Bystanders,protesters,journalists: A qualitative examination of different stakeholders’ motivations to participate in collective action

Supporters/protesters

Analysis of motivations to engage in collective action has primarily focused on supporters and protesters. A rich body of research has built on foundational work by Klanderman's (1997), demonstrating the importance of collective identity, collective efficacy, and perceptions of injustice in predicting collective-action intentions and behaviors. Extensions of this work suggest that individuals who engage in protest are motivated by shared identity, emotions, efficacy, and moral conviction through different pathways (Tausch et al., 2011; Van Zomeren, 2013; Van Zomeren et al., 2008). A common theme throughout the theoretical examination of supporters’ collective-action intentions is the importance of collective identity in motivating initial collective-action participation (Thomas et al., 2020), as well as differentiating initial and sustained participation (Van Zomeren et al., 2008). Collective identity occurs when individuals define themselves by various social categories (Tajfel et al., 1979). For Klandermans (1997), the sharing of a contentious collective identity drives individuals to move from consensus formation to consensus mobilization. Indeed, research indicates that frequent protesters are more likely to have high politicized collective identity perceptions (Turner-Zwinkels & van Zomeren, 2021).

Protesters may also be driven by efficacy perceptions; that is, they believe that their participation in collective action will enhance the likelihood of that action leading to success (Hornsey et al., 2006). The act of participation itself may also increase collective efficacy perceptions (Dixon et al., 2016), whether directly or through reevaluations of the collective-action goal (e.g., see Batel & Castro, 2015; Gulliver et al., 2021; Hornsey et al., 2006). The presence of violence in the collective-action context may also alter the influence of collective efficacy. For example, in some studies, higher collective efficacy has been positively associated with engagement in normative collective action, but negatively associated with non-normative actions (Tausch et al., 2011; Van Zomeren et al., 2013), whereas other studies have found higher collective efficacy perceptions may increase support for violent action (Setiawan et al., 2020). Thus, the role of efficacy appears complex, dependent on participants’ goals, levels of past participation and other factors, such as participants’ collective identity and belief in the fairness of their political systems (Gulliver et al., 2021; Louis et al., 2021; Scafuto & La Barbera, 2016; Setiawan et al., 2020).

In addition to these variables, a large body of research has demonstrated the contribution of emotions, injustice, and moral conviction (Thomas et al., 2009; Van Zomeren, 2013). For example, unjust treatment can elicit feelings of empathetic anger and moral outrage encouraging collective action to redress those injustices (Thomas & McGarty, 2009).

As demonstrated, the predictive utility of predictors of engagement in collective action have been demonstrated in past literature. In this study, we aimed to explore the ways in which these drivers manifest in the unique context of Hong Kong. This context presents key elements that may influence these drivers: a strong Hong Kong identity juxtaposed against the mainland Chinese identity (Au, 2017), the presence of escalating violence from different actors (Wang et al., 2020), and high issue salience across the community, where both bystanders and journalists were aware of the conflict and thus able to reflect upon their own participation or nonparticipation choices. For these reasons, the social unrest in Hong Kong provides a unique opportunity to examine and contextualize collective-action drivers amongst collective-action participants and two comparative stakeholders: bystanders and journalists.

Bystanders

Bystanders are individuals who are not actively participating in protests (Cristancho et al., 2019). While often operationalized as individuals who directly witness or are present at a situation of interest (Nelson et al., 2011), in this study we follow Saab et al.’s (2015) definition where bystanders are those who are neither direct targets nor perpetrators of group-based injustices. In the context of Hong Kong, bystanders are therefore individuals who are aware of the protest but chose to not engage in them in any way, neither in person via rallies or protests, nor through participating in online collective action.

Bystanders play a critical role in sustaining and legitimating social movements as well as influencing the outcomes social movements achieve. Given that movements continually seek to grow, attracting bystander attention and converting this to action is often a key goal for advocacy groups (Gulliver et al., 2021). Bystanders also exert considerable influence on public opinion, to which policymakers can be responsive (Andrews et al., 2010; Burstein, 2003). Democratically elected politicians are more likely to carry out public demands when media attention heightens bystanders’ perspectives (Burstein, 2003), with nonviolent protest more likely to enhance popular support for a movement (Wasow, 2020).

