At 7:30 am on 12 June 2019, it was already hot and sultry in Hong Kong. The early summer day had yet to reach rush hour, but a small group of protesters had already gathered, blocking Lung Wo Road near the Legislative Council (LegCo) where politicians were set to discuss a controversial bill that had sparked widespread opposition. Soon, more and more protesters arrived. They skilfully moved to stop the traffic on Harcourt Road, swiftly occupying the city’s major highway that connects court buildings, government branches, financial institutions, and the garrison army. By 8:00 am, Admiralty was teeming with anxious protesters. They heeded the call to strike from work and surround both the LegCo and the adjacent Hong Kong government headquarters, a towering glass-clad structure. Equipped with face masks, goggles, helmets, and umbrellas, they were prepared for potential clashes with the police, ready to face the situation head-on.

Admiralty – once a navy dockyard during the British colonial era and now the political and economic heart of China’s Special Administrative Region (SAR) – was an eventful place. Five years before the movement, in the autumn of 2014, a large section of it was occupied by protesters calling for democratic reforms to the city’s electoral system. Known as the Umbrella Movement, the largely peaceful occupation persisted for over two months but ultimately fell short of its objectives, as the government dismantled the encampments. Half a decade later, it seemed as though history was repeating itself. This time, however, protesters appeared more equipped and resolute in their actions. Moreover, they were driven by a more pressing objective: to halt the advancement of an amendment to the city’s extradition laws, which was scheduled for its second reading.

Several months earlier, the government put forth a controversial proposal to amend the city’s extradition law, allowing for the extradition of fugitives to mainland China. Despite Hong Kong’s status as an SAR of the People’s Republic of China (PRC) for the past two decades, there were no existing extradition agreements with the mainland. The amendment gradually ignited widespread public concern. Citizens feared that the law could be exploited to target political dissidents within the semi-autonomous city and erode the legal barrier that had traditionally distinguished Hong Kong from the mainland. Initial signs of dissent emerged in February 2019 as soon as the government tabled the amendment bill. However, it was not until May, after intense debates in the legislature and an online petition that garnered hundreds of thousands of signatures, that the opposition movement gained momentum and galvanized society.

June 9 marked a historic moment, as approximately one million people took to the streets in a peaceful demonstration against the imminent bill. The massive turnout made it the largest protest in Hong Kong’s history.Footnote ¹ Led by Civil Human Rights Front (CHRF), a coalition of prominent pro-democracy social movement organizations (SMOs), the June 9 rally was so massive that protesters found themselves stranded in Victoria Park for hours before they could commence the march. The rally officially ended at 10:00 pm – far later than typical marches. Many stayed on the streets for longer. But the Hong Kong government remained unmoved by the unprecedented turnout. It declared that it had no intention to suspend the bill and asserted that the second reading, the final step before the law’s implementation, would proceed despite the opposition.

Outraged by the government’s dismissal of public dissent, protesters once again assembled on June 12. This time, the demonstration unfolded quite differently. Unlike the organization-led rally on June 9, the day’s protest was not initiated by any leaders or organizers. Numerous organizations and trade unions had urged citizens to strike but did not provide specific plans of action. Instead, individuals primarily relied on online platforms to disseminate the call to mobilize. One such call was started by two ordinary Facebook users, who created an event page to invite others to ‘picnic alone’ at the government headquarters. Remarkably, within hours, over 10,000 citizens expressed their intent to participate. As protesters arrived in Admiralty for their ‘picnic’ the next day, there were no leaders or organizers to provide guidance, nor was there any plan or timetable to be followed. The event was entirely improvised. No one knew what to do next and how it would end.

A decentralized yet implicitly ‘organized’ form of mobilization quickly emerged with a spontaneous division of labour. Protesters tacitly assumed various roles. Some positioned themselves on the frontlines, directly engaging with riot police. Others played crucial supporting roles that aligned with their existing expertise. Doctors and nurses established impromptu first-aid stations to provide essential medical care to injured protesters. Van drivers used their vehicles to transport vital supplies such as water, saline solution, and surgical masks. Christian groups formed choir lines, physically separating protesters from the police, filling the air with calming religious hymns. Meanwhile, students used the encrypted messaging app Telegram to create channels to disseminate verified information about the protest. This decentralized and spontaneous protest set the stage for what was to come for the Anti-Extradition Law Amendment Bill Movement (hereafter referred to as the Anti-Extradition Movement).

Under immense public pressure, the government suspended the bill three days later on June 15. But protesters were not satisfied. They continued their mobilization and built upon the unfulfilled demands of the 2014 Umbrella Movement to incorporate political reforms in their demands. Adopting a ‘be water’ strategy, protesters emulated the tactics employed on June 12 and orchestrated numerous protest actions citywide. In the following months, guided by this decentralized logic, protesters paralyzed traffic, staged wild-cat demonstrations and airport sit-ins, broke into LegCo, organized neighbourhood protests, and turned university campuses into fiery battlegrounds. They boycotted pro-government businesses and actively supported those aligned with the movement, leveraging their economic influence to exert political pressure. Some activists even pioneered an ‘international front’ to mobilize support overseas.

Despite intensifying momentum, the Hong Kong SAR government stood firm and refused to make further concessions. As the Chinese authorities signalled their disapproval of the protests, local authorities shifted to a more hard-line stance. Riot police escalated their use of crowd control measures to suppress the unrest. The situation quickly radicalized the protests. Faced with a hardened government, protesters felt compelled to use increasingly transgressive and violent tactics to hold their ground. Yet, the movement managed to maintain a surprising level of cohesion despite tactical radicalization. While some protesters resorted to using weapons such as bricks, bamboo poles, and Molotov cocktails to confront the police, moderate protesters remained tolerant of these transgressive actions. This exceptional level of cohesion extended beyond the streets and translated into a significant political victory at the ballot box. In the local District Council elections held in November 2019, the pro-democracy camp won a landslide victory, securing 388 out of 479 seats.

The City of Protests

Despite the spectacle of 2019, mass protests were a familiar sight in Hong Kong. Since its handover to the PRC in 1997, the semi-autonomous city had earned a reputation as a ‘city of protests’ due to the regularity and scale of its protest activities (Dapiran Reference Dapiran2017). Operating under the One Country, Two Systems (OCTS) principle, Hong Kong citizens enjoyed freedoms and rights not available to their counterparts on the mainland. Despite maintaining a closed political system dominated by business and state interests, the semi-autonomous city afforded citizens the freedoms of speech and assembly. Empowered by these rights, opposition activists successfully organized a series of protest movements over the past two decades. These movements were sparked by a range of issues comprising heritage preservation, infrastructural development, education, and constitutional reforms. Other protests were more routine. On every June 4, hundreds of thousands would assemble for a candlelight vigil to commemorate the victims of the 1989 Tiananmen Movement. On every July 1, the anniversary of the sovereignty handover, hundreds of thousands would again participate in an annual anti-government demonstration, demanding greater government accountability and the implementation of universal suffrage.

