INTRODUCTION
Religious groups are powerful social forces; in addition to shaping people's daily lives, they also frequently engage in many issues often considered to be “secular,” including elections, political campaigns, and national identity building (e.g., Limaye et al. Reference Limaye, Malik and Wirsing2004; van der Veer and Lehmann Reference van der Veer and Lehmann1999). Even China — a country ruled by an authoritarian and atheist regime — has witnessed a great wave of religious revival since the state lifted the total ban on religious groups in 1982 (e.g., Ying Reference Ying2006; Lai Reference Lai2003). Today, despite the strict restrictions, religious groups are playing active roles in China. In fact, given the limited freedom of association, these groups are among the country's most resilient social organizations (Dean Reference Dean2003; Tsai Reference Tsai2007). This article examines the roles played by rural religious groups in China's local contentious politics. More specifically, it aims to explore whether religious groups stimulate or reduce collective contention and to make sense of the underlying mechanisms involved in such issues.
Although reports on state vs. religion conflicts in China are not rare (e.g., Wong Reference Wong2001; Tong Reference Tong2009), there is yet to be a comprehensive and systematic examination of the actual roles of religious groups in China's contentious politics. This article intends to serve as the first step in bridging this gap by presenting studies of original data and cases collected from rural China. In so doing, it hopes to engage with, and contribute to, the ongoing theoretical debates as to whether religious groups increase or decrease collective contention. According to the empirical findings, overall the occurrence of collective contention appears to exhibit no significant difference between villages with different amounts of religious groups; however, collective contention is less likely to occur in villages with religious groups that simultaneously overlap with secular social organizations and local authorities. Such findings highlight two issues that are often side-lined, if not outright neglected, in the existing literature. First, the relationship between religious groups and collective contention is diverse rather than uniform. Second, this relationship is shaped not only by religious groups but also by other important players in the local political arena. Elaborating on these findings, the rest of this article first reviews the relevant theoretical debates and operationalizes the competing arguments into two hypotheses. It then examines both hypotheses against representative survey data from rural China and finds that neither holds. A new explanatory framework is thus raised to disentangle the heterogeneous association between religious groups and collective contention, suggesting that religious groups with different relational statuses may play different roles in contentious politics. This new framework is subsequently verified through statistical tests. The article concludes with a brief discussion on its theoretical contributions and practical implications.
THEORETICAL DEBATES
The roles of religious groups in contentious politics have long been studied. However, despite the rich literature in this field, no clear consensus on the nature of these roles has been reached. While religious groups have been widely regarded as socio-political tools of control, used by the rulers to suppress collective contention, as Trejo (Reference Trejo2009) notes, “one of the most consistent and surprising” social scientific findings in recent years has been the extent of the involvement of religious groups in the founding of some of the most powerful secular social and political movements. The theoretical debates on the relationship between religious groups and collective contention are thus still ongoing. While many theorists suggest that religious groups help to maintain social stability, acting as tranquillizers that reduce the likelihood of collective contention, others argue that religious groups frequently challenge existing social orders, serving as agitators that stimulate collective contention.
Scholars who view religious groups as tranquillizers believe that these groups can discourage conflicts either by blurring the differences between different social groups or reducing people's awareness of them. Marx (Reference Marx1977), for example, famously claimed that religion served as the “opium of the people.” Although there are divergences of opinion about what Marx actually intended by drawing the comparison, it is generally agreed that he accused religion of reducing people's awareness of class conflict and suppressing people's willingness to engage in class struggle by providing them with a “fantastic realization of the human essence” that “has not acquired any true reality.” Similarly, in his discussions of the theodicy of good fortune and misfortune, Weber indicates that religion and religious groups can “mollify those at the bottom of the social structure” and thus “sanctify the status quo” (Pyle and Davidson Reference Pyle, Davidson, Swatos and Kivisto1998). In addition, Stark (Reference Stark, Thielbar and Feldman1972) suggests that the aspects of faith that serve as a relief for suffering is more likely to attract the poor, while Pope (Reference Pope1942) and Howe (Reference Howe and McNall1981) argue that the wealthy and powerful are able to control and manipulate belief systems, often appropriating religious ideas to legitimate existing social structures. In addition, many scholars consider religious groups as important sources of social trust (e.g., Smidt Reference Smidt1999; Veenstra Reference Veenstra2002; Wuthnow Reference Wuthnow2002; Welch et al. Reference Welch, Sikkink, Sartain and Bond2004) and thus contributors toward democracy (Putnam et al. Reference Putnam, Leonardi and Nanetti1994), social integration (Inglehart Reference Inglehart1997), prosperity, and social harmony (Fukuyama Reference Fukuyama1995). These theories give rise to the following hypothesis:
Hypothesis 1: Religious groups in contemporary rural China correlate negatively with collective contention; other things being equal, the more religious groups in a sample village, the lower the likelihood of collective contention.