There is a growing awareness of the importance of bystander responses to collective action. Some research indicates that bystander responses to collective action can be influenced by low motivation to take action (Montagno et al., 2021), uncertainty about government and protesters (Shadmehr & Boleslavsky, 2015), and low efficacy perceptions (Kern et al., 2015). Other studies highlight the importance of shared identities and values (Scafuto & La Barbera, 2016; Subašić et al., 2008).

Early work by Klandermans (1988) hypothesized that consensus, or ideological support, is an important factor in converting bystanders into activists. Montagno et al. (2021), for example, found that lack of motivation or disagreement with the perceived ideology or values of the activist group can also act as a barrier to action (see also Hartley et al., 2016). Similarly, Shadmehr and Boleslavsky (2015) examined the role of issue consensus in Hong Kong, finding that bystanders’ prior beliefs and uncertainty about government and protesters precluded their engagement in protest. Conversely, system-justifying ideologies can predispose people to not recognize injustice and decrease support for social change protest (Becker & Wright, 2011; Selvanathan & Lickel, 2019). In this study, we examined the extent to which bystanders demonstrated issue consensus regarding the pro-democracy cause and compared this against protester and journalist perceptions.

One factor that may have particular relevance regarding the extent to which bystanders may express ideological support for a cause is the presence or perceived likelihood of violence. In the main, normative, nonviolent collective actions receive more support from the general public than violent, non-normative actions (Thomas & Louis, 2014; Wasow, 2020). However, the influence of violence on bystander responses remains complex. Research in the Hong Kong context suggests that there are a range of psychosocial factors that influence attitudes toward violence (Chan, 2021). Studies of bystander participation in protests in repressive contexts demonstrate that police violence can in fact prompt strong emotional responses, which then stoke increased protest participation (Aytaç et al., 2017). Stott and colleagues(2021) found that coercive policing prompted identity change, empowerment, and radicalization in protesters; such changes may also occur in bystanders who witness police violence. Given the escalating incidence of violence in Hong Kong over the study period, particularly violence instigated by police, it may be that bystanders in this context express relatively greater issue consensus with the protesters. In this study, we explored how bystanders’ response to observed violence influenced the extent to which they share issue consensus with protesters.

Journalists

The Hong Kong pro-democracy protests generated substantial media coverage around the globe. Research has demonstrated how media representations play an important role in shaping public opinion (Andrews & Caren, 2010), creating momentum, and attention for protest (Wasow, 2020). Media play an important role in conveying messages of solidarity between protesters and the wider community, potentially increasing public support for the cause (Lee, 2020). However, as highlighted by Aytaç et al. (2017), ruling authorities often construct counternarratives seeking to justify their acts, such as labeling protesters as looters or claiming foreign influence and plots. A recent study (Wang & Ma, 2021), for example, demonstrates how media framing of the pro-democracy movement ranged from illegal riots (China Daily) to a fight against China's control (The New York Times). Perceptions around preferred frames can generate conflict for journalists, who individually help shape and frame movement-related discourse, while they and the platforms on which they share their work act as both gatekeepers of movement messages as well as actors in their own right (Gamson, 2004; Chan, 2014).

As well as navigating the challenge of balancing protesters’ and authorities’ movement-related narratives and perspectives, journalists may also experience pressure in their coverage of the cause. For example, activists often seek to obtain supportive media coverage by engaging with the identities, goals, and interests of media and other parties. Lee (2008) refers to this as protesters’ public relations, where protests are designed to attract reporters’ attention such as through undertaking dramatic newsworthy actions (see also Lee, 2014). However, journalists’ professional values require objectivity; a stance that has been shown in qualitative studies with journalists to be difficult to maintain (Shultziner & Shoshan, 2018; Tenenboim-Weinblatt, 2014). The influence of context on journalists’ objectivity has been shown in a range of studies. For example, in some authoritarian contexts, journalists have contested power in different ways, whether through maintaining opposition via anti-establishment reporting or working within the system (Repnikova, 2018). In the Hong Kong context, the anti-ELAB protests were notable for the increased use of social media, particularly live streaming (Fang & Cheng, 2022). These factors may all influence how journalists report on social unrest, as well their own engagement in the social unrest itself.