These protests shared a common feature, albeit one that would wane over time: the dominant role of SMOs. Both the annual July 1 rallies and the June 4 vigils were highly organized events, consistently led by pro-democracy parties or civil society organizations. They adhered to highly scripted and ritualistic formats – congregating in the same locations at the same times, marching the same routes, singing the same songs, chanting the same slogans, and following a familiar agenda. Participants in these demonstrations often engaged in a passive manner, following the familiar script laid out before them. The issue-driven movements, meanwhile, were less scripted and more improvised. They tended to be longer in duration and unpredictable in their evolution. For instance, the 2012 Anti-Moral and National Education Movement brewed for several months, eventually culminating in a week-long occupation of the government headquarters’ forecourt. The 2014 Umbrella Movement originally started as a civil disobedience campaign and class boycott but exploded into a city-wide occupation for eleven weeks. However, even in these increasingly spontaneous movements, a form of centralized leadership often emerged, composed of SMOs, opposition parties, and prominent activists. This leadership played a vital role in coordinating protest actions, handling logistics, negotiating with the authorities, and directing the protesters.

The 2019 Anti-Extradition Movement stood out from the earlier movements in several ways. The most obvious was its duration and turnout. Unlike most previous movements that normally lasted a day or no more than a week, the Anti-Extradition Movement lasted for seven months, even significantly exceeding the duration of the Umbrella Movement. It challenged the expectation that urban revolutions are often ‘limited in duration because they occur where the state’s coercive power is strongest and its nerve centers are concentrated’ (Beissinger Reference Beissinger2022: 204). The movement comprised several hundreds of protest actions dispersed across the city, different from how previous movements tended to concentrate in a few locations. Its cumulative participation rate also dwarfed previous movements. According to a poll, an astonishing 45% of Hong Kong’s population of seven million residents took part in at least one protest event during the movement, and 58% expressed support for its demands (Cheng et al. Reference Cheng, Lee, Yuen and Tang2022). This level of involvement surpassed the city’s largest protests after the handover – the 2003 July 1 rally and the Umbrella Movement – which reportedly mobilized 8% and 20% of the population, respectively (Cheng and Ma Reference Cheng and Ma2020). Compared with other notable mass mobilizations globally, the Anti-Extradition Movement also stood out. It surpassed the 12% participation rate in Ukraine’s Euromaidan of 2014 (Chupryna Reference Chupryna2021), the 16% in the Tunisian Revolution of 2011, the 8% in the Egyptian Revolution of 2011 (Beissinger, Jamal, and Mazur Reference Beissinger, Jamal and Mazur2015: 3), and the 10% in the United States’ Black Lives Matter protests of 2020 (Heaney Reference Heaney2022).

Perhaps the most distinctive feature of the Anti-Extradition Movement was its organizational form. In contrast to the routine and issue-driven protests, the movement did not have a centralized leadership. While it formally started with an organized rally led by a prominent SMO in early June, it swiftly became ‘leaderless’ as ordinary citizens, who were mostly unaffiliated with conventional organizations, planned and orchestrated protest actions in a decentralized and spontaneous manner, as described at the outset of the chapter. These protesters harnessed digital communication technologies and social networks to mobilize other fellow citizens and manage the intricacies of organizing. Instead of being passive participants following a pre-determined script, protesters became both the scriptwriters and the protagonists mobilizing on their own terms. While opposition parties and civil society organizations remained involved in organizing and coordinating some of the protest actions, they deliberately maintained a low profile and limited their role to resource provision and logistical support.

Leaderful Mobilization

Many scholars and observers have characterized the 2019 protests as a ‘leaderless’ movement (BBC 2019; Lai and Sing Reference Lai and Sing2020; Liang and Lee Reference Liang and Lee2023) to underscore its departure from traditional movements with a clearly identifiable leadership structure. However, the label can be misleading in two ways. First, by merely defining the movement as what it was not, it fails to inform us about the movement’s organizational structure and dynamics. Second, the term ‘leaderless’ could imply that leadership practices are absent, given that leadership is commonly understood as the defining quality of leaders. But this could misrepresent what occurred in the Anti-Extradition Movement. Leadership was certainly present, but it was not exercised by traditional leaders and organizations.

A new concept is needed to capture this unique organizational form. We argue that the 2019 Anti-Extradition Movement belongs to a class of movements that can be characterized as ‘leaderful mobilization’. Leaderful mobilization is defined as a form of mobilization where large numbers of protesters collectively and simultaneously exercise leadership without the traditional hierarchy that concentrates power in the hands of a few. Under leaderful mobilization, leadership is distributed across a wide array of actors, including ordinary citizens, latent networks, and conventional SMOs. In this sense, protesters are actively engaging in organizing and coordinating protest actions rather than merely attending them. They take on different leadership roles and tasks (Earl Reference Earl2007), such as planning, communication, information gathering, logistics, and advocacy, to sustain mobilization under a decentralized and horizontal structure. As such, leadership can be understood as a set of practices that are divided among participants and shared collaboratively, rather than as positions held by specific individuals or entities, or the attributes that they hold. In other words, the ‘mass’ in mass mobilization has developed a new life: they are no longer a collective entity being mobilized; instead, they are exercising agency and taking matters in their own hands. They are ‘masses’ of their own accord.

As leadership is distributed among participants, protesters who take leads can be considered to be informal leaders in these movements. These informal leaders do not necessarily have to reveal their identity to others. Moreover, they may not possess the same level of public profile or charisma as historical movement leaders like Mahatma Gandhi or Martin Luther King, and they usually do not wield the same authority over organizational bureaucracies and human resources as these traditional figures. Nonetheless, informal leaders have the capacity to develop strategic agency and make consequential decisions for the movement. For instance, they can craft and disseminate slogans, posters, or infographics via social media to frame issues and draw public attention. They can contribute ideas or devise innovative tactics that generate new political opportunities. Furthermore, they can utilize digital or latent social networks to mobilize their peers. To put ideas into action, they can assemble teams to strategize and execute specific initiatives. Although their leadership may be tied to specific actions and not be sustained over time, these informal leaders are capable of influencing and directing the movement through taking initiatives and collaborating with one another.

Another distinguishing feature of leaderful mobilizations is the intrinsic role of spontaneity. By spontaneity, we refer to events or happenings that are largely improvised without predetermined plans. David Snow and Dana Moss (Reference Snow and Moss2014: 1123) define spontaneity as a cover term for ‘events, happenings, and lines of action, both verbal and nonverbal, which were not planned, intended, prearranged, or organized in advance of their occurrence’. Benjamin Abrams (Reference Abrams2023: 3) further uses the term to characterize some of the recent protest movements, such as the Black Lives Matter Movement and the Arab Spring uprisings, as ‘spontaneous mass mobilizations’, which occurs when ‘large numbers of people partake in contentious politics without reliance on social movement organizations and their networks’. This is similar to how we conceptualize leaderful mobilizations. But there are three reasons why we prefer not to use ‘spontaneous’ to define such mobilizations. First, spontaneity is inherent to protests. Even protests that are meticulously planned can exhibit a significant degree of spontaneity once they unfold. Protesters may deviate from predetermined plans or routes, or the situation may unexpectedly escalate due to police actions. Second, even though mass mobilizations are inherently unpredictable, structural conditions matter in conditioning their onset and outcomes. Grievances, previous acts of mobilization or political opportunities often create a structured set of choice for actors, contributing to what Mark Beissinger (Reference Beissinger2011) calls ‘structure of contingency’. Third, elements of planning are often present within what appears to be spontaneous mobilization. For example, they still schedule and coordinate actions so that protesters can assemble at the same location and time, even though they might not have a detailed plan beyond that point. However, while we do not define such mobilizations primarily as ‘spontaneous’, it is still accurate to say that spontaneity is intricately woven into the fabric of leaderful mobilizations. Despite some form of planning, the absence of centralized leadership implies that protesters continually adapt and improvise their tactics according to the situation.

