Religious and Ethnic Discrimination: Differential Implications for Social Support Engagement, Civic Involvement, and Political Consciousness

Social identity threats, depending on the content of the identity targeted, may evoke varying socio-political responses. In this regard, religious discrimination may be especially threatening, challenging both the social group and its belief system, thereby promoting more active collective responses. This research examined how religious and ethnic identification differentially evoked engagement with support resources (ingroup and spiritual), civic involvement (including individual and collective action-taking), and political participation (voting or political consciousness) following group-based threats. Study 1 drew from the Canadian Ethnic Diversity Survey (N = 1806). Participants who reported religious discrimination demonstrated greater religious identification, ingroup social engagement, and civic involvement—comparable associations were absent for ethnic discrimination. Study 2 (N = 287) experimentally primed participants to make salient a specific incident of religious or ethnic discrimination. Although ethnic discrimination elicited greater ingroup support-seeking and political consciousness, religious discrimination was perceived as especially harmful and evokedmore individual and collective action-taking. Further to this, religious high-identifiers’ responses were mediated by engagement with ingroup or spiritual support in both studies, whereas no mediated relations were evident for ethnic identification. Findings are discussed in terms of distinct socio-political responses to threats targeting identities that are grounded in religious belief systems.

Although considerable research has examined responses to inequality within a specific sphere of identity (e.g., minority vs. majority ethnicities; Deaux, Reid, Martin, & Bikmen, 2006), little attention has been paid to whether active responses to discrimination might be differentially elicited across particular identities (e.g., ethnic vs. religious discrimination). Indeed, the nature and magnitude of responses to discrimination may depend upon the content of the identity and the unique meaning that it holds for group members (Livingstone & Haslam, 2008;Thomas et al., 2009). In this regard, like other identities, religious group membership provides a basis for social belonging.
Yet religious identities are founded on revered belief systems anchored in historical continuity and symbolic significance (Haslam, Reicher, & Platow, 2010). Thus, on one hand, these belief systems inherently provide a spiritual coping resource to deal with the challenges that individuals and groups encounter that is qualitatively different from other protective aspects of social identity. On the other hand, the impact of religious threats compared to other forms of discrimination might differ due to the broader ramifications of a threat that challenges both the social group and sacred or moral value system (Täuber & van Zomeren, 2013;van Zomeren, Leach, & Spears, 2012).
Indeed, intensified responses (e.g., moral outrage; Täuber & van Zomeren, 2013) to religious threats might even surpass threats to other belief-based social identities (e.g., political identity) due to the epistemological and ontological certainty inherent to religious ideals (Ysseldyk et al., 2010). As a result, analogous to responses linked to identification with social movements (Simon et al., 1998;Thomas et al., 2009), active efforts to protect the identity may be intensified when one's religious group membership, and by extension one's beliefs, are attacked.
Notwithstanding considerable overlap with ethnicity in some cases (see Chirot & Seligman, 2001 for an overview of ethno-political conflicts), religious identification has been posited to be a predominant factor in fueling ostensibly ethnic conflicts having political consequences, like those witnessed in Bosnia-Herzegovina, India, and Ireland (Leach & Williams, 1999;Muldoon, Trew, Todd, Rougier, & McLaughlin, 2007; see Ysseldyk et al., 2010, for an overview). Religious beliefs and practices have similarly served as a resource for political consciousness and mobilization among both African-Americans and Euro-Caucasians (Harris, 1994), as well as Muslim-Americans (Read, 2007). Even among mainstream members of the Muslim faith in the Netherlands (where the Islamic faith has been publicly condemned by some political leaders; Verkuyten & Zaremba, 2005), religious identification was positively linked to attitudes endorsing Muslim political organization (Phalet et al., 2010;Verkuyten & Yildiz, 2007), and to the justification of terrorism in the Commonwealth of Independent States when the perpetrator of the violence was believed to be a Muslim ingroup member (Fischer, Greitemeyer, & Kastenmüller, 2007). Thus, given that the goal of religious conflict is often to influence the number of people who have adopted and adhere to that set of beliefs (a conversion that is not possible where ethnicity is concerned; Worchel, 2004;Ysseldyk et al., 2010), along with a sense of absolute moral empowerment that sometimes fuels religious beliefs Wellman & Tokuno, 2004), religious identification may prove particularly powerful in motivating socio-political responses when that identity is threatened.

The Role of Engagement With Support Resources
Given the collective nature of political consciousness and civic action, social support from other members of the targeted ingroup should be especially vital for determining those responses to discrimination. Although one might expect discrimination to be most distressing to those who strongly identify with their group, such individuals are often least likely to report distress, especially in the case of groups that offer high levels of support to their members (Branscombe et al., 1999;Haslam, O'Brien, Jetten, Vormedal, & Penna, 2005). Social support can be both emotional (e.g., expressions of concern and care; attitudinal or opinion support) and/or instrumental in nature (e.g., material assistance; tangible support aimed at changing one's situation; Bertera, 1997;Dakof & Taylor, 1990;van Zomeren et al., 2004). It can act as a buffer against distress by moderating appraisals of discriminatory events from constituting a personal trial to perceiving a shared, group-based experience (especially among individuals whose group identity is central to their self-concept; Haslam & Reicher, 2006;Muldoon, Schmid, & Downes, 2009), or by attenuating negative psychological outcomes (Cohen & Wills, 1985;Foster, 2000). Moreover, it may not only protect against distress, but might also prompt group members to engage in collective responses to deal with the threat (Foster, 2000;van Zomeren et al., 2004). For example, coping with gender discrimination through social support-seeking has been associated with increased political and collective action aimed at enhancing women's group status (Foster, 2000;Foster & Matheson, 1995). Likewise, social support proved vital in promoting collective action tendencies among students dealing with perceived injustice within their university (van Zomeren et al., 2004).
Yet, it is conceivable that in the face of discrimination, highly-identified ethnic group members may well rely on support from other ethnic ingroup members as a basis for political consciousness and action-taking (see McAdam, 1982). Similarly, attendance at religious services (arguably reflecting engagement with social support resources derived from one's religious group) has been linked to enhanced support for action against a threatening religious outgroup, whereas more personal aspects of religion, such as prayer, were not (Ginges, Hansen, & Norenzayan, 2009). However, other research has suggested that both the extrinsic (social) and intrinsic (spiritual) aspects of religiosity were associated with intentions to take action against the perpetrators of religious discrimination (Ysseldyk, Matheson, & Anisman, 2011). In light of the unique, dual structure of religious identity (Wellman & Tokuno, 2004;Ysseldyk et al., 2010), it may be that highly religious individuals would be more likely to seek support from God and depend on their own spiritual capacity Stone, Cross, Purvis, & Young, 2003) to cope with difficult life experiences and rely less on support from their religious community. Nonetheless, given that regular attendance at congregational services may bring continuity and stability to religious support networks (Lim & Putnam, 2010;Stone et al., 2003), coupled with the resource of common values and faith, support from others sharing both group ties and beliefs could motivate highly-identified religious individuals toward greater political awareness and social action to deal with threats that directly target their religious affiliation. In effect, given the dual nature of religious identification, the support pathways through which religious high-identifiers may feel empowered to engage in civic and political action may be two-fold in nature, having both ingroup and spiritual components.

