Organized lobbying, including non-profit advocacy, is widely considered to be an essential factor of modern democracy. Following the Tocquevillian view of civil society, studies indicate that the participation of non-governmental organizations (NGOs) enhances democracy by building social capital, channelling grassroots demands, counterbalancing the state and providing citizens with daily opportunities for political participation.Footnote 1 In contrast, another group of scholars argues that NGO development can be compatible with authoritarian resilience. Comparative analyses have shown how autocrats exploit associational vibrancy to consolidate their power;Footnote 2 others have argued that the authoritarian legacy curtails the strength of civil society.Footnote 3 In Chinese politics, researchers focus on the growth of NGOs in the post-Mao era. Studies have shown that some social organizations have broken the previous model of the corporatist “transitional belt,”Footnote 4 can achieve autonomy in many aspects,Footnote 5 and advocate policies based on their expertise.Footnote 6 Although many scholars acknowledge the coexistence of autonomous NGOs and the authoritarian order, few have systematically examined the impact of this coexistence on the making of public policies, a situation where views of NGOs and the government confront each other.
To what degree can more autonomous NGOs change the policymaking process? Conversely, how do autocrats provide space for NGO autonomy and policy advocacy while eliminating the threat of outspoken criticism? An exploration of these two questions will contribute to the literature on authoritarian resilience from the perspective of civil society.
Authoritarian rulers encounter two paradoxical challenges in the governance of NGOs: to maintain social control and solicit cooperation.Footnote 7 On the one hand, policymakers contact NGOs to collect information, identify policy flaws and mitigate social conflicts.Footnote 8 On the other hand, NGOs’ participation in the public sphere may lead to more demands for political rights and institutional changes, threatening the foundation of the authoritarian order.Footnote 9
To manage this dilemma, authoritarian rulers adopt both repression and co-optation strategies. The Chinese Communist Party (CCP) constrains NGOs by imposing a strict registration system that only gives legal status to those deemed to be unthreatening. The registration system requires an NGO to find a “leading” Party or government branch which can oversee and be responsible for the NGO's daily activities.Footnote 10 However, at the same time, policymakers empower NGOs to provide public services, facilitate policy implementation and even identify potential discontent.Footnote 11
In this paper, we examine the CCP's combination of repression and co-optation through the lens of NGOs’ advocacy strategies. We stress that what matters to authoritarian stability is how NGOs design their advocacy strategies. We divide NGO advocacy into two categories. Political advocacy refers to an NGO's direct and exclusive negotiations with government officials, and social advocacy refers to an NGO's mobilization of social actors. Moreover, we employ an institutional framework to interpret the opportunities and constraints for NGO advocacy and to study how NGO autonomy influences their choice of advocacy strategies.
Our analyses of a survey of registered NGOs in Beijing, Zhejiang and Heilongjiang provinces yield three main findings. First, using the Confirmatory Factor Analysis (CFA), we show that NGOs’ advocacy strategies systematically differ from one another in terms of frequency: political advocacy is more widely adopted; moderate social advocacy is somewhat tolerated; and contentious social advocacy is rarely seen. Second, we find that NGO autonomy influences advocacy strategies in different ways. More autonomous NGOs tend to conduct more political advocacy by exclusively negotiating with the government. In contrast, gaining more autonomy makes NGOs hesitant to adopt moderate social advocacy approaches such as reaching out to the media. Finally, the variation in NGO autonomy does not impact the frequency of contentious activities. The results suggest that whereas the Chinese party-state allows NGOs with more autonomy to participate in the policymaking process, the participation is usually in the form of closed conversations with officials.
This study makes two theoretical contributions. First, the pre-existing works on NGO policy advocacy in China are mainly built on insightful case studies and qualitative evidence.Footnote 12 This study enriches the literature by conducting a quantitative analysis of a representative sample drawn in 2010, a period when NGOs enjoyed their most favourable political environment since 1990.Footnote 13 Second, the study reveals that the Chinese party-state uses social control (repression/self-censorship) and co-optation simultaneously to govern self-organized associations. The logic of repression is embedded in the party-state's cooperation with NGOs. The institutional setting pushes NGOs to advocate using non-threatening tactics, preventing overt challenges to the current regime.
NGO Advocacy under Authoritarianism
Literature on non-profit advocacy focuses on NGOs’ capacity to represent their constituencies, mobilize resources and interact with other players in the policymaking process. However, this approach may underappreciate the prominent role of political institutions in framing NGOs’ advocacy strategies. Following the state-centric approach,Footnote 14 we argue that political institutions shape NGO advocacy in that they define what activities and behaviour constitute policy advocacy and explain how the autonomy of an NGO impacts its advocacy strategies.
