Collective action is linked to activities which are intended to advance the situation of one’s in-group (see Stekelenburg and Klandermans, 2013). This include a wide variety of activities, like signing petitions, organizing marches, taking part in boycotts (Klandermans, 1997; Wright et al., 1990). Perceived injustice along with group effectiveness and group identification are important determinants of collective action intention (Agostini & van Zomeren, 2021; Van Zomeren et al., 2012). Group emotions (e.g. anger) and empowerment and self-efficacy grounded in social interactions with in-group members can facilitate development on the in-group engagement and willingness to act collectively. In our research we wanted to complement previous research and examine if perceived threat from the dominant out-group is related to collective action intension.

Thomas et al. (2009) advocated the Encapsulated Model of Social Identity on Collective Action (EMSICA) in which perceived injustice, along with group emotions linked to it, and group efficacy play an important role as determinants of identification with an in-group. Sharing the same emotions as other group members can have a significant impact on the formation of a group identity. These emotions influence the understanding of the social world, which defines the type of social relations in the group and between groups (Peters & Kashima, 2007). For creating a long-term commitment to social or political activities, developing a strong identification with the group is crucial. In addition, this identification must include a pattern of group norms regarding emotions, effectiveness, and behavior in relation to a specific group. If people outrage on injustice, group effectiveness, and involvement in collective action as the standards of in-group (in this case, a belief-based group), then at the same time they should feel exactly the same emotions, beliefs about effectiveness, and involvement in these activities.

According to Encapsulated Model of Social Identity on Collective Action, identification with a group mediates the relationship between affective responses to perceived injustice and group effectiveness, and collective action. Previous results showed that anger associated with perceived injustice is are significantly linked to developing group identification, and this is an important predictor of action’s intension (Thomas et al., 2016). EMSICA model is especially helpful in explaining the formation of emerging groups and group identities. Shared reactions (e.g. outrage because of unjust treatment) to a status quo can re-direct people’s motivation and facilitate group emergence (Thomas et al., 2012). The reason for this is that social identity development is preceded by interaction with and knowledge about similar others’ beliefs and opinions (e.g., Hardin & Higgins, 1996). Both anger and outrage contain a moral component and often appear in response to threats to moral values (Tangney et al., 2007). Thus, research presented in this article examines if perceived threat strengthens the perception of injustice, that evokes anger and outrage, and could lead to group identification and intention to protect the group and act on its behalf.

Perceived threat

Stephan and Stephan in their Integrated Threat Theory (ITT; 2000) separates threat in two types. The first one is realistic threat and the second one symbolic threat. Realistic threat is subjective perceived antagonism and disagreement over political, social or economic resources. Symbolic threat represents conflict regarding moral principles, norms, aspirations and beliefs between the groups (Riek, et al., 2006). Position of the group and the conflicts’ myths have a meaningful influence on functioning of the groups. They differently influence on perception of threats among the privileged and underprivileged groups. Underprivileged groups more often feel threatened, because fewer resources are controlled by them and they can be definitely overcome by the advantaged groups. High position and privileged groups can strongly react to seen threat, because more resources are in their hands (and they can lose them). Because of that situation, advantaged groups want to maintain the status quo (Stephan et al., 2002). Integrated Threat Theory shows that perceived threats have a significant influence on people’s behaviors, emotions, and cognitive functioning. Perceived threat is strongly linked to prejudice (Tausch, et al., 2007), anger (Renfro, et al., 2006), and aggression aimed at outgroup members (Maass, et al., 2003). What is important, these effects vary because threats themselves and the bonds with the in-group. As perceived threat is an inherent element of the intergroup relations, researches present that threats to different aspects of social identity, like positive distinctiveness (Wohl, et al., 2011) or general importance of group identity (Jetten, et al., 1997), could trigger group-defensive behaviors. Previous research indeed found that including observed threat as determinants of group action intention, facilitate the understanding of the motivation for pro-group behaviors (Çakal et al., 2016). Threats also increase the political involvement measured as the financial assistances made to the in-group (Miller & Krosnick, 2004).