The Present Research
Two studies assessed the relations between religious identification and socio-political responses among people who had experienced a religious identity threat (i.e., discrimination), and the extent to which those relations differed compared to threats targeting a comparable identity, namely ethnicity. To this end, we also assessed the potential mediating roles of engagement with ingroup and spiritual support resources. Study 1 determined the prevalence of, and factors associated with, experiencing ethnic or religious discrimination in Canada using a nationally-rep- Socio-Political Responses to Religious Discrimination 350 resentative sample. Study 2 primed participants to make salient a specific incident of either religious or ethnic discrimination, with the aim of more directly identifying support-seeking and socio-political responses that might differ based on the content of such threats.

Study 1
Previous research (e.g., Thomas et al., 2009;Ysseldyk et al., 2010) suggests that the relations between identification and socio-political outcomes may differ depending on the group targeted by an identity-relevant threat. To examine this proposition, we employed the Ethnic Diversity Survey (EDS;Statistics Canada, 2003), which included a large number of both ethnic and religious minority and majority respondents. These data allowed us to compare engagement with support resources and socio-political responses among those who reported ethnic versus religious discrimination, as well as those who reported no discrimination at all.
Canada was chosen as an ideal place for this evaluation in light of the rich religious and cultural diversity of its Given that religious discrimination may be especially threatening, responses across the threat types were expected to vary in that higher levels of socio-political responses (i.e., civic involvement and voting) as well as engagement with ingroup and spiritual support resources would be more evident among those who reported religious (compared with ethnic or no) discrimination (H1). In terms of identification processes in general, it was expected that individuals who highly-identified with their religious or ethnic group would report greater a) engagement with ingroup support, b) political participation, and c) civic involvement following discrimination compared with low-identifiers (H2).
However, support-seeking was expected to most strongly mediate the relations between identification and sociopolitical responses following religious (compared with ethnic or no) discrimination (H3), as those responses would be bolstered by spiritual strength and support from others who share group ties as well as beliefs when a religious identity threat was experienced.

Data
The data for Study 1 was from the EDS conducted by Statistics Canada in 2003. The aim of this survey was to better understand how people's backgrounds affected their participation in the social, economic, and cultural life of Canada (Statistics Canada, 2003). Respondents were interviewed by telephone in the 10 provinces, in the two official languages (English and French) as well as in several non-official languages. The survey used the 2001 census as a sampling frame with 42,476 people living in private households. Respondents provided detailed information on topics ranging from ethnic identity, religious participation, knowledge of languages, family background, social networks, and civic participation, to interaction with society, attitudes, satisfaction with life, trust, and socioeconomic activities (Statistics Canada, 2003). Of particular interest for the present study were measures relating to religious and ethnic identification, social networks, civic and political participation, and discrimination.

Participants
Given the focus of the present research on responses to ethnic and religious discrimination, a subset of the EDS sample was used (N = 1806). We included only those participants who indicated some religious affiliation (i.e., atheists and agnostics were excluded). Initial analyses were conducted to compare variables of interest between those who reported ethnic discrimination (n = 769), religious discrimination (n = 480), and a random sample of those who reported no discrimination of any kind (n = 557). The experience of ethnic or religious discrimination was determined by including those respondents who answered in the affirmative to the question "In the past 5 years/Since arriving in Canada, do you feel that you have experienced discrimination or been treated unfairly by others in Canada because of your ethnicity, culture, race, skin colour, language, accent or religion?", and indicated in a follow-up question (asking them to identify the basis of that discrimination) that it had been exclusively on the Similarly, identification with one's religious group was assessed: "Using a scale of 1 to 5, where 1 is not important at all and 5 is very important, how important is your religion to you?" (M = 3.85, SD = 1.29) (r = .10, p = .001).

Religious Engagement
The EDS assessed engagement with religious support networks in terms of both social interactions as well as the value placed on more intrinsic aspects of the identity. Engagement with one's religious social group was evaluated with the question: "In the past 12 months, how often did you participate in religious activities or attend religious services or meetings with other people, other than for events such as weddings and funerals?" Additionally, as an index of spiritual engagement, participants were asked: "In the past 12 months, how often did you do religious activities on your own? This may include prayer, meditation and other forms of worship taking place at home or in any other location." Participants' responses to both questions could range from 1 (at least once a week) to 5 (not at all), and were subsequently reverse-scored such that higher scores reflected greater religious ingroup (M = 3.44, SD = 1.52) and spiritual (M = 3.96, SD = 1.50) engagement (r = .48, p < .001).

