Advantaged groups play an important role as allies in achieving justice for disadvantaged groups, given their relatively greater influence in society (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010). Although much of the collective action literature focuses on disadvantaged group members’ efforts to improve their situation, research has increasingly focused on the role of advantaged group members or allies (Iyer & Ryan, Reference Iyer and Ryan2009; Iyer, Schmader, & Lickel, Reference Iyer, Schmader and Lickel2007; Leach, Iyer, & Pedersen, Reference Leach, Iyer and Pedersen2006; Mallett, Huntsinger, Sinclair, & Swim, Reference Mallett, Huntsinger, Sinclair and Swim2008; Shepherd, Spears, & Manstead, Reference Shepherd, Spears and Manstead2013; Stewart, Latu, Branscombe, & Denney, Reference Stewart, Latu, Branscombe and Denney2010; Subašić, Hardacre, Elton, Branscombe, Ryan, & Reynolds, Reference Subašić, Hardacre, Elton, Branscombe, Ryan and Reynolds2018; van Zomeren, Postmes, Spears, & Bettache, Reference van Zomeren, Postmes, Spears and Bettache2011). This is certainly a fertile area for investigation, given the unique position of the advantaged group. In advancing the interests of the disadvantaged group, privileged group members have to come to terms with their collective responsibility for the injustice experienced by the disadvantaged group, and act in a way that may violate their own group's interests (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010).
This is the case for men, who continue to hold more power in societies around the world and make decisions about policies that may either increase or reduce gender gaps (World Economic Forum, 2017). Likewise, most of the violence against women is perpetrated by men, putting men in a crucial position to end such violence (Flood, Reference Flood2011). This role of men's collective action for women has been recognised in social psychological research, which has examined variables such as gender self-esteem (Burn, Aboud, & Moyles, Reference Burn, Aboud and Moyles2000), in-group identification and emotions (Iyer & Ryan, Reference Iyer and Ryan2009), portrayals of feminist men (Wiley, Firnhaber, & Schilinsky, Reference Wiley, Firnhaber and Shilinsky2012), and moral concerns (Precopio & Ramsey, Reference Precopio and Ramsey2016) in predicting support for gender equality. However, men's allyship has yet to be investigated within an integrative model of collective action, such as the Social Identity Model of Collective Action or SIMCA (see van Zomeren, Reference van Zomeren2014). Additionally, the current study contributes to confirming or possibly even challenging the theory in Japan and the Philippines, contexts where it has yet to be examined, to our knowledge. Beyond replication of the SIMCA, the research extends the theory to include instances when advantaged group members’ ingroup identification may potentially inhibit collective action. These serve as contributions of the present article, which seeks to understand men's allyship both as something explicable from past models of collective action and as something that may challenge such models to illuminate new theoretical frontiers.
The Social Identity Model of Collective Action
According to the SIMCA (van Zomeren, Reference van Zomeren2014, Reference van Zomeren2016; van Zomeren, Leach, & Spears, Reference van Zomeren, Leach and Spears2010; van Zomeren, Postmes, & Spears, Reference van Zomeren, Postmes and Spears2008; van Zomeren, Saguy, & Schellhaas, Reference van Zomeren, Saguy and Schellhaas2012; van Zomeren, Spears, Fischer, & Leach, Reference van Zomeren, Spears, Fischer and Leach2004), collective action tendencies are predicted from four variables: politicised identification (one's sense of oneself as a member of a group), group-based anger at injustices or grievances faced by the group, and group efficacy (a sense that the group has the capacity to change the status quo), alongside moral conviction (perceptions of the moral importance of the issue). While the model was developed to predict action by disadvantaged groups for their own interests, van Zomeren, Postmes, Spears, and Bettache (Reference van Zomeren, Postmes, Spears and Bettache2011) found that for advantaged group members acting as allies, moral conviction predicted collective action tendencies through identification with the disadvantaged group, anger, and group efficacy. In contrast, they found no relationship between identification with the advantaged group and the other SIMCA variables. Thus, in its current form, the SIMCA does not include identification with the advantaged group within the model.
