereby the most responsibility in maintaining social stability.38 Their performance or even political careers are closely linked to their fullment of this responsibility.39 Top local leaders are expected to handle large-scale resistance in person. For example, in Anhui province in 2005, the party secretary of a city, who was also promoted to hold another

Kenneth Lieberthal, Governing China: From Revolution through Reform (New York: W.W. Norton, 1995). Pradeep Chhibber and Samuel Eldersveld, Local Elites and Popular Support for Economic Reform in China and India, Comparative Political Studies, 33 (2000), 35073. 35 Gabriella Montinola, Yingyi Qian and Barry Weingast, Federalism, Chinese Style: The Political Basis for Economic Success, World Politics, 41 (1996), 5081. 36 Kevin OBrien and Lianjiang Li, Selective Policy Implementation in Rural China, Comparative Politics, 31 (1999), 16786. 37 Yasheng Huang, Ination and Investment Controls in China: The Political Economy of CentralLocal Relations during the Reform Era (New York: Cambridge University Press, 1996). 38 At the county and township levels, it is common for the party secretary to sign the so-called responsibility contract with upper-level authorities to ensure the fullment of assigned tasks, including achieving social stability. City-level party secretaries may not be required to sign such contracts, but they are certainly assessed for this responsibility (Dong Xueqing, Zhang Heping and Zhang Zeyuan, The Life of County Party Secretaries, Perspectives, 9 November 2005, pp. 214; Maria Edin, State Capacity and Local Agent Control in China: CCP Cadre Management from a Township Perspective, China Quarterly, 173 (2003), 3572. 39 Kevin OBrien and Lianjiang Li, Rightful Resistance in Rural China (New York: Cambridge University Press, 2006); Edin, State Capacity and Local Agent Control in China.
34

33

Power Structure and Regime Resilience

417

important position of vice provincial governor, was removed because he was absent without justiable reasons when a riot involving 10,000 people occurred in his city.40 The Choice of the Mode of Response While local leaders are strongly motivated to prevent and stop citizens resistance, especially large-scale resistance, they face constraints in choosing the mode of response because of the different priorities between the central and local governments when they deal with citizens non-political resistance.41 Both levels of government may adopt one of the following modes of response: (1) concessions (i.e., citizens demands are met); (2) concessions with discipline (i.e., citizens demands are met, but some or all participants are punished); (3) tolerance (i.e., citizens demands are ignored, but the government also tolerates their resistance); and (4) repression (i.e., citizens demands are ignored, and some or all participants are punished). A crucial factor that makes the central government behave differently from local governments is its greater interest in protecting the regimes legitimacy.42 Legitimacy is about the political systems worthiness of being recognized.43 Given that the central government is more responsible for the operation of the political system or it largely represents the regime, it has a greater interest in protecting the regimes legitimacy. In contrast, local ofcials in China are more concerned with policy implementation or task fullment (such as maintaining social stability) and local issues, and thus legitimacy is not their main concern.44 A simple comparison of the incentive structures faced by the two levels of government reveals that the central governments more serious concern over legitimacy helps to make it more tolerant towards popular resistance than the local government. The central and local governments incur two types of costs when concessions are made: (1) economic and/or political costs; and (2) signs of weakness (see Table 1). Addressing citizens complaints may require the expenditure of nancial resources (economic costs). It may also require the government to correct its practices or policies or discipline state agents deemed responsible for the grievances or resistance (political costs). The cost of showing weakness is that making concessions may trigger more demands or actions. If the
In the riot, police cars were torched, six police ofcers were wounded and a nearby department store was looted. It was later discovered that the party secretary was with his mistress in another city for relaxation, and he was annoyed by the phone calls and turned off his cell phone. He was later arrested for corruption (Xiao Zhong, An Investigation of the Removal of He Minxu, a Vice Governor of Anhui Province, Inspections, 20 (2006), 212). 41 By non-political resistance, I mean that participants do not raise political demands that challenge the political authority of the party or the government. Such actions are normally based on legitimate or lawful claims or, as the government claims, are internal conicts within the people. To be sure, non-political resistance may be boundary-spanning or illegal according to the Chinese law. The government treats actions that aim to overthrow the party or the government or attack important facilities as unacceptable (see Long Xianlei, Sticking to Rule by Law and Appropriately Handling Instances of Collective Action, Public Security Research, 12 (2001), 503; Fan Fuming, Characteristics of Collective Action in Blocking Railways and Some Countermeasures, Journal of Shanghai Public Security Academy, 3 (2001), 304; Kevin OBrien, Neither Transgressive Nor Contained: Boundary-Spanning Contention in China, Mobilization, 8 (2003), 5164). 42 Ethan Michelson, Peasants Burdens and State Response: Exploring State Concession to Popular Tax Resistance in Rural China (unpublished, 2005). 43 Frank Michelman, Idas Way: Constructing the Respect-Worthy Governmental System, Fordham Law Review, 72 (2003), 34562. 44 OBrien and Li, Selective Policy Implementation in Rural China; Thomas Bernstein and Xiaobo Lu, Taxation without Representation in Contemporary Rural China (New York: Cambridge University Press, 2003).
40

418
TABLE

CAI

Costs and Benets in Concessions and Repression

Benets Costs a. Economic and/or political costs b. Signs of weakness a. Loss of legitimacy b. Risks in repressive measures

Concessions Repression

a. Stopping resistance b. Gaining legitimacy Deterring resistance

cost of showing weakness is considered to be the same for the two levels of government, the central government incurs a smaller cost than the local government does when making concessions. First, when the central government intervenes in a dispute and makes concessions, it often requires the local government to address citizen grievances with local nancial resources. Secondly, when malfeasant local ofcials are disciplined, the local government incurs a heavier loss because it relies more directly on these ofcials for local governance than does the central government. As far as the benets are concerned, if stopping resistance is equally desirable to the two levels of government, the central government gains more because legitimacy is more important to the central government than to the local government. In the case of repression, the two levels of government also incur two types of costs: (1) loss of legitimacy; and (2) risk from repression (Table 1). Repressing citizens resistance with legitimate claims damages the regimes legitimacy. For the reasons discussed above, the central government faces a higher cost of losing legitimacy than the local government when repression is used. The other type of cost incurred from repression is the risk arising from ineffective or failed repressive measures. For the central government, the risk means ineffective repression leading to more serious or regime-threatening resistance. This possibility (or the risk) is rather small given the power of the central party-state in China. In contrast, the local government faces a more serious risk in repressing resistance. For one, ineffective repression may cause the escalation of resistance, which signals the local governments failure in maintaining social stability. For another, forceful repression which causes serious casualties will damage regime legitimacy.45 In either situation, the central government will intervene, and local ofcials may be punished. However, the local government will face a small cost from repression if it is able to use the modes of repression that carry little risk (i.e., legitimacy is not its serious concern). Hence, when the risk from repression is low for both levels of government, the local government is more likely to use repression than the central government.