A range of potential factors have been shown to influence journalists’ reporting, many of which hinder journalists’ ability to remain objective. For example, journalists’ direct engagement with protesters can build a sense of common identity or chain of trust (Louis et al., 2020; Shultziner & Shoshan, 2018), which may manifest through news decisions (Patterson & Donsbagh, 1996), or overt participation in collective action or other forms of support for movement actors or their opponents (Shultziner & Shoshan, 2018; Tenenboim-Weinblatt, 2014). Some studies have also demonstrated that journalists may also engage in collective action itself through the course of their regular reporting, labeled by some as journalistic activism (Shultziner & Shoshan, 2018) or critical journalism (Repnikova, 2018). These forms of participation can convey alternative discourses from official state media, writing favorable opinion columns, assisting protesters outside of work hours, and coproduction of movement materials. These behaviors were argued by Shultziner & Shoshan (2018) to be strongly connected to journalists’ identification with the cause.

Journalists’ attitudes and behaviors not only reflect their own professional ethic of objectivity but can also influence those of bystanders. Journalists play a critical role in framing the perpetrators and victims of violence, which can influence bystanders and powerholders’ attitudes and responses (Wasow, 2020). For example, Luqiu (2021) noted that Hong Kong state media outlets criticized reporters for focusing negative coverage at police but not citizens. However, Lee's (2014) analysis of Hong Kong protest coverage indicated that journalists were likely to include comments from protest targets. More than one-third of articles included a response from a protest target, while articles on radical protest de-emphasized protesters’ voices. Given the complex yet important role of journalists in both experiencing collective action and shaping its narrative content for others, it is all the more striking that, to our knowledge, little social psychological analysis of journalists’ responses to protest has been conducted.

The present research

In this study we sought to expand the reviewed literature through exploratory qualitative analysis. We aimed to shed light on the psychological drivers of engagement in collective action amongst different three cohorts (protesters, bystanders, and journalists) who experienced highly turbulent times during protests in Hong Kong. We examined participation motivations across the three groups, starting with the 2014 Umbrella Movement and the subsequent 2019 Anti-Extradition Law Amendment Movement (Anti-ELAB). This period coincided with escalating political tensions. By 2019, mass mobilization expanded in response to a proposal enabling extradition of suspected criminals to mainland China. This Anti-ELAB movement represented a continuation of the largely nonviolent 2014 Umbrella Movement (Yuen & Cheng, 2017) but included increasingly violent confrontations with police (Ng & Kennedy, 2019). This volatile and complex context provides a unique opportunity to examine how different cohorts within the movement arena view political activism against perceived injustice, and the factors that influence their own engagement in it.

We undertook this analysis with three sets of qualitative data, enabling an exploratory, nuanced examination of individual and group-level factors associated with social justice orientations and actions at different time points across the prolonged period of social unrest in Hong Kong. This approach addresses Dixon and McKeown's (2021) urgent call for more in-depth qualitative studies of intergroup contact (see also Cristancho et al., 2019). Our comparative analysis of participation began through an analysis of the extent to which SIMCA variables align with protester motivations. We then compared these motivations to those of bystanders and journalists. Our analysis contributes to the literature in three ways. First, we examined differences between these groups to disentangle the influence of known predictors of collective-action engagement across distinct groups within an arena, seeking to identify shared motivations and perspectives on unrest generally, rather than on a specific incidence of unrest. Second, we analyzed differences between bystanders, protesters, and journalists to identify how motivators of engagement are linked to the different goals each group may have (Gulliver et al., 2021). Third, we examined the extent to which protest repertoires, especially violent collective action, influence attitudes toward movements seeking to challenge injustice.

Participants

We sought to interview different groups across the three-year period of social unrest in order to examine the underlying motivations regarding protest participation rather than focus specifically on participants’ views regarding one particular moment of social unrest. Subsequently, three sets of interviews were undertaken between 2018 and 2020, one for each of the groups of interest. Bystander data were collected in 2018, enabling participants to reflect on the past four years of escalating unrest. Protester data were collected through field interviews held during the June–July 2019 mass mobilizations, and journalist data were collected in late 2019 to 2020 to enable participants to reflect on their experiences of reporting on five years of social unrest. All interviewees were provided with information regarding the study and gave informed consent. Participants were debriefed and thanked after each interview.

In the protester group, 35 interviews were conducted by two researchers. Fourteen interviews were undertaken at a large-scale march on June 16, and 14 were undertaken at a blockade occurring on June 21, 2019. A further seven interviews were undertaken during the Legislative Council building protest on July 1, 2019. At each protest, the interviewer positioned themselves at the periphery of the protest and directly approached protest participants as they passed to request an interview. Each interview asked five specific questions (Supplementary Table 1). Given the volatile and stressful context in which they occurred as protesters were occupying streets and engaging in mass rallies, interviews did not deviate from the prepared questions, and were completed in around ten minutes to enable participants to continue their protest engagement.