Other researchers have confirmed the role of moral conviction in predicting advantaged group members’ collective action on behalf of disadvantaged groups within the German context (Barth, Jugert, Wutzler, & Fritsche, Reference Barth, Jugert, Wutzler and Fritsche2015). The role of group efficacy in advantaged group members’ collective action has likewise been investigated in relation to White American students’ support for African Americans in their university (Stewart et al., Reference Stewart, Latu, Branscombe and Denney2010). This research, in particular, demonstrated that higher perceived efficacy is associated with increased antidiscrimination action by advantaged group members. Furthermore, anger, whether towards others responsible for the injustice, or even one's advantaged ingroup, has been demonstrated to motivate European American students to engage in collective action for African American students by challenging those who perpetuate inequality, or providing restitution for victims (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010).
Politicised Identification: Focus on Feminism
The role of politicised identification may be complicated in matters of gender equality, especially when considered alongside men's identification with their own gender. Past research in Canada and the United States has shown that many men and women show ambivalence about and even hostility towards feminism (Yeung, Kay, & Peach, Reference Yeung, Kay and Peach2014). Additionally, both women and men from a US university were more likely to agree with feminist ideas or endorse covert feminism, rather than identify as feminists or endorse overt feminism (Burn et al., Reference Burn, Aboud and Moyles2000). Identifying with a feminist label may especially be a concern for men, given some stereotypes associated with being a feminist. For instance, Anderson (Reference Anderson2009) demonstrated in a US-based study that although feminist men tend to be evaluated positively, they are also seen as less likely to have conventionally masculine characteristics, and are considered less attractive. Thus, in our research, we focus on men's identification with the disadvantaged group of women, which can be construed as covert feminism, rather than identification as a feminist, or overt feminism. We return to this point in the discussion.
Identification with the Advantaged Group
Beyond the variables within the SIMCA, other researchers have paid greater attention specifically to identification with the advantaged group. Studies of allyship have shown that lower identification with the advantaged group is linked to higher collective guilt, which in turn increases the likelihood of collective action among American students (Mallett et al., Reference Mallett, Huntsinger, Sinclair and Swim2008) and willingness to compensate for historical injustices, especially when the group's history towards the disadvantaged group is made salient among Dutch students (Doosje, Branscombe, Spears, & Manstead, Reference Doosje, Branscombe, Spears and Manstead1998). These researchers suggest that low identifiers are less concerned about defending the ingroup (Mallett et al., Reference Mallett, Huntsinger, Sinclair and Swim2008) and feel less threatened when negative group characteristics are highlighted (Doosje et al., Reference Doosje, Branscombe, Spears and Manstead1998).
Research mainly with samples from North America and the United Kingdom focusing particularly on men's collective action intentions for women have found similar patterns: men who have low identification with their ingroup are more likely to perceive inequality (Iyer & Ryan, Reference Iyer and Ryan2009). However, in that research, highly identified men may still be moved to take action when they perceive the pervasiveness of the inequality and feel sympathetic towards the victim. In contrast, van Zomeren and colleagues’ (Reference van Zomeren, Postmes, Spears and Bettache2011) study using the SIMCA in the Netherlands and Hong Kong found no relationship between ingroup identification among advantaged group members and collective action tendencies or any of the other SIMCA variables.
These discrepancies may be related to the types of groups under examination: in the SIMCA study, the disadvantaged groups investigated were Dutch Muslims and Mainland Chinese, and the advantaged groups made up of Dutch non-Muslims and Hong Kong Chinese. Compared to women and men, these groups may have fewer interactions with each other, and may be seen as less of a threat to the privileges enjoyed by the advantaged group. Given the absence of a significant contribution of advantaged group members’ ingroup identification, van Zomeren and colleagues (Reference van Zomeren, Postmes, Spears and Bettache2011) excluded this variable from the SIMCA. However, we examine the possibility that in the case of men's collective action for women, their identification with their gender group may inhibit collective action tendencies, as the literature in this section has demonstrated. Thus, we argue for extending the SIMCA to include ingroup identification among advantaged group members as a possible factor inhibiting collective action when it comes to men's allyship. After all, intergroup relations between women and men differ slightly from other groups (e.g., racial or ethnic groups) that are typically examined in collective action research; women and men have greater degrees of interdependence, and men may see women as competition, with support for their advancement being seen as detracting from men's maintenance of their status (Fiske & Taylor, Reference Fiske and Taylor2013). Recognising these differences in the relationships and interactions between groups is especially important in collective action research to craft appropriate interventions. For instance, interventions to reduce inequality such as intergroup contact may be effective in addressing racial inequality (Tropp & Barlow, Reference Tropp and Barlow2018), but may not necessarily be as potent when groups have frequent and interdependent interactions to begin with.