Considering these results, in our studies we sought to analyze if perceived thread could be associated with collective action intention on behalf of the minority group (e.g., to protect the group). Specifically, we assumed that threat could heighten the perceived injustice, and this could strengthen the group identification and foster the involvement in pro group actions.

Social context of the research

Recent couple years in Poland were marked by emerged several social movements organized around opposing policies and laws that undermine democratic principles and ideas. Also, social groups fighting for a more inclusive society and women’s reproductive rights were growing. Equal treatment of LGBT + persons was one of the most discussed topics in Polish presidential campaigns in 2020. Right-wing politicians declared that LGBT + persons and their allies try to convince society that they are people, “but they are just ideology” (rmf24.pl, 2020). Other politicians dehumanized LGBT + people by saying that LGBT are not people (Wprost, 2020) or they are not equal to normal people (Kocejko, 2020). Conservative groups sponsored shuttle buses with signs that compare LGBT to pedophiles, which sprang demonstrations and voices of support for social activist opposing such treatments (Żelazińska, 2020).

The second context of collective action we analyzed is linked to actions for women’s reproductive rights. Black Protests, organized by women’s rights activists, left-wing parties and feminist groups, occurred in Poland in 2016 (Bielińska-Kowalewska, 2017). These demonstrations involved thousands of people on the streets in many Polish cities and towns, expressing their disagreement with plans to restrict the abortion law in Poland. They were successful (as these changes in law were abandoned), but only in the short term. In October 2020, Polish Constitutional Court has announced their verdict about abortion law in Poland. According to the ruling, severe and irrevocable impairment of the fetus or an incurable life-threatening disease are no longer reasons for legal abortion (Gardocka, 2020). It means that abortion law in Poland became even more restricted, as it resulted in near-total ban on termination of pregnancy. As the result, many mass demonstrations were organized to express their disagreement with this change (Jurszo, 2020) and hundreds of thousand people took the streets.

Situations presented above are just two examples of the contexts for collective actions in Poland. The year 2020 showed that fighting for equality and for in-group rights is important for people and cannot be ceased even by the pandemic. Even though presented examples of collective actions were directed at different social issues, we can find a common features. Importantly, for the current studies, both movements presented above sprang from the perceived threat to the in-group, be it dehumanization, discrimination or changes in laws that were perceived as a threat to Polish women. Thus, perception of threat, both symbolic and realistic, has played an important role in mobilizing people and directing them to the streets.

Current research

Researches presented in this article are aimed at examining the role of perceived threat (realistic and symbolic) in encouraging people to act on behalf of the group. We tested if perceived injustice and group identification mediate the relationship between perceived threat and collective action intention. Following hypothesis were tested:

H1: perceived threats (realistic and symbolic) is related to collective actions on behalf of in-group.

H2: perception of injustice and group identification facilitate the association between perceived realistic threat and collective actions.

H3: perception of injustice and group identification facilitate the association between perceived symbolic threat and willingness to engage in collective actions.

We conducted four studies to examine those predictions. Two of them examined motivation to act collectively in supporting LGBT + equality (Studies 1 and 2), and two (Studies 3 and 4) was conducted to explore the correlates of collective actin intention in the context of the newly formed social movement, that opposes near-total ban on abortion in Poland. For conducting this research, we obtained formal approval from Ethical Committee at Department of Social Sciences, University of Gdansk.

Study 1

Participants

The Study 1 was conducted among Polish Internet users in the context of LGBT + rights. We aimed to recruit participants from the general population. 182 participants took part in this study. Participants were recruited by nationwide research panel Ariadna and completed an online survey; 47.8% of them were women and 52.2% were men in age range from 19 to 72 (M = 41.37; SD = 12.80). As research was conducted in the context of LGBT + rights, we asked participants about their sexual identity. 87.4% of them were heterosexual, 4.9% - bisexual, 2.9% - homosexual and 4.9% of respondents identified themselves by other sexual orientations.

Measures

If not mentioned otherwise, we include a Likert scale with answers oscillating from 1 (definitely not) to 7 (definitely yes).