Ethnic Engagement
To assess ethnic ingroup engagement, participants were asked: "As far as you know, how many of your friends have [participant's ethnic background] ancestry?", to which responses could range from 1 (all of them) to 5 (none of them) (reverse-scored, such that higher scores indicated more ethnic social ties; M = 2.66, SD = 1.17). In addition, as a measure of engagement with the "intrinsic" aspects of ethnic identity, they were asked "Using a scale of 1 to 5, where 1 is not important at all and 5 is very important, how important is it for you to carry on [participant's ethnic background] customs and traditions, such as holidays and celebrations, food, clothing or art?" (M = 1.62, SD = 0.95) (r = .29, p < .001).

Political Participation
The index of political participation combined responses to three voting questions: "Did you vote in the last federal/provincial/municipal election?". Responses were re-coded (no = 0, yes = 1) and mean scores served as an index of political participation (M = 0.71, SD = 0.39; α = .87). Analyses involving political participation only included those respondents who were eligible to vote, namely Canadian citizens aged 18+ years.

Civic Involvement
Civic involvement was assessed with the question: "Are you a member of, or have you taken part in the activities of, any groups or organizations at any time in the past 12 months?". The total number of civic groups in which individuals participated was used as an index of civic involvement (M = 0.76, SD = 0.88).

Assessing the Role of Discrimination on Identification, Engagement, and Socio-Political Responses
Initial analyses were conducted to assess the extent to which religious and ethnic identification, engagement (ingroup, spiritual, and traditions), and social action (political and civic) differed as a function of self-reported ethnic discrimination, religious discrimination, or no discrimination (H1). iii Multivariate analyses of covariance (MANCO-VAs; controlling for age, gender, household income, and education) revealed several significant differences as a function of the type (or presence) of discrimination. With regard to the religious variables, Pillai's Trace = 0.108; F(6, 2924) = 27.72, p < .001, η 2 = .054, individuals who experienced religious discrimination were more highlyidentified with their religious group, and reported more religious ingroup social and spiritual engagement than did those who had experienced ethnic or no discrimination (see Table 1). However, individuals who experienced ethnic discrimination also appeared to engage in spiritual support-seeking to a greater extent than those who reported no discrimination, corroborating previous findings that suggest individualized religion (e.g., prayer) may be an important coping mechanism even in the context of threats targeting other group memberships.  .74*** 12 Note. Means in the same row that do not share subscripts differ at p < .05; values for covariates (age, gender, household income, education) were set at their mean values. Analyses involving political participation exclusively include Canadian citizens of 18+ years. *p < .05. **p < .01. ***p < .001.
Few differences as a function of discrimination emerged with regard to the ethnic variables, Pillai's Trace = 0.015; F(6, 1466) = 1.89, p = .08, η 2 = .008. As seen in Table 1, neither ethnic identification nor ethnic ingroup engagement differed as a function of discrimination. However, those who reported ethnic discrimination were significantly least likely to value their ethnic group's customs and traditions (compared to the religious discrimination or no discrimination groups).
Finally, civic, but not political, participation differed as a function of discrimination. As seen in Table 1, individuals who reported religious discrimination also reported more civic involvement (i.e., greater participation in organized groups) than individuals who experienced ethnic or no discrimination, while individuals who reported ethnic discrimination reported more civic involvement than those who did not report any discrimination as well.

The Mediating Role of Support Engagement
In addition to the role of threat type (i.e., religious or ethnic) in participants' levels of identification, engagement with support networks, and social participation, the relations among those variables also varied as a function of the type of discrimination (H2; see Table 2). Thus, it was of interest to assess the potential mediating role of engagement with various sources of support (H3; i.e., ingroup, spiritual, tradition) for participants who reported religious, ethnic, or no discrimination, with a focus on variables that were related to both the independent variable (i.e., identification) and the outcome(s) (i.e., political or civic participation). To assess these mediated models, we conducted regression-based bootstrapping analyses with 10,000 bootstrap samples and 95% confidence intervals (CIs), controlling for age, gender, household income, and education. These analyses provide evidence of mediation if the significant direct relationship between the predictor and outcome variable becomes non-significant and the CI does not include zero (Hayes, 2012(Hayes, , 2013  Note. Covariates include age, gender, household income, and education level. Correlations reflecting religious discrimination (n = 480) are above diagonal, ethnic discrimination (n = 769) below diagonal; correlations relevant to proposed mediating models are bolded. Analyses involving political participation exclusively include Canadian citizens of 18+ years. Partial correlations for respondents who did not report any discrimination are presented in Appendix A. *p < .05. **p < .01. ***p < .001. Table 2 (above the diagonal), among respondents who reported religious discrimination, religious identification was associated with greater ingroup social engagement, spiritual engagement, and greater civic (but not political) involvement. Similarly, religious ingroup engagement was related to both civic and political involvement, whereas spiritual engagement was associated with greater civic involvement alone.

Religious Discrimination -As seen in
When both religious ingroup and spiritual engagement were controlled in the relation between religious identification and civic involvement, the direct relation was reduced to non-significance (see Figure 1). Indeed, the 95% CI for religious ingroup engagement's mediated path did not include zero, suggesting that greater social engagement with members of one's religious ingroup fully accounted for the relation between religious identification and civic involvement among those who reported religious discrimination. Conversely, spiritual engagement did not mediate this relation. Thus, religious ingroup social (but not spiritual) engagement appeared to be vital in facilitating civic participation. It should be noted that when alternative directional models were assessed, he mediating role of ingroup engagement with civic participation as the predictor and religious identification as the outcome was significant.
Thus, it is plausible that civic participation fostered religious identification (direct β = .15, p = .001; controlled β = .03, p = .394) through engagement with other members of one's religious ingroup, B = .08, SE = .02, 95% CI = .04, .13. Figure 1. The mediating role of engagement with support resources between religious identification and civic involvement among respondents reporting religious discrimination in Study 1.
Note. Standardized regression coefficients (βs) between variables are reported in ovals; bootstrapped indirect effects (Bs; SEs) and 95% CIs for mediators are reported in rectangles and significant mediators are shaded; analyses control for demographic variables; dotted lines indicate where a previously significant relationship is rendered non-significant; n = 480. *p < .05. **p < .01. ***p < .001. Table 2 (below the diagonal), among respondents who reported ethnic discrimination, ethnic identification was positively associated with both ingroup social engagement and civic participation. However, these two variables were unrelated to each other; thus, social engagement did not account for the increased civic involvement associated with ethnic identification. Likewise, ethnic identification was neither related to engagement with ingroup traditions nor to political participation, and neither type of ethnic engagement (i.e., ingroup or tradition) was associated with political or civic participation. Thus, no mediated models were assessed among those who reported ethnic discrimination.