The central government prohibits the use of force in dealing with peaceful resistance. According to the 1998 directive issued by the central government, state authorities are required to settle instances of social unrest based on considerations of whether or not participants demands are political, there is organized violence, there is intentional confrontation, and there is support for the resistance from overseas. In dealing with organized attacks that are consciously directed at party and government agencies and other important facilities, force can be used, but it should not be used to deal with non-political and peaceful action by citizens with legitimate demands. Police ofcers who directly face participants can bring equipment but not weapons with them, while back-up ofcers can bring weapons with them if necessary (see Long, Sticking to Rule by Law and Appropriately Handling Instances of Collective Action).

45

Power Structure and Regime Resilience

419

Yet, although the central government is more tolerant of citizens resistance, its concessions are conditional. When the central government decides to make concessions to citizens, it often means that it will intervene in the conict between citizens and a local government. This is likely when the central government feels the pressure to stop the resistance and to protect regime legitimacy. In collective resistance, a few factors often make the central government regard intervention as necessary: (1) casualties from the resistance, (2) media exposure, and (3) the number of participants in the resistance. Casualties and media exposure can generate serious pressure on legitimacy, while the large number of people indicates the scope of resentment and the potential disruptive consequences. Hence, a combination of casualties with media exposure or with a large number of participants is very likely to invite intervention from the central government. Conversely, the central government is very unlikely to intervene in disputes that are peaceful and small scale simply because the pressure for intervention is small. The central government may also fail to intervene in some large-scale incidents because of the political costs incurred from the necessity of disciplining local governments or their ofcials. Punishing local ofcials, especially high-ranking ones, can be a complex political issue. Specically, the central government is less willing to intervene in cases that involve high-ranking local ofcials (for example, at the provincial level), as opposed to low-ranking local ofcials (for example, at the township or county levels).46 These conditions for the central governments intervention affect the local governments responses. When facing intervention or a threat of intervention from the central government, local governments will have to use concessions or concessions with discipline to stop citizens resistance quickly. However, in the absence of intervention or when the likelihood of intervention is negligibly small, local governments assume considerable autonomy in choosing the mode of response. Concessions will be difcult if (1) the local government faces high economic or political costs; and (2) partial concessions or tolerance results in persistent resistance which threatens social stability or local leaders images. In these circumstances, repression is a low-cost option, but the local government will also take the blame for repression.
DEALING WITH POPULAR RESISTANCE IN CHINA

To examine how the government deals with popular resistance is largely to examine the outcomes of the resistance. Popular resistance in China has received much academic attention in recent years,47 but how the government at each level responds to it has not been adequately examined. This is largely because cases of collective resistance are rarely reported in a way that makes systematic comparisons feasible. To reduce this problem, this research tries to tap as many sources as possible, which is achievable due to the increase of reports in and outside China. This study is based on a collection of seventy-eight cases (see Table 2). A basic criterion for including a case is that there should be information about the outcome (i.e., whether or not the protesters demands were met and how the
Li Zhilun, Strengthening the Discipline of Party Members and Governance and Doing a Good Job of Enforcing Anticorruption Measures, China Discipline Inspection, 4 (2000), 57. 47 See, among others, OBrien and Li, Rightful Resistance in Rural China, chap. 2; Yongshun Cai, State and Laid-Off Workers in Reform China: The Silence and Collective Action of the Retrenched (London: Routledge, 2006), chap. 7; Bernstein and Lu, Taxation without Representation in Contemporary Rural China, chap. 5; Elizabeth Perry and Mark Seldon, eds, Chinese Society: Change, Conict and Resistance, 2nd edn (New York: Routledge, 2003).
46

420
TABLE

CAI

Number of Cases Collected

Number Number Results of action Concessions Concessions with discipline Repression Tolerance Total Punishment (people) Detained Arrested Sentenced to jail Total 10 10 54 4 78 308 21 255 584

Modes of action Petitions Protests/demonstration Confrontations Attacks on state agencies Total Intervention by The central government Provincial authority only None Total Source: Authors collection.

29 17 17 15 78 5 10 63 78

government dealt with the participants, see Appendix).48 Certainly, this number does not allow us to draw conclusions on the frequency of successful resistance or the frequency of repression in China. Nonetheless, our collection includes cases that occurred in twenty-one provinces between 1995 and 2006, and they do shed light on the states response to popular resistance. Our data provide evidence for the rationale behind government response analysed in the earlier section. They show that the outcome of popular resistance is largely determined by the cost of concessions and the forcefulness of resistance. For analytical convenience, we divide participants claims into high-cost demands and low-cost demands. A demand is seen as costly to meet if it creates a zero-sum game between the government and protesters in the sense that the protesters gain is the governments loss (i.e., economic or political cost); otherwise, a demand is seen as less costly to meet (for example, addressing citizen grievances caused by a third party). The political hierarchy in China implies that intervention or the threat of intervention from the central authorities remains the most serious constraint on local governments or a crucial condition for successful resistance.49 Based on the three conditions for intervention discussed earlier, we see an action as forceful if it has more than 500 participants, involves casualties or receives media coverage; otherwise, an action is seen as less forceful.50 As Figure 1 shows, repression is most likely to be used when the demands are costly while the resistance is not forceful. Of the forty-two cases of less forceful resistance with high-cost demands, forty were repressed. By comparison, concessions or concessions with discipline are commonly used to deal with forceful resistance. Only three of the twenty instances of forceful resistance were repressed. Yet, success in resistance with high-cost

48 It must be acknowledged that this criterion for case collection has a limitation because cases whose results were not reported were excluded. This is also the reason why our collection only includes four cases that were tolerated or ignored by local governments. 49 OBrien and Li, Rightful Resistance in Rural China, chap. 2; Bernstein and Lu, Taxation without Representation in Contemporary Rural China, chap. 5. 50 According to the police department, a collective action involving more than 500 is seen as a large-scale action, and an action involving more than 1,000 participants is seen as an especially large-scale action (see Chen Jinsheng, Research Report on Instances of Collective Action (Beijing: Mass Press, 2004), p. 32).