The bystander group consisted of ten university students recruited via email and printed advertisements. Interviews were undertaken in November 2018 and took approximately 40–60 minutes. Participants were asked questions regarding their participation and perspectives on social movements since 2014. One participant had actively participated in protests during this period and was thus excluded from the bystander study population. All other participants were aware of the protests but were neither targets nor perpetrators of group-based injustice, nor did they engage in online or offline collective behaviors regarding the protests, thus qualifying them as bystanders for this study (Saab et al., 2015.

Seven journalists were interviewed from November of 2019 to January of 2020, as social unrest peaked and garnered international attention. Participants were recruited through the authors’ networks, followed by snowball sampling. Journalists interviewed worked on a range of print and online forms of media and engaged in activities ranging from live-streaming protests to compiling media updates. All respondents had reporting experience covering the social movement, and all but one had over five years of experience in their journalism roles. The interviews each took approximately 40–60 min.

Interview timing, location, length, and participants were selected to align with our research aims. Interviews sought to examine individuals’ motivations regarding collective action in general, rather than on the specific issues of contention at the time. As a result, interviews were timed and held in a location that best suited gathering insights of salience to that group at the time. Bystanders were interviewed at length in neutral meeting rooms, during a period after the Umbrella Movement protests but prior to the Anti-ELAB protests. Similarly, journalist participant interviews were timed to provide a deeper discussion of their motivations and perspectives on collective action as the Anti-ELAB protests escalated. In contrast, protester interviews were undertaken directly in the field during a series of large-scale protests. Participants were randomly selected and the interviews were short and succinct, all necessitated by the chaotic context in which they occurred (see also Maguire et al., 2020). A benefit of the field-based data collected from participants was a larger sample (n = 35) in comparison with that of bystanders (n = 9) and journalists (n = 7). Furthermore, most data on protester motivations are obtained outside of the protest context, whereas our field interviews allowed us to gain insight of what people felt at the particular moment of protest. Biographical details were not requested as most protests were not legal; however, bystanders and protesters were primarily university students. This selection was particularly important given that university students have played a pivotal role as both organizers and participants (Au, 2017; Ng & Kennedy, 2019). Journalist participants were in their 20s or 30s, while bystanders were in their 20s. All interviews were conducted in Cantonese. The list of questions used to guide each interview are provided in Supplementary Table 1.

Analysis approach

Participants in the protest group were asked seven identical questions. Our research goal was to compare well-established protester collective-action motivations with those of different actors within the protest arena. As a result, the bystander and journalist interviews were longer and followed a semi-structured interview schedule. Each bystander and journalist participant was asked a series of open-ended questions, which provided the opportunity to freely express their views and feelings about the recent social movements in Hong Kong. In each group, more specific questions related to their group were also asked; responses to questions unrelated to participation choices were excluded from this analysis.

All interviews were transcribed in Cantonese and translated into English. Thematic analysis was undertaken by three coders from Hong Kong and Australia in three stages over a six-month period. For each of the three groups, a first round of coding was undertaken in Cantonese by the interviewer, using a deductive (theory-driven) approach to analysis based on the known predictors of protesters’ collective-action intentions and behavior (following Braun & Clarke, 2021). Cantonese text excerpts in each theme were then translated into English and reviewed by the first author. Following this, the first author undertook a reflexive thematic analysis on the English translations using NVivo, following Braun and Clarke's (2020) six recommended steps: familiarization, coding, generating initial themes, reviewing and developing themes, refining, and defining and naming themes.

Once the themes and associated text excerpts were compiled, a research assistant reviewed text excerpts and codes in both English and Cantonese, guided by a codebook. Disagreements were examined and resolved via discussion. This process concluded with a final list of text excerpts for each theme. Following Mazzoni and Cicognani (2013), we then used a comparative analysis approach to analyze text excerpts in themes that occurred across all three datasets and examine their reciprocal relationships. This process of close reading of the text by multiple analysts at multiple times in both Cantonese and English ensured triangulation of findings and that translation peculiarities or context-specific information were identified and considered during the analysis process.