The aforementioned view, however, has recently come under increasing fire. Many researchers argue that religious groups actually agitate rather than tranquillize collective contention, offering three explanations for this finding. First, religious doctrines and values do not always perpetuate the status quo; instead, they may “serve a prophetic function, promoting social action to redress society's ills” (Pyle and Davidson Reference Pyle, Davidson, Swatos and Kivisto1998). For example, Lam (Reference Lam2006), Morris (Reference Morris1986), and Woodberry and Shah (Reference Woodberry and Shah2004) all argue that the participatory and civic attitudes embedded in Protestantism make Protestants more likely to engage with collective contention. Similarly, Nepstad (Reference Nepstad2004) demonstrates how some Christian clergy skillfully employ religious rituals, stories of martyrs, and biblical teachings to establish a link between faith and activism, while many scholars find that the “liberation theology” contributed much to pro-democratic protests in and beyond Latin America (e.g., Adriance Reference Adriance1986; Neal Reference Neal1987; Berryman Reference Berryman1987). Second, religious groups may cause or reinforce a so-called clash of civilizations and the associated or underlying socio-political conflicts. Huntington (Reference Huntington1996) argues that the primary axis of conflict in the post-Cold-War world falls not along ideological or economic lines but cultural and religious ones, and his thesis is supported by empirical evidence drawn from across the globe, including Southeast Asia (Searle Reference Searle2002; Houben Reference Houben2003), South Asia (Cady and Simon Reference Cady and Simon2006; Gould Reference Gould2011), the Middle East (Baumgartner et al. Reference Baumgartner, Francia and Morris2008), Africa (Ukiwo Reference Ukiwo2003; Ellis and Haar Reference Ellis and ter Haar2004) and the Balkan Peninsula (Bax Reference Bax2000). Last, religious groups may provide organizational structures and personnel for collective contention. Empirical research in Eastern Europe and the Arabic world demonstrates that powerful collective protests that eventually overthrew the ruling order often benefitted from pre-existing organizations and the leadership of religious groups (Salehi Reference Salehi1988; Osa Reference Osa and Smith1996). These theories imply an alternative hypothesis:
Hypothesis 2: Religious groups in contemporary rural China correlate positively with collective contention; other things being equal, the more religious groups in a sample village, the higher the likelihood of collective contention.
DATA AND THE INITIAL TEST
Contemporary China is an ideal setting to test the competing theories summarized in the previous section thanks to its richness and diverseness of both religious groups and collective contention (Cai Reference Cai2010; Goossaert and Palmer Reference Goossaert and Palmer2011; Yang Reference Yang2012). To systematically study religious groups and collective contention in rural China, working with colleagues I conducted a national survey covering 119 sampled villages (in 59 townships, 30 counties, and six provinces) selected through a stratified random sampling strategy in 2008. In addition to information on religious groups and collective contention in each sample village, the data also cover a list of demographic, socioeconomic, and political indictors, allowing us to better understand the background situations of our sample.Footnote 1 Table 1 reports the descriptive statistics of the sample villages.
Table 1. Background Information of the Sampled Villages
N = 119.