Power Structure and Regime Resilience

Forcefulness of Action Cost of concessions Low Forceful a. Concessions ( 5 )* b. Repression c. Tolerance a. Concessions ( 12 ) b. Repression ( 3 ) c. Tolerance Not Forceful a. Concessions ( 2 ) b. Repression ( 6 ) c. Tolerance ( 3 ) a. Concessions ( 1 ) b. Repression ( 40 ) c. Tolerance ( 1 )

421

High

Fig. 1. State response to popular resistance Notes: *Including two cases of concessions with discipline; Including eight cases of concessions with discipline; N 73; ve cases were excluded because they did not have clear demands.

demands can be costly because the local government commonly uses concessions with discipline to deal with such incidents. Our data thus reveal local governments constraints and autonomy in dealing with popular resistance. By demonstrating the circumstances under which concessions and repression are used by local governments, the remainder of this section examines how the political arrangement has helped to maintain social stability in China while protecting the central governments legitimacy. Intervention from the Central Government and Concessions Among the seventy-eight cases we collected, local governments made concessions or concessions with discipline in eleven of the twenty successful cases because of the intervention from the central government (ve cases) or provincial governments (six cases). Those cases also show that the intervention is highly conditional or selective. Usually, once the central government has decided to intervene, it has to make concessions to citizens and sometimes punish malfeasant agents to show how your sins have provoked the wrath of the fanatics and have brought this punishment upon yourselves.51 The central government tends to intervene when the pressure of regime legitimacy mounts, especially after media exposure. For example, in Jiahe county in Hunan province, the local governments carried out housing demolition in 2003 without providing reasonable compensation to the homeowners and arrested three people who resisted. After Chinese Central Television disclosed the case, the central government sent a team to the county. In the end, the homeowners received more compensation and the three people were released. Four ofcials, including the county party secretary and the county magistrate, were removed and a fth was given party discipline.52 At other times, popular resistance invites intervention when protesters pressure the legitimacy of the central government with innovative tactics. In Linquan county in Anhui province, the government imposed heavy taxes and fees on the already poor peasants by ignoring the central governments regulations.53 From 1992 to 1995, peasants from a village made repeated but fruitless appeals to state authorities at the central and local levels.
Gamson, The Strategy of Social Protest, p. 82. Southern Urban News, 6 June 2004, p. 2. 53 Chen Guidi and Chun Tao, Survey of Chinese Peasants (Beijing: Peoples Literature Press, 2003), pp. 82120.
52 51

422

CAI

Two peasant leaders were sentenced to jail for the petitions. Under the leadership of one peasant who was released from the prison, seventy-four peasants went to Beijing to make another desperate appeal. On Sunday 29 October 1995, the seventy-four peasants went to Tiananmen Square one after another to avoid being noticed as a group. After all of them had arrived, they suddenly knelt down. Their action immediately attracted attention from the central government, which required the local ofcials from Anhui to travel to Beijing that night. The peasants problems were addressed and the county party secretary was transferred. The central government may also intervene because the scale of resistance is too large to ignore and the resistance involves casualties. In October 2004, about 100,000 peasants in Hanyuan, a county in Sichuan province, protested against low compensation for the loss of farmland and homes due to the construction of a dam. After repeated but fruitless petitions to higher-level authorities, tens of thousands of peasants approached the construction site on the night of 27 October 2004 to stop the construction. The government sent about 1,500 police ofcers to maintain order. Violent confrontations occurred between the peasants and the police ofcers, resulting in the deaths of both citizens and police ofcers. Given the vast number of peasants, the local government sent more than 10,000 militiamen to help establish order. Reportedly, the provincial party secretary who went to the site to address the problem was surrounded by the peasants for hours. When the central government learned of the news, it sent a work team headed by a leader of the State Council to the county. The team announced the central governments decision that the event was a large-scale gathering of migrants who did not know the truth about the cause of confrontation (i.e., exempting the responsibility of most participants). Meanwhile, the provincial government promised to increase the compensation and to have their homes relocated. A deputy city mayor who had been promoted from being the countys party secretary, together with at least seven senior county ofcials, was detained on charges of corruption. The countys party secretary was also removed.54 These cases show that local ofcials who abuse power may face serious consequences. An intervention by the central government can both enhance regime legitimacy and warn local ofcials against the abuse of power. Such concessions are, however, conditional. The central government faces costs when the intervention leads to the punishment of local ofcials. In the above cases in Jiahe, Linquan and Hanyuan, the highest-level ofcials who were disciplined were county-level ofcials who were in the low rungs of the political hierarchy. When high-ranking local ofcials are involved in a dispute or its settlement, the central government is less willing to intervene.55 In these circumstances, the central government would reduce the damage to its image and regime legitimacy by prohibiting the media from covering the cases. In December 2005, a bloody confrontation occurred in Dongzhou, a village in Shanwei city in Guangdong province. In this village, the city government built a power plant that occupied a plot of land. The villagers complained about inadequate compensation for the land and pollution of their water. A deadly confrontation between villagers and police ofcers occurred on 6 December during which, according to the ofcial account, three peasants were killed and eight were wounded. Shooting unarmed citizens has been very

South China Morning Post, 4 January 2005, p. 4. In several legal cases, the central government knew the problems but refused to intervene because those cases involved some provincial leaders (21st Century Global Report, 2 March 2003, p. 1).
55