Intention to participate in moderate and radical collective actions

We choose items constructed for the scale of in-group action intention by van Zomeren et al. (2004). The scale comprised 5 questions about normative (α = 0.96; ‘I am willing to participate in protests and demonstrations for the equality of non-heterosexual people’) and 3 non-normative (α = 0.93; ‘I am willing to engage with others in the occupation of buildings when it is necessary for defending equal rights’).

Perceived threat

For measuring perceived threat to the in-group, we include items from the scale introduced by Stephan et al. (2002). We choose 3 items for realistic threat (α = 0.76; ‘Those who oppose LGBT + equal rights have too much political power in a country’) and 2 items for symbolic threat (α = 0.78; e.g. ‘Homosexual people and heterosexual people have different values’).

Injustice appraisal

To measure the sense of injustice (α = 0.87), scale choose items form the van Zomeren et al. (2004) measure. The respondents were to determine the level of their emotions regarding discrimination against LGBT group on a scale from 1 (strongly disagree) to 5 (strongly agree). Sample item: ‘Opposition to equality is unjust’.

Group identification

We used 14 items of the Polish version of the group identification scale by Leach et al. (2008) (α = 0.97; ‘I feel committed to LGBT community’).

Results

To test the hypothesis on the association between perceived threat and collective action intention, and on mediation role of perceived injustice and group identification, we conducted the Pearson’s correlation analyses (Table 1) and mediation analyses. Realistic (but not symbolic) threat was positively linked to both moderate and radical collective action intentions.

Table 1 Zero-order correlation analyses with Pearson’s r for Study 1
Full size table

To test our predictions, series of mediation analyses (PROCESS macro, model 6, with 10,000 bootstraps; Hayes, 2013). Results of mediation analyses show that perceived injustice and group identification mediate the relation between realistic threat and both moderate (R2 for the model 0.54, B = 0.23; 95% CI [0.12, 0.36]), and radical collective action intentions (R2 for the model 0.37, B = 0.19; 95% CI [0.09, 0.33]). That is for LGBT equality moderate collective action, perception that opponents of equal rights are threatening for LGBT community and receiving too much support from powerful organizations, was related to injustice appraisal, one’s group as more efficient in pursuing common goals (B = 1.06). Participants who more strongly perceived injustice to one’s group were strongly identified with that group (B = 0.44), and identification link to tendency to act collectively on behalf of equal rights (B = 0.48) (Fig. 1). For radical collective action, results were similar, with a link between identification and willingness to act collectively B = 0.41 (Fig. 2). No relationship between symbolic threat and collective action intention was observed.

Fig. 1
figure 1

Mediation analyses with willingness to engage in moderate pro-equality collective action as depended variable; Study 1

Full size image
Fig. 2
figure 2

Mediation analyses with willingness to engage in radical pro-equality collective action as depended variable; Study 1

Full size image

Study 2

By conducting Study 1 we verified our hypotheses with correlational design. We conducted experimental Study 2, which focused on real-life examples of anti-LGBT public statements and measure the perceived thread after these statements have been made cognitively accessible. Moreover, comparing to Study 1, two changes were administrated. First, we have changed symbolic threat items to not focused on value differences between heterosexual and homosexual people, but between proponents and opponents of equal marriage rights for LGBT people. Second, as strength of the correlations in Study 1 could be driven from inclusion of participants who oppose equal rights for LGBT people (they probably would not perceive their own action as threat to LGBT people and will not willing to take part in collective action for equality), in Studies 2, 3 and 4 we concentrated only on supports of given social movement (pro-LGBT equal rights in Study 2 and pro-choice in Studies 3 and 4). This allow as to examine if among allies and supporters, perceived threat is linked to intention to act on behalf of the in-group.

Participants

Participants were recruited by Polish nationwide research panel Ariadna and completed an online survey. As questions of the survey where design for supporters of LGBT equal rights, in final analyses, we included only participants who declared this support. In total, 119 from 175 participants fulfilled these criteria and were included in this study. 56,3% of them were women and 43,7% men (M = 42,29; SD = 14,13).