The Policy Advocacy of NGOs in China
Over the past four decades, China's rapid economic growth has brought dramatic social changes, of which a notable aspect is the “associational revolution.”Footnote 15 Since the beginning of the post-Mao era, there has been a huge surge in the number of Chinese NGOs. The market transition has broken the state's domination of economic resources, creating room for the pluralization of social interests. In addition, while maintaining the authoritarian structure, the Chinese party-state has made significant institutional adaptions in response to the rise in social demands.Footnote 16
The “associational revolution” creates opportunities for NGOs to become involved in the policymaking process, but it has not yet led to an “advocacy explosion.”Footnote 17 Since the early 1990s, the (re-)organization of society has started to influence, if not constrain or challenge, government policies. In the beginning, most NGOs tried to stay politically neutral and intentionally avoided conflicts.Footnote 18 However, after years of development, they became more frequently and intensively engaged in the decision-making process through advocacy and appeals in the new century.Footnote 19 For instance, some environmental NGOs (ENGOs) organized petitions to challenge local policies in order to protect natural resources and the interests of residential communities.Footnote 20 Industry associations, allying with large enterprises, played an active role in setting the agenda, collecting potential policy options and monitoring the implementation of final decisions.Footnote 21
The above-mentioned examples provide a glimpse into what political opportunities are available to NGOs and their corresponding behaviour. However, the existing works on Chinese civil society have some limitations. First, many studies focus on specific types of NGOs that enjoy the privilege of autonomy and expansion, such as ENGOs or trade associations.Footnote 22 Second, although scholars have pointed out several novel advocacy platforms for the non-profit sector, few have systematically examined the composition of NGOs’ advocacy strategies under authoritarian rule. Third, while attributing policy advocacy to NGOs’ political resources and the state's selection, researchers might have left an important aspect underexamined: NGOs’ independence from the state. To develop from the existing literature, this paper addresses the diversity of the advocacy strategies used by Chinese NGOs and the relationship between NGO autonomy and advocacy, based on large-sample representative data. An institutional framework is adopted to explain the variations.
Typology of NGO Advocacy
A broad definition of advocacy includes any direct or indirect attempt to influence public policy.Footnote 23 However, the scope of “any” attempt varies according to the institutional context. In Western democracies, advocacy is loosely defined as “the act of pleading for or against a cause, as well as supporting or recommending a position” and “the act of addressing legislators with a view to influencing their votes.”Footnote 24 NGOs’ advocacy behaviour includes, but is not limited to, activities such as organizing, convening, analysing, educating, writing, petitioning and lobbying to change public policies “with the goal of securing or retaining social justice.”Footnote 25 Much of the existing literature looks into the various advocacy venues used by social actors, providing a non-exhaustive list “of the political activities in which non-profits might engage”:
Nonprofit organizations might use letter-writing and telephone campaigns directed at legislators to advocate on behalf of an interest group; they might testify at legislative hearings or mobilize individuals to demonstrate at a state capitol to influence policy on behalf of a client population that they serve; they might promote a particular mission by supporting or resisting particular policies or legislative agendas; they might engage in non-partisan public education around political issues; they might endorse political candidates; or they might sponsor demonstrations and marches.Footnote 26
The first step is to categorize the above-mentioned tactics into subsets. Dongshu Liu constructs a two-by-two table that groups policy advocacy channels in the dimensions of formality and consistency, arguing that the four channels are selected based on ENGOs’ political resources.Footnote 27 This typology captures the two key features of how NGOs interact with the state, but somehow leaves out NGOs’ interaction with society. Departing from Liu's definition of policy advocacy, we include the strategies NGOs use to approach society and further consider the features of societal channels.
In this paper, we distinguish political advocacy from social advocacy. The former refers to NGOs’ efforts to target policymakers directly, while the latter describes indirect attempts to influence and mobilize social actors, including individuals, corporations, media and other non-profit organizations. Accordingly, political advocacy is realized in the form of legislative advocacy, administrative advocacy and electoral advocacy. Media advocacy, research and public education, coalition building, grassroots advocacy and other collective actions constitute social advocacy.Footnote 28
Political advocacy, sometimes called “rights-based advocacy,” includes legislative advocacy to define rules and procedures, programme advocacy to monitor government programmes, and direct negotiation with officials. More specifically, political advocacy tactics also differ in degrees of formality. NGOs may deliver their opinions in well-organized and systematic forms, such as attending policy meetings and writing official reports. On the other hand, some negotiations with the government are less regular and formal. For example, NGOs can take advantage of their personal connections and talk to officials via telephone or email. We categorize attending policy meeting and writing reports efforts as formal political advocacy and less formal negotiations through personal connections as informal political advocacy.
