|Home | About | Journals | Submit | Contact Us | Français|
Social identity in Northern Ireland is multifaceted, with historical, religious, political, social, economic, and psychological underpinnings. Understanding the factors that influence the strength of identity with the Protestant or Catholic community, the two predominate social groups in Northern Ireland, has implications for individual well-being as well as for the continuation of tension and violence in this setting of protracted intergroup conflict. This study examined predictors of the strength of in-group identity in 692 women (mean age 37 years) in post-accord Northern Ireland. For Catholics, strength of in-group identity was positively linked to past negative impact of sectarian conflict and more frequent current church attendance, whereas for Protestants, strength of in-group identity was related to greater status satisfaction regarding access to jobs, standard of living, and political power compared to Catholics; that is, those who felt less relative deprivation. The discussion considers the differences in the factors underlying stronger identity for Protestants and Catholics in this context.
In Northern Ireland, identity is “at the heart of the schism,” (Cassidy & Trew, 1998, p. 1). However, the extent to which identity is rooted in religious ideology or socio-political factors including the past impact of intergroup struggles or current comparisons between groups resulting from the changing nature of power following the peace process remains an open question. Moreover, whether similar or different processes are at work in strengthening in-group identity for members of the two predominant social groups is unknown. It is important to better understand the antecedents of strength of in-group identity as it may play a significant role in the reproduction or resolution of protracted intergroup conflict, as well as individual well-being and family processes (Hammack, 2010; Merrilees et al., 2011). The current study simultaneously examines the extent to which past impact of sectarian conflict, current group status satisfaction, and current church attendance contribute to an individual's strength of identity, and whether these relations vary between Protestants and Catholics.
The conflict in Northern Ireland has historical, religious, political, social, economic, and psychological elements (Cairns & Darby, 1998). There are several causes of tension within Northern Ireland. These include divergent opinions regarding the constitutional position of Northern Ireland within the United Kingdom, disagreements over the causes and consequences of conflict related violence, and issues of cultural expression and public space (Shirlow & Murtagh, 2006). Contrary positions tend to align with religious background and as noted by Garry (2011) in his analysis of the 2011 Assembly election: “religion remains a very strong predictor of vote choice with almost no Catholics giving a first preference vote to the UUP (pro-British) or DUP (pro-British) and almost no Protestants giving a first preference vote to the SDLP (pro-Irish unification) or Sinn Fein (pro Irish unification)” (p3). Annually since 1998, the Northern Ireland Life and Times has surveyed adults' changes in attitudes, values and beliefs on social policy issues. In 2012, the survey indicated an enduring link between religion and national/cultural identity. It was found that 78% of Catholics and only 5% of Protestants considered themselves to be either “Irish not British” or “more Irish than British,” while 73% of Protestants and 5% of Catholics defined themselves as “British not Irish” or “more British than Irish” (NILT, 2012). Accordingly, we will use the terms “Protestant” and “Catholic” to distinguish these social groups.
Social identity theory (SIT) has been applied to Northern Ireland in an attempt to explain the attributions and behaviors of the two prominent social groups (Cairns, Kenworthy, Campbell, & Hewstone, 2006; see Leach & Williams, 1999). In its original expositions, SIT provides insights into some dimensions of what are often complex intergroup relations contexts. SIT proposes that identity is derived from a number of sources, including social comparisons related to power, status, salient group attributes, or the distribution of scarce resources (Tajfel, 1982), and that individuals have a goal of positively differentiating their own group from another group in order to achieve a positive self-image (see Tajfel & Turner, 1986 for review). When unable to join a higher status group in order to increase personal self-esteem, individuals may respond to threat, including sectarian violence or shifting social status, by strengthening in-group identity. Members of social groups of perceived lower status, but who challenge that subordinate classification, may create social movements with the aim of changing the social order (Hammack, 2010). At the same time, higher status groups are largely concerned with maintaining a status quo of power and privilege. Thus, lower and higher status groups, or minority and majority groups, have different interests, motivations, and concerns which shape their behavior.
The conflict in Northern Ireland is, and has long been, a dynamic situation with implications for group status. Historically, Catholics were considered the structurally disadvantaged group, suffering greater inequities with regard to human rights, including less access to education, social housing, jobs, and political power (Cairns & Darby, 1998). During the Troubles, Catholics also experienced greater social disadvantages and hardships, such as higher unemployment and welfare dependency, and longer waiting times regarding the allocation of social housing compared to Protestants (Cairns & Darby, 1998). The 1998 Belfast/Good Friday agreement called for power sharing, protection of human rights, and a re-examination of educational and employment opportunities, with implications for group status resulting from a potential shift in social and political power.
However, adding complexity to these issues around status, some have suggested that both groups in Northern Ireland may consider themselves to be in the minority: Catholics the minority in Northern Ireland and Protestants the minority on the island of Ireland (Jackson, 1970). This double minority model complicates the examination of group differences in these processes. Understanding the bases for strength of identity and group differences in status in this changing and complex context may help identify the conditions which may reduce or exacerbate intergroup conflict. Fuller understanding of these underlying principles has relevance both within Northern Ireland and in other areas of prolonged conflict.