Measures and procedure

People participating in the study were initially asked to express their support or opposition for current social issues (e.g. the death penalty, environmental initiatives, and accepting refugees), including support for LGBT equal rights. Then, to present anti-LGBT sentiment that exist in public discourse, the randomly assign half of the respondents (experimental group) were presented with statements of politicians from previous months, concerning non-heterosexual persons (statements that appeared in the media or political mass-gatherings). Immediately after introducing the experimental priming, participants responded to questions about their perception of those statements and whether they could lead to negative consequences for non-heterosexual people. An example of a question is: “The presence of such opinions in the media may increase discrimination against LGBT people.” Control group did not receive any statements to read and evaluate. In the next steps, the respondents were requested to reply on subsequent questions (on a scale ranging from 1, definitely not, to 7, definitely yes; if not mentioned otherwise):

Intention to participate in moderate and radical collective actions

We used the scale which is based on the scale of engaging in collective actions by van Zomeren et al. (2004). This scale comprised questions about normative (α = 0.95) and non-normative (α = 0.95) collective actions. Perceived threat. As in Study 1 we used a scale of realistic (α = 0.75) and symbolic (α = 0.68) threat created by Stephan et al. (2002). Perceived injustice. To measure perceived injustice (α = 0.92), we used the questions based on scale created by van Zomeren and co-workers (2004) (e.g. ‘I think the negative opinion about LGBT people is unjust and unfounded’). Group identification. We used the scale created by Leach et al. (2008) (α = 0.96) (e.g. ‘Being part of supporters of LGBT equal rights is an important part of how I see myself’).

Results

Primary analyses

To verify whether the applied manipulation was effective, a one-way ANOVA was performed for each type of threat. No significant manipulation effect has been demonstrated for any type of threat. For the realistic threat, effect equals F (1,117) = 0.15; p = .70 (M = 4.90 in the experimental and M = 4.80 in the control group). For the symbolic threat, effect was F (1,117) = 0.27; p = .60 (with mean results in experimental M = 5.00, and M = 5.11 in the control group). As no differences between groups were recorded, in the next steps, we did not compare groups as planned.

We have decided – for exploratory purpose – to conduct correlational and mediation analyses separately for control an experimental groups. Pearson’s correlations analysis (Table 2) indicated that realistic threat were positively correlated with normative collective actions. Similar results were observed for non-normative collective actions. Thus, although priming participants with anti-LGBT statements did not result in stronger perceived threat, realistic threat was linked to moderate collective action intention in both control and experimental groups.

Table 2 Zero-order correlation analyses with Pearson’s r for Study 2
Full size table

Study 3

As the previous two studies were conducted in only one social context (i.e. LGBT rights), we aimed at examining if the observed results will generalize to different collective action context. We conducted Study 3 in women’s collective actions against a near total ban on abortion.

Participants

Undergraduates students were asked to recruit their family members to take part in a study on social attitudes and support for pro-choice demonstrations. Only people who oppose near-total ban on abortion that was introduced in Poland (i.e. (support pro-choice movement), were considered for analyses. In sum, 451 participants that fulfilled these criteria participated in the study. Study 3 was conducted via an online survey, with 64,1% of participants self-described as women and 35,5% as men, with 0.4% choosing ‘other’ option. Age range from 16 to 64 (M = 27.32; SD = 9.73).

Measures

If not mentioned otherwise, all items were answered on a scales with ranges from 1 (definitely not/definitely do not agree) to 7 (definitely yes/definitely agree). We used same measures as in previous studies, but the context of the questions has been changed (i.e. target group and context of collective action). Intention to participate in moderate and radical collective actions. We used the scale which is based on the scale of engaging in collective actions by van Zomeren et al. (2004) (e.g. ‘I am willing to organize events against the total ban on abortion’). This scale consisted of questions about normative moderate collective actions (α = 0.71). Perceived threat. Each kind of threat was measured by one question from realistic (‘People seeking a complete ban on abortion have too much political power in Poland’) and symbolic (‘People determined to completely ban abortion have different values than I do’) threat scales developed by Stephan et al. (2002). Reliability of single question’s scales finds it confirmation in previous research (e.g. Atroszko et al., 2017). Perceived injustice. To measure perceived injustice, we used single question based on items developed by van Zomeren et al. (2004) (‘I consider the actions of people who want to completely ban termination of pregnancy in Poland unjust’). Group identification. We used 3 items from the scale created by Leach et al. (2008) (α = 0.72) (‘I feel a connection with people who oppose the total ban on abortion’, ‘I am in solidarity with group of people who oppose the total ban on abortion’, ‘I feel at one with those who oppose the total ban on abortion’).