Ethnic Discrimination -As seen in
No Discrimination -Given the associations between identification, engagement with one's ingroup, spirituality, or traditions, and socio-political involvement among those who reported experiencing ethnic or religious discrimination, it was also important to assess those relations among those who did not report any discrimination (see Appendix A). Correlations revealed that ethnic identification was not associated with any other variables of interest.
However, like the relationships found among those who reported religious discrimination, religious identification was associated with both ingroup and, spiritual engagement, yet not with civic involvement or political participation.
Thus, given the lack of significant relationships between identification and the key outcome variables of interest, no mediated models were assessed among those who did not report any discrimination.

Summary
Taken together, the results of Study 1 suggest that religious discrimination was more likely to evoke a variety of responses (e.g., greater identification, civic involvement) when compared to participants who had not experienced any discrimination, but even compared to those who reported ethnic discrimination (H1). However, in contrast to previous research (Simon & Klandermans, 2001;Verkuyten & Yildiz, 2007), neither type of identification was associated with political participation, perhaps due to a loss of political trust in the wake of discrimination (Michelson, 2003). Nonetheless, in addition to the direct association between identification and civic involvement (H2), identifying with one's religious group was also indirectly linked to greater civic involvement through the tendency for religious high-identifiers to engage with other members of their religious ingroup (H3). Importantly, the same mediating roles of engagement were not observed with ethnic identification and threat.
These findings suggest that religious identification may have a unique role in fostering civic involvement, especially among those who have experienced religious discrimination, in part because religious identity itself serves as a platform for greater engagement with ingroup support networks (Putnam & Campbell, 2010).
Despite these outcomes, Study 1 has some limitations in terms of the present paper's aims. Although serving as a nationally-representative sample, participants' responses were not directly framed in the context of dealing with their discrimination, but rather more general responses regarding engagement with support networks and sociopolitical involvement were measured. This may have also contributed to the relatively low effect sizes observed.
In addition, the measures used to assess identification and engagement varied for religious versus ethnic identities (e.g., religious spirituality and ethnic traditions), and thus might not be directly comparable. Finally, although responses across those who reported religious and ethnic discrimination were examined, many respondents identified their foremost ethnic ancestry as Canadian; although this was not a surprise in the Canadian context, it is, of course, also a national identification. Thus, Study 2 was designed to expand on Study 1 as well as to address these shortcomings.

Study 2
Study 2 expanded on Study 1 by assessing whether the same patterns would be evident (a) when participants were primed to recall a specific incident of religious or ethnic discrimination, (b) when a more nuanced measure of political consciousness (rather than simply voting) was included, thereby including non-citizens as well (Ramakrishnan & Bloemraad, 2008), and (c) when more detailed measures of social support-seeking and actiontaking in response to the specific event (rather than general engagement with support networks and civic involvement) were employed . In addition, a direct measure of appraisals of the stress associated with the discriminatory event was included in Study 2, to better determine whether religious discrimination was perceived as a threat to the individual. Analogous to Study 1, religious (compared with ethnic) discrimination was expected to more strongly evoke ingroup and spiritual support-seeking, action-taking, and political consciousness (H1). Also like Study 1, it was anticipated that identification would be positively associated with support-seeking and those same socio-political responses (H2), but that important variations would be evident as a function of the type of discrimination (religious or ethnic). Specifically, the role of support-seeking was expected to play a greater role in accounting for the relations between religious (compared with ethnic) identification and socio-political outcomes (H3).

Participants and Procedure
A community sample of Canadian adults (62 men, 225 women) ranging in age from 17 to 65 years (M = 32.41, SD = 11.44) participated in an online study described as assessing group membership and ways of coping with stressful life events. Individuals were recruited through several religious and community-based organizations (e.g., churches, community centres) and websites (related to cultural and religious diversity). Data validity checks were conducted (e.g., whether reverse-scored items were answered in an appropriate direction compared with other items) and data from participants who seemingly responded at random were deleted. As in Study 1, a large portion of participants identified their religious affiliation as Christian (Catholic, n = 76; Protestant, n = 81). The remaining participants self-identified as Muslim (n = 43), Bahá'í (n = 29), Jewish (n = 21), Buddhist (n = 21), Hindu (n = 13) or Sikh (n = 3). Just over half of participants indicated that they were Euro-Caucasian (n = 151), whereas the remainder self-identified as South-Asian (n = 42), Asian (n = 37), Arab/West-Asian (n = 31), Black (n = 7), Hispanic (n = 1), or as being of mixed ethnicity (n = 15). Three participants also reported their ethnicity as "other", specifying "Jewish"; however, those three participants also self-identified their religious affiliation as Jewish.
After being provided with a written overview of the study and giving informed consent, participants completed demographic questions and were randomly assigned to one of two conditions. In one condition, participants completed a measure of religious identification (to increase the salience of this identity) and then wrote a short description of an incident in which they experienced discrimination or felt threatened on the basis of their religious affiliation (based on Sahdra & Ross, 2007). Following this written threat recall, participants completed measures assessing religious ingroup social support, spiritual support, political consciousness, and actions they had taken to deal with the threatening event.
iv The second condition entailed identical measures, but pertained to ethnic identification and threat. Given that the assessment of responses used to deal with a specific threat situation (rather than general tendencies) was a primary goal, in contrast to Study 1, the design of Study 2 precluded the inclusion of a no-discrimination condition. Participants received compensation in the form of $10 gift cards, were debriefed, and given a list of contacts for counseling services should they experience any distress. Finally, in an effort to alleviate any adverse emotional effects that might have occurred as a result of recalling a discriminatory event, participants completed a written (online) task designed to elicit positive recollections of interactions with members of their own and other religious or ethnic groups.