Based on the original survey data, Hypotheses 1 and 2, which share the same independent and dependent variables but vary in the direction of the correlation, are tested through the following ordinary least squares (OLS) regression model.Footnote 2
$$IfCC_i = {\alpha} + NRG_i {\beta} + CV_i {\gamma} + TD_i {\theta} + {\mu} _i $$
In Model A, IfCC is the dependent variable, which refers to occurrences of collective contention involving 30 or more participants in the sample village since 2000; NRG is the independent variable, which refers to the “number of religious groups” in a sampled village; CV refers to a vector of controls; TD refers to the township-level dummy variables; α is the intercept; β is the coefficient of NRG; γ is the coefficient of CV; θ is the coefficient of TD; μ is the residual error of the model; and i is the village code.Footnote 3 The set of controls, which are the same in all models, include 10 variables measuring the capability of local authorities, the incentives of village cadres, the quality of village elections, the extent and frequency of land requisition, and the geographic, demographic, and economic backgrounds of the sample.Footnote 4 The set of township-level dummy variables are included to ensure that comparisons are actually made within each pair of the sampled villages in the same township. In so doing, it is possible to eliminate the impacts that cultural traditions or local government policies have on collective contention, which are otherwise difficult to control for.
The statistical results of Model A are reported in Table 2. The insignificant coefficient of the independent variable suggests that the occurrence of collective contention appears to be on the same level in villages with and without religious groups. That is to say, neither of the two hypotheses formed from the established theories holds true against the empirical survey data. A new framework is thus necessary for understanding the actual relationship between religious groups and collective contention in rural China.
Table 2. Statistical Results for Model A
N = 119. Robust standard errors are in parentheses. Significance at the 10% level is represented by *, at the 5% level by **, and at the 1% level by ***.
THE DIVERSE ROLES OF RELIGIOUS GROUPS IN CONTENTIOUS POLITICS
The insignificant statistical results in Table 2 could be interpreted as implying that religious groups and collective contention are not related to each other in rural China. However, my field experience suggests that at least some groups are closely associated with collective contention in one way or another. Moreover, while religious uprisings may only form a very tiny proportion of the collective contention in contemporary China,Footnote 5 it is not uncommon for religious groups — the most resilient independent social organizations in the country — to get involved in contentious politics triggered by non-religious issues (Liu et al. Reference Liu, Liu, Tao and Tao2010). To understand the apparent mismatch among the theories, the statistical results, and my field experience, more attention should be paid to the special features of collective contention in authoritarian countries like China.
The Pathology of Collective Contention in Authoritarian Countries
The skyrocketing increase of collective contention in China is widely regarded as a result of the public anger and discontent fuelled by corruption and government failures (O'Brien Reference O'Brien2002a). While such an observation appears compelling, on its own it fails to explain two features of China's collective contention. For one, corruption and rent-seeking seem inevitable due to the absence of effective institutional arrangements against the abuse of power (Lü Reference Lü2000; Sun Reference Sun2004), so these problems are ubiquitous across China (Manion Reference Manion2004). Collective contention, however, is distributed quite unequally across different localities (Thornton Reference Thornton2012; Hurst Reference Hurst2004). In addition, collective contention often starts from trivial events that could happen anywhere, and participants of collective contention do not always have a clear picture of the events that triggered the entire episode (Liu et al. Reference Liu, Liu, Tao and Tao2010).
The solution to such a puzzle, I suggest, emerges when we follow Melucci (Reference Melucci1989) by shifting attention from why collective contention occurs to how it occurs. Grievances are prevalent among citizens in many societies (Oberschall Reference Oberschall1978; McCarthy and Zald Reference McCarthy and Zald1977), but only under some circumstances can they trigger collective contention. In democracies, the mandates of governments are granted through elections, and citizens can express their grievances through votes. Facing the pressure of an election, politicians have to pay serious attention to citizens' claims. As a result, negotiation channels between the state and society are generally sufficient and efficient, and citizens find it relatively easy to express their grievances through institutionalized means.Footnote 6 Moreover, the boundary between contained and transgressive behaviors is normally explicit and credible in democracies thanks to the rule of law, so citizens are less likely to participate in collective contention when they have legitimate channels through which to express positions and demands. Officials in autocracies and authoritarian regimes, however, often turn a blind eye to people's grievances since their power normally comes from above (Bendix Reference Bendix1977). Absolute autocracies can still maintain stability through strict social controls (Lau Reference Lau1990), but authoritarian countries like contemporary China are particular vulnerable to collective contention because their citizens often use the rights granted by the constitution and laws as weapons of “rightful resistance” (O'Brien and Li Reference O'Brien and Li2006).Footnote 7 Moreover, due to the unclear and unstable boundary between contained and transgressive contention, authoritarian states have to face the constant risk that “more concessions beget more contention” (O'Brien Reference O'Brien2002b).