The term ‘leaderful’ was originally developed in organizational studies to address the evolving needs of the workplace and the inadequacies of the traditional leadership model, which typically revolves around a single leader. Joseph Raelin (Reference Raelin2003, Reference Raelin2011) famously proposes the idea of creating ‘leaderful organizations’, in which he views leadership as a collective practice that can be distributed among members of an organization. The idea of leadership as a relational and distributive practice, rather than as a stable set of attributes inherent in individuals, has also been adopted by organizational scholars and applied to social movement settings (Sutherland, Land, and Böhm Reference Sutherland, Land and Böhm2014; Western Reference Western2014). Sasha Costanza-Chock (Reference Costanza-Chock2012: 9) described the Occupy Movement of the early 2010s as ‘leaderful’ to illustrate how any participant could learn to interact with the press without designating specific individuals as official spokespeople. While Costanza-Chock initially used the term in the context of the movement’s media strategies, it is a fitting characterization of the Occupy Movement as a whole. Occupy protests around the world often adopted decentralized and horizontal structures with no identifiable leaders, with participants engaging in direct democracy and collective decision-making through participatory assemblies (Juris Reference Juris2012; Smith and Glidden Reference Smith and Glidden2012). A similar characterization could be applied to the Arab Spring uprisings, which became a major inspiration for the Occupy Movement. These revolts saw civil society activists and ordinary citizens spontaneously mobilizing against authoritarian rulers without central leadership, relying heavily on social media for planning and coordination (Tufekci Reference Tufekci2017). To an extent, one could also regard the New Social Movements – especially the feminist and environmental movements of the 1970s – and the anti-globalization movement of the 1990s and 2000s, as having leaderful elements, given how they rejected formal leaders and adopted horizontal organizational structures (Freeman Reference Freeman1972).

In other words, leaderful mobilization is not a completely new political phenomenon. However, there is something distinct and intriguing about Hong Kong’s leaderful mobilization in 2019. First, even with an established political opposition, the Anti-Extradition Movement embraced a leaderful structure, eschewing the leadership of political parties and civil society organizations. This contrasts with other contexts where leaderful mobilizations emerged: in authoritarian regimes such as Egypt and Syria, the political opposition was either weak or non-existent before the uprisings. It was no surprise that protests had to take leaderful forms in such contests. But why did Hong Kong’s opposition organizations, which were highly institutionalized with strong mass support base, not take a leading role in the Anti-Extradition Movement? Second, unlike leaderful movements elsewhere that often involved occupying public spaces or setting up protest encampments, the Anti-Extradition Movement adopted a more fluid and adaptable approach. Embracing Bruce Lee’s ‘be water’ dictum, protest actions were spatially dispersed, eschewing fixed protest routes and avoiding the occupation of public spaces. Yet, even in the absence of a designated focal point – such as Tahrir Square during the Egyptian revolution, Zuccotti Park during Occupy Wall Street or Plaza Mayor during the Spanish Indignados Movement – the Anti-Extradition Movement maintained its resilience and carried on for more than six months. Furthermore, in contrast to other recent leaderful mobilizations that typically faced ‘tactical freeze’ (Tufekci Reference Tufekci2017), where protesters found it difficult to develop new strategies after the initial stages, the Anti-Extradition Movement saw continuous innovation of strategies and tactics. Moreover, protesters also managed to keep themselves organized and maintain a high level of cohesion despite the lack of traditional organizational structures.

Our Puzzle

Hong Kong’s evolution from organization-led and scripted protests to the 2019 Anti-Extradition Movement presents an intriguing puzzle. What contributed to the rise of leaderful mobilization in a semi-authoritarian context? What explains the waning influence of traditional organizations in its trajectory of mass mobilizations? What catalysed the transformation of ordinary citizens from passive adherents of established groups to engaged protesters motivated by convictions? How did this leaderful mobilization manage to attain such an extensive scale, sustain itself for months, and maintain a relative degree of organization and unity without centralized leadership or the occupation of a fixed space?

This book aims to explain the ascent of leaderful mobilization from a historically established paradigm of organization-led protests in post-handover Hong Kong. We present a theory of mediated threat to elucidate how perceived threats to civic freedoms and institutional autonomy gave rise to changing forms of mass mobilizations. Our central argument posits that threats do not instantly trigger protests; rather, they must be perceived and socially processed among citizens to spark mobilization. Different groups of citizens may perceive the same threat in different ways, resulting in a spectrum of mobilizational responses and the formation of new organizations, groups, and networks. This process gradually alters the relational dynamics of the opposition through which new threats are assessed, precipitating new mobilizing structures from which future mobilizations will arise, and ultimately altering their organizational forms.

Existing Explanations

Existing social movement theories offer useful insights for some of these questions, yet they seem inadequate for addressing our broader puzzle. In what follows, we will explore three strands of literature, each centred on a concept that has significantly influenced the study of contentious politics and provides some degree of analytical leverage for our case.

A Theory of Mediated Threat

To address these limitations, we present a theory of mediated threat that aims to explain the evolution of contentious politics and emphasizes how perceived threats mobilize political challengers while simultaneously influencing their relational dynamics. Our theory builds on the foundation of existing explanations, incorporating political opportunities, threats, social networks, and digital communication technologies as components. This approach is inspired by the seminal Dynamics of Contention (McAdam, Tarrow, and Tilly Reference McAdam, Tarrow and Tilly2001), which critiques the political process model for analysing political contexts and actors as given. The dynamics of contention framework proposes to identify the recurring causal mechanisms that constitute, in different combinations and sequences, processes of mobilization and demobilization, through which different episodes of contention emerge. Additionally, it also aims to devote more attention to agency and incorporate mechanisms, such as threat attribution and identity shift, into structural approaches. The theory of mediated threat follows the guidance of this framework. On the one hand, it seeks to delineate the process through which mass mobilizations in Hong Kong emerged episodically but became gradually less reliant on organizations. On the other hand, it also seeks to identify the mechanisms that constitute the process – such as the mechanisms that changed the mobilizing structure and those that sustained leaderful mobilization.

At the centre of our theory is the concept of threat. In the contentious politics literature, threat is primarily conceptualized in two ways. The first approach, outlined by Goldstone and Tilly (Reference Goldstone, Tilly, Aminzade, Goldstone, McAdam, Perry, Sewell, Tarrow and Tilly2001: 183), defines threat as the cost of action or inaction – the ‘costs that social groups will incur from protest or that it expects to suffer if it does not take action’. Based on rational choice theory, this definition portrays threats as negative rewards or what resource mobilization scholars refer to as negative selective incentives (Oliver Reference Oliver1980). Goldstone and Tilly further categorize threat into two types: ‘current threat’, which pertains to the costs incurred by not taking action to prevent the harms imposed by a regime; and ‘repressive threat’, which relates to the costs of facing repression when action is taken. Consequently, current threat can motivate individuals to engage in pre-emptive actions to avoid more severe outcomes, whereas repressive threat can deter mobilization that could otherwise be harmful. Individuals must then weigh these costs to determine the necessity of action. When the cost of inaction (current threats) outweighs the cost of action (repressive threats), collective action becomes more likely.