The Japanese and Philippine Contexts
Japan and the Philippines are two Asian countries with very different contexts and histories of both collective action and gender inequality. For instance, collective action features more prominently in contemporary Philippine history, having overthrown two presidencies through mass protests (Liu & Gastardo-Conaco, Reference Liu and Gastardo-Conaco2011). Stark contrasts can also be seen in terms of the gender gap in these two countries. In the 2017 Gender Gap Index (World Economic Forum, 2017), the Philippines ranked 10th while Japan was much lower at 114th place. Looking at the measures within this index, the Philippines had higher rates of women's economic participation and opportunity, educational attainment, and political empowerment compared to Japan, whereas Japan had the advantage in maternal health. Recognising this persistent inequality and its impact on their economy, the Japanese government has been pushing for more equal employment and advancement opportunities for women in their policies in recent years, even creating a Gender Equality Bureau Cabinet Office (MacNaughtan, Reference MacNaughtan2015; Nae, Reference Nae2017; Schieder, Reference Schieder2014). Although the Gender Gap Index may not completely capture the experiences and conditions of individual women, it is nevertheless indicative of the relative positions of women and men in the broader sociocultural context.
Going beyond the macroeconomic level, gender role inequality persists in Japan (North, Reference North2009), with more traditional gender role attitudes among both women and men (Nguyen et al., Reference Nguyen, Morinaga, Frieze, Cheng, Li, Doi and Li2013). Anti-feminist backlash has even been observed in conservative organisations that engage in collective action against gender equality (Suzuki, Reference Suzuki2017). Japanese women also continue to experience violence: 19.8% report having experienced physical violence, 16.8% experienced psychological violence, and 9.7% experienced sexual violence (Suga, Reference Suga2018). This is not to say that women are no longer disadvantaged in the Philippines, as discrimination and sexism continue to be problematic. For one, Filipino women still experience sexism in their day-to-day contexts (Torre, Reference Torre2016). Traditional gender role attitudes and even sexism are likewise normalised by remarks such as rape jokes from the president himself (Curato & Ong, Reference Curato and Ong2018). Moreover, violence against women remains persistent: in the 2017 National Demographic and Health Study (Philippine Statistics Authority & ICF, 2018), 20% of women reported having experienced emotional violence, 14% said that they had experienced physical violence, and 5% reported experiences of sexual violence from their current or latest husband or partner. Given this situation of women in the two countries, we consider whether the SIMCA, along with men's identification with their ingroup, can help provide directions for promoting men's collective action to reduce discrimination.
The Present Study
Our research examines collective action among advantaged group members for the disadvantaged group using the SIMCA, with the inclusion of identification with the advantaged group in the model. The literature we reviewed thus far demonstrated inconsistent evidence in relation to the role of identification with the advantaged group. We suggest that this may be a function of the types of groups under consideration. In our research, we will be examining gender groups, which have higher degrees of interdependence and possible competition compared to previous SIMCA research on advantaged groups’ collective action. We therefore hypothesise the following: (1) consistent with the SIMCA, higher levels of moral conviction, identification with the disadvantaged group, group efficacy, and anger will be positively related to willingness to engage in collective action; and (2) greater identification with the advantaged group will be linked to lower collective action. Although this inhibiting effect of ingroup identification has been demonstrated in previous research on men's collective action for gender equality (Burn et al., Reference Burn, Aboud and Moyles2000; Iyer & Ryan, Reference Iyer and Ryan2009; Subašić et al., Reference Subašić, Hardacre, Elton, Branscombe, Ryan and Reynolds2018), this has yet to be examined using the SIMCA framework and controlling for other relevant variables. In situating ingroup identification with SIMCA, we wish to understand whether identification with the advantaged and disadvantaged groups may conflict with each other. This will contribute to the still-limited research on the model in relation to advantaged groups.
Aside from the inclusion of identification with the advantaged group, our study also contributes to the body of work on the SIMCA by extending the research beyond Western, Educated, Industrialised, Rich, and Democratic or WEIRD populations (Henrich, Heine, & Norenzayan, Reference Henrich, Heine and Norenzayan2010). By testing the model in these two different countries with varying histories of collective action and situations of discrimination against women, we seek to provide stronger support for the universality of the SIMCA. That said, we do not hypothesise any difference between the two samples. Nevertheless, we do remain open to the possibility of finding challenges to the model to open up potential areas for further exploration. In doing so, we contribute to nuancing the application of the theory depending on the social context.