Participants' Identity Threat Descriptions
Common themes among the religious identity threats that Christians (i.e., Protestants and Catholics) reported included experiences of discrimination while in foreign countries, insults and demeaning jokes at Christianity's expense, and acts of church vandalism. Threats recalled by religious minorities shared many commonalities with those reported by Christians, such as insulting jokes or demeaning remarks at the expense of one's religion, and acts of vandalism toward religious landmarks. However, as reported by others (Byng, 2008), those experiences often varied based on the characteristics of the religious group. Some experiences reported by religious minorities, for example, were especially severe in nature, including physical attacks and killing of family members or friends on the basis of their religious belief (e.g., Bahá'í persecution in Iran).
Among those who were asked to recall an ethnic identity threat, common themes among Euro-Caucasians included feelings of reverse-discrimination in the workplace, insults or discrimination in contexts where Caucasians were not the majority (e.g., foreign countries), and threats or physical attacks by those of other ethnic origins. In addition, discrimination experiences based on European heritage (e.g., "German Nazis") and accusations of being racist  & Garcia, 2005). Common themes among the threats that ethnic minorities described included race-based insults and derogatory name-calling, experiences of ostracism, and dealing with acts reflecting stereotypical attitudes (e.g., Arabs being extensively searched at airports; Blackwood, Hopkins, & Reicher, 2013).
In effect, the kinds of threats experienced among majority or minority groups were quite similar, whether they were referring to religious or ethnic group identities.

Measures Identification
Identification (Cameron, 2004) was measured in terms of religious or ethnic group membership, depending on the threat condition to which the individual had been randomly assigned (e.g., "Being a member of my religious/ethnic group is an important part of my self-image"; see Appendix B for all items included in Study 2). Participants rated nine items, ranging from 0 (strongly disagree) to 5 (strongly agree) (M = 2.93, SD = 1.03; α = .93).

Discrimination Appraisals
Appraisals of the discriminatory event were assessed along three dimensions (with three items each), namely  (Peacock & Wong, 1990). Participants rated each item from 1 (not at all) to 5 (extremely).

Ingroup Support
The Consumption of Social Support Scale (Bertera, 1997) assessed perceptions of religious or ethnic ingroup support, depending on the threat condition to which the individual had been randomly assigned. Participants rated how often they had used nine forms of social support (e.g., "You received some information from others which helped you understand the situation") from other religious or ethnic ingroup members in terms of dealing with their recalled threat, from 1 (not at all) to 5 (about every day) (M = 1.54, SD = 0.98; α = .93).

Spiritual Support
Three positive religious coping items from Pargament el al.'s (1998) Brief RCOPE were used to assess the extent to which participants sought spiritual support (e.g., "Sought help from God in letting go of my anger"). Respondents rated each item from 0 (not at all) to 4 (a great deal) in terms of how much or how frequently they engaged in each behaviour to deal with their discriminatory event (M = 1.32, SD = 1.13; α = .96).

Political Consciousness
To assess political attitudes related to one's religious or ethnic group, participants reported their agreement with four statements (e.g., "Members of my religious/ethnic group have to start to work together in order to gain political influence in Canada"; Verkuyten & Yildiz, 2007) on a scale ranging from 1 (strongly disagree) to 7 (strongly agree) (M = 4.37, SD = 1.78; α = .87).

Actions
Participants indicated the extent to which they took four types of action (based on Wright et al., 1990), ranging from 0 (not at all) to 6 (extremely) in response to their discriminatory event. These included requests for mediation either as an individual ("Sought advice from a governmental or non-governmental organization (

Results and Discussion
Assessing the Role of Discrimination Type As in Study 1, to examine whether responses to discrimination varied as a function of the group targeted (i.e., ethnic or religious; H1), MANCOVAs were conducted (controlling for age and gender). v As can be seen in Table   3, group identification did not vary across religious versus ethnic discrimination. However, analysis of appraisals of the discrimination event, Pillai's Trace = 0.016; F(3, 281) = 1.76, p = .154, η 2 = .019, indicated that religious discrimination was perceived as being especially central (i.e., having serious implications). Likewise, religious discrimination was also perceived as more threatening and challenging than ethnic discrimination, although these latter differences did not reach significance (Table 3). vi

Socio-Political Responses to Religious Discrimination 360
Variations were also evident in the support (ingroup, spiritual) sought or experienced as a function of religious or ethnic discrimination, Pillai's Trace = 0.091; F(2, 267) = 13.03, p < .001, η 2 = .066 (Table 3). Specifically, individuals who recalled ethnic discrimination reported more support from ingroup members. Conversely, individuals who recalled a religious threat reported more spiritual support-seeking than did those asked to recall an ethnic threat.
With regard to socio-political responses, political consciousness was greater among those who reported on ethnic discrimination (see Table 3), yet individuals who recalled religious discrimination indicated taking more solutionseeking and confrontational actions (at both individual and collective levels) in an effort to deal with the threat, Pillai's Trace = 0.03; F(4, 273) = 2.54, p = .04, η 2 = .043, although only the difference in collective solution-seeking actions reached significance.