Protesters

We found that protesters’ motivations for engaging in collective action mirrored the known psychological drivers of engagement in collective action (Van Zomeren, 2013). Protesters cited identity and identity enactment and injustice as their primary drivers, supported by a unanimous consensus about the issue. These variables appeared to be linked to an underlying sense of moral obligation to participate. Lack of protest efficacy and the presence of violence did not pose a barrier to participation for protesters.

Identity as a Hong Konger was spontaneously mentioned by many protesters and was linked to the perception of injustice:

I can see injustice in Hong Kong, so I have to stand up in support as a Hong Konger. (P6)

While it was difficult to tease apart injustice, moral conviction, and moral obligation in protesters’ comments, the perception that authorities’ actions were unjust or morally transgressive appeared to be directly connected to a feeling of moral obligation to take action.

Moralization of the issue appeared to be related to both the broader demands for the preservation of existing democratic rights as well as authorities’ responses to these demands:

I think what the government has done is not right. As such, I think I have to stand up since the only thing that I can do is to express my appeal. (P23)

An obligation to express solidarity with other protesters was also demonstrated by some protesters:

It is our responsibility [to come out]. When I see so many people come out, I won’t let them walk alone. (P23)

While bystanders consistently argued that protest actions were ineffective, protesters argued that they felt obliged to join despite fears that they would not achieve any positive outcomes:

I think I will regret it for the rest of my life if I did not [come out]. I would rather do an unsuccessful thing than not try and regret in the future. (P10)

For some protesters, the hope that the goal could still be achieved bolstered their feelings of obligation to attend alongside perceptions of the potential efficaciousness of the action:

I have to come out even [though] I think my power is little. The power of a person is really little, but the power of one million and two million people can be a lot. Also, I believe that we will succeed one day, I won’t give up. (P34)

Even if seen as inefficacious, many protesters redefined the goal of the protest into a more achievable outcome. For example, although many declared previous protests had been inefficacious, they would then state that their efforts helped send a message to Hong Kong authorities:

It can produce noise. At least the government knows that we have an opinion. We have to make it known to them. (P28)

Other protesters argued that the protests were effective at garnering international attention:

Foreigners started noticing. The New York Times also reported on the “black cops,” now the international society knows about it. (P34)

Protesters also reevaluated the goal toward building the movement:

When others see that there are so many people coming out, they will think about the reasons causing so many people to care about an issue. In short, letting more people know about our determination on the issue is good for it. (P24)

Protesters did not change their engagement choices when faced with the use of violence. They were largely accepting of the use of violence (although few stated willingness to directly use violence themselves). They argued that violence was a response to inefficacious conventional nonviolent actions, and therefore a valid response to the political inaction that had greeted past protests.

Hong Kong people have tried joining parades with peaceful and rational attitudes. Then, they realized that the government wouldn’t change …. After that, they tried to escalate their actions … Hong Kong people also want to know under what conditions the government will respond to us. (P18)

The distinction between violence perpetrated by protesters as opposed to that perpetrated by the police was very important to many protesters. Many argued that violent action was often the last resort or simply a justifiable and unavoidable response to aggressive police actions:

When we are … bare-handed without any weapons while the police fire their guns, who is more violent? (P7)

Bystanders

Bystander perspectives on collective action were strikingly different from those of protesters, both regarding general motivations to engage as well as views on violent collective action. Unlike protesters, bystanders seldom used language conveying a sense of collective identity (connected with participation), emotional response, or perceived injustice. Instead, their choice to remain as bystanders and not participate in the social unrest was most closely linked to three themes: the lack of issue consensus, perceived movement inefficacy, and a negative view of violent action. We consider each of these themes in turn.

No bystanders expressed a strong collective identity that was associated with action (e.g., such as a politicized or opinion-based identity). Two bystanders described perceived threats to what they saw as unique characteristics of Hong Kongers, but did not connect these with a need to take action:

Cantonese is …… the characteristic of Hong Kong people, i.e., using Cantonese together with English in sentences. I don’t think that I will find it in the next generation, so I consider our generation to be the last. (B3)

Bystanders expressed ambivalence over collective action and the identities they associated with it, as well as few indications of any emotional connections to the protests or protest goals. They also did not convey any strong sense of injustice or moral outrage in their interviews:

I don’t have anything that really offends me, because I don’t have a clear stance …. I don’t have anything which makes me want to express my voice, or to [go] against. (B4)

One explanation for the lack of identification, emotion, and injustice themes in this group may be related to the low degree of issue consensus that prevailed. While most bystanders stated some support for what they perceived to be the protesters’ goals, there was no clear consensus on the issue. Some bystanders were unsure of their position on the issue:

I didn’t totally agree with a particular side … I want to understand both sides more first, I don’t want to be influenced by them. (B2).