In a context of protracted conflict, one salient factor is the impact the struggle had on the individual and their community in the past. During the most recent period of violent sectarian conflict known as the “Troubles” from 1968 to 1998, over 3,600 people were killed. Around half of those killed were combatants: either state security forces (n=1023) or paramilitaries (n=440). Of the civilian victims, nearly two-thirds were Catholic (Shirlow, 2012). In addition, approximately 35,000 people were forced from their homes due to intimidation and an additional 50,000 sustained conflict-related injuries (Shirlow & Murtagh, 2006). Many citizens experienced high levels of militarization, house searching by the army and multiple and unreported physical, verbal and emotional threats and abuse until the paramilitary ceasefires of 1994.
Experience of violence and victimization may have significant consequences for intergroup conflicts as well as identity and well-being, particularly for members of lower status, minority group (Noor, Shnabel, Halabi, & Nadler, 2012; Vollhardt, 2012). Victimhood, particularly when viewed as undeserved, may serve to grant higher moral status to the group, promoting identification (Leach, Ellemers, & Barreto, 2007). Relatedly, collective trauma may shape future identity and action, as with Germans and Jews after the Holocaust (Barkan, 2000) and political unification among African Americans following the end of slavery (Eyerman, 2004). That is, recollection of the traumatic events and discrimination in the past may mobilize lower status groups to pull together to protect themselves from possible future reoccurrence of violence (Bar-Tal, 2007).
Both Protestants and Catholics were victimized during the Troubles, therefore we expect past impact will predict strength of identity for both groups. However, as the structurally-disadvantaged group experiencing greater negative effects in the past, the impact of sectarian conflict may play a more powerful role for Catholic than Protestant social identity.
Whereas judgments of impact may be especially salient for those who considered themselves or their group to have suffered disproportionally in the past, relative deprivation or gratification may be a particularly salient processes for understanding strength of identity with regard to perceptions of current group status, particularly during periods of social change. Relative deprivation theory (Crosby, 1976; see Walker & Smith, 2002) focuses on individuals' subjective evaluation of inequality, with these feelings of unjust disadvantage stemming from social comparison with other groups, or with the in-group at a different point in time (Brown & Middendorf, 1996). Relative gratification, the opposite of relative deprivation, which indicates satisfaction with the status quo or distribution of power between groups, has similarly been shown to result in ingroup bias (Guimond & Dambrun, 2002).
Following the 1998 Belfast Agreement, some Protestants perceived their in-group to be losing ground to the Catholic community relative to their historically dominant position (Cairns et al., 2006; Hargie et al., 2008). Catholics benefited from significant social and legislative change, including the near cessation of religious discrimination in the labour market, political power-sharing, public funding of the Irish language and culture, and significant growth in a Catholic middle class (Southern, 2007). For example, by 2003, 29% of Protestants perceived their own cultural tradition to be the underdog, compared to 18% of Catholics (Mac Ginty & du Toit, 2007).
In the zero-sum culture of Northern Ireland (Leach & Williams, 1999), in which the success of one group (i.e., Catholics) is considered to come at the expense of the other (i.e., Protestants), post-accord shifts in power and status may be particularly salient for the strength of identity for Protestants, some of whom consider their community to be losing their privileged status.
A third factor hypothesized to contribute to social identity in Northern Ireland is religion itself, which may be differentially predictive for Catholics than Protestants. Church attendance has declined for both groups, but Catholics in Northern Ireland are still more likely than Protestants to be frequent attenders at church services (Hayes & Dowds, 2008; McAllister, 2005). Considering the highly segregated nature of housing in Belfast, it is likely that individuals attend religious services in their local community. Church attendance then may foster a sense of connectedness and provide opportunities to give and receive social support, further strengthening identity for Catholics who are more likely to attend services than Protestants. Furthermore, in broad terms Catholicism, to a greater degree than Protestantism (which in Northern Ireland is linked to significant intra-denominational diversity), is seen as emphasizing the importance of the collectivity or community over the individual/personal (Mitchell, 2004; McAllister, 2000). On the other hand, it has been pointed out that Catholicism “played heightened and symbolic roles” (Mitchell, 2004, p. 249) at certain crucial times in the conflict, such as the hunger strikes. Given its significant role in this religious society, it is hypothesized that church attendance will be a significant predictor of strength of identity in this context, and may be an even stronger predictor among Catholics than Protestants.
Women's role in Northern Ireland during the Troubles had stereotypically been relegated to tending the home, raising children, and caring for their husbands, rather than as political subjects (Shirlow & Dowler, 2010). Although some women were active combatants, or organized as part of the Northern Ireland Women's Coalition contributing to peace negotiations, women on the whole were largely excluded from the formal political processes between 1974 and 2007 (McEvoy, 2009). Moreover, many were personally impacted by the conflict, but their needs have been largely ignored in psychological research. As such, women's ideas and concerns about priorities and agendas in post-accord society have often been neglected, with implications for distribution of resources and intergroup relations (Cockburn, 2007). Investigating predictors of women's strength of social identity in this context is important for full understanding of the complex issues, such as individual well-being (Merrilees et al., 2011), and ultimately for the success of the peace process (McEvoy).