54

Power Structure and Regime Resilience

423

rare in China since 1989, but only the deputy head of the city public security bureau was given administrative discipline. The central government kept a very low prole in this case. It was not until 10 December that the Xinhua News Agency released a short news piece on this event, accusing a few peasants of inciting the riot. The central government sent a team to the province (but not to the village), and it apparently accepted the explanation of the local government, which blamed a few peasants for the confrontation. It is not very clear why the central government did not punish any other local ofcial in this case. One very likely reason is that provincial leaders in Guangdong had directly intervened in the settlement of this case. The shooting occurred on 6 December. The provincial party secretary went to the village the next day. The provincial party secretary was a member of the politburo, the paramount party body in China, and his views might have had to be respected even by the central authorities. The brief news released by the Xinhua News Agency on 10 December was provided by the local government. Hence, even if the central government found that the released report was awed, it would be almost impossible for it to change the statement later.56 The central governments intervention is not only conditional but also limited in that it does not intervene in local governments punishment of some participants, especially in drastic actions, after concessions are made. This is understandable because unconditional concessions will make the local government vulnerable. Tarrow writes, One of the most remarkable characteristics of collective action is that it expands the opportunities of others. Protesting groups put issues on the agenda with which other people identify and demonstrate the utility of collective action that others can copy or innovate upon.57 By punishing certain participants, the local government shows that a victory can be very costly, thereby reducing the encouraging effect. In the Hanyuan case, the county and city government imposed serious punishment on some participants ex post. In 2005, twenty-eight participants were tried in local courts and found guilty, and one was executed on the charge of killing a police ofcer.58 In the Linquan and Jiahe cases, participants were jailed before intervention from the central government. Intervention from the Provincial Authority and Concessions The above discussion does not mean that citizens will always fail without intervention from the central government. The local governments worry about a possible intervention from the central government may also lead to concessions. Most conicts between citizens and the government in China have been caused by grassroots governments or their agencies, especially at the county level.59 The highest-level local authority in a place is the provincial-level authority. Hence, an intervention from the provincial government is the most important way for citizens to have their voices heard in the absence of attention

After the bloody confrontation, the city government exercised strict control over the village, preventing a free ow of information between villagers and external parties for months. Police searched for and arrested those who had participated in the riots. Not surprisingly, this incident was not allowed to be reported in the country (South China Morning Post, 12 December 2005, p. 4; South China Morning Post, 3 March 2006, p. 6). 57 Tarrow, Power in Movement, p. 97. 58 Yanan Daily, 27 March 2006, p. 4 59 Xie Li, An Analysis of Administrative Reconsiderations in the Country in 2000, Administration and Law, 6 (2001), 1819.

56

424

CAI

from the central government.60 Those factors (i.e., casualties, media exposure and number of participants) that affect the central governments intervention also shape the provincial authoritys reaction. When a serious confrontation occurs in a place, the provincial government will naturally react because its response is likely to be interpreted by the central government as the local leaders ability to govern. In addition, as regulated, serious confrontations should be reported to the pertinent central authorities within a regulated period of time (for example, two hours).61 If such cases are disclosed in the media or are reported to the central government, provincial leaders will be under pressure to solve the conict as soon as possible. They may, therefore, punish those ofcials deemed responsible for the confrontation in order not to disappoint the central government. In one case in Dingzhou city (i.e., at the county administrative level) in Hebei province, the local government constructed a power plant, and the plant needed a piece of land from a village to store the coal ash from it.62 The construction of the ash-storage facility was contracted to a businessman who received strong support from the local government. In March 2004, when villagers learned that they had been seriously under-compensated, they began to make petitions to different levels of authorities but were ignored. A few villagers were sentenced to jail for their resistance. Before July 2004, the city government and the contractor made more than ten attempts to construct the ash-storage facility with force by sending police ofcers to the site, but they met the strong resistance of the peasants. The peasants then built dozens of tents on the land to prevent the construction, with 100 to 200 villagers on duty each night. Such peasant resistance upset the local government and the contractor. In the early morning of 11 June 2005, more than 300 thugs suddenly attacked the peasants who were sleeping in the tents, killing six peasants and wounding another forty-eight. One peasant managed to videotape the ghting at the cost of a broken arm. The ve-minute footage was later shown on the internet as well as by foreign television networks. The casualties and media coverage generated serious pressure on the local authorities, including the provincial government. The provincial public security bureau immediately formed a taskforce to solve the case, and it found that the contractor had hired the thugs. On 13 June 2005, two days after the event, the provincial authority in Hebei removed the city party secretary and the mayor. In December, twenty-seven suspects were tried. Four suspects who organized the attack were sentenced to death, three were sentenced to the death penalty with a reprieve, and another six, including the party secretary and the contractor, were sentenced to life imprisonment. The party secretary was accused of participating in the planning of the attack (their plan was not to inict fatal injuries on the villagers). The provincial government also decided not to use the villages land for the project.63 Nevertheless, whether or not the provincial authority will intervene and how it will intervene in a confrontation are also conditional. The provincial authority may not intervene or only intervene in a symbolic way if it believes that doing this will not draw blame from the central authorities as in the Shanwei case. In addition, as provincial leaders punish malfeasant ofcials mainly because they want to distance themselves from those
60 In addition, because the top leaders (i.e., the party secretary and the magistrate) in a county are directly under the management of the provincial authority, these leaders will not be effectively disciplined without intervention from the provincial authority. 61 Chen, Research Report on Instances of Collective Action, pp. 27597. 62 China Economic Times, 20 June 2005, p. 3. 63 Overseas Chinese News, 16 December 2005, p. 2.