Little comparative research on allyship has been conducted, and by examining the SIMCA model for men's collective action against gender discrimination in these two different countries, our research can contribute to further strengthening support for the theory in contexts where it is not as commonly explored — or to providing novel challenges to the model.
Method
Design and Procedure
The study used a correlational design, utilising surveys for data collection. It is part of a larger research project comparing the impact of emotions on collective action in three countries. Participants were recruited in exchange for course credit from large undergraduate psychology classes and told that they would be answering a survey on their opinions about social issues. They were briefed about the nature of the study and asked for consent for participation. They rated their emotions towards their given situation of discrimination against women, and a number of collective action measures. Afterwards, they were debriefed about the objectives of the study.
The instruments were administered using pen-and-paper surveys. The questionnaires were written in English for the Philippine sample, and translated into Japanese for the Japan sample (after back translation had confirmed the appropriateness of the translated scales).
Participants
A total of 234 male undergraduate students (M age = 19.58, SD = 1.82) from the Philippines (n = 131) and Japan (n = 103) were included in the study.
Measures
SIMCA measures
We measured the variables within the SIMCA, with items rated on a scale from 1 (strongly disagree) to 7 (strongly agree). Four items each measured identification with men (α = .86) and identification with women (α = .81) and were written in parallel (e.g., ‘I identify with other men’ and ‘I can identify with the struggles of women’; van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011). Four items were also used to measure moral conviction (α = .82; e.g. ‘Ending discrimination against women is part of the core of my moral convictions’), adapted from Zaal and colleagues (Reference van Zomeren, Postmes, Spears and Bettache2011), and group efficacy (α = .92; e.g. ‘As men, I think we can change this situation of discrimination against women’), from van Zomeren et al. (Reference van Zomeren, Postmes and Spears2012). We measured anger using a three-item scale (α = .93) that asked participants to rate how angry, furious, and outraged they felt about the situation of discrimination against women, on a scale from 1, not at all, to 7, very much (Iyer, Schmader, & Lickel, Reference Iyer, Schmader and Lickel2007; Scherer, 2005; van Zomeren et al., Reference van Zomeren, Spears, Fischer and Leach2004; van Zomeren et al., Reference van Zomeren, Postmes and Spears2012).
Collective action measures
Eight items assessed participants’ willingness to engage in collective action on behalf of women (α = .87). Two questions adapted from van Zomeren and colleagues (Reference van Zomeren, Postmes and Spears2012) asked about more general collective action tendencies (e.g., ‘I would like to do something against this situation’). The next six items pertained to specific collective action behaviors in response to the question, ‘Thinking about the things you would do to eliminate discrimination against women, how willing are you to engage in the following actions? (e.g., “Sign a petition”)’.
Results
The overall mean scores of the variables (see Table 1) reflect responses close to the midpoint, except for ratings of group efficacy (M = 5.86, SD = 1.13), with scores showing a comparatively high sense of efficacy among men. The Philippine sample appeared to have higher means on identification, efficacy, moral conviction, and action intentions than the Japanese sample, but not more anger, where both countries had similar ratings (M JP = 4.65, SD = 1.73; M PH = 4.89, SD = 1.70). However, direct comparisons between means could not be performed due to the lack of measurement invariance. Bivariate correlations (refer to Table 2) demonstrate a positive relationship between willingness to engage in collective action, and all of the variables in the model. Notably, all variables were significantly correlated at the zero-order level with willingness to engage in collective action, except for identification with men (r JP = −.12, p = .21; r PH = .10, p = .25).
Table 1 Descriptive Statistics
Table 2 Correlations
Note: *p < .05, **p < .01, ***p < .001.
Measurement invariance was not established between the two countries (see Supplementary Tables 1–6 in the Supplementary data). Due to the lack of invariance, we ran separate hierarchical regressions for Japan and the Philippines to examine the replication of SIMCA in each country, with the addition of identification with men as a predictor. Moral conviction was entered in the first step, as this is theorised to drive identification with the disadvantaged group (van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011). This identification, along with identification with the advantaged group, was entered in the second step, followed by group efficacy and anger, as identification with the disadvantaged group is said to influence group efficacy and anger.
