The Institutional Foundation of Countermobilization: Elites and Pro-Regime Grassroots Organizations in Post-Handover Hong Kong

Abstract Countermobilization has been a common strategy for autocrats to counteract the threat of opposition. Although the use of countermobilization has drawn scholarly attention, research on the mechanisms that enable countermobilization remains limited. This article underscores the role of political institutions in allowing autocrats to carry out countermobilization through incentivizing elites to serve as a bridge between the state and the masses. Focusing on the case of Hong Kong, where pro-government countermobilization is rising along with pro-democracy challenges against the hybrid regime, the article argues that countermobilization is enabled because societal elites are incentivized through political institutions to organize the masses and develop mobilization capacity through grassroots organizations. Using original elite biographical data and organizational data, the article shows that elites with more ties with grassroots organizations are more likely to remain in office in the Chinese People's Political Consultative Conference. The findings offer an institutionalist explanation of how authoritarian rulers enact countermobilization by leveraging elite intermediaries and their grassroots networks. In this light, political institutions can serve as a conduit for the state to extend social control.

nationalist groups to orchestrate large-scale pro-government rallies in the face of rising anti-Putin protests. In Venezuela, both the Chavez and the Maduro regimes have called on supporters, and at times armed militias, to countermobilize against anti-government protests. During the Arab Spring uprisings, it was not uncommon to see regime incumbents resorting to the use of countermobilizationthe most notorious of which was perhaps Egypt's Battle of Camels, when the Mubarak regime sent thugs on camels to quell pro-democracy protests in Cairo. While the context, form, scale of mobilization and regime contexts may differ, these examples have attested to the fact that countermobilization is a strategic option in the contemporary authoritarian playbook (Frantz and Kendall-Taylor 2014;Svolik 2012). They suggest that autocrats can, in fact, play an active role in contentious politics, a domain where they are typically treated as the target.
While the use of countermobilization among autocrats has attracted growing scholarly attention (Ekiert et al. 2020;Hellmeier 2020;Hellmeier and Weidmann 2019;Robertson 2011), relatively little is known about the mechanisms that enable countermobilization. Two theoretical puzzles stand out. First, although it is often assumed that the state plays a central role in orchestrating countermobilization, there has been limited research on how the ruling incumbent builds up the mobilizing structure that connects them to the masses and enables swift countermobilization. Second, despite its crucial role, the state may not want to be always seen as the mastermind 'pulling the strings'. To achieve a credible display of power, countermobilization should have some level of perceived authenticity of the people's support towards the state. Pro-regime mobilizations should consist of 'real', 'ordinary' people acting voluntarily and showing 'worthiness, unity, numbers and commitment ' (Tilly and Wood 2013), or at least the semblance of these elements. While it is clear that authoritarian rulers often mobilize through a variety of social organizations to provide a veneer of societal support, how they induce more organic and bottom-up mobilization has remained very much unexplained.
How can the state build up a mobilizing structure that can swiftly mobilize the masses while creating the impression that these supporters are not simply puppets of the state? This article argues that one mechanism to facilitate countermobilization is to institutionally incentivize pro-regime societal elites who can organize and mobilize the masses. By rewarding such elites with offices in formal political institutions, the state can leverage its networks to build a robust organizational capacity in preparation for regime-threatening moments. One testable hypothesis is that societal elites with higher organizational and mobilizational capacities, measured by their leadership ties to grassroots organizations, are more likely to be rewarded with offices in political institutions.
We test this institutional mechanism of countermobilization in Hong Kong, a semi-autonomous region of the People's Republic of China. As a hybrid regime underpinned by a liberal authoritarian order, at least until recently, Hong Kong is a revealing case because of the dynamic influxes of a developed civil society with consistent calls for democracy and an increasingly authoritarian state with a growing number of pro-regime grassroots organizations that actively defend the political status quo (Cheng 2020;Lee 2020). 1 Episodes of pro-democracy mobilization and proregime countermobilization have become a regular pattern in the city's politics. This offers researchers a unique opportunity to examine the mechanisms that enable countermobilization. This article first documents how grassroots organizations play an increasingly salient role as the vehicles of countermobilization by projecting proregime voices and counteracting pro-democracy challenges. We then conduct regression analyses using original elite biographical and grassroots organizational data. We demonstrate that societal elites with more leadership ties with grassroots organizations are more likely to remain in office in the Chinese People's Political Consultative Conference (CPPCC), an important elite institution at the national level.