In contrast to the individual-level focus on incentives to join protests, Paul Almeida (Reference Almeida, Snow, Soul, Kriesi and McCammon2019) offers a structural approach. He suggests that despite a wealth of literature addressing structural political opportunities, such as elite conflict or institutional access, the concept of structural threats remains underdeveloped. Almeida defines structural threat as ‘negative conditions intensifying existing grievances and creating new ones in stimulating collective action’ (Reference Almeida, Snow, Soul, Kriesi and McCammon2019: 45). This definition differentiates threat from the well-established concept of grievances. In his view, grievances are pre-existing internal conditions experienced by aggrieved individuals, while threats denote external encroachments that aggravate these conditions or generate new ones. Almeida (Reference Almeida, Snow, Soul, Kriesi and McCammon2019) further identifies four forms of structural threats – (1) economic-related problems, (2) public health/environmental decline, (3) erosion of rights, and (4) state repression. He also demonstrates how they give rise to various types of mobilizations.

While these efforts have revived the concept of threat in contentious political research, the conceptualization of threat remains over-simplified. Viewing threats solely as costs implies an individualistic approach that considers protests as outcomes of individuals weighing costs and benefits. This individual-centric framework has been heavily critiqued by POS theorists. Moreover, it is unclear how individuals assign specific values to particular threats. This difficulty is especially evident in the evaluation of ‘current threats’ – gauging the cost of inaction is challenging since inaction tends to have a social rather than personal impact. Take, for instance, the case of Hong Kong’s extradition bill. Most ordinary citizens are unlikely to be directly affected by the possibility of extradition. They only perceive the bill as personally costly when they can relate its social consequences to their individual lives. This process of connecting the collective impact to the personal level is not automatic. It requires individuals to see themselves as part of the collective even if they are not personally under threat.

Viewing threats as structural is also problematic. The structural view assumes that threats are objectively negative conditions that automatically prompt individuals to mobilize in protest. This view, however, ignores how structural threats also need to be perceived and understood as such by social actors (Leenders and Heydemann Reference Leenders and Heydemann2012). This distinction mirrors the difference between structural and perceived opportunities. Focusing on the 1979 Iranian Revolution, Charles Kurzman (Reference Kurzman1996) argues that structure and perception do not always align. People might fail to perceive opportunities, and they might sometimes perceive opportunities wrongly even when they do not exist. For Kurzman, Iranians protested not solely because they observed the weakening of the state (structural opportunities) but because they perceived the opposition as stronger (perceived opportunities). Similarly, in discussing the structural bias of political opportunity, William Gamson and David Meyer (Reference Gamson, Meyer, McAdam, McCarthy and Zald1996: 276) assert that opportunities are ‘subject to framing processes and often serve as a source of internal disagreements within movements regarding appropriate action strategies’. This is particularly evident when political opportunities are volatile or involve ‘relative opportunities’ between institutional and extra-institutional actions.

Like opportunities, threats must be framed or socially constructed as harmful to spur mobilization. This process typically necessitates the mediation of the social structures within which actors are embedded, which provide them with the interpretive frameworks and cultural resources to comprehend threats and offer responses (Shesterinina Reference Shesterinina2016). To illustrate, take the example of climate change. The increasing average temperatures of our planet, sea level rise, and the increasing frequency of extreme weather events are typically seen as structural threats. However, these climate phenomena themselves may not compel individuals or societies to act. They must be socially constructed as threats, and this process requires the mediation of existing social structures. For example, academic institutions may contribute by conducting research that scientifically validates climate change as a threat. The media then disseminates this information to the public, framing climate change in a way that highlights its catastrophic impact on human societies and ecosystems, thereby inducing a sense of urgency and danger. On the other hand, there are also groups that perceive climate change not as a threat, but as a conspiracy. These groups, leveraging their own media outlets and influential figures, construct an alternative reality where climate change is seen as a false threat propagated by vested interests. This shows that social structures and their interpretive frameworks do not always lead to a consensus on what constitutes a threat but instead create divergent or opposing perceptions.

In the context of Hong Kong, the perception of the extradition bill as a threat to Hong Kong’s civic freedoms and autonomy did not emerge immediately. Rather, it required continuous learning and social interactions, as well as experiences with previous threats, for such a perception to coalesce among citizens. Moreover, it required persistent effort from opposition actors to frame the bill – initially seen as a legal matter – into a political issue with profound repercussions. In short, threats are not merely assessed individually or imposed structurally. They demand perception and mediation by social actors within their cultural context and relational configurations before they are viewed as harmful to well-being. This mediation process is crucial because it not only determines the possibility of protest mobilization but also shapes the organizational forms that the mobilization assumes.

Reconceptualizing Threats

Before outlining the threat mediation process, we first propose our own conceptualization of threats to bridge individualistic and structuralist views. We define threats as both the actual and potential harm directed at either institutions or individuals. This definition goes beyond our everyday understanding of threats – the potentiality to cause harm – to encompass the actual harm being inflicted. By considering both institutions and individuals as potential victims, we aim to reconcile individualistic and structuralist perspectives.

We distinguish threats along two dimensions. The first concerns the scope of the harm. Threats can be either generalized or particularistic, a distinction similar to the one between generalized and particularistic trust (Luo Reference Luo2005). Generalized threats pose harm to institutions, which encompass formal and informal rules and norms that organize social, political, and economic relations (North Reference North1990). Examples of institutions include government accountability, protection of social rights and civic freedoms, elections, rule of law, and due process. The term ‘generalized’ refers to the non-targeted nature of these threats. They inflict harm not on specific individuals, but on the wider institutional structures, which affect individuals indirectly and non-specifically. On the contrary, particularistic threats are those that inflict harm on selected individuals, based on their actions or attributes. This targeted nature makes them particularistic. These threats often manifest as targeted repressive actions such as harassment, surveillance, spying, bans from public office, arrests, torture, and mass killings (Davenport Reference Davenport2007).

The second dimension is the temporality of harm. Threats can manifest in different temporalities. Some could be long-term, inflicting slow and gradual harm to institutions or individuals. For example, state actors could gradually expand executive powers and erode legislative or constitutional oversights; and they could harass opposition activists and prevent them from obtaining any positions of political power. These acts do not happen overnight but are slow, protracted, and gradual. We refer to them as ‘recurrent threats’. Conversely, threats can also be imposed suddenly. State actors could introduce policies, legislations, or court rulings viewed as threatening to established institutions, or use coercive forces against individuals, such as violently repressing protesters or imprisoning activists. These threats occur in a much shorter time frame, often driven by events with a sense of urgency. Hence, we characterize them as ‘contingent threats’. Table 1.1 presents a typology of threats based on these two dimensions.