Social advocacy covers civic engagement such as grassroots lobbying (for example, the mobilization of constituents to pressure the legislators), attempts to influence public opinion and educational efforts to encourage mass participation.Footnote 29 Some social advocacy tactics are more contentious than others.Footnote 30 In the Chinese context, the level of contentiousness varies according to the scale and form of social mobilization. Therefore, mobilizing NGO members, using media and allying with other NGOs fall in the sub-category of moderate social advocacy. Unlike in the West, using judicial tools to impact policies is also seen as social advocacy, as it resorts to attracting public attention to challenge the existing rules or current situations.Footnote 31 The sub-category of contentious social advocacy includes actions such as organizing social protests or collective petitions which expose discontent on the streets and are less tolerated by the government.
To sum up, there are significant differences between political advocacy and social advocacy. However, the distinction between the two should not fully occupy our attention since there are also noticeable and meaningful variations within each category regarding the degree of formality and level of contentiousness. These within-type differences also reflect NGOs’ rational responses to the co-option and repression imposed by the authoritarian state.
The Institutional Opportunities and Constraints
While democracies have established formal channels for the participation of social groups, in a single-party authoritarian state with weak party competition and electoral representation, NGO advocacy can serve as a major mechanism for policy input and interest intermediation.Footnote 32 We find that three institutional characteristics influence NGOs’ choice of political or social advocacy: a relatively open and fragmented authoritarianism system; the deliberate ambiguity of laws and policies combined with extreme sanctions; and a dual-management system that constrains NGO autonomy.Footnote 33
China has been depicted as a fragmented authoritarian regime where the policymaking and implementation processes are dominated by bureaucratic bargaining without the participation of social groups.Footnote 34 However, this model has changed since the 1990s. From a top-down perspective, the Chinese government has gradually loosened its control of the agenda-setting and allows social actors to voice their opinions, as long as the state holds the last word on policy.Footnote 35 From a bottom-up perspective, decades of rapid economic growth and social changes have incubated more NGOs, media agencies and individual activists. Some of these groups can successfully enter the policymaking sphere by adopting strategies compatible with the structural and procedural constraints of the fragmented authority (FA) framework.Footnote 36 The government welcomes the engagement of NGOs if it fits with socioeconomic goals and is organized under the state's control.Footnote 37 The policy advocacy of NGOs can sometimes become the most important reference for officials when designing policies.Footnote 38 A recent study describes this operating environment as organizing under duress: while the party-state “has permitted the growth of civil society, it continues to repress organizational activism.” There are two dimensions at the local level, fragmented control and competitive control of NGOs, which lead to the mobilizing without the masses strategies of grassroots labour NGOs.Footnote 39
While the Chinese state grants social actors more space to advocate for policies and provide social services, it does not hesitate to repress these activities when necessary, especially in recent years.Footnote 40 Since the goal of authoritarian regimes is to demobilize and maintain social stabilityFootnote 41 activities that involve large-scale mobilization or contention are more likely to be repressed.Footnote 42 However, the boundaries between political and social affairs in China are ambiguous. There is uncertainty about what kind of social advocacy is regarded as “disorder” or “illegal.” “The combination of subtly shifting political winds and the absence of detailed rules for every situation (and the inconsistency of different government branch policies) means that even long-time journalists, editors, and lawyers can get into trouble for actions they thought were acceptable or that had gone unnoticed in the past.”Footnote 43 Thus, authoritarian regimes generate high levels of uncertainty, fear and risk.Footnote 44 As a result, NGOs, especially the registered ones, self-censor. They seek credible information from government and limit their activities to orderly participation. In such ways, they restrain themselves from contesting the government and do their best to avoid being regarded as threats to regime stability.Footnote 45 In other words, Chinese NGOs are likely to participate without contestation.
These two institutional features may influence NGOs’ policy advocacy strategies, but their impact may also be conditional on the degree of autonomy enjoyed by the NGOs, which is shaped by the dual-management system, another important aspect of the corporatist institutions.