The research seeks to make two contributions to the study of authoritarian politics. First, instead of studying the pervasiveness or outcomes of countermobilization, we use an in-depth case study and diverse sources of qualitative and quantitative data to demonstrate the mechanism through which countermobilization could be regularly performed. In doing so, we uncover the complex political process that reveals the importance of a robust network of grassroots organizations and the critical intermediary role of societal elites who organize and mobilize the masses. The findings join recent scholarly works (Ekiert et al. 2020;Hellmeier and Weidmann 2019) that focus on the proactive role of the state in mobilizing movements but at the same time emphasize the interaction between state and non-state actors in jointly enabling countermobilization. Second, we highlight the dynamic interaction between state institutions and social organizations. The state's strategic use of institutional incentives is predicated on the latter's embeddedness in social organizational networks. The provision of such incentives encourages competition among elites to further their organization and mobilization work, which in turn helps the state extend its infrastructural penetration into society (Mann 1984). Seen in this light, institutions are important not only as an intra-elite device for making policy compromise and inducing cooperation but also as a conduit for the state to extend its control over society.
Countermobilization as a ruling strategy Despite earlier assumptions that authoritarian regimes are characterized by political demobilization and pervasive apathy, plenty of evidence has pointed to the contrary. While authoritarian regimes and their hybrid variants are often challenged and destabilized by mass protests (Almeida 2003;Bunce and Wolchik 2010), it is also common to observe regime incumbents adopting countermobilization as a strategy to cope with opposition threats and demonstrate regime strength. Examples abound from all over the world and across regime types, ranging from outright authoritarian regimes such as China and Egypt to hybrid regimes like Russia, Venezuela and Turkey. Alongside repression (Bellin 2012;Davenport 2007) and co-optation (Frantz and Kendall-Taylor 2014;Truex 2017), which are the two major components of the standard authoritarian toolkit, countermobilization has offered an alternative strategy for authoritarian incumbents to shore up their power in the face of challenges and crises.
The increasing salience of countermobilization in various contexts has attracted growing attention among scholars of authoritarian and contentious politics. This body of research has followed two major lines of inquiry. One strand of research involves in-depth, country-specific case studies that focus on why and how autocrats adopt the strategy of countermobilization in different contexts (Hawkins 2010;Horvath 2011;Robertson 2011;Smyth et al. 2013). These works have demonstrated various rationales that motivated countermobilization and the multifarious vehicles of mobilization involved, including party organizations (Levitsky and Way 2010;Porter 2002), citizen groups (Handlin 2016;Robertson 2011) and government-hired thugs (Ong 2018). Grzegorz Ekiert et al.'s (2020) recent edited volume, Ruling by Other Means, well exemplifies this approach. Drawing upon a wide array of cases of what they call 'state-mobilized movements', which 'encompasses an array of collective social and political actions instigated or encouraged by state agents for the purpose of advancing state interests' (Ekiert et al. 2020: 5), contributors have painted a complex picture of why, when and how authoritarian regimes seek to activate countermobilization as well as what results from these actions. The dizzying array of practices has led Ekiert et al. to conclude that 'neither the motives nor the modes of [state-mobilized movements] are easily explained by regime type' (Ekiert et al. 2020: 8).
Another line of research focuses on cross-national comparisons, often using large data sets to examine the patterns and outcomes of countermobilization. Sebastian Hellmeier and Nils Weidmann (2019), for instance, use a large-n protest data set to explain the time of occurrence of pro-regime rallies in autocracies. Their analysis shows systematic increases in pro-government mobilization during episodes of large domestic and regional opposition mobilization, when there are high risks of a coup and before elections. Leveraging the same data set, Hellmeier (2020) further shows that foreign diplomatic pressure, namely in the form of threats and sanctions, significantly increases mobilization in support of incumbent autocrats. The evidence also lends support to the moderating role of media freedom in that regimes with stronger media censorship are more likely to see pro-government mobilization. Using qualitative data and a comparative historical approach, Dan Slater and Nicholas Smith (2016) compare five former British colonies and explain the causal mechanisms through which counter-revolutions, defined as 'collective and reactive efforts to defend the status quo and its varied range of dominant elites against a credible threat to overturn them from below', have produced durable dominant parties and political orders. In other words, counter-revolutions have long-lasting political implications.
These works have generated important insights for the study of authoritarian politics while confirming the role of countermobilization as a strategy in the authoritarian playbook. However, despite the growing body of scholarship, there has been a lack of discussion of the technologies and mechanisms through which authoritarian regimes mobilize their supporters. No doubt, it has been well established that the state plays a crucial role in the process by creating the organizational infrastructure and providing the resources to sponsor countermobilization efforts. Indeed, states are 'skillful and resourceful mobilizers' (Ekiert et al. 2020: 11). They can rely on a variety of instruments and partners, including party organizations, mass organizations, citizen groups, ersatz social movements, militias and thugs, to recruit large numbers of participants for initiating countermobilization (Handlin 2016;Kasza 1995;Robertson 2011). Even though the state itself does not necessarily possess these mobilizational capacities, it can mobilize through these groups in a way that Hellmeier and Weidmann (2019) characterize as 'pulling the strings'. As such, it is easy to gain the impression that the state can simply activate countermobilization at its own disposal and at a rather low cost. This leaves us with an outstanding puzzle: how does the state create a more stable and durable mobilizing structure that connects it to the masses and enables mass mobilization more efficiently?