Our conceptualization seeks to amend the distinction proposed by Goldstone and Tilly (Reference Goldstone, Tilly, Aminzade, Goldstone, McAdam, Perry, Sewell, Tarrow and Tilly2001). As discussed, Goldstone and Tilly differentiate between current threats and repressive threats: current threats pertain to the ‘harms that are currently experienced or anticipated’ (184), while repressive threats refer to the harm posed by repression. Although logical within their definition of threat, it fails to encompass the full spectrum of threats. Both current and repressive threats can exist in the background or emerge afresh, and both can be directed at institutions and individuals. By distinguishing between temporality and scope, our conceptualization aims to provide a more nuanced categorization of the threats that society and people could encounter. Furthermore, by moving away from a simplistic cost-based conceptualization, we diverge from Goldstone and Tilly’s assertion that repressive threats deter people from mobilizing in protests. Indeed, extensive research has shown that repression can also increase mobilization (Francisco Reference Francisco1996; Khawaja Reference Khawaja1993; Moore Reference Moore2000; Olivier Reference Olivier1990; Rasler Reference Rasler1996; Schock Reference Schock1999).

It is important to differentiate between types of threats with greater nuance because not every type of threat has the potential to catalyse protest mobilizations. We expect that contingent threats, whether generalized or particularized, are more likely to spur mass mobilizations because they create shocks and surprises. Meanwhile, recurrent threats seldom spark mobilization directly because citizens are more used to them; however, they are still prone to the accumulation of grievances, which raises the likelihood that contingent threats in the future will spark mobilization. Furthermore, it is important to note that the boundaries between these four types of threats are not rigid. Contingent threats can evolve into recurrent threats over time, losing their initial sense of urgency. Moreover, contingent threats can bring attention to the existence of underlying recurrent threats. For instance, a crackdown on mass protests can expose the unchecked nature of police powers. While particularistic threats primarily target individuals, they can also indirectly damage norms and procedures. For example, unchecked police powers on individual cases can create precedents to undermine due process and the rule of law. Conversely, generalized threats can also harm individuals indirectly. For instance, barring opposition candidacy not only restricts the freedom to run for election but also involves the exclusion of specific candidates.

Relational Dynamics of the Opposition

After differentiating between various types of threats, our next step is to outline the threat mediation process through which threats result in mass mobilization. The previous section posits that contingent threats have greater mobilizing potential than recurrent threats. But how do contingent threats trigger mass mobilizations and shape different organizational forms? Figure 1.1 illustrates this process. A key component here is the relational dynamics of the political opposition, which conditions how threats are perceived, framed, and constructed. This component determines two outcomes: first, whether a threat can spark mobilization; and second, what kind of mobilizing structure from which protests will occur. By mobilizing structure, we refer to the ‘collective vehicles, informal as well as formal, through which people mobilize and engage in collective action’ (McAdam et al. Reference McAdam, McCarthy, Zald, McAdam, McCarthy and Zald1996: 3). But before turning to that, let us focus on two key factors that shape the opposition’s relational dynamics: (1) the relative strength of the institutionalized opposition to the state, and (2) the level of fractionalization within the opposition.

Figure 1.1 Mediation process of threat-induced mobilization.

Organizational Forms of Mass Mobilization

Figure 1.2 illustrates four ideal-types of organizational forms in a 2x2 typology. The vertical axis represents the strength of the institutionalized opposition, defined as the opposition’s capacity to contest political power. A strong opposition is more likely to trigger organization-led mobilization in response to a contingent threat, maximizing the opposition’s challenge to the incumbents. Conversely, when the opposition is weak, threat likely gives rise to mass mobilizations guided by networks or non-hierarchical structures. Meanwhile, the horizontal axis represents the level of fractionalization within the opposition. When fractionalization is low, implying that opposition groups are more unified, their perceived threats would align more closely, leading to a more cohesive mobilizing structure. However, when fractionalization is high, different opposition groups would perceive a threat differently and propose different tactical solutions, leading to a more fragmented mobilizing structure.

Figure 1.2 Relational dynamics of the opposition and the organizational forms of mobilizations.

‘Brokered mobilization’ arises when the institutionalized opposition is strong and there is minimal fractionalization among opposition groups. In this scenario, mass mobilization is orchestrated by a primary SMO or a coalition of SMOs acting as brokers. On the other hand, ‘factional mobilization’ occurs when a strong institutionalized opposition is marred by significant internal divisions. Here, multiple opposition organizations or SMOs lead and organize movements concurrently, each with its distinct goals and tactics. ‘Fragmented mobilization’ emerges when the institutionalized opposition is weak and fractionalization is pronounced. Under these conditions, networks, informal groups, and individuals engage in spontaneous mobilization without formal organization, yet with low levels of cooperation and cohesion. Such mobilization is characterized by multiple centres of authority and action, which may not necessarily align around a unified goal. Lastly, ‘leaderful mobilization’ unfolds when the opposition is weak but exhibits low levels of fractionalization. In this situation, networks, informal groups, and ordinary citizens in the opposition camp can become united in response to a contingent threat. As a result, they can mobilize spontaneously but cohesively without centralized leadership. Although multiple centres of authority and action exist, protesters in leaderful mobilizations always collaborate towards a shared objective. Figure 1.3 provides the graphical representation for each organizational form: circles, dots and lines represent organizations, actors, and ties respectively.

Figure 1.3 The four organizational forms of mobilization.

Threat-Induced Changes

It is important to note that the opposition’s relational dynamics is by no means a static structure. It can evolve over time. While contingent threats are more likely to trigger mass mobilizations, recurrent threats can continuously reshape opposition’s relational dynamics. First, recurrent threats can erode the strength of the institutionalized opposition. For example, executive aggrandizement and the erosion of legislative oversight present generalized threats that can isolate opposition figures or parties from political institutions, thereby weakening the opposition’s standing against the incumbent (Thompson Reference Thompson2021). Similarly, restrictions on resources or curtailment of civil society organizations affiliated with the opposition can undermine their organizational infrastructure, hindering their ability to effectively organize dissent (Cavatorta and Durac Reference Cavatorta and Durac2010; Wiktorowicz Reference Wiktorowicz2000). Second, recurrent threats can foster fractionalization within the opposition. Selective repression targeting radical opposition figures or parties can deepen the chasm between them and their moderate counterparts, heightening fractionalization within the opposition (Ash Reference Ash2015; Sika Reference Sika2023). Conversely, undifferentiated repression may have the inverse effect, consolidating previously fractionalized opposition groups due to the shared experience of widespread repression (Nugent Reference Nugent2020).

Third, recurrent threats can stimulate the development of new organizations, groups, or networks, providing fresh mobilizing structures. For example, during the rule of General Hosni Mubarak in Egypt, longstanding grievances against the political and economic status quo, combined with the perceived ineffectiveness of the institutionalized opposition, led to the rise of a new opposition movement named the Kefaya in the 2000s. This movement served as a loosely structured umbrella for various opposition groups and activists (Clarke Reference Clarke2011). However, the emergence of new formal or informal structures can sometimes lead to increased division within the opposition. We see this in conflict-ridden regions like Northern Ireland during the Troubles or present-day Palestine, where ongoing repressive threats from British and Israeli authorities respectively led to the creation of radical offshoots from the opposition, further fractionalizing it (Alimi, Bosi, and Demetriou Reference Alimi, Bosi and Demetriou2012; De Fazio Reference De Fazio2013). Still, regardless of the degree of fractionalization, the rise of threat-induced groups introduces new relational dynamics within the opposition. The threats that prompt their creation also shape their structures, goals, and tactical preferences, making them more responsive to future threats.