Power Structure and Regime Resilience

425

ofcials, they do not have an incentive to prevent lower-level ofcials from punishing certain participants who are believed to have damaged the image of local authorities, including the provincial government. The provincial government may even require local governments to punish certain participants. Therefore, it is common for both local ofcials and some participants to be punished when serious confrontations occur. Of the seventy-eight cases we collected, the provincial authorities intervened in ten cases, and participants were detained, arrested or put in jail in eight of them. In a high-prole case in Zhejiang province, peasants succeeded in forcing local governments to accept their demands but paid a high cost. The government in Dongyang, a county-level city, set up an industrial park in one of its townships in 2001. The chemical factories in the park produced serious pollutants that threatened the villagers health and damaged their farmland. The villagers appealed to the local and central authorities repeatedly, but to no effect. In 2001, more than ten villagers were arrested by the local government for their assault on the factories in the park. After four years of fruitless efforts, the villagers built tents at the industrial parks entrance in 2005. For two weeks in March and April 2005, about 200 elderly villagers lived in the tents to keep a 24-hour vigil to block the entrance to the industrial park. In the early morning of 10 April 2005, the city government organized about 3,500 police ofcers and government ofcials to disperse the villagers and dismantle the tents. With a rumour ying around that two elderly women had been killed in the confrontation (reportedly, nobody was killed), thousands of angry villagers clashed with the police and government ofcials, beating them and smashing dozens of cars and buses that had ferried the police ofcers and ofcials. Local ofcials and police did not dare to use more force and ran away.64 After the confrontation, the local government was eager to end the situation, and six of the thirteen factories were ordered to move out of the township for good. The provincial authority imposed serious punishment on the major local ofcials deemed responsible for the event. The former city party secretary, who had been promoted as the head of the propaganda department of the higher-level city party committee, was removed from his position. The deputy city party secretary, who was also the mayor, was removed from both positions, and the township party secretary was also removed. Another ve ofcials, including a deputy mayor, were given administrative discipline.65 The peasants, however, also paid a high price. Eight peasants were given sentences of between eight months and ve years in jail. One ofcial said to them, If youre not guilty, then the government is. And the government cannot be guilty. As a result, the case of Huaxi illustrates that even when farmers win, they lose.66 Local Governments Autonomy and Repression While intervention from above does not always entirely favour citizens who stage resistance, their chance of success is signicantly reduced if there is no such intervention at all. Given the conditions for intervention from above discussed earlier, most cases of citizens resistance will not trigger intervention from above. First, most instances of
In the end, more than 140 people were sent to hospitals, a majority of them police ofcers and government ofcials, including a deputy city mayor (South China Morning Post, 3 March 2006, p. 2). 65 The other four were the head and two deputy heads of the environmental protection bureau and the township head (Jinhua Daily, 1 January 2006, p. 1). 66 South China Morning Post, 3 March 2006, p. 2.
64

426

CAI

resistance tend to be peaceful and do not involve serious casualties because the use of violence is not a common option on the part of citizens.67 What Tarrow writes is also true in China: Where violence occurs or is even likely, this gives authorities a mandate for repression and turns non-violent sympathizers away.68 As discussed below, citizens use of violence justies local governments repression and increases the risks for themselves. Secondly, citizens disruptive actions make sensitive news, and such stories are rarely reported in the media within China before they end. Thirdly, most instances are relatively small scale. The numbers of participants in disruptive actions range from dozens to tens of thousands in China. An account of about 3,600 instances that occurred in six provinces shows that the average number of participants was eighty-one.69 Nationwide, about 730,000 people participated in 8,700 instances of collective action in 1993, or on average eighty-four in each; in 2004, about 3.76 million people participated in 74,000 instances of collective action, or an average of fty-one in each.70 According to the police department, an action with more than 500 participants is regarded as a large-scale action, and an action with more than 1,000 participants is an especially large-scale one.71 If this criterion is used, most instances of collective action in China are small scale. Since 1994, collective actions with 100 or more participants have not exceeded 16 per cent of the total number of instances, and it was 12 per cent out of 58,500 in 2003.72 An account of 1,117 instances of social unrest shows that cases with more than 500 participants accounted for 7 per cent, and those cases with more than 1,000 participants accounted for 2.2 per cent, whereas those cases involving less than 100 participants accounted for 76 per cent.73 Despite these facts, the actual number of large-scale actions is not small. Nationwide, the number of instances with 100 or more participants in each increased from 1,400 in 1994 to 7,000 in 2003.74 However, it is impossible for the central and provincial governments to intervene in most of these cases. Therefore, local governments at the county and city levels, which handle the most incidents of resistance, assume signicant autonomy. The Rationale Behind Repression As mentioned earlier, local governments in China do not have to use repression if concessions are not costly. For example, existing research on worker resistance to economic restructuring nds that some unpaid workers succeeded in receiving some allowance by approaching the government.75 Nevertheless, repression does remain an important option, which suggests that popular resistance can be costly. In the seventy-eight cases we collected, 255 people were sentenced to jail for up to fteen years, and one was
Cai, State and Laid-Off Workers in Reform China, p. 95. Tarrow, Power in Movement, p. 104. 69 Chen, Research Report on Instances of Collective Action, pp. 634. 70 He, Properly Handling Conicts in Our Country, p. 9. 71 Chen, Research Report on Instances of Collective Action, p. 32. 72 It was 13.6 per cent (out of 10,000) in 1994 and 14.5 per cent (out of 30,000) in 2000 (He, Properly Handling Conicts in Our Country, p. 10). 73 Chen, Research Report on Instances of Collective Action, p. 62. 74 He, Properly Handling Conicts in Our Country, p. 9. 75 Ching Kwan Lee, The Revenge of History: Collective Memories and Labor Protests in Northeastern China, Ethnography, 1 (2000), 21737.
68 67