Dealing with such problems is not a straightforward matter in authoritarian countries, as the leading newspapers and TV channels — typical sources of credible information in democracies — are controlled by the state, such that “more reliance is placed on oral and unofficial means of communications” (Fandy Reference Fandy2000). Moreover, the recent commercialization of media in China has provided more space for journalists to “push the envelope of what the regime considers off-limits by investigating stories about local corruption and abuses of power” (Nathan Reference Nathan2003). Therefore, the news media in China today, as Chen and Shi (Reference Chen and Shi2001) argue, have “negative effects on people's attitudes toward political institutions in general” and “make people distrust government.” Lacking credible communication channels, misperceptions between the state and the society, which are fuelled by widespread rumors, can easily turn trivial incidents into triggers of collective contention (Liu et al. Reference Liu, Liu, Tao and Tao2010). Moreover, without effective channels of communication, discontented citizens and the government find it impossible to credibly and explicitly demonstrate their true position and the price they are willing to pay to defend them. Under such circumstances, mutual agreements are hard to reach and collective contention becomes unavoidable.
However, some Chinese villages are still less vulnerable to the aforementioned problems than their neighbors. Despite the lack of trustworthy formal institutions for information exchange, citizens may still consider the signals and messages from the authoritarian state as credible if they personally believe in the political figures around them (Fewsmith Reference Fewsmith2001; Gibson Reference Gibson2001). In rural China, local cadres are the political figures who have the closest relations with villagers (Tao and Liu Reference Tao, Liu, Zang and Kou2013). Therefore, if a religious group can facilitate credible communications between local cadres and villagers, it has a better chance of preventing everyday grievances and conflicts from fermenting into collective contention.
Bridging Religious Groups with Secular Social Forces
Religious groups are widely regarded as important sources, containers, and maintainers of trust among their members (e.g., Smidt Reference Smidt1999; Veenstra Reference Veenstra2002; Wuthnow Reference Wuthnow2002; Welch et al. Reference Welch, Sikkink, Sartain and Bond2004). However, the mutual trust felt among religious group members is not necessarily extendable to the rest of the local community (Larsen et al. Reference Larsen, Harlan, Bolin, Hackett, Hope, Kirby, Nelson, Rex and Wolf2004). As a result, not every religious group is willing and able to serve as the communication channel between cadres and villagers, and the roles played by religious groups in contentious politics may thus vary.
To better understand this difference, we first look at Stream East (SE) and Stream West (SW), two adjacent villages in Southeast China. The two villages are separated by a narrow stream and were in fact artificially divided from the same settlement in the 1950s for purely administrative purposes. Thanks to their close location and shared history, SE and SW are very similar in cultural, historical, demographical, economic, and general policy terms. There is a temple in each village, which is the focal point of each villages' only religious group. Both temples worship the same deity and each is managed by a board consisting of a dozen villagers. Cadres in both villages regularly participate in the collective ceremonies and public projects organized by the local temple. Yet the two temples behaved very differently after similar policy decisions made by the two village authorities in the late 1990s, when cadres in both SE and SW launched ambitious projects to covert public lands into industrial zones.