Organizational Autonomy and NGO Advocacy
In the civil society literature, a defining feature of NGOs is their self-governing capacity, which pluralists believe will act as a check and balance to state power and foster democracy.Footnote 46 By definition, an autonomous NGO should have the self-governing capacity to organize its own activities and determine its internal governance procedures.Footnote 47 As the resource dependence theory suggests, organizations’ autonomy should be understood in terms of their interdependence from their external constraints.Footnote 48 Measurements of autonomy should incorporate a broad examination of all sorts of actors including government officials, private enterprises and other NGOs. A 2016 study puts forward some important but non-deterministic indicators of organizational autonomy:
The capacity to own assets, incur liabilities, or engage in transactions in its own right;
The control over the selection of all or most of the organizations governing officials;
A meaningful degree of financial autonomy, including the ability to refuse funding from an external source; and
The ability to determine the basic mission and purpose of the organization.Footnote 49
Some points about NGO autonomy need further clarification. First, organizational autonomy should be measured within a multidimensional framework rather than a unitary spectrum: it should be associated with an NGO's discretion in managing financial and human resources as well as its subordination to other types of organizations. Second, among the organizations NGOs work with, the government is one of the most important players. The interactions between the government and NGOs are constrained by political institutions that set the rules by which civil society functions. Third, and consequently, the supervisory unit system is established both to grant NGOs space and resources to fulfil their roles in a pluralistic way and to limit their autonomy and strengthen state control in a corporatist way.Footnote 50
Some scholars regard mass organizations as an important tool for enhancing authoritarian durability owing to their role in maintaining social control and mobilization support functions.Footnote 51 Likewise, we argue that some strong authoritarian regimes can strategically use NGOs to make policy concessions by selectively incorporating some social organizations and repressing others. More importantly, the uncertainty of punishment presses social actors to self-censor their political behaviour. We use the case of China to explore NGOs’ policy advocacy and examine the party-state's co-optation of NGOs. Compared to mass organizations, which have very limited autonomy, NGOs’ role in sustaining or weakening an authoritarian regime is an equally, if not more, important topic, since NGOs are numerous and have different (and higher) degrees of autonomy.
In general, Chinese registered NGOs enjoy more de facto autonomy than the formal institution grants. There are two reasons for this. First, the formal design of policy is not equivalent to the actual practice. The government and supervisory unit may not have enough human and financial resources to monitor NGOs. Second, the government's support and commitment of its resources are not equated to control.Footnote 52
Several mechanisms work simultaneously to link an NGO's autonomy with its choice of advocacy venues. Autonomy may have a positive impact on an NGO's political advocacy efforts. An autonomous NGO might conduct more political advocacy because it does not have as many alternative channels as a less autonomous NGO. On the one hand, the opportunity for political advocacy is more valuable to an autonomous NGO; a less autonomous NGO may have other ways to influence policies. Indeed, it is likely that an NGO with strong links to the government does not even need to resort to political advocacy. Therefore, unlike the existing literature, which argues that policy advocacy is favourable for NGOs with strong political ties and connections (in other words, less autonomous NGOs), we argue that policy advocacy occurs more often among NGOs without strong political connections. On the other hand, a small proportion of autonomous NGOs, with abundant financial and personnel resources and expertise, are more active in their advocacy, as the pluralism theories predict. These combined, autonomous NGOs tend to use political advocacy, or the orderly political participation that the CCP encourages.
We hypothesize that NGO autonomy impacts social advocacy in a negative way. Compared with political advocacy, social advocacy is more threatening to regime stability and legitimacy, so the regime might only feel it acceptable if social advocacy (1) comes from the less autonomous NGOs that have a closer relationship with the regime or (2) is contained within a relatively low level of mobilization that will not get out of control. Social advocacy by autonomous NGOs, by contrast, might be out of control and dangerous to the regime.Footnote 53 Switching the perspective from the state to NGOs, we can see another part of the theory. The political environment that NGOs are faced with is highly uncertain: the regime does not spell out the criteria for ordered social participation, and the punishments for crossing the red line can be severe. This is especially the case for the autonomous NGOs that lack credible insider information, guidance and support from government branches (or the supervisory units); they are more likely to exercise self-censorship.Footnote 54 This explains why many autonomous NGOs choose to conduct their activities within the safe zone, limit any aggressive claims and help to promote the social welfare goals of the state. The strong self-censorship of autonomous NGOs in social advocacy resonates with the theory of control parables: “uncertainty over the limits of political tolerance amplifies repression and pushes people to control themselves.”Footnote 55
So, who then conducts social advocacy? Similar to recent studies that find that insiders with better knowledge of government red lines are more outspoken,Footnote 56 we contend that NGOs with a close relationship with the government – and consequently less autonomy – are better at understanding to what extent the government allows criticism and tolerates contestation. Therefore, social advocacy, or social participation in general, is carried out by less autonomous NGOs who are close to the state because (1) they know where the red line is and how to avoid crossing it; and (2) the regime will only tolerate sensitive or risky social advocacy by organizations it can control.