Power Structure and Regime Resilience

427

sentenced to death. Another 308 were detained, and twenty-one were arrested, some of whom were later tried in courts and were likely to be sentenced to jail. The cases we collected suggest several reasons for the use of repression. One is the cost of concessions. Popular resistance is likely to be suppressed when the participants demands are regarded as difcult to meet (see Figure 1). In our collection, forty-two of the seventy-eight cases concerned economic demands. However, not all economic demands are easy to meet. For example, twenty of the forty-two cases concerned disputes over land use in rural China, predominantly, the non-agricultural use of farmland. An ofcial in the Ministry of Land and Resources said in 2006, Almost all the serious law violations in land use involve the government or leaders.76 As local governments tend to reap benets using farmland for non-agricultural activities, they are reluctant to pay peasants adequately. Repression becomes their common option when farmers resist. Secondly, repression is used because citizens resistance threatens the image or authority of local governments or local leaders. Take collective petitions as an example. Participants were punished in nineteen of the twenty-nine petition cases we collected. In one case in Gansu province in 2005, more than 200 villagers appealed to the county government for the release of a person who provided legal advice to the villagers in a land dispute. However, the petitioners were accused of attacking state agencies, though they did not even enter the ofce compound. Eleven petitioners were sentenced to jail for up to ve and a half years.77 Sometimes, petitioners who have a high stake in their action refuse to give up their resistance easily. Yet, when their resistance becomes persistent, local governments are upset because such resistance may escalate into more disruptive actions or be directed to upper-level authorities. In twelve of the nineteen cases, citizens were punished because they made repeated petitions. In one case in Hebei province in 2000, three villagers were sentenced to jail for up to ve years because they led villagers to appeal to the city government repeatedly for a solution to a land dispute.78 Local governments are especially reluctant to see their local people present petitions to provincial or central authorities. In ve cases, the participants were punished because they appealed to the central and provincial authorities in an unacceptable way. In one case, ve petitioners from Shanxi province went to the central authorities in Beijing in 2005 but were accused of violating the rules. All of them were ruled guilty in the local court in Shanxi and were sentenced to jail for three to ve and a half years.79 A provincial-level ofcial thus said, Protection of and sympathy for petitioners is possible only to a limited extent. Arresting petitioners often reects the will of the major party and government leaders.80 Thirdly, repression is used because the political risk is small. Most incidents of resistance are not forceful enough to deter repression. While intervention from above is an important condition for successful resistance, it was absent in sixty-three of the seventy-eight cases we collected. Nationwide, the percentage of the cases ignored by the central and provincial governments could be much larger. The lack of intervention has to do with the scale of resistance, which partly indicates its forcefulness. In the fty-six failed cases (excluding two riots) we collected, only in two cases did the number of participants reach 1,000. Local
76 77 78 79 80

New Beijing Post, 16 April 2006, p. 2. Western Business News, 9 February 2006, p. 2. China Youth Daily, 31 August 2001, p. 3. Shanxi Daily, 13 September 2005, p. 1. Li Junde, Is Petitioning A Crime? Perspective, 7 April 2003, pp. 256.

428

CAI

governments made concessions without intervention from above in nine of the twenty successful cases. In four of the nine cases, the number of participants exceeded 5,000. In another four cases, the number of participants ranged from 1,000 to more than 3,000. The nal case was a small-scale one in which more than 200 villagers appealed to the city government, complaining about the corruption of their village party secretary. These cases do not mean that small-scale resistance never succeeds, but its chance is much smaller because both the central and local governments feel less pressure.81 The risks from repression are reduced also because pertinent laws provide convenient bases for local governments to punish certain participants who are commonly accused of holding illegal assemblies and demonstrations, inciting the masses to attack a government agency, inciting the masses to disrupt social order, or assembling a crowd to disrupt trafc. For example, according to criminal law, those who organize the masses to attack state agencies can be sentenced to up to ten years in prison. In addition, a collective petition to the government with more than ve participants is also illegal. Participants who intend to form independent organizations or have connections with overseas organizations may be accused of subverting the government, which is a political crime. Whether or not popular resistance violates the law can be subject to the interpretation of the local government. Certainly, some participants have been punished because their behaviour seriously violated the law and caused damage. In a district in Guangdong province, a few villagers who had conicts with the township government, together with students who were resentful about the high tuition fees, held a demonstration in February 2001, which also attracted other participants. After they had marched to the township government, the demonstration turned into a riot. The participants attacked the township government, burned down 107 rooms and took more than twenty motorcycles and other property belonging to government employees. As a result, twenty-three participants were punished, with one sentenced to jail for ten years, two sentenced to six years and another two sentenced to ve years.82 Repression is commonly used to deal with such riots. Given local governments autonomy, participants have to be cautious in choosing the mode of resistance in order to reduce the risks. As elsewhere, the non-violent demonstrators need more courage; the militants have to be wiser.83 In China, when violence occurs, it is rarely planned; instead, it is often triggered by factors that are not anticipated by the participants or the government. All the same, as mentioned earlier, many participants may still be punished for peaceful resistance because local governments have discretion to interpret their actions, especially when some actions are boundaryspanning.84 In forty-nine of seventy-eight cases, participants resorted to protests, demonstrations, confrontations with ofcials or police ofcers or attacks on state agencies (as claimed by the government). In thirty-six cases (excluding the cases of concessions with discipline), the participants were punished. In one case in a county in the Guangxi Autonomous Region in 2005, more than 200 villagers approached the county authorities because of a dispute
Xi Chen, Chinese Petitioners Tactics and Their Efcacy (paper presented at the Conference on Grassroots Political Reform in Contemporary China, Fairbank Center, Harvard University, October 2004). 82 Pingdingshan Daily, 2 January 2002, p. 3. 83 Donatella della Porta, Protest, Protesters, and Protest Policing: Public Discourses in Italy and Germany from the 1960s to the 1980s, in Marco Giugni, Doug McAdam and Charles Tilly, eds, How Social Movements Matter (Minneapolis: University of Minnesota Press, 1999), pp. 6696, at p. 82. 84 OBrien, Neither Transgressive Nor Contained.
81

Power Structure and Regime Resilience

429

over land ownership. They were accused of disrupting the social order and about 110 peasants were detained although they had not caused any damage. Eighty-two were released after each had paid between 2,000 to 8,000 yuan in bail, and ten were sent to labour camps for up to two years. Another seventeen were sentenced to jail for up to eight years. This case may not be typical, but it indicates the degree to which the local government can abuse its power.85 Such cases of repression have no doubt damaged the image of the local government. Nevertheless, the people may not blame the central or even the provincial government as much as they blame the local government because the former governments are not involved in the repression.86 In forty-three of the seventy-eight cases in which some participants were put in jail, only in one case was the rst trial conducted in a city court (i.e., two worker leaders in Liaoyang were accused of subverting the government), whereas all the other cases were tried in county-level courts. This implies that almost all the lawsuits (i.e., forty-two cases) against the participants were led by state authorities at the county level or lower. Hence, it may not be an exaggeration to state that the repression in most instances has been carried out without the knowledge of the central government. Although Chinese citizens trust in the central government easily tends to disappear after they have made fruitless appeals to the central authorities,87 it is very likely that participants in most of the tens of thousands of cases have not appealed to the central government before they fail or are repressed. Yet, the central government will face a serious problem if an increasingly large number of citizens realize that it is unwilling or unable to address their grievances.
CONCLUSION