Both aiming at “making the villagers rich,” the industrial zone projects in SE and SW started at almost the same time. However, the project in SE soon reached a premature end due to frequent collective contention, whereas the project in SW was highly successful. Today, while SE is often described by locals as “poor and chaotic,” SW has become one of the richest villages in the county thanks to its industrial zone. However, things were not so promising for SE at the beginning: as well as their colleagues in SE, cadres in SW were also accused by some villagers of being arbitrary and fractious. Luckily, thanks to the mediation of the temple, villagers in SW eventually came to believe that they could benefit from the industrial zone. They accepted the decision of their cadres without launching any collective protest, and many even voluntarily moved the graves of their ancestors to clear the fields for the industrial zone. While the temple in SE also has a close relationship with local cadres, it failed to deal with the emerging divergence between the cadres and some villagers as to whether or not the public land should be converted into an industrial zone. Without credible negotiations between the two parties, rumors soon spread that the project was simply taking place in order for cadres to embezzle public money. Even though the official investigation from the township government found no evidence of corruption or maleficence, villagers in SE nonetheless launched several collective protests against the industrial zone project and eventually forced their village authority to withdraw the plan. However, few villagers in SE today seem not to envy what their neighbors have achieved.
The two temples are very similar in religious belief and organizational structure. Why, then, was only the one in SW able to successfully mediate in the conflict between cadres and villages? The answer to this puzzle, I suggest, lies in the relationship between religious groups and secular social organizations: the head of SW's Elderly Citizens' Association — a vibrant secular social organization — also sits on the temple board, whereas the temple in SE is completely isolated from any secular social organization. The overlap with the secular Elderly Citizens' Association enables the temple in SW to approach a wider public beyond its own boundaries, and it also reduces the temple's political sensitivity by allowing it to put forward its agendas through the overlapping secular social organization. On the other hand, without the cover of secular social organizations, temple leaders in SE were afraid that becoming involved in a source of contention between cadres and villagers might entail trouble for them, given that the government is highly suspicious of the influence of religious groups in the secular domain. Moreover, without the assistance of secular social organizations, the temple in SE was hardly able to reach out to those villagers who do not share the religious faith of the group.
The difference between the temples in SE and SW reveals the importance of the relationship between religious groups and secular social organizations. Overlapping membership domains with secular social organizations such as elderly citizens associations, folk culture societies and local dancing groups, as demonstrated by the case of SW, allows religious groups in rural China to extend their influence to a wider public despite the hostile political environment. Religious groups that overlap with secular social organizations are, therefore, more likely to have vigorous interactions with other groups and individuals in the local community. As a result, they have a better chance of meaningful involvement in secular affairs, including mediating in conflicts between local cadres and discontented villagers. Moreover, religious groups that overlap with secular social organizations have a better chance of acting as “nodes” in the local nexus of power and governance. Existing research has confirmed that social stability and good governance are often associated with strong cooperation and efficient coordination between different social organizations, which could integrate divergent opinions, foster consensuses, gestate general social trust, and discourage “Jacobinical” populist mobilization by increasing mutual understanding among different social actors (Stone Reference Stone2003; Tocqueville Reference Tocqueville2000; Kornhauser Reference Kornhauser1959; Fukuyama Reference Fukuyama1995). In other words, when religious groups overlap with secular social organizations, they bridge hierarchical and horizontal social relations, thus fitting their members into the “structural holes” of local interpersonal networks (Burt Reference Burt1992). In so doing, such religious groups can simultaneously exchange information and resources with multiple networks, and they are thus more capable of facilitating credible negotiations between different actors in the local community.
It should be underlined that a religious group needs to overlap with secular social organizations rather than other religious groups in order to increase its chances of proactive engagement with local public affairs. The reason for this lies in the country's strict religious restrictions. Religious groups remain politically sensitive in China today. A considerable proportion, if not the vast majority, of Chinese rural religious groups are not registered with the government, either because they are unwilling or not allowed to do so or simply because their beliefs do not fit into the five official religions. In order to survive and grow in such a hostile environment, religious groups without connections with secular social organizations have to deliberately distance themselves from sensitive issues like contentious politics so as to avoid unnecessary attention and possible suppression by the state. Overlapping with other religious groups could not reduce, and may even increase, the suspicions of the government. Therefore, only religious groups overlapping with secular social organizations can push their agendas through secular partners and engage with public affairs without significantly arousing the suspicions of the government.