Based on literature and fieldwork, we suspect that the relationship between NGO autonomy and contentious social advocacy, the riskiest among the four subtypes, might be contingent and less clear. Given the fragmentation of government power and the possibility for an NGO to ally with some government bureau(s) against others to make “rightful” claims, which can reduce the risk of repression, there might still be limited space for contentious social advocacy. Following the same logic of social advocacy discussed above, this would suggest that the more connected (less autonomous) NGOs are more likely to pick up contentious social advocacy. However, as some case studies reveal, whether this relationship works is highly dependent on the role of policy entrepreneurs and other contextual factors not open to ex-ante measures.Footnote 57 Furthermore, since in our survey only a very small proportion of NGOs reported using contentious means to challenge the government, statistical models testing this relationship might be less effective.
In summary, political and legal institutions not only restrict NGOs’ advocacy activities but also shape the relations between NGO autonomy and policy advocacy efforts. The CCP encourages the orderly political participation of NGOs and regards it as an important element of “deliberative democracy,” given the Party and government's need to gather feedback and information from society. The levels of NGO autonomy may influence the adoption of advocacy activities in different ways. More autonomous NGOs, lacking corporatist channels through which to lobby, may have to conduct more political advocacy than the less autonomous ones; however, without credible information, more autonomous NGOs may be more concerned about the uncertainty and political risk of social advocacy and therefore avoid it.
Based on the discussion in this section, we propose the following hypotheses:
Hypothesis 1: Chinese NGOs with more autonomy tend to make more efforts in both formal and informal political advocacy.
Hypothesis 2: Chinese NGOs with more autonomy tend to make fewer efforts in moderate social advocacy and contentious social advocacy. Footnote 58
Empirics
Data
The data come from a survey conducted by the Center for Civil Society Studies, Peking University, from 2010 to 2012. The survey targeted NGOs registered in Beijing, Zhejiang and Heilongjiang provinces using a stratified random sampling method. The sample included three prefectures in Zhejiang, two in Heilongjiang and four prefecture-level districts in Beijing. From each prefecture, two counties or county-level cities/districts were randomly selected. With the assistance of local civil affairs bureaus, the research team acquired a full list of the registered NGOs in each sampled unit and randomly selected ones to survey. Trained graduate and undergraduate students collected the data through face-to-face interviews with NGO leaders or self-administered questionnaires. The result was a valid sample of 1,195 NGOs, with an acceptable response rate of 56 per cent. Figure 1 displays the response rates by locality and administrative level.Footnote 59
Figure 1: Response Rates by Region (1) and by Administration Level (2)
The sampled areas reflect some regional variations that may impact local NGOs’ behaviour in the eastern part of China. As the political and economic centre of the country, Beijing attracts a diversity of NGOs and has the attention of the central state. Zhejiang is one of the most developed areas of China, with a highly active private sector which can empower its NGOs. By contrast, Heilongjiang represents the middle-income provinces where the legacy of the planned economy plays an essential role in shaping social interactions. In 2011, Beijing, Zhejiang and Heilongjiang ranked 2nd, 6th and 17th in terms of GDP per capita among the 31 provinces of mainland China. The NGOs in the three sampled provinces were all active although challenged by differing socioeconomic conditions.
Admittedly, the sample is not representative of the whole of China. The three sampled provinces are in the east of the country, where the economy is more influenced by marketization and opening-up policies and the population is less ethnically divided. The sample may not capture NGOs registered in the central and western regions of the country. Moreover, there are numerous unregistered NGOs, which remain underground in China. It is difficult, if not impossible, to acquire large-sample data of unregistered NGOs. Ethnographic work and case studies have provided valuable analyses of China's unregistered NGOs.Footnote 60
Measurement
Dependent variables. The dependent variables are indicators of an NGO's political advocacy or social advocacy behaviour.Footnote 61 Political advocacy refers to an NGO's endeavours to impact decision makers through inclusive and corporatist channels. In particular, there are two ways for an NGO to negotiate with the government: through formal institutions such as internal meetings or policy consultation reports, or through informal institutions such as personal connections with the government. Social advocacy refers to an NGO's mobilization of social actors and enunciation of its position via exclusive and pluralistic channels. Some social advocacy strategies are more contentious than others.