As hypothesized, moral conviction significantly predicted collective action intentions in both countries (Model 1: Japan β = .57, p < .001, Philippines β = .51, p < .001), with these effects reduced upon adding in identification (Model 2: Japan β = .48, p < .001, Philippines β = .39, p < .001), and group efficacy and anger (Model 3: Japan β = .35, p = .001, Philippines β = .19, p = .048) in the model. Similarly, group efficacy was a significant predictor of willingness to engage in collective action (Japan β = .37, p < .001, Philippines β = .27, p < .001). Contrary to the SIMCA, anger did not predict men's willingness to engage in collective action significantly in either country (Japan β = .02, p = .80, Philippines β = .16, p = .05).
Country differences were apparent in the role of identification. Identification with women was significant only in the Philippines (β = .24, p = .008), even after factoring in group efficacy and anger (β = .22, p = .011). However, identification with the disadvantaged group was not a significant predictor in Japan (Model 2 β = .15, p = .16; Model 3 β = .08, p = .44). Instead, identification with men predicted lower collective action intentions in Japan (β = −.16, p = .048), even when efficacy and anger were controlled (β = −.17, p = .031). This form of identification had no such effect in the Philippines (Model 2 β = −.01, p = .88; Model 3 β = −.02, p = .80). The final SIMCA model including identification with men predicted 46.6% of the variance in collective action intentions in the Japanese sample, F(2, 97) = 9.37, p < .001, and 38.2% of variance in the Philippine sample, F(2, 125) = 8.31, p < .001. Post hoc power analyses revealed that the model as a whole was well powered (PH = .99; JP = .99). However, the test of the role of identification specifically was lower in power for both countries (PH = .43, JP = .33); larger samples would be recommended for future research as the effect may be relatively small.
Discussion
We examined whether the SIMCA, along with the additional variable of identification with the advantaged group, predicted men's willingness to engage in collective action on behalf of women in Japan and the Philippines. Consistent with the SIMCA and previous research on allies’ collective action, moral conviction was associated with greater willingness to engage in collective action in both countries (Barth et al., Reference Barth, Jugert, Wutzler and Fritsche2015; Precopio & Ramsey, Reference Precopio and Ramsey2016; van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011), as was group efficacy (Stewart et al., 2011; van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011). Contrary to expectations, however, anger did not significantly predict willingness to engage in collective action on behalf of women, while the role of identification varied between the two countries. Identification with the disadvantaged group of women predicted greater willingness to engage in collective action for men in the Philippines, while identification with the advantaged group of men was associated with lower willingness to engage in collective action for women among Japanese men, thus providing only partial support for our hypotheses in each case.
The SIMCA and Men's Allyship
To our knowledge, the present research provides the first test of the SIMCA in Japan and the Philippines, as well as being among few studies of men's allyship; it is a contribution of the article to extend the scholarly literature in these areas beyond WEIRD samples, and reveal nuances between the universal and particular elements of theories such as the SIMCA. The extended model accounted for approximately 40% of the variance in men's allyship intentions in each case, which is an impressive and important achievement, demonstrating the robust ability of the SIMCA to model advantaged group action on behalf of disadvantaged groups. Moreover, consistent with van Zomeren and colleagues (Reference van Zomeren, Postmes, Spears and Bettache2011), as noted above, the role of moral convictions in allyship collective action was replicated, and so was the role of efficacy and (for the Philippine sample) the role of identification with the disadvantaged group.
However, there were also challenges to the model; in particular, with anger a non-significant predictor for both samples. The lack of significance in the role of anger appears to be inconsistent with previous SIMCA research on allies’ collective action from Europe and Hong Kong (van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011). One possibility to explain the inconsistency is difference in the measurement of anger. In our research, anger was measured in relation to a specific example of discrimination against women recalled by the men in the sample, whereas van Zomeren and colleagues’ (Reference van Zomeren, Postmes, Spears and Bettache2011) work focused more generally on anger towards perceived injustice against women. A second methodological explanation is that, as other researchers have demonstrated, the motivating effect of anger may depend on the availability of a relevant target (van Zomeren, Reference van Zomeren2016; van Zomeren, Saguy, Mazzoni, & Cicognani, Reference van Zomeren, Saguy, Mazzoni and Cicognani2018). This may especially be the case among advantaged group members, as the target of anger may vary (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010; Iyer & Ryan, Reference Iyer and Ryan2009; Leach, Snider, & Iyer, Reference Leach, Snider, Iyer, Walker and Smith2002) — some men may experience anger as moral outrage towards the injustice experienced by women, while others may express this as moral indignation towards specific men only, or even what they perceive to be undeserved advantages received by women. Underlying these differences in the forms of the advantaged group's anger are their appraisals of the stability and legitimacy of the current social structure (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010; Iyer & Ryan, Reference Iyer and Ryan2009; Leach et al., Reference Leach, Snider, Iyer, Walker and Smith2002; van Zomeren, Reference van Zomeren2016). Future research may thus examine these different types of anger in promoting and inhibiting collective action among advantaged groups.