Results

Pearson correlations analysis showed that peaceful collective action intension linked positively with realistic, symbolic and physical threat (Table 3). To test our predictions, one more time mediation analyses were conducted (PROCESS macro, Model 6, with 10,000 bootstraps; Hayes, 2013). Mediation analysis showed that perceived injustice and group identification mediate the association among realistic and symbolic threats and moderate peaceful collective action intention. For realistic threat, R2 for the model 0.38, B = 0.04; 95% CI [0.02, 0.07]; and for symbolic threat, R2 for the model 0.37, B = 0.05; 95% CI [0.02, 0.07] (see Figs. 3 and 4).

Table 3 Zero-order correlation analyses with Pearson’s r for Study 3
Full size table
Fig. 3
figure 3

Mediation analyses with willingness engage in moderate pro-choice collective action as depended variable and realistic threat as predictor; Study 3

Full size image
Fig. 4
figure 4

Mediation analyses with willingness engage in moderate pro-choice collective action as depended variable and symbolic threat as predictor; Study 3

Full size image

Study 4

In Study 4, we aimed at replication of the Study 3 results and accordingly we conducted this research in a context of abortion law restriction in Poland. We focused on collective actions against abortion ban, advocated by pro-choice movements in Poland.

Participants

229 participants – undergraduate students and their family members - completed the online survey. We included only persons that declared they are against more restricted abortion law (154 participants; 58,4% women, 40,9% men and 0,6% people with different gender) in age range 16–67 (M = 31,96; SD = 11,89).

Measures

If not mentioned otherwise, all questions were responded on a scales oscillating from 1 (definitely not/definitely do not agree) to 7 (definitely yes/definitely agree). As in study 3 we used 3-item group identification scale (Cronbach α = 0.98). We used two items to measure perceived injustice (r = .73; ‘I feel this group is being treated unfairly by those in power’, ‘I feel that the public’s opinion of this group is unfair’). Intention to be a part of group actions was measured by seven items (four for moderate, α = 0.92, and three for radical collective action, α = 0.92). As for perceived threat, we used two item for symbolic threat (r = .82) and three for realistic threat (α = 0.80).

Results

Pearson’s correlation analysis showed that in a group of people who are against the abortion law restrictions moderate collective actions are correlated with realistic and symbolic threat (Table 4), and preparedness to participate in radical action was link to realistic threat only.

Table 4 Zero-order correlation analyses with Pearson’s r for Study 4
Full size table

Mediation’s analysis confirmed previous results and show that injustice appraisal and connection with a group mediate the link amongst threats and collective action intention. Indirect effect for realistic threat was B = 0.17; p = .06; LCI = 0.07; UCI = 0.29);, and for symbolic threat B = 0.13; p = .05; LCI = 0.05; UCI = 0.23 (see Figs. 5 and 6). For non-normative collective actions results were very similar for realistic (B = 0.11; p = .05; LCI = 0.04; UCI = 0.22) and symbolic (B = 0.09; p = .04; LCI = 0.03; UCI = 0.18) threat.

Fig. 5
figure 5

Mediation analyses with willingness engage in moderate pro-choice collective action as depended variable and symbolic threat as predictor; Study 4

Full size image
Fig. 6
figure 6

Mediation analyses with willingness engage in moderate pro-choice collective action as depended variable and realistic threat as predictor; Study 4

Full size image

Discussion

Presented studies was conducted to extend the existing models of engaging in collective actions with perceived threat (realistic and symbolic). Tested hypotheses were aimed to check if perceived threat increases the tendency to engage in collective actions. We tested mediation models, exploring if perceived injustice and group identity mediate the link between observed threat and collective actions. Presented studies were conducted in two social contexts: women’s rights with the focus on abortion law and LGBT rights. For realistic threat, hypotheses were confirmed. The more members of in-group and their allies perceived out-group as threatening, the stronger feeling of unjust treatment and identification with the discriminated group, and identification is linked to readiness to act jointly with others. For symbolic threat, that designate differences in values between in-group and out-group, similar relationship was established (Study 2, 3 and 4) when out-group was defined as people who oppose giving equal rights to LGBT community and more choice for women (but not when value differences between LGBT and heterosexual people was measured; Study 1).