The Mediating Role of Support
As in Study 1, several of the relations between identification and the other variables of interest varied as a function of the type of discrimination (H2; see Table 4 for correlations). Thus, the mediating role of support (i.e., ingroup, spiritual) was assessed for participants who reported religious or ethnic discrimination (H3), with a focus on variables that were related to both the independent variable (i.e., identification) and the outcome(s) (i.e., political consciousness or action) (Hayes, 2012(Hayes, , 2013. Note. Covariates include age and gender. Correlations reflecting religious discrimination (n = 144) are above diagonal, ethnic discrimination (n = 143) below diagonal; correlations relevant to proposed mediating models are bolded. + p < .10. *p < .05. **p < .01. ***p < .001.
Religious Discrimination -Among participants who reported religious discrimination, identification was associated with greater ingroup social support, spiritual support, political consciousness, and greater solution-seeking action at both individual and collective levels. Similarly, ingroup support was positively associated with political consciousness and both (individual and collective) solution-seeking and confrontational action, whereas spiritual support was associated with these same outcomes with the exception of individual confrontation (see Table 4, above the diagonal). Given this pattern of findings, the mediating roles of ingroup and spiritual support were assessed in the relations between religious identification and the outcomes of political consciousness, individual solution-seeking, and collective solution-seeking action. As seen in Figure 2a, when both religious ingroup and spiritual support were controlled in the relation between religious identification and political consciousness, the direct relation, although remaining significant, was reduced. Furthermore, the 95% CI for ingroup support's mediated path did not include zero, suggesting that ingroup support partially accounted for the relation between religious identification and heightened political consciousness following recollection of a discriminatory event that targeted one's religious group. Conversely, spiritual support did not mediate this relation. Note. Standardized regression coefficients (βs) between variables are reported in ovals; bootstrapped indirect effects (Bs; SEs) and 95% CIs for mediators are reported in rectangles and significant mediators are shaded; analyses control for demographic variables; dotted lines indicate where a previously significant relationship is rendered non-significant; n = 144. *p < .05. **p < .01. ***p < .001.
When ingroup and spiritual support were controlled in the relation between religious identification and individual solution-seeking, the direct relation became non-significant (Figure 2b). Indeed, the 95% CIs for ingroup and spiritual support's mediated paths did not include zero. Likewise, although the direct relation between religious identification and collective solution-seeking remained significant when ingroup and spiritual support were controlled, these relations were reduced and the 95% CIs for the mediated paths also did not include zero (Figure 2c). In effect, both ingroup and spiritual support-seeking partially, if not fully, mediated the relations between strong religious identification and greater inclinations to take both individual and collective solution-seeking actions.   Alternative directional models were also assessed and, akin to the patterns found in Study 1, the reverse was also evident. Specifically, the direct relation with individual as the predictor and religious identification as the outcome (β = .17, p = .024) was reduced to non-significance (β = .00, p = .998) when both ingroup support, B = .03, SE = .02, 95% CI = .01, .07, and spiritual support, B = .09, SE = .02, 95% CI = .05, .14, were controlled. Likewise, when collective action-taking served as the predictor and religious identification as the outcome, the direct relation (β = .49, p < .001) was reduced in significance (β = .23, p = .005) when both ingroup support, B = .05, SE = .02, 95% CI = .02, .08, and spiritual support, B = .09, SE = .03, 95% CI = .05, .15, were controlled. Table 4 (below the diagonal), among participants who recalled an incident of ethnic discrimination, identification was associated with greater ingroup social support, political consciousness, and solution-seeking actions both individually and collectively. Similarly, ethnic ingroup support was positively associated with individual and collective solution-seeking, but not political consciousness. Thus, it was possible that ingroup support would mediate the relations between ethnic identification and (both individual and collective) solution-seeking action. However, when ingroup support was controlled in the relation between ethnic identification and individual solution-seeking, the direct relation, β = .18, p = .013, remained significant, β = .15, p = .035, and the 95% CI for ingroup social support's mediated path included zero, B = .04, SE = .03, CI = -.01, .12. Similarly, when ingroup support was controlled in the relation between ethnic identification and collective solution-seeking, the direct relation, β = .22, p = .002, remained significant, β = .19, p = .008, and the 95% CI for ingroup social support's mediated path included zero, B = .06, SE = .05, CI = -.01, .18. In effect, as in Study 1, support from ingroup members did not account for the increased individual or collective solution-seeking action associated with strong ethnic identification following discrimination. Although alternative directional mediated models were assessed (e.g., the role of ethnic identification in accounting for greater action-taking among those with higher levels of ingroup support), no alternative mediators were significant (i.e., the 95% CIs for indirect effects all included zero).

Summary
Taken together, the results of Study 2 revealed that, as expected, religious discrimination was perceived as especially central to group members and evoked more spiritual support-seeking, individual and collective action-taking (H1). Although ethnic discrimination evoked greater support from ingroup members and greater political consciousness, that support appeared to do little to motivate ethnic high-identifiers' socio-political responses of any kind.
However, religious identification was associated with a broad array of responses (H2), and support from ingroup members for those who were more highly identified with their religious group at least partially accounted for their increased political consciousness, while spiritual support-seeking further contributed to religious identifiers' actiontaking in response to religious discrimination, both as individuals and as a collective (H3). This said, given that the reverse causal direction was also evident, individual and collective-action taking in the face of discrimination may also foster religious identification by means of enhanced perceptions of support from ingroup members and reliance on spiritual resources. Importantly, as in Study 1, the same mediating roles of support were not evident with regard to ethnic identification and threat.

General Discussion
Despite suggestions that religious and ethnic identities may serve similar group-based functions (Kinnvall, 2004;Verkuyten & Yildiz, 2007;Worchel, 2004), the present research demonstrated that responses to discrimination Socio-Political Responses to Religious Discrimination 364 on the basis of these distinct identities show important variations. We had suggested that ethnic and religious identities are qualitatively different based on their content, in that religious identity is founded not only in one's group membership but additionally reflects a system of fundamental and sacred beliefs. Indeed, in two studies, religious discrimination elicited tangible, active behaviors (e.g., civic involvement), which seemed to be motivated by perceived spiritual strength or engagement with support from ingroup members. Although responses to ethnic discrimination were by no means benign, such threats appeared to evoke more moderate attitudinal responses (e.g., political consciousness) compared with those following religious threats, and support did not play a mediating role in this regard. Such was the case irrespective of whether relations among these variables were assessed using general indices derived from a large nationally-representative survey (Study 1), or after priming participants to recall a specific incident of, and responses to, ethnic or religious discrimination (Study 2). These differences in responses could thus have important implications for understanding intergroup conflicts rooted in religious versus ethnic identities.