Linking Religious Groups to Local Authorities
Overlapping with secular social organizations allows some religious groups to better engage with contentious politics; however, such a feature alone cannot guarantee that these groups will always successfully mediate rather than intensify contention. Thanks to their close relationship with secular social organizations, religious groups that overlap with secular social organizations are more capable of launching collective protests as well as mediating in conflicts. Clearly, there is another factor that serves to differentiate those religious groups that are effective in preventing collective contention from others.
To identify this factor, we look at the temples in another pair of villages not far from SE and SW. Grand Rock (GR) and Little Rock (LR) are two villages separated by a narrow street. As was the case with SE and SW, GR and LR had been one settlement until their recent separation for administrative purposes. Their close geographical location and common history make GR and LR highly similar in their cultural, historical, and demographic backgrounds, and the economic and general policy situations of both villages remain highly similar today. Moreover, the village temples in GR and LR worship the same deity, have similar organizational structures, and overlap with the local Elderly Citizens' Association. However, the temple in LR includes local cadres as members and hence overlaps with the village authority — this difference led to the contrasting roles played by the two temples in the same contentious episode.
In 2010, GR and LR almost simultaneously received an order to construct fire barriers in their public woods from the County Fire Department, and both villages had to rise funding for this from among their villagers. Many residents in both villagers became quite unhappy. They regarded the order as unreasonable and unfair, and some even planned to protest against the Fire Department, hoping to force it either to withdraw its order or to provide sufficient funding for the construction. The temple in GR became a keen organizer of protests. Under the cover of the overlapping Elderly Citizens' Association, it not only called for its members to resist the order but also tried to blockade the woods when the village authority sent a team to clear the site for the fire barriers. The order from the Fire Department was interpreted as a conspiracy in which village cadres would spend public money in exchange for personal benefits or opportunities. The protests caused significant trouble for the village cadres, and GR was heavily fined for not finishing the fire barriers on time. The order from the Fire Department, first issued in 2010, had not been rescinded when I visited GR in 2012. The village authority had to borrow money to pay for both the fire barriers and the fines. Rumors spread that village cadres embezzled much of the construction funds, and collective protests became a constant issue in GR.
The situation in LR, however, was quite different. A trustee of the village's Elderly Citizens' Association came up with an innovative idea to solve the problem. Instead of resisting the order to build fire barriers, he proposed the building of a walking trail that could serve as both a fire barrier and a footpath. Such a project obviously required much more funding, and the temple in LR played a vital role in persuading villagers to support it. Thanks to the members who also hold positions in the village authority and the Elderly Citizens' Association, the temple board firmly stated that its followers and other villagers would benefit from the project, and it thus mobilized believers to donate to the project and framed the donation as a virtuous action. In addition, through the Elderly Citizens' Association, it also persuaded other villagers, including some Christians, to support the project. Moreover, the temple, in tandem with the Elderly Citizens' Association, supervised the process and the accounts of the project, making sure that no donation was misused. At the time of my visit in 2012, although the walking trail in LR was yet to be completed, it was already attracting more than 200 visitors each weekend. Villagers in LR are proud of the project. Many seemed to have completely forgotten how unpopular the order to build fire barriers had been in the first place.
Thus, the temples in GR and LR, although highly similar in many ways, reacted to the same exogenous shock in completely different ways. Such a phenomenon highlights the importance of the relationship between religious groups and local authorities. Religious groups in China overlap with local authorities when they include incumbent or retired cadres as members. As in the case of LR, despite the strict religious restrictions in the country, it is possible for incumbent or retired Chinese township and village cadres to become religious group members in some parts of the country. These cadres play important roles in creating mutual understanding between the religious groups to which they belong and the local governments for which they work. Thanks to the dual memberships held by these incumbent or retired local cadres, religious groups with a membership that overlaps with the local authorities' membership are more likely to serve as creditable negotiation channels between the state and society.