Unlike most transitional economies, China is simultaneously undergoing market reforms, socio-economic transformation and urbanization. The society is replete with conicts and citizens resistance. In the pre-reform period, mass movements (such as the Cultural Revolution) were initiated by the party-state or top leaders, and they could be stopped when the elite decided to do so. In comparison, during the reform period, citizens undertake incidents of resistance based on their own interests. Therefore, how to manage social conict is a serious issue faced by the Chinese party-state. Economic development can be a solution to some of the problems that have triggered resistance, but it is far from adequate. As some of the cases presented in this article have shown, certain modes of development (such as the occupation of farmland, privatization and pollution) have caused conicts directly. This article demonstrates that the power structure within the Chinese political system adds signicant resilience to the regime in addressing social conicts by granting conditional autonomy to local governments. Decentralization in China has institutionalized the strong power of local governments, but local ofcials also have to full assigned responsibilities. This political arrangement has important implications for the management of citizen resistance. One is that it affects the regimes legitimacy. Kuran shows that the sudden collapse of socialist regimes in Eastern Europe has to do with citizens who chose to pursue their true preferences when opportunities (for example, crises) arose, with these

85 86 87

Takungbao, 20 January 2006, p. 4. Lianjiang Li, Political Trust in Rural China, Modern China, 30 (2003), 22858. Yu Jianrong, Criticisms of the Petition System in China, China Reform, 2 (2005), 268.

430

CAI

true preferences differing from what the governments desired because the regimes had lost legitimacy.88 In China, decentralization helps to protect the legitimacy of the central government and the regime in two ways. First, the decentralized power structure allows the central government to distance itself from blame-generating situations when local governments use repression, which is a basic method of avoiding blame or protecting legitimacy.89 Moreover, when large-scale incidents or incidents involving casualties occur, the central governments intervention shifts the responsibility and blame to local governments. This arrangement also allows the central government to pretend not to know of the repression when it does not want to intervene. Secondly, given that local governments assume considerable power and autonomy, there can be variation in their attitudes towards disgruntled citizens. Such variation may reduce citizens blame of the political system (i.e., they attribute the problems to their local governments), and citizens perceptions of the regimes legitimacy may vary across the country.90 Divided state power also promotes the regimes resilience by improving information ow. Reliance on repression makes the government less sensitive to citizens problems and complaints, which further damages the regimes legitimacy.91 In China, by tolerating resistance from citizens that does not threaten the regime, the central government is better informed about the social situation and the sources of citizens grievances,92 which allows the party-state to address some of the problems to protect its legitimacy. Therefore, popular resistance in China has not only helped to force the central government to strengthen the implementation of policies favouring citizens but has also contributed to the adjustment of national policies disfavouring citizens.93 However, local governments autonomy and power also prevents what can be called abby exibility or lack of discipline.94 Local governments power determines that citizens resistance in China is short lived, often isolated and does not challenge the regime. Unlike the communist government in Poland, both the central and local governments in China are intolerant of organized resistance. Therefore, the many instances of resistance by different groups of people who have been affected by the reform or socio-economic changes remain largely unconnected in China.95 If effective challenges to a regime are made possible often by the presence of organizational networks that can mobilize participants,96 what happened in communist Poland (i.e., regime-threatening resistance)

Kuran, Now Out of Never. Weaver, The Politics of Blame Avoidance. 90 See, for example, Jie Chen, Popular Political Support in Urban China (Palo Alto, Calif.: Stanford University Press, 2004). 91 In the former East Germany, the populace was by and large discontented with its standard of living and lack of political freedom, and it grew increasingly so in the period 197589. However, there had been little action taken until 1989 because the violent suppression of the 1953 mass uprising with the help of Soviet military forces rendered the cost of anti-regime protest prohibitive for all but a small minority. Nonetheless, repression also led the people to lose condence in the communist system. See Lohmann, The Dynamics of Information Cascades. 92 OBrien and Li, Rightful Resistance in Rural China, chap. 2. 93 Chen and Chun, Survey of Chinese Peasants; Bernstein and Lu, Taxation without Representation in Contemporary Rural China. 94 Barghoon, Politics in Russia, p. 308. 95 Bernstein and Lu, Taxation without Representation in Contemporary Rural China, chap. 5. 96 McAdam, Political Process and the Development of Black Insurgency, 193070; Tilly, From Mobilization to Revolution.
89

88

Power Structure and Regime Resilience

431

is unlikely to take place in China at least in the near future. The reason is simple: when threats [from the government] occur without a preexisting organizational infrastructure, they will likely deter challengers from sustained contention.97 Nevertheless, given the multiple levels of authority in China, resistance is likely to persist as long as citizen rights are violated, and the central government is constrained by regime legitimacy. Tensions between citizens intentions to protect their rights and the limited political space will continue to serve as an important driving force for the party-state to institutionalize statecitizen interactions by restraining its power and getting the struggle off the streets and into the courts.98
APPENDIX: DATA COLLECTION