This puzzle is further highlighted by a paradoxical feature of countermobilizations. To showcase organic support for the regime incumbents, countermobilizations should appear to be authentic and initiated by civilians despite varying degrees of state involvement. As Graeme Robertson (2011: 195) notes, 'the state has to design a movement that people would actually want to join'. This implies that countermobilizations, in order to serve as a credible display of state power, should be supported by some degree of voluntary participation and organic, genuine social interests rather than relying solely on patronage ties or paid incentives. One way to achieve voluntary and organic mobilizationor at least the appearance of itis to build a plethora of organizations which create multiple frames and collective identities that genuinely resonate with the masses in terms of their affective ties or social interests, such as birthplace ties, gender and age groups, and economic sectors. These frames and identities can be ostensibly unrelated to politics in normal times but can be activated to form an 'amalgam' of pro-regime mass in times of regime crises. But these networks of organizations that reflect a certain plurality of interests and framing are costly to build and maintain, and are difficult to engineer from a top-down approach without the vital societal embeddedness of state actors. As such, in understanding how authoritarian regimes build durable mobilizing structures for countermobilization, it is also important to look at how these structures can cultivate networks of plural social organizations to foster or at least give some semblance of voluntary participation on the part of ordinary citizens.

Institutional foundation of countermobilization
This article explains how authoritarian regimes create a more durable mobilizing structure for countermobilization by underscoring the role of political institutions. By focusing on the role of elites in connecting the ruling incumbent and the masses, we argue that political institutions can structure incentives for elites to perform tasks desired by autocrats, which, in our case, is to become leaders in grassroots organizations through which they can mobilize the masses, thus enabling countermobilization. Specifically, autocrats can manipulate the implicit promotion criteria in political institutions in a way that societal elites who excel at organizing and mobilizing the masses will enjoy a better chance in holding office in those institutions, where they can benefit from the resulting streams of political and economic rents. In other words, elites are absorbed into political institutions not only because they can represent particular societal interests, but also because they excel at performing certain tasks deemed important by autocrats.
A plethora of studies have shown that political institutions are important tools through which autocrats enable credible power-sharing among elites or co-opt opponents. Many autocrats rely on quasi-democratic elections to incorporate citizens' preferences into policymaking (Gandhi and Lust-Okar 2009) and, in some cases, to create a 'punishment regime' in which loyal supporters are rewarded while disloyal ones are punished (Diaz-Cayeros et al. 2003;Greene 2007). Meanwhile, partisan legislatures produced by elections allow autocrats to make policy concessions to elites, thus ensuring their compliance and cooperation (Gandhi and Przeworski 2007). Sometimes, they can also create a 'divided structure of contestation' which undermines the ability of the opposition to mobilize against the regime (Lust-Okar 2005). To summarize, core to this body of literature is the idea that political institutions induce credible commitment among elites and forestall their defection. While building on this literature, this article posits that political institutions not only co-opt elites but also serve as incentive devices to motivate elites to perform tasks favoured by the autocrats. Given that their admission or promotion criteria are often unwritten and subject to autocrats' discretion, political institutions may serve as crucial instruments for autocrats to adapt to ever-changing political circumstances and suit their political needs. In our case, because the semi-authoritarian regime needs a robust supply of pro-regime supporters to countermobilize against opposition threats, such institutions can incentivize those societal elites who demonstrate the capability to perform such tasks by favouring their promotion. In short, societal elites are institutionally rewarded for their capacities to consistently maintain the organization of the masses and effectively mobilize them at times of heightened opposition challenges.
In doing so, this article also seeks to highlight the conceptual overlap between countermobilization and the traditional authoritarian strategy of co-optation. Extant literature tends to portray countermobilization as a distinct strategy from repression and co-optation, despite recognizing the fact that they can work in combination with one another. While some scholars have already pointed out that countermobilization can have repressive manifestations (Hellmeier and Weidmann 2019; Ong 2018), we aim to emphasize how countermobilization can also be connected to co-optation. To be precise, mechanisms of co-optation can serve as building blocks for countermobilization. This gives us a more complex view of the conceptual linkages between different strategies in the authoritarian playbook.

The case of Hong Kong
Hong Kong has been characterized as a hybrid regime combining democratic traits with autocratic ones (Fong 2017;Ma 2011). The city has robust rule of law and strong civil liberties on a par with democratic countries. However, its political system remains semi-authoritarian for two reasons. First, it has maintained an 'executive-led' government inherited from the colonial era, with the chief executive elected by a small selectorate, a powerless legislature and a corporatist system favouring business and pro-Beijing elites (Fong 2014;Ma 2007). This 'power elite' comprising tight-knit business-state networks has constituted a systemic barrier against further democratic development (Ho et al. 2010). Second, on top of the local political system, Hong Kong is also nested within the national institutional framework of the Chinese party-state. This embeddedness guarantees China's control and influence, which is partly channelled through multiple political institutions that bridge national and local politics. For example, Beijing has actively co-opted local elites into national institutions such as the National People's Congress (NPC) and the CPPCC. Beijing also operates, along with other agencies, a Central Liaison Office in the city to cultivate pro-China support in different social sectors and to influence elections, largely through clientelism (Lee 2020).