Responding to Religious Identity Threat
As hypothesized, stronger religious identification was associated with greater civic involvement (Study 1) and action-taking at both individual and collective levels (Study 2) in response to discrimination targeting one's religious group. Despite the severity of some of the discrimination experiences reported in Study 2, however, identification only appeared to promote solution-seeking responses (e.g., seeking mediation) rather than confrontational actions.
These findings might suggest that the ingroup ties inherent to identification itself provided a coping resource to effectively deal with discrimination (Haslam & Reicher, 2006;Muldoon et al., 2009). Further to this, such responses are also in line with previous research suggesting that mainstream (i.e., non-extremist) religious beliefs generally facilitate prosocial attitudes and behaviors (Shariff & Norenzayan, 2007). Indeed, spiritual support-seeking featured most prominently in accounting for the relation between stronger religious identification and the greater likelihood of action-taking in response to discrimination. In effect, engagement with other ingroup members appeared to provide the impetus necessary for highly-identified religious individuals to be involved in civic activities (Study 1) and, when coupled with their ability to "call on God for help" , provided the necessary strength to take conciliatory action aimed at alleviating the injustice toward their group (Study 2). These findings perhaps suggest the need for physical and spiritual alternatives akin to instrumental (i.e., material, tangible) and emotional support (i.e., concern and care, albeit from a "higher power"; Pargament et al., 1998), as suggested by others (Dakof & Taylor, 1990;van Zomeren et al., 2004), as two distinct pathways to civic involvement and actiontaking in the aftermath of religious discrimination.
The absence of a strong relation between religious identification and confrontational action-taking in Study 2 might also be argued as evidence that such events were insufficiently threatening to warrant more extreme responses, or that some group members simply felt powerless to directly challenge the perpetrators of their injustice. Alternatively, the lack of relation is consistent with previous research suggesting that such confrontational action-taking to overcome the disadvantaged status of one's group is relatively rare (Matheson & Anisman, 2009;Wright et al., 1990) and so, even following religious discrimination, such extreme responses might be motivated by factors other than group identification. Indeed, had our sampling sought fundamentalist members of religious sects (rather than the general population), more extreme reactions might have been reported (Friedman & Rholes, 2009 Ysseldyk, Talebi, Matheson et al. 365 1999) and can promote collective action in an effort to resolve such inequalities (Deaux et al., 2006), this finding supports our contention that religious discrimination may elicit stronger responses because it is particularly threatening. Indeed, direct appraisals of the stress associated with the identity threat reported in Study 2 also indicated that religious discrimination was perceived as being especially central in terms of its implications for the individual and, although to a lesser extent, was more challenging and threatening than was ethnic discrimination.
Perhaps for this reason both ingroup and spiritual support were associated with a broad array of actions following religious threats, whereas such relations were largely absent (with the exception of ingroup support and solutionseeking action) among those reporting ethnic discrimination.
In contrast to the role of spiritual support-seeking in relation to action-taking, ingroup support alone partially accounted for the association between religious identification and political consciousness following religious discrimination (Study 2). Thus, in line with earlier theorizing (Foster, 2000;Simon & Klandermans, 2001), high levels of group identification coupled with support from ingroup members appeared to result in the politicization of religious group identity. It is possible that those who identified highly with their religion experienced spiritual support in a manner that was autonomous from political concerns. Alternatively, highly-identified religious individuals may have believed that their spiritual strength could be relied upon to contend with challenges, deeming the political arena irrelevant. This reasoning might offer an explanation for the lower levels of political consciousness reported following religious discrimination relative to those associated with ethnic discrimination in Study 2. On one hand, given that some of the injustices reported took place outside of Canada (e.g., persecution in the 'homeland' or a foreign country), it could be that Canadian political participation was deemed irrelevant. On the other hand, it may be that the collective mobilization aspects of religious group membership predominated, relative to the support offered by the ingroup sharing one's system of beliefs to promote social change within the political sphere.

Responding to Ethnic Identity Threat
Identification with one's ethnic group is often important to the self-concept and has been shown to buffer the negative effects of discrimination (Branscombe et al., 1999). We have suggested, however, that ethnic and religious identities differ conceptually in that the latter is linked to a belief system that not only enhances coping efforts Rutjens et al., 2012), but may also serve to intensify responses when threatened (Thomas et al., 2009;Ysseldyk et al., 2010Ysseldyk et al., , 2011. Not to exaggerate those differences, in the present study, responses to discrimination among individuals who highly-identified with their ethnic group in many ways mirrored those of religious-identifiers. Ethnic discrimination was associated with increased ingroup support-seeking and political consciousness (Study 2). And, like religious identification, ethnic identity appeared to evoke civic and solutionseeking action at both the individual and collective levels, as well as greater political consciousness. Nonetheless, ethnic ingroup engagement or support was largely unrelated to socio-political efforts to deal with ethnic threats in both studies, with the exception of solution-seeking action in Study 2. It may be that the negative connotations of being perceived as ethnocentric (especially as stereotype-confirming), thereby potentially exacerbating collective threat against (and within) the ingroup (Cohen & Garcia, 2005), diminished individuals' desires to turn group-based support into political or social action based on ethnicity. In effect, whereas defending one's beliefs might have been deemed as vital, defending one's ethnicity may have been perceived as less imperative. Interestingly, however, spiritual support-seeking was positively associated with civic involvement (Study 1) and both collective solution-seeking action and greater political consciousness (Study 2) following ethnic threats. Thus, even when one's ethnic group was targeted, some individuals appeared to call on religious resources in an effort to cope.