In villages with religious groups that overlap with the local authority then, trivial conflicts between the state and society are far less likely to grow into collective contention for several reasons. First, thanks to incumbent and retired cadres who simultaneously occupy positions in local authorities and religious groups, local authorities and ordinary villagers have a better chance of communicating with each other, and both parties are more likely to regard the information and signals from each other as credible. As a result, both local authorities and ordinary villagers can better understand the actual demands and true position of the other side, and therefore enjoy better chances of reaching compromises and agreements before getting trapped in collective contention, which is usually a lose/lose outcome in the Chinese context.Footnote 8 Moreover, the existing literature also suggests that collective contention in contemporary China is often triggered by rumors, many of which have later been proven to be heavily exaggerated or simply untrue (Liu et al. Reference Liu, Liu, Tao and Tao2010). Such rumors, however, are less likely to spread in villages such as LR, where residents can efficiently communicate with cadres thanks to the credible channels provided by religious groups overlapping with local authorities.
A New Framework
The foregoing comparisons suggest important lessons for better understanding the roles played by religious groups in contentious politics in authoritarian countries such as China. The different actions taken by similar religious groups during similar episodes of contention in two pairs of adjacent villages remind us of two issues that are often side-lined, if not completely neglected, in the existing theories. First, it seems more reasonable to consider the relationship between religious groups and collective contention as diverse rather than as uniform. Second, instead of assuming that such a relationship is simply determined by the stance of religious groups, it seems more reasonable to also look at other important actors in local communities.
Prior research has shown that political trust in authoritarian countries often exists in terms of multiple layers and in various dimensions, and citizens in authoritarian countries often trust the central government more than local cadres (Li Reference Li2008; Reference Li2013). While the mismatch of political trust in rural China provides some “breathing space” for the central government because “dissatisfaction with lower levels does not immediately generate demands for fundamental political reforms” (Li Reference Li2004), it also highlights that the key to preventing trivial conflicts from growing into collective contention is the mutual understanding between local cadres and ordinary citizens. Religious groups, as many have observed, have strong potential to bridge different local social forces and to link them with the state (e.g., Hillman Reference Hillman2004; Chau Reference Chau2005). However, like the temples in SE and GR, not every religious group is able to be accepted by both villagers and cadres as credible communication channels.
To make full sense of religious groups' roles in contentious politics, I suggest that we need to take their relational status into consideration. First, when religious groups overlap with secular social organizations, they can better reduce political sensitivities and reach a wider public by presenting agendas through their secular partners. As a result, religious groups that overlap with secular social organizations are more likely to get involved in public affairs and have a positive impact in terms of reducing the likelihood of collective contention. Moreover, among religious groups that overlap with secular social organizations, those that also overlap with local authorities (i.e., that include incumbent or retired cadres as group members) are more likely to be able to reduce the occurrence of collective contention, because these groups are more likely to be trusted by both ordinary villagers and local officials and hence have a better chance of serving as negotiating channels between the state and society.
Figure 1 illustrates the new explanatory framework. Drawing on the two criteria highlighted in the figure, we can categories religious groups into four groups: Type 1 groups overlap with secular social organizations but are isolated from local authorities; Type 2 groups are isolated from both secular social organizations and local authorities; Type 3 groups overlap with local authorities but are isolated from secular social organizations; and Type 4 groups simultaneously overlap with both secular social organizations and local authorities. The descriptive statistics of the four types of religious groups are reported in Table 3.
Figure 1. A new explanatory framework on the relationship between religious groups and collective contention in rural China.
Table 3. Descriptive Statistics on the Four Types of Religious Groups
As illustrated in Figure 1 and Table 3, among the four types of religious groups, only Type 4 groups are capable of facilitating credible communications between different actors in the local community and hence reducing collective contention. Therefore, collective contention should be less likely to occur in villages with more Type 4 groups, whereas other religious groups should have no significant impact on the occurrence of collective contention. This argument gives rise to the following testable hypothesis:
Hypothesis 3: Other things being equal, the likelihood of collective protest significantly correlates with the frequency of Type 4 religious groups but does NOT significantly correlate with the frequencies of Types 1, 2 and 3 religious groups, respectively.