Against this background, cycles of mass protests emerged and pressed for local autonomy and democratic reforms throughout the past two decades (Cheng 2016). Notable episodes include the rally on 1 July 2003, the Umbrella Movement in 2014, and, more recently, the Anti-Extradition Law Amendment Bill Movement in 2019. The mobilizational strength of Hong Kong's civil society, coupled with its strong calls for democratic changes, has not only threatened regime stability but also constrained the state's strategic options. On the one hand, clientelism in electoral politics did not seem to have substantially turned the pro-government population into the majority (Wong 2015). On the other hand, overt and harsh repression can be costly as it may intensify opposition unity and jeopardize the city's status as an international financial centre. Co-optations through the granting of policy concessions are no longer effective in placating social groups that are now seeking a fundamental political overhaul. Nor can the democracy movement be effectively quenched by the co-optation of opposition leadersthat very act would be seen as desertion and betrayal by the pro-democracy masses. Moving beyond traditional repression and co-optation, a new pattern of countermobilization emerged in response to the incessant pro-democracy challenges.

Institutionalizing countermobilization
It is in this context that the regime turned to the playbook of countermobilization, which began with the emergence of numerous small-scale pro-regime groups around 2010, such as Caring Hong Kong Power, Voice of Loving Hong Kong and the Defend Hong Kong Campaign. These groups fashioned themselves as citizen-based protest organizations and frequently staged protests against the prodemocracy camp or in support of the government (Cheng 2020; Lee 2020). The naming of these organizations reflected their loyalty and political beliefs, symbolizing the idea that they were self-mobilized to act out of their love towards the city. However, despite attracting media attention, the undisciplined and melodramatic style of their actions proved to be an unnecessary embarrassment for the regime.
These newly emerging protest organizations have led the regime to rethink the strategy and replace them with a more organized and disciplined pro-regime force. Rather than simply grooming pro-government forces through clientelist relationships, the aim is to expand them and better institutionalize countermobilization through a robust organizational network that can effectively organize and mobilize a broad base of constituents. In Hong Kong, such a role is crucially assumed by a wide array of grassroots organizations, such as community groups, women and youth groups, and native-place associations. Some of these organizations have served as a part of the Chinese Communist Party (CCP)'s united front work to 'work on the masses' (Ma 2007); others are either newly established or newly incorporated into the network. Figure 1 displays the organizational proliferation by showing the number of newly established native-place associations, a major type of grassroots organization, after 1984. As shown, more than 200 such organizations were formed after 1997, compared with only 230 throughout the 156 years of colonial rule. The growth was particularly obvious after 2008, especially in 2012 and 2014, and the momentum was maintained in the subsequent years. Table 1 summarizes the number of affiliated organizations and individual members of the major pro-regime grassroots organizations.

Mobilization efforts by pro-regime grassroots organizations
The overall trend is that pro-regime grassroots organizations have proliferated in number while forming connections with one another through shared directorships. The result is an extensive organizational network that provides the organizational infrastructure for countermobilization. Indeed, since around 2013, these grassroots organizations have become more active in contentious politics. They began to venture beyond electoral mobilization for pro-Beijing candidates and started organizing counter-protests against their pro-democracy counterparts on a more regular basis.
According to data from the Hong Kong Police (see Table 2), counter-protests have been increasing since the early 2010s. From 2012 to 2017, the number of legally endorsed protests organized by pro-Beijing groups on 1 July and 1 Octoberrespectively, the Hong Kong's Handover anniversary and China's Note: Federations are distinguished from associations in that the former are supra-organizations that consolidate scattered associations for coordination and resource-sharing. To compile a comprehensive list of native-place associations, we conducted a thorough search using Wisenews, an online news portal, and the list of organizations registered under Hong Kong's Societies Ordinance, using keywords such as tongxianghui, shetuan lianhui and lianyihui. We also looked up newspaper advertisements posted by major native-place associations, which sometimes list their member associations. After that, we looked for their year of establishment from the official websites or annual reports of native-place federations and associations.
National Day, when anti-government protests were commonincreased significantly, whereas those organized by pro-democracy groups decreased. Table 3 further summarizes the major pro-Beijing demonstrations since the Umbrella Movement in 2014. Although the turnouts were often inflated, the fact that these numbers are growing, and that counter-protests have become more frequent, highlights the regime's interest in projecting the image of its widespread social support in the face of recurring pro-democracy challenges.