However we know that perception of in-group threats, both realistic and symbolic, can affect intolerance toward out-groups (e.g. asylum seekers; Mancini et al., 2020), in our studies we focused on exploring how threat motivates people form in-group to act on behalf of in-group members. We can assume that perceived threats are vital elements associated with the tendency to be involved in collective actions among groups who support more liberal legal regulations. Our studies complement previous research on Encapsulated Model of Social Identity on Collective Action (Thomas et al., 2009), adding a possible factor that awakens injustice appraisal. Presented results are also are also compliant with research on collective angst, understood as a negative emotional response to the observed threat to the survival of one’s in-group (Wohl & Branscombe, 2009). For example, this perception of fundamental threat is lined to in-group defenses and pro in-group attitudes and actions. These relationships have been confirmed for more conservative (Wohl & Branscombe, 2009; Wohl et al., 2011) and liberal values and group (Jaśkiewicz et al., 2022). Moreover, the presented results align with recent advancements in the field of relative deprivation theory. Research has demonstrated that group relative deprivation can amplify tendencies towards violent collective action, with hostile feelings acting as mediators of this effect (Su et al., 2022). Our findings directly correspond to these outcomes, as we observed that perceived threat, which is presumptively linked to the perception of relative deprivation, is associated with collective action, while injustice appraisal, presumptively linked to anger and hostile emotions directed towards the source of the threat, mediates this relationship. This model is further supported by a study conducted on terrorism detainees from Indonesia (Yustisia et al., 2020). In this study, the authors focused on the perception that Islamist principles were under threat from secularism and the West. Despite the vastly different cultural context, they obtained results similar to ours: perceived threat was linked to pro in-group actions, and this association was mediated by perceived injustice and social identity.

Limitations of presented studies should be highlighted as well. First, we concentrated on two contexts of collective action, but both of them are rather progressive ones and on behalf of the underprivileged group. Members of groups that are more engaged in actions for more traditional values and policies are often supported by the powerful allies (the government and the Church). As the result, these groups might perceive less realistic threat (but possibly still symbolic threat as, progressive groups would be perceived as fundamentally differ in values and their social goals). More studies should explore if perceived threat works differently when actions for status quo or traditionalists’ social change are considered. Second, our experimental manipulation in Study 2 was inefficient in priming stronger threat for in-group. The context of the study was linked to the ongoing political campaigns and we used real-life example of statements that were threatening for LGBT community. As the topic of LGBT rights and attacks on them by conservatives politicians and catholic bishops were often discussed in Poland in last years, it is possible that participants in the control group were also aware of these attacks via media exposure. Thus, the media exposure and knowledge about these statements could be factors that influenced participants’ responses. We did not measure any of these confounding variables. The third limitation is lack of a national representative sample. We examined willingness to act on behalf of progressive social change, and policies that a are in core of so called “cultural wars”. For this reason, a larger sample, adjusted to the demographic profile of the population in Poland, would be useful to assess generalizability of the tested theoretical model. Finally, fourth limitation is linked to the strength of the threat perceived. We did not measure the consequences of extremely strong threat that could be felt by social activists in more authoritarian countries and cultural contexts. That is, we did not explore the boundary conditions (i.e. moderators) that influence the positive relation between threat and action intinction. It is plausible that extreme threat would lead to ‘fly’ or ‘freeze’ reactions instead of ‘fight’. Thus, more research with examination of variables that could change presented relationships are needed.

Limitations notwithstanding, we presented possible pathways to extend a previous model of collective action. Our results showed that the perceived in-group threat could - at least in some cultural contexts - trigger injustice appraisal and this negative sentiment is linked to stronger in-group identification and behavioral intentions.