$$IfCP_i = {\alpha} + NT1_i {\psi} _1 + NT2_i {\psi} _2 + NT3_i {\psi} _3 + NT4_i {\psi} _4 + CV_i {\gamma} + TD_i {\theta} + {\mu} _i $$
Hypothesis 3 is tested through Model B, which includes four independent variables: NT1, NT2, NT3, and NT4 refer to the frequencies of Type 1, Type 2, Type 3, and Type 4 religious groups in a sample village, respectively. ψ1, ψ2, ψ3, and ψ4 are their coefficients, respectively. The meaning of the other symbols is the same as those explained earlier in relation to Model A.
The statistical results of Model B are reported in Table 4, which shows that, among the four independent variables, only the coefficient of NT4 is significantly negative. In other words, while collective contention is less likely to occur in villages with more Type 4 religious groups, the frequency of other types of religious groups in a sample village does not have a significant impact on the occurrence of collective contention. That is to say, the results in Table 3 fully confirm the new explanatory framework illustrated by Figure 1.
Table 4. Statistical Results for Model B
N = 119. Robust standard errors are in parentheses. Significance at the 10% level is represented by *, at the 5% level by **, and at the 1% level by ***.
It is worth noting that some recent studies have also implied that the roles of religious groups in local politics may vary or change over time. Trejo (Reference Trejo2009), for example, finds that Catholic clergy in Latin America were traditionally in favor of the status quo but became major institutional promoters of rural indigenous causes when “confronted by the expansion of U.S. mainline Protestantism.” In addition, Tsai (Reference Tsai2007) finds that temples are more likely to push local cadres to provide public goods in Chinese villages than churches are. Therefore, the occurrence of collective contention might possibly vary between villages with different levels of religious competition or between those with groups practicing different religious faiths. However, neither alternative hypothesis holds when tested against the same data set used to validate the new explanatory framework. In other words, the key factor that differentiates the religious groups that can influence collective contention from those that cannot is indeed the relational status of these groups rather than the competition they face or the faith they have.Footnote 9
CONCLUSION
This article systematically examines the roles played by Chinese rural religious groups in contentious politics. According to the empirical findings, collective contention is less likely to occur in villages with a higher number of religious groups that simultaneously overlap with secular social organizations and local authorities, while other types of religious groups have no significant impact on the occurrence of collective contention. That is to say, at least in rural China, the relationship between religious groups and collective contention is multifaceted, and this relationship is determined not only by religious groups but also by their relations with other important actors in local communities, particularly secular social organizations and local authorities.
These findings invite us to reflect on the ongoing debate on the roles of religious groups in contentious politics. Given that religious groups may act as both social agitators and tranquilizers, each of the competing theories seems to have captured certain aspects of the relationship between religious groups and collective contention. However, to make better sense of such a relationship, we need to understand the factors that differentiate the roles of different religious groups in contentious politics, especially the political settings of, and the important actors in, local communities. For students of Chinese politics, this paper confirms that the variation in local cadres' social relations and everyday actions may lead to different types of state-society relations in Chinese villages, despite the apparently powerful party-state machine (e.g., Tsai Reference Tsai2007; O'Brien and Li Reference O'Brien and Li1999). It also suggests that social groups with different relational statuses may take up different attitudes and actions vis-à-vis the state.
In practical terms, this paper suggests that the Chinese government would benefit from more flexible and sophisticated policies toward religious affairs. The current coercive policy is failing to stop the rapid development of religious groups. Moreover, it will probably undermine the local state/religion relationship in the long run and may eventually become a cause of contention. Indeed, homogeneous suppression of religious activities may unintentionally force religious groups with different faiths and practices to unite into a powerful oppositional social force. Therefore, to better maintain social stability, the party-state may want to adopt different policies to manage its relationship with different religious groups. Finally, lifting the ban on religious believers joining the ruling party would offer the authoritarian regime the chance of more informal but credible communications with citizens, and hence make it better able to prevent trivial conflicts from developing into collective contention.
Supplementary materials and methods
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1755048314000789.