The Role of Advantaged Group and Disadvantaged Group Identification
Our results on the role of identification likewise are partially inconsistent with past research. As with previous research on the SIMCA among advantaged groups (van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011), identification with the disadvantaged group, but not with the advantaged group, predicted willingness to engage in collective action in the Philippine sample. On the other hand, the results from the Japanese sample reflect past research on men's support for gender equality, where higher identification with the advantaged group has been linked to lower support for women (e.g., Burn et al., Reference Burn, Aboud and Moyles2000; Iyer & Ryan, Reference Iyer and Ryan2009; Subašić et al., Reference Subašić, Hardacre, Elton, Branscombe, Ryan and Reynolds2018). To understand these contrasts, we may again turn to appraisals of the relative positions of the advantaged and disadvantaged groups.
Although the current study did not measure perceived legitimacy, Japan's low rankings in the Gender Gap Index (World Economic Forum, 2017), particularly in women's economic participation and opportunity, educational attainment, and political empowerment, may be reflective of long-standing structures that legitimise gender discrimination. Recognising this persistent inequality and its impact on their economy, the Japanese government has been pushing for more equal employment and advancement opportunities for women in their policies in recent years (MacNaughtan, Reference MacNaughtan2015; Nae, Reference Nae2017; Schieder, Reference Schieder2014). Such initiatives may make salient the instability of the advantaged position of men, and be seen by high identifying men as a threat (Leach et al., Reference Leach, Snider, Iyer, Walker and Smith2002), lowering their willingness to engage in collective action. Given these factors, it may be that in contexts where differences between advantaged and disadvantaged groups are perceived as legitimate and unstable, higher identification with the advantaged group may be associated with lower willingness to engage in collective action. Indeed, system justification, or the belief in the fairness and legitimacy of one's social system, has been associated with a backlash against feminism (Yeung et al., Reference Yeung, Kay and Peach2014). Future researchers may thus investigate how system justification and perceptions of the socio-structural relationships between men and women may interact with advantaged group identification to predict collective action.
In contrast to the Japanese context, gender discrimination and inequality may be more likely to be considered illegitimate and non-normative in the Philippines, which is more egalitarian according to the Gender Gap Index (World Economic Forum, 2017). Within such a context, identification with the disadvantaged group may play a more potent role for men's willingness to engage in collective action, as men could feel less threatened by advancing women's rights given the latter's present visibility in workplace, educational, and political settings. This possible interaction between appraisals of legitimacy and identification with the advantaged and disadvantaged groups can be an area for further examination in subsequent research. As van Zomeren (Reference van Zomeren2016) suggested, links between structural and individual or group level factors may be investigated to better understand their contributions to understanding collective action.
Although our research reflects different patterns in these two countries, it is important to stress that the present samples are not nationally representative. We would not want to overstress national differences based on our data, given that they are confounded with other factors. For example, the results in the Philippines could also be associated with having recruited from a large state university in the capital city of the country known for its progressive views and activism, versus having recruited in a private Catholic university in Japan, which might be expected to be more socially conservative. For our purposes, the key point is that differences in egalitarianism within the social context might moderate the role of identification with the advantaged group (with more discriminatory contexts making men's identity more of an inhibiting factor opposing feminist collective action). Men's identification with the disadvantaged group, women, may be more likely to emerge as a predictor of their action to oppose discrimination in contexts where women are more equal to men and there is a culture for men of solidarity with women. Future research should test these interactions empirically.
Implications, Strengths, and Limitations
As the present data confirm, moral conviction and group efficacy play key roles as predictors of men's collective action on behalf of women. Given this, men may be encouraged to promote gender equality by communicating opposition to discrimination as an absolute moral issue, and increasing their sense that men, as a group, are a potent force in advancing women's rights.