In these protests, a common feature was that pro-regime grassroots organizations were the key vehicles of mobilization. For instance, in response to the 2014 Occupy Central pro-democracy campaign, the regime established a group known as the Alliance of Peace and Democracy, which brought together numerous grassroots organizations to counter-protest against pro-democracy forces.  The Alliance initiated a petition campaign that claimed to have collected signatures from nearly 1,600 organizations and 1.8 million individuals. Besides the well-known pro-Beijing parties, signatories included a wide range of grassroots organizations -26% of which were native-place associations, while the rest include community organizations, resident associations, alumni groups, youth groups and women groups, as well as cultural and sports associations. This unprecedented parade of pro-regime forces would not have been possible without the already existing grassroots organization networks. Shortly after that, the alliance organized a mass rally in mid-August 2014, claiming a turnout of 130,000, making it the largest pro-regime mobilization in the city at that time. Many pro-regime grassroots organizationsespecially native-place associations and community organizationsmobilized their members to participate. The Fujianese federation played a particularly crucial role in the event. Its honorary president, businessman Hung Chao Hong, served as the 'chief commander', while its president, Thomas Wu, served as the deputy. Both Hung and Wu were reported to have donated millions of dollars to organize protests (Apple Daily 2014).
After two pro-independence candidates were elected in the 2016 Legislative Council election, grassroots organizations and pro-Beijing parties formed a coalition known as the 'Alliance against Insulting China and Hong Kong Independence'. Modelled upon the Anti-Occupy Central Alliance, the new alliance comprised 25 founding organizations, including seven large native-place associations and three major community organizations. The honorary president of the Fujianese federation, Hung, again served as the convener, showing that leaders of grassroots organizations had an important role to play in the mobilization process. Like the 2014 countermobilization, a petition campaign was launched to call for the disqualification of the pro-independence lawmakers, which claimed to have collected endorsements from 470 pro-regime grassroots organizations. In addition, the anti-independence Alliance staged a protest in October that year to showcase its popular support, followed by another one a month later to support the NPC Standing Committee's interpretation of the Basic Law on oath-taking.

Countermobilization during the anti-extradition bill protests
The 2019 Anti-Extradition Law Amendment Bill Movement marked another critical point in which pro-regime grassroots organizations played a salient role in mobilizing counter-protests. Sparked by a government-proposed amendment to the city's extradition laws, a series of mass protests broke out across the territory and morphed into the city's most serious political crisis since the Handover. While enormous public pressure forced the government to suspend the amendment bill, local police continued to use heavy-handed tactics to quell the continuing protests, which further intensified anti-government mobilization. In response, the central government adopted a more proactive strategy from late July and launched a coordinated campaign to counter the protests. Indicative of that is a meeting in early August, when more than 500 business leaders and pro-Beijing politicians, many of which are NPC and CPPCC members, were summoned to Shenzhen, where they were given further instructions from Beijing officials to 'safeguard Hong Kong's prosperity and stability'. Beijing's unambiguous stance mobilized the pro-Beijing camp. Three mass rallies were staged after late July to support the Hong Kong government and back the police's handling of the protests. The biggest was the 'No Violence, Save Hong Kong' protest on 17 August, which claimed a turnout of 476,000. An analysis of the 72 pro-Beijing co-organizers shows that 41 of them hold directorships at the pro-regime grassroots organizations included in our analysis. The author attended that protest and counted 75 grassroots organizations, 39 of which were native-place associations. The author also conducted an onsite survey (n = 157) on the participants, which showed that 24% of the respondents had participated in activities organized by native-place associationssignificantly higher than that of professional associations (17%), churches (10%) and political parties (9%). Evidence collected by the author indicates that these groups mobilized their members through messaging apps such as WhatsApp and WeChat. One example was the United Zhejiang Residents' Associations, which disseminated a recruitment message through its network and asked different county units across Zhejiang province each to mobilize a required number of protesters.

Promotion in the CPPCC as institutional incentive
The construction and expansion of this network of grassroots organizationswhich enables timely and effective countermobilizationis facilitated by a group of elites who act as the intermediary between state and society. As shown in the above examples, by serving as directors of grassroots organizations, societal elites, many of whom are business people and professionals, are in a position to mobilize members en masse, often through stoking their nationalistic identity or conservative orientation, whenever there is a political need to do so (Yuen 2021). While members can be mobilized through shared values and native-place or other social ties, elites also provide selective incentives through their donations to foster members' participation, such as by providing transport or offering sumptuous meals after protest events. Interviews with directors suggest that grassroots organizations typically (with some exceptions) do not pay members in cash to protest (since that would incur a huge financial cost), but they would make sure that members are treated well. Members often see such protest events as an occasion for social gathering with other fellow members.