Although recurrent threats have the potential to reshape the opposition’s relational dynamics, the agency of opposition actors also plays a pivotal role. These actors can proactively establish new organizations, groups, or networks in response to threats. Without their willingness to collaborate and take risks, new mobilizing structures would not materialize. Additionally, opposition actors can strive to lessen intergroup fractionalization by setting aside their differences. While this is often challenging under normal circumstances, it becomes more achievable when the opposition faces immediate and critical threats, especially when these threats do not differentiate between opposition groups. Furthermore, opposition actors can endeavour to frame impending threats in a way that aligns their threat perceptions more closely. This framing can create collective identities based not on existing political divisions, but on more fundamental social identities, such as occupation. While fractionalization may not entirely vanish, it can be concealed or temporarily brushed aside.

Scope Conditions

Our analysis focuses on the evolution of contentious politics in post-handover Hong Kong, which had been governed by a hybrid regime. This hybrid regime is distinctive in its blend of liberal and authoritarian elements. Governed by the OCTS principle, the city’s historical legacy of civil liberties, coupled with the influence of capitalist elites, has fostered a degree of openness not found in mainland China’s one-party socialist rule (Fu Reference Fu2018). It is within this context that opposition movements and mass mobilizations had arisen. However, the structure of this regime was not predetermined; it evolved in response to shifting political circumstances and developed in tandem with contentious politics. Thus, our analysis adopts a within-case, temporal approach to trace the dynamics of contention from the late colonial period to the early post-handover years, spanning from the 1980s to the 2010s. Throughout this time, the hybrid regime’s institutional openness, civil liberties, and media freedoms provided a space for opposition and dissent, albeit within an increasingly constrained political environment under a more assertive state.

The scope conditions of our theory should thus be set within the context of non-democracies or democracies experiencing backsliding. It is primarily relevant to competitive authoritarian regimes that maintain a certain level of political openness, allowing for the presence of opposition, though not necessarily its flourishing. There are many examples of such countries: Egypt, Tunisia, Algeria, Ukraine, Thailand, Myanmar, Malaysia, and Sri Lanka during the 2010s. In these places, mass protests often emerged as political negotiations and electoral processes were increasingly viewed as insufficient for preventing electoral fraud, enabling power-sharing among diverse social groups, or rectifying adverse injustice (Weiss and Aspinall Reference Weiss and Aspinall2012; Sidel Reference Sidel2021; Ketchley Reference Ketchley2017; Moss Reference Moss2022). Moreover, our theory could also be applicable to backsliding democracies, where recurring threats are undermining the representativeness and accountability of political institutions, thereby diminishing their capacity to respond to those threats effectively. Some recent examples include Hungary, Poland, Chile, India, Indonesia, and the Philippines, where debates surrounding immigration, identity, religion, and environmental sustainability have heightened politicization, undermining trust in established elites and respect for formal institutions.

Our theory is less applicable to consolidated and functional democracies, in which power can alternate between incumbents and the opposition, and where threats are typically managed through the judiciary, electoral competitions, or civic actions (Vanessa et al. 2021). In these regimes, the interplay between incumbent and opposition would allow the energies from the streets to be channelled into formal institutions. However, even in consolidated democracies, mass mobilizations do happen occasionally to counteract perceived threats. For example, Taiwan’s Sunflower Movement of 2014 and South Korea’s candlelight protests of 2016–2017 showcased how even democracies could experience threat-induced mobilizations (Ho Reference Ho2018; Jung Reference Jung2023). In the United States, despite having strong democratic institutions and an independent judiciary, threats associated with immigration, abortion, identity, and electoral procedures have sometimes triggered massive protests where rival political groups are pitted against one another (Tarrow Reference Tarrow2021). Thus, our theory can shed light on such contexts, even though they have better institutional systems to address popular demands. Finally, our theory also does not apply to full-fledged autocracies, where there are scant political opportunities for intermediate associations to form, latent networks to develop, or leaderful structures to emerge. Nonetheless, some aspects of leaderful mobilization resonate with the Belarus protests in 2019–2020, the anti-Covid lockdown protests in mainland China in 2022, and the Iranian protests in 2021–2022.

The case of Hong Kong is distinguished by its liberal oligarchy structure, which does not conform to the conventional models of competitive authoritarianism or backsliding democracies. Nonetheless, it does share elements from both types of regimes, as evident in its power dynamics and contentious politics. Thus, despite being a single-case study, Hong Kong’s distinctive political context, coupled with its changes along the spectrum between a backsliding democracy and a competitive authoritarian regime, enables us to conduct within-case comparisons and engage in inductive theory building (George and Bennett Reference George and Bennett2005; Pepinsky Reference Pepinsky2019).

Case Description

To showcase the analytical power of our theory, let us place it in the context of Hong Kong. In the early post-handover years, the city has faced various recurrent and contingent threats across a spectrum of issues, including constitutional reforms, introduction of new legislations, plans for infrastructure development, and the impact of inbound tourism. Some of these issues originated domestically, while many others were a result of efforts to promote economic and political integration with mainland China. From the early 2000s to the latter half of the decade, the democratic opposition was relatively strong and cohesive. Pro-democracy parties and organizations adopted a two-pronged approach to counteract contingent threats. On the one hand, they leveraged their institutional power in the legislature and other advisory bodies to bargain with the authorities. On the other hand, they organized mass protests from time to time that were often well-ordered and ritualistic to showcase the popular support that they could mobilize. In such protests, these established organizations often formed the mobilizing structure and served as its leadership, providing both strategic guidance and logistical support. Mass mobilizations, in this sense, could be said to be ‘brokered’ by these organizations.

The anti-government demonstration on 1 July 2003, which mobilized half a million citizens, serves as a perfect example of what we call ‘brokered mobilization’. The demonstration was sparked by a national security legislation, known as Article 23, which was framed by the opposition as a dire threat to the city’s civic and political freedoms (Ma Reference Ma2005). Threat perceptions among opposition groups were promptly coalesced under two demands: the retraction of the national security legislation and the resignation of the then Chief Executive, Tung Chee-hwa. With the presence of a robust institutionalized opposition, these groups united within a coalition spearheaded by the newly established CHRF. The coalition eventually orchestrated a centrally coordinated, meticulously scripted demonstration in which the CHRF was responsible for everything from the claims to picketing arrangements. Protesters simply attended the rally, adhering to the prearranged plan.

Entering the late 2000s, while contingent threats like Article 23 legislation diminished, recurrent threats persisted. These recurrent threats were arising from issues such as urban redevelopment and economic integration with the mainland. Viewing the established opposition as incapable of counteracting these threats, new opposition actors and groups began to emerge outside of the institutionalized opposition, advocating a more confrontational approach that gave emphasis to street protests. These actors, mostly consisting of young activists, gained inspiration from the anti-globalization movements abroad and promoted more spontaneous and disruptive forms of mobilizations. Meanwhile, the democratic opposition began to split. The emergence of new opposition parties, such as the Civic Party and the League of Social Democrats, led to intense competition for votes, ultimately eroding opposition unity. On the other hand, pro-government parties and organizations became increasingly strengthened under Beijing’s intensified state-building efforts after the 2003 mass rally, gradually eroding the overall institutional power of the opposition.