The lack of a significant association between anger and men's collective action for women in both contexts suggests that the targets and experience of anger need more nuanced examination. Our study was not able to distinguish among different types of anger that may be experienced by the participants based on their appraisals of the legitimacy and stability of the relative positions of their group with that of women (Iyer & Leach, Reference Iyer, Leach, Stürmer and Synder2010; Iyer & Ryan, Reference Iyer and Ryan2009; Leach et al., Reference Leach, Snider, Iyer, Walker and Smith2002) or the target of this anger (van Zomeren et al., Reference van Zomeren, Saguy, Mazzoni and Cicognani2018). This is one area that future researchers may pursue in understanding advantaged groups’ collective action for the outgroup. Alternatively, it may be the case that anger simply plays a less consistent role for allies, or for respondents from Asian contexts (cf., van Zomeren et al., Reference van Zomeren, Postmes, Spears and Bettache2011). Future research examining allyship and non-WEIRD samples will illuminate these issues.
The role of group identification, on the other hand, may be better understood by investigating appraisals of the legitimacy and stability of the relative positions of women and men, as well as group norms based on these appraisals, systematically across cultural contexts. This is one limitation of the present study, which researchers can investigate further. Moreover, we note the small sample sizes in the two countries and relatively small effects of the identification variables. Thus, future work with larger samples may examine the role of identification with both the advantaged and disadvantaged groups alongside appraisals of legitimacy, stability, and group norms in different cultural contexts.
With these initial, tentative directions in relation to group identification, researchers may explore the impact of different targets for interventions to reduce gender discrimination. In some contexts such as the Philippines, where identification with women matters more, collective action may be encouraged by increasing identification with women and feminist men. For instance, Wiley and colleagues (Reference Wiley, Firnhaber and Shilinsky2012) suggest more positive portrayals of feminist men, as current associations with feminism tend to be quite negative (Anderson, Reference Anderson2009).
Meanwhile, in the Japanese context, and others where identification with men appears to inhibit collective action, interventions may consider changing definitions of the male identity and attempting to reduce the threat associated with pushing for equality. Other possibilities focusing on advantaged group identity may also be considered: category inclusion or treating the disadvantaged group as part of a common ingroup (Barth et al., Reference Barth, Jugert, Wutzler and Fritsche2015; Reicher, Cassidy, Wolpert, Hopkins, & Levine, Reference Reicher, Cassidy, Wolpert, Hopkins and Levine2006; Subašić et al., Reference Subašić, Hardacre, Elton, Branscombe, Ryan and Reynolds2018), category norms or communicating helping as part of the ingroup's core identity, and category interest or creating the perception that persecuting the outgroup will bring harm to the ingroup (Reicher et al., Reference Reicher, Cassidy, Wolpert, Hopkins and Levine2006). However, cultural and structural factors may need to be examined further to understand this impact of identity on willingness to engage in collective action (van Zomeren, Reference van Zomeren2016; van Zomeren & Louis, Reference van Zomeren and Louis2017).
Conclusions
Drawing on two samples from Japan and the Philippines, the present data highlight the power of moral convictions and efficacy to mobilise men to support feminist collective action. The findings make a contribution to the literature by attesting the utility of using a general model of social identity and collective action (SIMCA; van Zomeren, Postmes, & Spears, Reference van Zomeren, Postmes and Spears2008; van Zomeren, Postmes, & Spears, Reference van Zomeren, Postmes and Spears2012) to investigate allyship in novel contexts with non-WEIRD samples. But the studies also provide three contrasts with previous research. They demonstrate null findings for anger — which was not mobilising for activism for either Japanese or Filipino men. They also yielded inconsistent findings for identification, where identification with men was demobilising for Japanese men but not a barrier to feminism for the Filipino sample, and identification with women was mobilising for Filipino men but not their Japanese counterparts. In challenging ongoing discrimination against women in the 21st century, activists as well as researchers may profit from closer scrutiny of the normative or political factors that moderate the antecedents of feminist collective action, in order to design or test specific motives that are mobilising or demobilising.
Table 3 Willingness to Engage in Collective Action Predicted by Identification With Men and the Social Identity Model of Collective Action in Japan
Note: *p < .05, **p < .01, ***p < .001.
Table 4 Willingness to Engage in Collective Action Predicted By Identification With Men and the Social Identity Model of Collective Action in the Philippines
Note: *p < .05, **p < .01, ***p < .001.
Acknowledgments
We would like to acknowledge support from the K.S. Yang Award for the Asian Association of Social Psychology Summer School Group Research.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/prp.2018.30
Open access