These elites are nominated to the board of directors of grassroots organizations for different reasons. Some are nominated because of their existing social status, others due to their experience as members of the organization, their length of service or their generous donations (Yuen 2021). 2 However, to leverage them as bridges to the masses, these elites must be incentivized to become willing to incur the costs of those mobilization tasks. The incentives, as we argued earlier, come from the institutional rewards of holding offices in the CPPCC, a valuable conduit into sources of policy goods or private economic rents.
The CPPCC is China's top consultative assembly, comprising delegates from the CCP and other elites from various political parties and organizations, and offering policy advice and suggestions to the party-state. Given its prestigious status, the assembly has served as a major united front platform through which the party co-opts elites (Sagild and Ahlers 2019). In post-Handover Hong Kong, the CPPCC, along with the NPC, has also been a major mechanism to co-opt local elites, such as wealthy businessmen, professionals and politicians, into the party's united front (Fong 2014). In 2018, around 10% of the 2,158 delegates were from Hong Kong. A seat in the CPPCC helps the elite signal their status and relationship with party officials, which facilitates access to policy goods and economic rents. Our interviews with two directors of native-place associations suggest that one of the motivations for joining the board of directors is to become eligible for nomination for national or local CPPCC positions. 3 These benefits are especially vital for elites who have businesses in China or those who are hoping for career advancement. They also come with other perks and privileges: easier entry to Mainland China, special channels for immigration clearance and VIP lounges in airports, among other things.
Unlike the NPC, whose members are elected by a locally formed committee, CPPCC seats are selected through an opaque process. For the Hong Kong division, candidates are first nominatedsome by the Central Liaison Office, some by the United Front Work Department and some by powerful pro-China local elitesbefore they are selected by the CPPCC national committee. Thus, despite the lack of elections, the quest for CPPCC seats among elites is highly competitive. In some ways, the opacity has made the competition more intense, as elites need to speculate on the selection criteria. This in turn creates the incentive structure for elites to execute certain tasks desired by the party-state, while giving the party-state the flexibility to decide what such tasks should be.

Data and method
To test whether elites with higher organizational and mobilizational capacities of the masses are more likely to be awarded CPPCC offices, we gathered two types of data: CPPCC elites' biographical data and grassroots organizations' leadership data. We focus on 205 Hong Kong representatives of the 12th CPPCC  and exploit the transition to the 13th session (in January 2018) to examine the determinants of elites' office retention in the 13th CPPCC. 4 Table A1 in the online appendix provides summary statistics of these members. Most of them are men, and over half are entrepreneurs. The mean age is 63.9 and the mean number of organizational ties, a variable which we will explain below, is 2.84. While 108 of them were successfully reselected to the 13th session, 97 failed to secure another term of service. The dependent variable of our analysis is therefore whether an elite remains in office in the 13th CPPCC. This transition is particularly significant because, since 2014, pro-democracy forces have posed many challenges to the regime by staging numerous episodes of protest mobilization. This is the same period during which we observed the rise of countermobilization initiated by pro-Beijing grassroots organizations (Table 2 and Table 3). If the theory is correct, rewards for those countermobilization efforts should be observed in the next selection to the 13th CPPCC.
Given the prominent role of grassroots organizations in initiating countermobilization, we operationalize the organizational and mobilizational capacities of elites as the number of leadership positions they hold at various grassroots organizations. Each grassroots organization has a central body of leadership, usually called the 'board of directors'. Crucially, not only do most grassroots organizations publicize their leadership information, but such information is also often advertised on the party's mouthpiece newspapers, accompanied by interviews with those leaders, elaborating on their 'work on the masses'. We interpret these instances as signalling devices through which those holding leadership positions show their competence and strategic value to the regime in organizing and mobilizing the masses.
Using annual reports, newsletters and advertisements published in statecontrolled newspapers, we collected leadership data on 43 major grassroots organizations, 5 whose sessions fall within the period from 2010 to 2021. The data set comprises 7,398 distinct individuals. This information allows us to create our main explanatory variable by adding up the number of leadership positions held by each CPPCC elite at various grassroots organizations, which we call 'organizational ties'. An elite who has a leadership position at one such organization will have one organization tie; those who have leadership positions at three organizations will have three organizational ties, and so on. We restrict our focus to a subset of grassroots organization appointments based on two criteria. First, the starting years of the appointments have to be before 2018 such that we do not capture the fact that newly selected CPPCC members might attract more organizational affiliations after the 13th CPPCC selection in 2018. Second, the ending years of the appointments have to be on or after 2018 such that the elites were functioning members of the organizations during the selection. These criteria help increase the validity of our explanatory variable in that they capture meaningful organizational ties at the time of the selection process.
In addition, we collected biographical information of the elites, such as age, native places, primary occupation, political party affiliation, past CPPCC tenure and other information that can be used as control variables to guard against alternative explanations. We then employ a series of logit regression models to examine the effect of organizational ties on the probability of remaining in office in the 13th CPPCC. The descriptive statistics of those variables are shown in Table A1 in the online appendix. Table 4 presents the logit model results with different specifications. Model 1 shows the simple bivariate relationship between organizational ties and being elected to the 13th CPPCC. The positive and significant coefficient suggests that holding more leadership positions in grassroots organizations is positively associated with elites' chances of holding office in CPPCC.