The descriptive results suggest that the sampled NGOs adopted a wide range of ways to advocate, yet the frequency of these strategies varied. While inclusive and corporatist negotiation with the government (political advocacy) was encouraged, pluralist mobilization from the bottom up (social advocacy) was constrained. Figure 2 presents the percentages of NGOs that had used each of the aforementioned ways to impact public policies in the previous year. The two most commonly adopted means were attending policy meetings organized by the government (72.10 per cent) and writing policy reports (58.33 per cent). By comparison, the sampled NGOs rarely organized contentious activities such as petitions (3.18 per cent) or protests (4.18 per cent). The frequency of moderate social advocacy tactics falls in the middle. Contacting the media (35.17 per cent) and allying with other NGOs (37.71 per cent) were more common tactics than contentious social advocacy activities, and less commonly adopted than political advocacy strategies.
Figure 2: Percentage of NGOs Adopting Each Advocacy Tactic
Independent variables. A broad definition of NGO autonomy is based upon an NGO's relationship with its political, economic and social actors. In this study, we narrow down the measurement of NGO autonomy to the extent that an NGO is free from government intervention on a daily basis.
Many studies of Chinese NGOs have proposed ways to measure an NGO's autonomy. For example, Scott Kennedy finds that the autonomy of Chinese business associations will “be compromised if they are: (1) initiated by the state; (2) required to register and be affiliated with a government agency; (3) staffed by government officials; and (4) dependent on government financing.”Footnote 62 However, other scholars have challenged the idea of using an NGO's organizational and resource dependence on the government as an approximation of its autonomy. Owing to the dynamic nature of state–society relations in China, whether the government was involved in the setting up of an NGO does not necessarily lead to an NGO's dependence on the government in later years. In a similar fashion, based on the observation of cross-national cases, Wang Shaoguang finds financial dependence may not be a good indicator of NGO autonomy in many circumstances.Footnote 63
To measure the de facto autonomy of an NGO, we need to examine how the government participates in the daily operation of an NGO. By institution design, the supervisory unit is the government branch that interacts with the affiliated NGO directly and takes responsibility for overseeing its activities. Therefore, our measurement of NGO autonomy stems from the three major ways through which a supervisory unit can collect information and influence an NGO's daily decisions. The supervisory unit can (1) staff the affiliated NGO with officials or send agents to monitor its activities; (2) appoint the NGO's leaders; and (3) attend the NGO's important work meetings. We specify NGO autonomy using these three indicators.Footnote 64 Figure 3 presents the descriptive statistics of NGO autonomy based on the survey.
Figure 3: Percentage of NGOs Having Gained Organizational Autonomy
Empirical strategies
Our goal is to examine the marginal effect of an NGO's autonomy on its advocacy, conditioning on other related factors. The dependent variables are the predicted scores of formal/informal political advocacy and moderate/contentious social advocacy based on the CFA Model.Footnote 65 The independent variables are whether an NGO has gained autonomy by organizing its own working meetings, appointing its own leaders and keeping the monitors away.
The control variables include (1) an NGO's size and fiscal capacity, measured by the number of members (logged) and the amount of annual expenditure (logged); (2) an NGO's working field, measured by whether the NGO is interested in political, judicial or diplomatic issues; (3) an NGO's domestic political resources, measured by whether central, local or people's congress officials have attended the events organized by the NGO; (4) NGO leaders’ political resources, measured by whether the NGO leaders have worked for the government or joined the CCP; (5) an NGO's international connections, measured by whether the NGO has received funds from abroad or invited foreign guests to its important events.
To address the unobserved confounding effects, we take the regional fixed effects into account. The sampled NGOs were registered with 25 different local civil affairs bureaus, which ranged from county-level to provincial-level agencies. By adding the fixed effects, we examine the impact of NGO autonomy on NGO advocacy within each locality. In theory, the regional fixed effects absorb the effect of the unobserved regional confounders such as different socioeconomic conditions, exceptional local policies or local officials’ political ideologies and preferences. We use robust standard errors assuming the clusters in each locality.
Results and Discussions
The configuration of advocacy tactics
First, we examine the systematic grouping of NGO advocacy tactics using the principal component analysis (PCA) and confirmatory factor analysis (CFA). Political scientists have used the two methods to reduce dimensions and explore the structure of multiple variables.Footnote 66 According to Figure 4, the eigenvalue of the third component drops significantly, indicating that an NGO's advocacy tactics are correlated with each other based on some structures.