What occurred during this phase was an amalgam of mass protests that coexisted with contrasting forms – what we term ‘factional mobilizations’. The traditional opposition carried on with their routine mass protests, such as the annual July 1 rallies and the June 4 candlelight vigils, even as turnouts dwindled. Despite continuing to call for a faster pace of democratization, these routine protests failed to address the emerging threats that citizens were facing. In response to these threats, new opposition groups staged protests that adopted more decentralized and spontaneous forms, such as the heritage preservation campaigns between 2006 and 2008, the Anti-Express Rail Link (Anti-XRL) Movement of 2009, and the Anti-Moral and National Education (Anti-MNE) Movement of 2012. They shared a similar group of participants as the routine protests, but these participants became much more involved in the mobilization process rather than simply show up as an attendee. Also, although conventional pro-democracy parties and organizations remained relevant in supporting these protests, they were often marginalized from the leadership core.

As the state continued to erode the institutional bargaining power of the opposition, pro-democracy supporters increasingly questioned its effectiveness as the primary broker for mass mobilization. Meanwhile, the rise of new opposition groups and networks, which differed in their perception of threats from the traditional democrats, further fragmented the opposition, leading to ongoing internal conflicts. This set the stage for ‘fragmented mobilization’, characterized by loosely organized, spontaneous protests that were often regarded as ‘leaderless’. These movements were ‘fragmented’ because the opposition groups often disagreed on strategies to counteract perceived threats. One major point of contention was the role of political institutions: while moderates still considered parliamentary politics and electoral competition as viable means to counteract threat, radicals found these institutional approaches inadequate and leaned towards more assertive street mobilizations. The latter approach was promoted by a burgeoning faction within the pro-democracy camp, known as the localists. They viewed the increasing political and economic integration with the mainland – exemplified by the surge of Chinese tourists and immigrants – as a recurrent threat to Hong Kong’s local identity. For the localists, the only recourse was to push for greater autonomy or even independence from China, employing more militant protest tactics to confront these challenges.

The anti-mainlandization protests that occurred between 2012 and 2014 were the precursors of ‘fragmented mobilization’, when localist protesters organized repeated actions against mainland tourists and parallel traders. But the Umbrella Movement of 2014 stood out as a perfect illustration of this mobilization form. During the movement, pro-democracy protesters occupied several city areas and called for universal suffrage. The movement’s organizational structure was unique: it was led by a diverse coalition of civil society groups, traditional SMOs, and pro-democracy parties, while day-to-day activities were self-managed by regular protesters. Despite initial unity among the protesters in response to the police’s use of tear gas, the movement saw a growing divide in threat perception across different groups and factions, notably between the localists and the traditional democrats. The traditional democrats were concerned that sustained, widespread occupation might trigger a violent crackdown by the authorities, whereas radical protesters viewed the occupation and the adoption of militant tactics as potent symbols of defiance against the state. The radical stance gained traction among the young student leaders, who felt pressured to heed the demands of the more committed occupiers. Such dynamics led to a split within the movement’s formal leadership, crippling the pro-democracy movement as a whole.

This does not imply that the opposition remained perpetually fragmented. In the years following the Umbrella Movement, opposition groups began to see a greater alignment in their threat perception, spurred on by Beijing’s state-building efforts (Lee Reference Lee2025). These efforts were coupled with indiscriminate crackdown on pro-democracy activists, which shifted from targeting just the radicals to increasingly affecting the moderates as well. This shift led to a significant convergence in the threat perceptions among opposition groups, especially between localists and traditional democrats.

The proposed extradition bill in 2019 proved to be the tipping point that unified these factions. Although the bill initially garnered little attention, it soon mobilized significant opposition once the ‘early risers’, including activists and pro-democracy politicians, portrayed it as a severe threat to Hong Kong’s political and legal autonomy. A massive online petition campaign in May 2019 further united a diverse array of alumni, professional, recreational, religious, and community networks against the bill. These threat attribution mechanisms synchronized the perceived threats among citizens, galvanizing numerous existing and newly established groups and networks. This gave rise to a leaderful mobilizing structure that encompassed SMOs, networks, and individual citizens, but without a singular figure or entity taking the lead. Leveraging both digital communication technologies and offline networks, informal leaders and protesters engaged in what we describe as ‘peer collaboration’ despite the absence of centralized leadership. This overarching mechanism encompasses a range of micro-mechanisms, including tactical innovation, information curation, task coordination, and discursive negotiation – all of which were crucial in sustaining mobilization. Figure 1.4 outlines the relationship between different actors and the key mechanisms of ‘leaderful mobilization’.

Figure 1.4 Mechanisms of leaderful mobilization

To summarize, the theory of mediated threat aims to illuminate the dynamics of contention that brought Hong Kong to the moment of leaderful mobilization. Instead of viewing threats as objective realities imposed externally, our theory conceptualizes them as socially constructed, requiring perception and interpretation by oppositional actors through specific cultural or political lenses. Moreover, threats do not merely trigger mobilizations; they also continually reshape the relational dynamics in the democratic opposition. This process gives rise to new forms of political agency and influences the mobilization strategies used by both actors and supporters involved in protests. In other words, threats shape not only the likelihood of mobilization, but also its organizational form. It is this dynamic process that contributed to the eventual rise of leaderful mobilization.

Theoretical and Empirical Contributions

This book aims to contribute to three spheres of knowledge: the conceptualization of threats; decentralized and spontaneous mobilizations; and hybrid regimes. First, our theory of mediated threat seeks to enrich how threats are conceptualized in the study of contentious politics by offering a more nuanced categorization of threats. While extant literature only offers broad categorizations of threats (Almeida Reference Almeida, Snow, Soul, Kriesi and McCammon2019; Goldstone and Tilly Reference Goldstone, Tilly, Aminzade, Goldstone, McAdam, Perry, Sewell, Tarrow and Tilly2001), we distinguish different types of threat based on their target (institutions or individuals) and temporality (recurrent or immediate) – a typology that we will elucidate in Chapter 2. This typology allows us to go beyond the predominant focus of the literature on the impact of immediate repressive threats on mobilization (Francisco Reference Francisco1995; Hess and Martin Reference Hess and Martin2006; Lichbach Reference Lichbach1987; Opp and Roehl Reference Opp and Roehl1990; Rasler Reference Rasler1996; Steinert-Threlkeld, Chan, and Joo Reference Steinert-Threlkeld, Chan and Joo2022) and to also concentrate on non-repressive and longer-term threats. Moreover, we focus on the process through which threats are mediated before inciting potential mobilizations. Rather than presuming a direct mobilizing effect of threats, we consider how threats are perceived and interpreted by political actors, altering the relational dynamics of the opposition by fostering new political agencies and the emergence of new groups and networks. Lastly, while existing theories usually focus on how the source of discontent triggers the occurrence of mobilization, our theory further links it with the organizational form of the mobilization – not just in terms of its structure, but also its participatory nature. This bridges the gap between why people mobilize and how they mobilize – a connection that has been under-theorized in the current literature (Simmons Reference Simmons2016).