Results
The association between organizational ties and office retention in the 13th CPPCC can be endogenous to longer past CPPCC tenure: members who have served multiple terms in the past might acquire more organizational ties on the one hand, and be more likely to continue holding office in the new session. Model 2 investigates this alternative explanation by controlling for the total number of past CPPCC sessions an elite has served (12th session included). The coefficient of organizational ties remains positive and significant. Interestingly, past CPPCC tenure has a negative effect on office retention in the 13th CPPCC, indicating that experienced members might be being replaced and relatively novice elites being favoured.
Relatedly, there is an implicit norm of rotation in Chinese elite politics, known as 'seven up, eight down'. This means that candidates who are at the age of 67 or below can be promoted to high office, while those at the age of 68 or above are highly discouraged, if not prohibited, from seeking another term. Whether an elite is over the age limit correlates, again, with the chance of office retention in the 13th CPPCC and the number of organizational ties accumulated. Model 3 examines this possibility by including a dummy variable indicating whether a candidate was at or over the age of 68 in 2018. The estimated coefficient of organizational ties remains positive and significant; the effect size even increases. The estimated coefficient of the over-age limit dummy exerts, as expected, a strong negative effect on the chance of remaining in office in the next session. Model 4 further controls for the effects of gender and college education, which, again, bolster the effect of organizational ties.
Models 5 to 7 examine three crucial alternative explanations: native place, occupation and local political party affiliation. First, a large portion of the grassroots organizations we investigate are built upon common native places in Mainland China. The relationships between different native-place groups can be complementary but also competitive. For example, it is speculated that elites of Fujian origin might enjoy a political advantage over other southern Chinese groups because Xi Jinping, who became the 'paramount leader' in 2012, spent a substantive portion of his tenure in Fujian. To control for this potential native-place effect, Model 5 includes dummies for the major native places -Guangdong (as reference category), Fujian, Hakka, Zhejiang and Chaozhoumost active and relevant in Hong Kong politics. The results do not suggest any systematic advantage of any native place, and the estimated coefficient of organizational ties remains positive and significant.
Similarly, Model 6 controls for the effects of occupation. Certain occupations, such as entrepreneurs, can be actively sought after by grassroots organizations   for their financial resources, and enjoy a systematic advantage of promotion in CPPCC for the strategic values of having them co-opted. The estimated coefficient of organizational ties, however, remains robust after the inclusion of various occupation dummies (with entrepreneurs as the reference category). Lastly, affiliation with important local political parties can be driving both leadership ties to grassroots organizations and the chance of retaining office in the 13th CPPCC. Model 7 investigates this explanation by controlling for elites' affiliations to major local political parties. 6 The inclusion of local political party affiliation does not change the direction of the estimated coefficient of organizational ties, although it becomes statistically insignificant. We are less concerned by this change because grassroots organizations' leaders are indeed embedded within the larger network of both political parties and other local political institutions (such as the District Council or the Legislative Council). Furthermore, in the subsequent analyses that exclude honorary appointments (more below), the organizational ties variable remains statistically significant even if a political party is controlled for, which might suggest that those elites with political party affiliations received mostly honorary appointments in grassroots organizations. Apart from various control variables included above, another important source of endogeneity is that elites receive leadership appointments in grassroots organizations simply due to the fact they were incumbent CPPCC members. Appointing those elites to their leadership helps increase the prestige of the organizations; having those appointments helps elites to further signal their social status or fame which, in turn, helps them to be more competitive in the CPPCC selection. In other words, it is the signalling of social status that drives both organizational ties and holding office in the CPPCC.
The nature of grassroots organizations' leadership data allows us to tease out the signalling effect of organizational ties. There are two types of leadership positions in our data: honorary and non-honorary positions. Honorary appointments do appear to be devices for fame signalling as they include a wide and rather indiscriminate array of important politicians and social elites. We construct an alternative measure of organizational ties by excluding all honorary appointments and including only non-honorary ones. We rerun the same seven models specified above (results reported in Table A2 in the online appendix). The estimated coefficients of organizational ties, on the whole, have larger effect sizes (compared to Table 4) and remain statistically significant through out all specifications (including Model 7 that controls for political party affiliation).
To aid the interpretation of the substantive effects of organizational ties and other covariates, we simulate the first differences of the probabilities of remaining in office in the 13th CPPCC given some changes in the covariate values, based on the full model that excludes honorary appointments in the analysis (Model 7 in Table A2 in the online appendix). Figure 2 presents the results of the first differences analyses. For example, the first entry shows that when an elite is over the age limit (compared with one within the age limit), his or her probability of remaining in office in the 13th CPPCC substantively decreases by 56.7% (90% CI: 41.8% to 69.5%). Similarly, longer previous CPPCC tenure (three years compared with one year) also decreases the probability of remaining by 17.0% (90% CI 6.4% to 29.9%). Most crucially, having more organization ties (four ties compared with none, which correspond to the maximum and median values of that variable) increases the probability of retaining office in the 13th CPPCC by 18.8% (90% CI 2.40% to 36.9%).