Figure 4: The Goodness of Fit Test (PCA)
Next, we need a method to identify the latent factors behind the ten advocacy strategies and build linkages between empirical tests and theories. A CFA model testifies how well the observed variables fit into the hypothesized grouping. We propose five models and compare their goodness of fit using the CFA: (1) Model 1 – the one-dimensional model; (2) Model 2 – the two-dimensional model separating political advocacy from social advocacy; (3) Model 3.1 – the three-dimensional model including political advocacy, moderate social advocacy and contentious social advocacy; (4) Model 3.5 – the three-dimensional model including formal political advocacy, informal political advocacy and social advocacy; (5) Model 4.1 – the four-dimensional model, as discussed above. Table 1 presents the goodness of fit for the five models. In general, Model 4.1 outperforms the others in terms of absolute and relative fitness. The comparative fit index (CFI) of Model 4.1 reaches close to 0.90 and the RMSEA comes below 0.60, indicating an acceptable level of model fitness.Footnote 67
Table 1: The Goodness of Fit Test (CFA)
The impact of NGO autonomy on advocacy
Based on Model 4.1, we construct the factor scores of the four types of advocacy strategies. Next, we explore the differences in the advocacy factor scores between NGOs with and without autonomy in appointing leaders, holding meetings and keeping monitors away, controlling for an NGO's size, working fields, domestic political resources, international connections and the regional fixed effects.
Figure 5(1) reports positive relationships between the three indicators of NGO autonomy and the factor score of formal political advocacy under the 95 per cent confidence level. We find that when an NGO has no agents from the supervisory unit participating in its working meetings or monitoring its activities, the formal political advocacy score increases by 0.222 and 0.234. This means, all else being constant, an NGO with the autonomy to run its own work meetings with only its members attending will participate in policy meetings or write policy reports one or three times more a year than one without this form of autonomy. The empirical result also indicates that an NGO with local political resourcesFootnote 68 has more opportunities to participate in policy meetings or consultations.Footnote 69 It should be noted that an NGO's local political resources can move up the formal political advocacy score only by 0.114, which is half the impact of NGO autonomy.
Figure 5: The Impact of NGO Autonomy on Formal Political Advocacy (1) and on Informal Political Advocacy (2)
The effect of NGO advocacy on informal political advocacy is similar. The three indicators of NGO autonomy move up the factor score of informal political advocacy by 0.081, 0.433 and 0.261 (see Figure 5(2)). The growth by 0.433 can be interpreted as that an NGO with the autonomy to run its working meeting will conduct five more trials to impact policies a year compared to those without this autonomy. The effect of NGO autonomy is stronger than all other control variables included in the analysis.Footnote 70
As for social advocacy, NGO autonomy works in two different ways. According to Figure 6, NGOs without their supervisory units appointing their leaders are expected to lower moderate social advocacy scores by 0.087. Similarly, keeping the monitors away brings down the moderate social advocacy score by 0.111. This means that an NGO that is not monitored by a supervisory unit will reduce its annual frequency of media usage, allying with other NGOs, mobilizing its members or using judicial tools to impact policies. In contrast, the effect of having central/local political resources on moderate social advocacy is significantly positive, although its impact is smaller than the NGO autonomy indicators.Footnote 71
Figure 6: The Impact of NGO Autonomy on Moderate Social Advocacy
Figure 7 shows that NGO autonomy does not have a significant impact on contentious social advocacy. There are two possible reasons for this. First, only a tiny proportion of the surveyed NGOs actively organized social protests and petitions. Second, as discussed in the previous section, contentious social advocacy is more contingent on some non-institutional factors such as an NGO leaders’ political entrepreneurship.
Figure 7: The Impact of NGO Autonomy on Contentious Social Advocacy
Conclusion
While NGO advocacy is an vital ingredient of democracy, it can also play important co-optation and consultation roles under some authoritarian regimes,Footnote 72 helping authoritarian rulers remain in power.Footnote 73 The establishment of institutions for political participation, and their appeal among the public at large, is listed as one aspect of the CCP regime's institutionalization.Footnote 74 The CCP politburo also emphasizes “deliberative democracy” as an important part of “democracy with socialist characteristics,” declaring that the government should consult with the relevant mass organizations, social organizations and people's representatives for policymaking.Footnote 75
In this paper, we add to the existing research by studying how NGOs advocate under an authoritarian regime and how autonomy affects their advocacy strategies. The institutional structure greatly influences the extent to which NGOs can participate in the policy process by framing the decision-making rules.Footnote 76 Although China has undergone an “associational revolution” in the past decades, NGO advocacy is at an early stage of development with limited scope, and an “advocacy explosion” is not currently within our sight. This reflects contemporary state–society relations in China, conforming to what is called “contingent symbiosis”Footnote 77 and “consultative authoritarianism,”Footnote 78 and the increasing pluralistic policymaking, which in turn creates conditions for better governance under authoritarianism.