The data used in this study were collected from several sources. Forty-nine of the seventy-eight cases were collected from formal publications (i.e., permitted by the government) in China, and sixteen were collected from government websites or government-permitted sources. The remaining thirteen cases were collected from the authors eldwork in China and overseas media. Sources Petitions (twenty-nine cases). Chen Guidi and Chun Tao, Zhongguo nonmin diaocha (Beijing: Renmin wenxu chubanshe, 2004), pp. 82120; Shenzhen Yearbook 1998, p. 209; Fazhi ribao, 2 December 2000; Zhongguo qingnian bao, 31 August 2001; Fazhi ribao, 3 July 2000; Lian Xinqiao, Miandui baimin jifang qunzhong, Minqing yu xinfang, no. 5 (2001), p. 24; The Research Group, Zhongguo zhuanxingqi quntixing tufashijian duice yanjiu (Research on measures to deal with mass incidents) (Beijing: Xueyuan chubanshe, 2003), p. 316; http://news3.xinhuanet.com/focus/200304/16/content 833579.htm , accessed 21 March 2006; Zhongguo jingji shibao, 2 January 2003; Beijng yule xinbao, 7 January 2004; Philip P. Pan, Washington Post, 7 August 2004; Li Li, Baimin qunzhong jifang, fasheng zai huanjie xuanju hou, Minqing yu xinfang, no. 11 (2003), pp. 1617; Shengyang wanbao, 22 October 2004; http://www.sd. xinhuanet.com/news/200508/17/content 4902146.htm , accessed 21 March 2006; Shaanxi ribao, 13 September 2005; Huashangbao, 19 January 2006; Jinhua ribao, 14 March 2006; Huashangbao, 19 January 2006; Lanzhou chenbao, 31 December 2005; Shanxi fazhibao, 24 November 2004; http://www.pingyin.cn/2006/412/100111 755 72.html , accessed 22 April 2006; http:// www.yinzi.cn/sx/ShangLuo/news/2006/03/1910031421.html , accessed 25 March 2006; Nanyang ribao, 20 April 2006; Jinghua shibao (Jinghua Times), 9 March 2005; Yongshun Cai, State and Laid-off Workers in Reform China: The Silence and Collective Action of the Retrenched (London: Routledge, 2006), pp. 969; Mingpao, 7 March 2001; interviews with the author held in China in November 2005. Protests (seventeen cases). We want to eat, Far Eastern Economic Review, 160 (26 June 1997); http://dailynews.sina.com.cn/s/143443.html , accessed 4 February 2006; Nanfang ribao, 27 July 2000; Jinhua ribao, 12 June 2001; http://jcy.bjtzh.gov.cn/gzdt/jcyw/jcyw (03) 09.htm , accessed 2 April 2006; Eric Eckholm, Where workers, too, rust, bitterness boils, New York Times, 20 March 2002; Beijng yule xinbao, 7 January 2004; Nanfang duoshi bao, 18 May 2004; Nanfang duoshi bao, 18 May 2004; http://www.cq. chinanews. com.cn/newsview.asp? nid 52739 , accessed 2 April 2006; Fazhi ribao, 5 September 2004; Kathy Chen, Chinese protests grow more frequent, violent, Wall Street Journal, 5 November 2004; Yin zhoukan, 14 October 2004; Tang Jianguang, Ningxia Yinchuan chuzuche tingyun fengbo (The incident of the taxi drivers strike in Yinchuan of Ningxia), Xinwen zhoukan (Newsweekly), 8 September 2004, pp. 1113; http://www.lawyer-china.com/comhom/stilaw/791.asp , accessed 5 March 2006; http://news.sdinfo.net/72340168526266368/20010430/18722.shtml , accessed 22 March 2006; Shenzhen shangbao, 5 April 2006.

Almeida, Opportunity Organizations and Threat-Induced Contention, pp. 3878. Samuel Huntington, Political Order in Changing Societies (New Haven, Conn.: Yale University Press, 1968), p. 359.
98

97

432

CAI

Confrontations (seventeen cases). Qin Guolin, Chen Hong and Shi Dezhou, Qian ren jifang, Minqing yu xinfang, no. 3 (2002), pp. 401; Dai Gang, Hongjiang quntixing shijian chuzhi yinfa de xikao, Gongan yanjiu, no. 7 (2001), pp. 658; http://www.qtfy.gov.cn/show.asp?id 320 , accessed 3 March 2006; http://www.bbzy.org/detail.jsp?title , accessed 2 April 2006; Jinghua shibao, 2 April 2006; Shi Jiangtao, Peasants in upstream ght to halt dam, South China Morning Post, 4 January 2005; Joseph Kahn, For China masses, an increasingly short fuse, International Herald Tribune, 31 December 2004; Chizhou ribao, 27 June 2005; http://www. Longhui. net/news/local/200 405/14458.htm , accessed 21 March 2006; http://www.sjw.gov.cn/common/Show Info. aspx?id 145942 , accessed 2 March 2006; Edward Cody, For Chinese, peasant revolt is rare victory, Washington Post, 13 June 2005; Takungpao, 19 January 2006; Wang Keqin and Qiao Guodong, Hebei Dingzhou cunmin canlie duikang buming wuzhuang xiji an diaocha (An investigation of the armed attack on the villagers in Dingzhou in Hebei province), Zhongguo Jingji shibao (China Economic Times), 20 June 2005; Minnie Chan, Shanwei villagers seethe at ill-treatment in land grab, South China Morning Post, 3 March 2006; Mingpo, 13 October 2006; Howard French, Anger in China rises over threat to environment, New York Times, 19 July 2005; Yu Jianrong, Nongmin youzuzhi kangzhen jiqi fenxian, Zhanlu yu guanli, 3 (2003), 116; reports from district courts. Attacks (fteen cases). Nanfang ribao, 28 August 2000; Nanfang dushi bao, 24 November 1999; Jiancha ribao, 11 June 2001; Beijing wanbao, 5 January 2003; Beijng yule xinbao, 17 January 2005; Chengdu shangbao, 2 November 2004; Sanqin dushi bao, 29 October 2004; Yangzhi wanbao, 10 May 2005; Wuhan Wanbao, 27 September 2005; Xibu shangbao, 9 February 2006; http://www.investchina.com.cn/chinese/ law/92993.htm , accessed 15 April 2006; Chende ribao, 21 October 2003; http://hnfy.chinacourt.org/ public/detail.php?id 50328 , accessed 27 March 2006; Jiancha ribao, 19 December 2005; http://www.lz160.net/show news.asp?new id 3905 , accessed 10 April 2006.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.