Most bystanders argued that the protests were ineffective, a perception that—unlike protesters—appeared to strongly influence their choices to not participate. Bystanders also did not appear to differentiate between short-term (e.g., high attendance at a protest) and long-term goals (e.g., full democracy):

The demonstrations in recent years were more like a mere formality rather than having any practical use. Therefore, I think that they were not very effective. It becomes an annual gathering. (B3)

Bystanders’ arguments regarding the inefficacy of protests were also often closely linked to a belief that political and systemic changes were impossible in the Hong Kong context:

The autonomy of Hong Kong will have to be returned to China … this change makes Hong Kong more and more similar to China. It cannot be stopped unless Hong Kong stops following the system of China. However, this will never materialize. (B3)

The third theme that emerged as a strong influence on bystanders’ choices regarding participation was action type. As highlighted above, protesters largely accepted the occurrence of violence. In comparison, bystanders argued that violent action influenced their participation choices for three reasons. First, they saw it as morally wrong and in violation of the principles of harmony:

There are many different demands in society. You cannot act violently just because your demand is not accepted, because it is important for society to reach a consensus. It is a value of the general public and the society … everything should be based on the greatest benefit to the society. Benefits to the minority cannot override those of the others. (B5)

Bystanders also saw violent action as ineffective in either achieving their perceived movement goals, or attracting new protesters to the movement:

In my opinion, using violence is not a good method, because, firstly, it allows the regime to use politics to suppress you. Secondly, when you use violence, it causes some citizens who support or sympathize originally to think: is that right to fight for something in this manner? They may become more reserved about it or even object to it. (B8)

Finally, some bystanders argued that they personally could not engage in action if it had violent components, either because it clashed with their self-identity or their fear of potential consequences such as arrest:

I am not a person who likes radical forms of protest … I tend to be peaceful, rational, and nonviolent, and I’m afraid that I will bear some consequences, like being arrested. (B5)

Journalists

Unlike bystanders and protesters, journalists unanimously stated that their role identity as a journalist precluded any engagement in collective action. This journalist identity was also directly linked to their goal in the collective-action arena, which all journalists said was to communicate what was happening to the general public in an unbiased fashion:

My profession is to report the truth: to report what is happening live. (J2)

All journalist participants highlighted the critical importance of neutrality and nonparticipation in the protests as the core values of their role identity. However, they recognized that they held a role as actors within the protest arena that other groups—primarily protesters—sought to ally with:

In the Anti-ELAB movement, many have high expectations of journalists or media companies and expect certain media to stand by the side of protesters. (J4)

These expectations impacted their ability to remain impartial in their work on the ground. For example, a number of journalists noted incidents of being asked to favor a side:

Many protesters would ask that the reporters act as their sentry and help scout out where the police were. I refuse directly and state that I cannot be a participant—I am a reporter and my job is to capture the news. … Other than towards my company, a reporter should not side with anyone. (J2)

Others noted the complex role journalists navigated attempting to maintain a neutral stance in highly volatile situations.

I was just standing there, but the police officers suddenly surged forward towards me. Two “raptors” [the Special Tactical Squad] pointed at me and accused me of breaking through their cordon. But in reality, I never moved a single step. Sometimes, we also told the protesters not to use journalists as a shield. (J5)

These events transformed some journalists’ professional boundaries, suggesting that journalists’ perceptions about professionalism may be not static but rather a social construct that varies in reaction to critical events. For example, despite the strong imperative to maintain impartiality, journalists did participate in protests in discreet behaviors. This engagement took different forms. The first of these related to providing aid to police or protesters, and seemed primarily prompted by emotional responses such as sadness, fear, and anger:

While I was reporting on the Anti-ELAB movement, I did help police officers and protesters to wash their eyes and dress their wound because I cannot bear to see anyone get hurt. (J5)

The second form of engagement relates to information content. Some journalists chose not to report on protesters' strategies or tactics. Others mentioned obscuring or hiding protesters' identifying information:

Journalists do sometimes take sides with demonstrators, for example, to not capture certain actions of protesters on camera. (J6)

These engagement behaviors were also linked to emotional responses. Some journalists highlighted how the human response of wanting to assist others could override the strong motivations to remain an impartial nonparticipant:

A protester fell right behind me and I instinctually helped him up. However, in fact, as a journalist, we weren’t allowed to do so … A riot policeman had raised his baton to hit me and yelled, “what are you doing?” I had no choice but to let go of the protester's hand and witness the riot police arrest the protester. I was upset. (J4)

Perceived power imbalances between protesters and police alongside feelings of injustice, anger, and a shared identity as a Hong Konger appeared to influence these decisions to engage:

If I were not on the scene of a live broadcast or doing follow-up, I would indirectly let the [protester] groups on the scene know there are police stationed ahead and not to move forward … but I won’t directly let the protesters know—that is my bottom line. It is my opinion that the police and protesters do not hold the same level of weaponry. Once the police subdue them, they will be subject to head injuries. As a citizen of Hong Kong, it pains me to see protesters battered or injured. (J2)

Journalists did not state whether they perceived collective action to be efficacious or not. Their primary concern was their own ability to undertake their role and get a good story:

Journalists … all want a break. But we all want our own exclusive news story too. This is the goal of a reporter. (J1)

Action type did not directly influence journalists’ choices regarding engagement. However, the impact of violence again prompted emotional responses that threatened to override the values of impartiality and non-engagement tied to their role identity. Several journalists noted how conditions of extreme violence led to tension between their professional role and personal values:

You hate the police but you will still try hard to save the vanishing life in front of you. This is humanity. Nothing is more important than lives. (J1)

Finally, some journalists highlighted how issue consensus impacted their decision-making. This was exacerbated by conflicts between a journalist's and their employer's ideological stance, which was particularly difficult to manage during periods of high personal risk:

With the difference in political stance between my company and I, I feel helpless. Because apart from my job at my company, is there anything else I can do? This helplessness makes me feel pessimistic and discouraged … Slowly I started questioning my job as a reporter, because the public has given the role of helping people to reporters, and when it comes to the Anti-ELAB Movement, I always think, what else can reporters do? What can we help with? And when I get hit by tear gas, I also think, what if one day I cannot return home? (J7)

Strengths and limitations

This study undertook an exploratory comparative analysis to compare motivations to participate in collective action amongst three distinct groups who play important roles in collective-action arenas. Our approach and findings have a number of strengths. First, our qualitative analysis approach has enabled a deep examination of motivations and the identification of patterns across the three cohorts (Braun & Clarke, 2020). Second, this approach has responded to calls for more qualitative studies of intergroup contact (Dixon & McKeown, 2021), particularly in varied cultural contexts where social problems are the focal research topic (Bernardo & Liu, 2015; van Zomeren, 2019). Through doing so, this study has provided additional support for SIMCA predictors within a violent collective-action context, as well as demonstrated the importance of issue consensus and identity in protest behavior trajectories from nonparticipants to participants.

We also note some limitations. In understanding the generalizability of the results, we acknowledge that our small convenience samples may have attracted unusual respondents or circumspect responses that biased the results. For example, bystanders who were ready to boldly comment on the protests in interviews may be unrepresentative of the broader population of non-actors, who may be even more disengaged or fearful of involvement than those represented here. Furthermore, bystanders and journalists may have been unwilling to voice a sense of common identity with protesters or agreement with their aims and methods, in the context of an ongoing crackdown against dissent by the state. Moreover, a fear of consequences from engaging in collective action may have influenced some bystander or journalist participants’ identity attributions as a means to justify their nonparticipation.

We also note that contextual factors, such as radicalization before and after moments of unrest during the Anti-ELAB protests (e.g., Lee et al., 2021), may have influenced participants’ opinions about engaging in social unrest and increased their tolerance of protest-related violence more broadly. The changing context could also influence participants’ role identities over time, which our data indicate is an important factor linked to protest participation choices. Our data were collected across 2018–2020, and thus could have been influenced by these factors. Future research could aim to capture the groups’ perspective synchronously and longitudinally in order to examine differential responses to the changing context over time. In addition, while the face-to-face interviews here offer rich data that speak to important research questions, future research could seek comparative data using written responses to anonymous surveys, in order to examine whether respondents more readily engage the topics, and whether they disclose different or similar motives, perceptions, and identities. These studies could also seek to examine protesters’ perceptions around the types of violence that they believe are justifiable and the contexts in which this violence would be tolerated. Future research could seek to examine these perceptions across separate groups more substantively, and then build upon this work with experimental and longitudinal analysis.