While limited NGO policy advocacy could be attributed to the institutional arrangements of a strong single-party regime, we go further to discuss how this regime shapes and constrains different types of NGO advocacy. We find that the majority of Chinese NGOs prefer political advocacy – mainly cooperative, peaceful and constructive tactics that meet the needs and willingness of the government. Moreover, more autonomous NGOs tend to use political advocacy and avoid social advocacy, since social advocacy is more likely to be regarded as contesting the government and threatening the regime. These findings resonate with “consultative authoritarianism”; the simultaneous expansion of a fairly autonomous civil society and the development of more sophisticated and indirect tools of state control over this civil society.Footnote 79 Our study contributes to this literature with a quantitative analysis of the advocacy strategies of NGOs from the institutional perspective.
The findings of this study have important theoretical implications for the literature on authoritarian resilience and change. First, while the civil society literature tends to regard NGOs as a potential threat to authoritarian rule, we find that NGOs could serve as a co-optation tool for authoritarian rulers. A strong authoritarian regime can selectively make policy concessions with some NGOs or extract policy information through consultation. This differs from studies of NGO policy advocacy in other (electoral) authoritarian regimes such as Russia where NPOs build “vertical ties” with political elites to secure funds and lobby in a “managed democracy,” leading to an intertwined NPO–state relationship resembling co-optation.Footnote 80 We find that authoritarian co-optation could be more diversified and institutionalized.
Second, this paper enriches our understanding of the role of grassroots NGOs, registered NGOs and mass organizations, each with different levels of autonomy, in sustaining authoritarian regimes.Footnote 81 Chinese NGOs participate in policymaking but do not contend or confront the government. Compared to political advocacy, NGOs’ social advocacy is relatively limited. Moreover, NGOs with various degrees of autonomy choose different advocacy venues and strategies. Unlike the existing literature, we find that autonomous NGOs tend to use political advocacy because these opportunities are more valuable when attempting to impact policy. Conversely, the relatively risky social advocacy is done by those who are close to the regime because they know where the red line is and avoid crossing it so that the regime will have effective control and feel comfortable.
The control parables work well at generating fear and lead to self-censorship among the NGOs. Compared to repression and even differentiated control, control parables have the advantage of reducing confrontation and ensuring participation. NGOs participate but do not contend, and a collaborative relationship, not a confrontational one, characterizes the model of state–society relations in China. All of these factors help the authoritarian regime to stay in power as it promotes organizations that can assist the state while keeping threatening organizations at bay.Footnote 82 Therefore, we find the two paradoxical functions, social control (repression/self-censorship) and cooperation (co-optation), are made to complement each other under authoritarian rule. Meanwhile, the NGOs currently have limited capacity to mobilize members, social actors and other social organizations. The horizontal social self-organization is still underdeveloped as mobilization with the masses is strictly prohibited,Footnote 83 and the civil society, if there is one, faces dreadful challenges.
Further studies are needed to understand the bifurcating effects of NGO autonomy on advocacy efforts. First, besides the institutional approach we take here, future research may apply mediation analysis to probe the conditions under which organizational autonomy increases or decreases NGO advocacy. Second, advocacy strategies such as horizontal alliances, especially with whom NGOs choose to join hands, are worth more attention. Third, the situation has changed since Xi Jinping took power. It is now characterized by crackdowns on many types of activism,Footnote 84 stricter monitoring over “sensitive” NGOs, and strengthened Party control as Party branches are established within NGOs.Footnote 85 These changes can diminish an NGO's autonomy or make the political environment more repressive. NGO advocacy strategies, as well as the relationship between an NGO's autonomy and advocacy efforts, may change substantially. Therefore, it would be enlightening to conduct a comparative study between 2010 and 2020 in the near future.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0305741022000510
Acknowledgements
The authors thank Professor Yuan Ruijun and Guo Qiang for their valuable feedback. This research project was funded by Institute of Public Governance, Peking University (北京大学公共治理研究所 TDXM202107), the Ministry of Education of China Youth Fund Project of Humanities and Social Sciences (教育部人文社科研究青年基金项目 No. 18YJC810010), and The National Social Science Fund of China (中国国家社科基金 No. 19CZZ009).
Conflicts of interest
None.
Biographical notes
Zheng SU is an associate professor in the School of International and Public Affairs at Shanghai Jiao Tong University. His research interests include state–society relations, public opinion and political economy of development in China.
Shiqi MA is a PhD candidate in the government department at Cornell University. Her research focuses on Chinese urban politics seen through the lens of state control and social stability.
Changdong ZHANG is a research associate at the Institute of State Governance Studies and an associate professor of political science at the School of Government at Peking University. His research interests include taxation politics/fiscal sociology, state and society relationships, and institutionalism. Professor Zhang is the author of Governing and Ruling: The Political Logic of Taxation in China (University of Michigan Press, 2021).