Second, our findings enrich the knowledge of decentralized, networked, and spontaneous mobilizations, which have become prevalent worldwide over the past decade. By conceptualizing them as leaderful rather than leaderless, we introduce a perspective that illuminates their inherent qualities, instead of defining them by what they lack. Moreover, while existing research often concentrates on the emergence and initial phases of mobilization (Abrams Reference Abrams2023; Castells Reference Castells2012; Tufekci Reference Tufekci2017; Pearlman Reference Pearlman2021), our work delves into how such movements sustain themselves and maintain unity without central leadership, and how they innovate tactics and coordinate actions. Also, moving beyond the conventional focus on digital communication technologies, we stress the interplay between online and offline realms, examining how protesters leverage digital technologies and real-world social networks. Our analysis transcends a simplistic parallel view of online and offline networks, instead emphasizing their interplay, which shapes the overall organizational landscape of leaderful mobilizations. Furthermore, we explore the impact of contingent events and the absence of centralized leadership on protest radicalization, an aspect that existing scholarship has not thoroughly investigated. Our findings suggest that solidarity played a critical role in both propelling and restraining radicalization. This shows that although the dynamics of leaderful mobilizations are inherently unpredictable, they can generate endogenous mechanisms to regulate their development.

Third, our book contributes to the growing literature on hybrid regimes by examining how mass mobilizations can alter the political dynamics within such systems. Existing scholarship primarily focuses on how autocrats preserve their power by selectively incorporating democratic elements, which ostensibly creates a ‘balance’ that bolsters their rule (Gandhi Reference Gandhi2008; Levitsky and Way Reference Levitsky and Way2010). Our findings from Hong Kong illustrate the inherent instability of such a ‘balance’ (Carothers Reference Carothers2018; Dresden and Howard Reference Dresden and Howard2016). While Hong Kong’s liberal oligarchy, which limited citizens’ suffrage but allowed a degree of contestation through alternate channels, maintained a delicate power balance between the ruling coalition and pro-democracy opposition, escalating mass protests gradually unsettled this balance. This prompted the Chinese party-state – the de facto patron of the liberal oligarchy – to expedite state-building efforts in the semi-autonomous region. Paradoxically, these state-building measures only served to further amplify mass protests and deepen the distrust of conventional pro-democracy organizations among their supporters, thereby further destabilizing the balance. Our findings thus emphasize the crucial role of contentious politics in disrupting and reshaping the dynamics of hybrid regimes. Even if elite and opposition actors may not intend to alter the balance, their interactions, coupled with mass mobilizations, could result in unforeseen and unintended consequences, undermining the hybridity that underpins hybrid regimes.

Methods and Data

This book is the result of eight years of extensive and on-the-ground research. Our journey began with the 2014 Umbrella Movement, which sparked our curiosity to delve into the complex dynamics of Hong Kong’s contentious politics. Initially, our focus was primarily on individual episodes of protest. However, as we witnessed successive events unfold, we recognized the limitations of this episodic approach and pivoted towards a broader examination of the spaces between these episodes and the historical contexts that gave rise to them. Our research scope expanded to include the evolution of the political landscape, mobilizing structures, and modes of organizing in post-handover Hong Kong. Adopting a long-term perspective on these dynamics (McAdam, Tarrow, and Tilly Reference McAdam, Tarrow and Tilly2001), we shifted our analysis from focusing on the visible elements of social movements, such as prominent organizations and individual activists, towards examining the relational dynamics between various types of challengers, powerholders, and bystanders.

We employed a mixed method approach in this book, utilizing multiple types of data to address our research question. Here, ‘mixed method’ does not merely denote the juxtaposition of quantitative and qualitative data. It emphasizes the potential of both to complement and reinforce one another, thereby enhancing the overall validity and reliability of our findings (Ivankova and Wingo Reference Ivankova and Wingo2018). This approach allows us to explore diverse perspectives on the same issue, leading to a more holistic understanding. Our data sources include onsite protest surveys, in-depth interviews with key individuals, participant observations, and social media data. We also used secondary data, such as official statistics, newspaper articles, video footage, and other protest-related records. In the following section, we will provide an overview of our data sources and illustrate how their integration enabled a comprehensive study of contentious politics in Hong Kong and its broader implications.

Roadmap

The rest of the book is structured into 10 chapters. Chapter 2 analyses the formation of Hong Kong’s entrenched liberal oligarchy from a historical-institutional perspective. We trace the emergence of a tripartite coalition consisting of the Chinese party-state, civil servants, and business elites. We also delve into the complex dynamics between this coalition and the burgeoning opposition, examining how protests have been managed and contained since the 1980s, until shortly after the handover. Chapter 3 examines the emergence and evolution of a new cycle of contention during the mid-2000s. We highlight how the deepening threat perceptions resulting from the regime’s state-building advances spurred mass mobilizations. Meanwhile, we underscore how the creation of new civil society groups and how the normalization of new repertoires of contention contributed to changes in the episodes and modes of protest mobilization.

Chapter 4 explores the dynamics and consequences of the 2014 Umbrella Movement. We illustrate how its spontaneous eruption led to a hybrid organizational structure comprising a formal leadership and self-mobilized protesters who saw themselves as leaderless, and how such dynamics eventually led to the division and fragmentation of the opposition. Chapter 5 elucidates how the 2019 Anti-Extradition Movement erupted despite the lack of political opportunities in the post-Umbrella period. We demonstrate how abeyance networks from previous mobilizations and an online petition campaign transformed the idea of extradition into a widely perceived existential threat, galvanizing popular support for the movement and leading to a leaderful mobilization.

Chapter 6 looks at how sectoral networks propelled the movement. By focusing on religious groups, legal professionals, and medical practitioners, we demonstrate how social identities informed protesters what roles to take up during a leaderful movement, and how these sectoral networks provided expertise and resources to facilitate the movement’s organizing efforts. Chapter 7 examines how secondary school action groups, established by students in their respective schools, played a crucial role in mobilizing teenagers against the extradition bill by tapping into and leveraging their latent social capital. Utilizing Instagram as a platform, these groups facilitated connections among students within and across schools, often by capitalizing on their schools’ identities and leveraging various sources of social capital tied to those identities. This enabled loose and fragmented social networks can be mobilized in social movements, provided that they can activate their latent social capital.

Chapter 8 explains how the Anti-Extradition Movement was organized and sustained in the absence of a centralized leadership. We illustrate how protesters, networked in both online and offline realms, collaborated on a spontaneous, horizontal, and many-to-many basis to generate a continuous stream of actions – a mechanism that we term ‘peer collaboration’. Chapter 9 investigates the emergence of political consumption as a protest tactic in the later stage of the movement. We examine how this innovative tactic, encompassing boycotting and buycotting, emerged by utilizing the market logic. We also highlight the significance of political consumption as a movement consequence. Chapter 10 delves into the process of tactical radicalization observed in the 2019 movement. By adopting a relational approach, we analyse radicalization as a result of dynamic interactions across multiple arenas. We explore how discursive negotiations among protesters served as both the driving force and the limit to radicalization. It induced moderate protesters to extend tacit support to their more militant counterparts while acting as a restraint mechanism to curtail excessive measures.

Chapter 11 discusses the implications of Hong Kong’s contentious politics within the global context of democratic backsliding and leaderful mobilizations. We highlight the contributions of our theoretical framework and the implications of Hong Kong’s contentious pathways for hybrid regimes and beyond.