Limitations
Notwithstanding the patterns of findings generated by the present research, some potential limitations should be noted. As with variations among the measures within Study 1 used to assess religious or ethnic support (e.g., spiritual support-seeking vs. importance placed on ethnic traditions), the operationalization of some of the variables of interest also varied across the two studies. Although Study 2 was designed intentionally, in part, to expand on Study 1 with measures more specific to participants' discrimination experiences, the differences in measures may also underlie some of the differences in the findings observed across the two studies. For example, in Study 1, religious ingroup engagement entailed participating in religious activities or services-although attendance at such events can be motivated solely by social factors (e.g., extrinsic religiosity; Ysseldyk et al., 2011) it often includes a spiritual component as well. In Study 2, the assessment of religious ingroup support (e.g., receiving help from other ingroup members) was distinguished from spiritual motivations. As noted earlier, this may help to explain some of the low effect sizes observed, but might also be key to understanding why only engagement with ingroup support networks (potentially having both social and spiritual components) mediated the relations between religious identification and civic involvement in Study 1, whereas both ingroup and spiritual support mediated the relationships between religious identification and solution-seeking action in Study 2.
Despite these mediated relations and the experimental design employed in eliciting religious versus ethnic identity threat recall in Study 2, relations among the variables in both studies were correlational in nature. Indeed, previous research has demonstrated that levels of religious identity salience (primarily Catholic, Protestant, and Jewish) were significantly higher among American university students four days following the September 11 th , 2001 terrorist attacks compared to six months later (Moskalenko, McCauley, & Rozin, 2006), suggesting that the relation between identification and socio-political events may be reciprocal. Indeed, in the present study, although supportseeking significantly accounted for the relations between religious identification and action-taking, alternative directional models also revealed that civic involvement (Study 1) and solution-seeking action (Study 2) promoted greater support perceptions, predicting increased group identification. Nonetheless, several theoretical models also support the direction of relations as hypothesized and observed in the present investigation (e.g., Foster, 2000;van Zomeren et al., 2004), making it equally likely that initial religious group identification fostered the support that led to action. Moreover, the retrospective, qualitative recounting of previous ethnic and religious discriminatory events in Study 2 allowed for valuable insight into the types of threats faced by those groups, as well as an account of responses following those events (i.e., rather than response intentions). Following from earlier research, several other variables (e.g., emotions, coping strategies; Livingstone, Spears, Manstead, Bruder, & Shepherd, 2011;van Zomeren, Leach, & Spears, 2012) were assessed in the present research yet did not mediate in the same way that ingroup and spiritual support did here. Of course, there remain other factors (e.g., political sophistication or orientation, Choma & Hafer, 2009; past versus future temporal focus, Rabinovich & Morton, 2012; moral threat or outrage, Täuber & van Zomeren, 2013) that could also be important to account for in high-identifiers' responses to discrimination and might be considered in further research.
Finally, it could be argued that Canada is a country in which experiences of discrimination are tempered given its official policy of cultural recognition through multiculturalism (Bloemraad, 2006) and political support for religious pluralism (Kunz, 2014), coupled with the rich experiences that such pluralism brings (Talebi & Desjardins, 2012).
It is thus arguably a lower-bound context for the effects we examine. Nevertheless, the 2014 attack on Canada's Parliament Hill (CBC News, October 22, 2014), ostensibly motivated by religious extremism, represented a groupbased threat at the national level. Likewise, Canada's citizens have endured group-based threats rooted in religious ideological differences at the individual level, as evidenced in the present investigation. Indeed, although some injustices reported took place outside of Canada (Study 2), numerous discriminatory threats took place within Canada as well (Study 1). Discrimination exists in Canada, as evident in self-reports in the EDS as well as audit studies of job-seeking, which show that those with "ethnic" names (including names of Indian or Chinese origin, which might also signal non-Christian religions) are less likely to receive employment callbacks than similar resumes sporting "Anglo" names (Oreopoulos & Dechief, 2012). Still, given the Canadian context, generalizations to other countries should be made cautiously.

Conclusion
Taken together, our findings extend previous research by demonstrating that although both religious and ethnic discrimination evoke strong socio-political responses, ethnic discrimination appeared less likely to promote reactions relative to religious discrimination, which was more likely to elicit active responses that were encouraged by spiritual strength and support from ingroup members. This divergence across types of group-based threats is in line with previous theorizing (i.e., Ysseldyk et al., 2010) and might be attributed to religious threat as reflecting an assault on the sacred belief system itself. Thus, the impacts of religious identity threats may exceed those of threats to identities that are of more secular importance. At the very least, these findings bring us toward a greater understanding of the factors that may propel the socio-political responses that so often occur in the midst of injustice or conflict rooted in religious diversity.
Notes i) Solution-seeking and confrontational action may be thought of as analogous to normative and non-normative action , or constructive and unconstructive action (Kamans, Otten, Gordijn, & Spears, 2010), as noted in earlier research.
iii) Given potentially qualitatively different experiences of discrimination among minority and majority participants, the moderating role of group status (i.e., minority vs. majority ethnicity and religious affiliation) was also assessed (Preacher, Rucker, & Hayes, 2007). However, few differences in response patterns emerged as a function of group status. Details of these analyses are available for the interested reader upon request from the first author.
iv) Measures of emotional reactions (Watson, Clark, & Tellegen, 1988) and coping strategies (Matheson & Anisman, 2003) were also included in the present study; however, given their lack of significant relationships with the remaining variables of interest, they are not presented here.
v) Given the considerably smaller sample size in Study 1 (N = 287) compared with Study 2 (N = 1806), only age and gender were included as covariates.
vi) As in Study 1, the moderating role of minority-majority status (with the targeted group) was assessed, with few differences in response patterns emerging as a function of group status. Details of these analyses are available upon request from the first author.

Funding
This work was supported by the Canadian Institute for Advanced Research, the Canadian Institutes of Health Research, and the Social Science and Humanities Research Council of Canada.