Lastly, we address the concern that not every grassroots organization is equally important; those grassroots organizations vary in their membership sizes and political importance. While it is difficult to collect precise information on those metrics, we construct an ordinal variable which reasonably captures the relative size and importance of each grassroots organization. 7 With this ordinal variable, we construct an alternative explanatory variable, organizational scores, in which the counts of organizational ties are weighted by the relative size and importance of those grassroots organizations captured by the new ordinal variable. Results are presented in Table A3 (which includes honorary appointments in the analysis) and Table A4 (excluding honorary appointments) in the online appendix. The results remain substantively unchanged: the organizational scores are positively and significantly associated with office retention in the 13th session of the CPPCC across most specifications.

Discussion and conclusion
Authoritarian rulers do not always rely on repression and co-optation to maintain social control. When civil society is organizationally strong and when citizens demand fundamental institutional changes, these classic strategies may lose their effectiveness. In these situations, many authoritarian rulers have resorted to the strategy of countermobilization to counteract opposition groups. But how can the state build up a mobilizing structure that can swiftly mobilize the masses while creating the impression that supporters are not simply puppets of the state? The baseline hypothetical scenario is one of a Guangdong male entrepreneur within the age limit, holding a college degree, with average past CPPCC tenure (2.17 years) but no organizational tie nor any local political party affiliation. Each circle represents the estimated first difference in probabilities given some changes in covariate values. Horizontal lines show 90% confidence intervals.
This article has provided an institutionalist explanation of how formal political institutions can play a role in incentivizing societal elites to organize countermobilization on behalf of the state. By studying the case of post-Handover Hong Kong, our research has revealed the importance of a robust network of grassroots organizations, on the one hand, and the critical intermediary role of societal elites who organize and mobilize the masses, on the other. It has explained how the semiauthoritarian regime implemented this strategy on a routine basis by providing institutional incentives to those societal elites: pro-Beijing elites with more leadership ties to grassroots organizations are more likely to remain in office in the CPPCC, which in turn offers them access to rents and privileges in the mainland. By institutionalizing political rewards for local elites that are connected to the masses, the regime enjoys a greater ability to countermobilize against prodemocracy threats and dampen demands for political changes in its hybrid system. Our analysis has two major theoretical implications for the study of authoritarian politics and social control. First, while extant scholarship has shed light on the use of countermobilization as a strategy for autocrats to buttress their rule, our findings delineate the institutional mechanism through which countermobilization can be implemented. We show that societal elites play a critical intermediary role in the process by bridging the state with myriad grassroots organizations. Although the state retains a crucial role in signalling the need for countermobilization, the involvement of elites makes grassroots organizations look less like top-down 'transmission belts' and more like social organizations that tap into a variety of social relations, which in turn enables the infrastructural penetration of the state into the society. This might also contribute to the perceived authenticity of countermobilization and its representation of genuine social interests. In short, countermobilization is not simply a state-centred story, as has been typically portrayed. It is made possible at the confluence between the state and society, where the role of the state might be covert and indirect.
Second, extant literature on authoritarian institutions often focuses on the ability of political institutions, such as legislatures and elections, to induce cooperation and facilitate credible commitment among elites. Our findings indicate that institutions can also serve as an incentivizing mechanism for autocrats to build up their social bases and mass support through the mediation of power-seeking elites. Unlike mainstream institutionalist approaches that look at how elites are incorporated into political institutions because of their established social power, our evidence reveals a reverse mechanism in that it is the institutionsand the access to rents and privileges that incentivize themthat build up social power. One may even argue that people hold office in institutions not because they are elites to begin with; they only become elites by gaining positions in such institutions because they are willing to execute certain tasks desired by the state.
Although Hong Kong is in some ways unique as it is a subnational semiautonomous polity nested within a larger authoritarian state, our case study has broader theoretical implications. Given its initial lack of a strong mass base, the Hong Kong regime bears similarities to other hybrid regimes that also lack a hegemonic ruling party (e.g. Venezuela), as well as military regimes that seized power through revolutions or coups d'état (e.g. Egypt). In these regimes, autocrats similarly do not have ready access to the mass base at the inception of their tenure, and they may also need elite intermediaries to do the bridging work. The case of Hong Kong thus offers a possible explanation for why these regimes were able to countermobilize against opposition threats despite their inherent weaknesses. To test for generalizability, more cross-national comparisons are needed to identify whether similar mechanisms are at work and what allowed such mechanisms to take root.