Given the growing evidence for diverse predictors of identity in a setting of protracted conflict, the current study will investigate how past impact of sectarian conflict, contemporary relative deprivation, and church attendance influence strength of in-group identity for Catholic and Protestant women in Northern Ireland fifteen years after the Belfast Agreement. In contrast to past work that has focused on differentiating the categories and labels ascribed to individuals (Cairns & Mercer, 1984; Cassidy & Trew, 1998; Muldoon, Trew, Todd, Rougier, & McLaughlin, 2007), the current study aims to understand the processes affecting strength of women's in-group identity in Northern Ireland. Specifically, Catholic/Protestant category will be tested as a moderator to see if the group with which one identifies changes the predictors of social identity. In other words, does the foundation for the strength of in-group identity look different for Catholics and Protestant women in this historically divided society?
Participants included 692 women from Belfast, Northern Ireland, who were part of a longitudinal study examining children and families in the context of political violence. In addition to conceptual interest in women's experiences and attitudes, women were included as the adult respondents because of their availability for data collection and the predominance of single female led households in socially deprived areas in Belfast. Forty-three percent of participants were married or living as married, 24% were separated or divorced, 2% were widowed, 5% were dating, and 26% were never married. Mean age was 37.15 years (SD = 6.12). Approximately 60% of participants reported growing up in Protestant communities and 40% in Catholic communities during the Troubles.
Participants were selected using a stratified random sampling strategy in socially deprived communities in Belfast. An expert demographer in the region selected areas to have a proportional representation of Catholic and Protestant communities and variability in the intensity of historical political violence. See (author identifying citation) for more detail about the sampling procedure and areas.
Surveys were conducted in the home in 2006-2008 by an established market research firm employing interviewers native to the region. All questionnaires were administered as an interview lasting approximately one hour. Consent was obtained prior to the interview and participants received £20 for their participation. The research project was approved by the committees for the protection of human subjects at all corresponding universities.
To measure the impact of political violence (i.e., the Troubles), participants responded 0 (none), 1 (a little) or 2 (a lot) to three questions asking “When you were growing up, what was the impact of the Troubles on: you/your family/the area where you lived” (O'Reilly & Stevenson, 2003). Participants responded to these questions at Wave 1 and scores were summed across items (possible range 0-6) with higher scores reflecting greater negative impact. Cronbach's alpha for the current sample was .91.
To assess group relative deprivation, in Wave 2 participants reported the degree to which they were satisfied with the standard of living/the way jobs are distributed/and the way political power is distributed between the [Catholic/Protestant] communities in Northern Ireland on a scale from 0 (extremely unsatisfied) to 6 (extremely satisfied) (Pettigrew, Christ, Wagner, Meertens, van Dick, & Zick, 2008). These questions were reverse scored and summed (possible range 0-18) with higher scores reflecting greater relative group deprivation or dissatisfaction with the distribution of power and resources between the groups. Cronbach's alpha for the current sample was .86.
Church attendance at Wave 1 was assessed with the following question: “Apart from special occasions such as weddings, funerals, baptisms, or so on, how often nowadays do you attend services or meetings connected with your religion?” Protestants and Catholics, respectively, responded as follows: 2.4%/28.5% “once a week or more”, 1.0%/12.2% “at least once in two weeks”, 2.9%/13.7% “at least once a month”, 7.2%/9.6% “at least twice a year”, 5.5%/5.9% “at least once a year”, 11.6%/4.8% “less often than once a year”, 49.2%/17.0 “never or practically never”. In addition, 20.2%/8.1% responded either “varies too much to say” or “don't know” and were coded as missing.
Participants reported the social group to which they identify (Protestant or Catholic) and then answered five items in Wave 2 designed to measure the strength of their in-group identity (Brown, Condor, Mathews, Wade, & Williams, 1986). Sample items include “Would you say you are a person who considers the ___ group important?” and “Would you say you are a person who identifies with the ____ community?” Participants responded to statements on a scale ranging from 1 (never) to 5 (very often) (possible range 5-25), with higher scores reflecting stronger identity with the in-group. The reliability and validity of the scale has been established (Brown et al., 1986), and the measure has been widely used in Northern Ireland. Cronbach's alpha was .93 for the current sample. A latent variable was estimated using the items of the scale as indicator variables.
Table 1 includes the overall means, standard deviations and bivariate correlations for all study variables separately for Protestants and Catholics. Independent t-tests were conducted to compare Catholics and Protestants on all of the study variables. Catholics reported significantly more religious attendance, greater past impact of sectarian conflict, and higher scores on all individual strength of identity items.
The proposed SEM model is shown in Figure 1 and was tested with AMOS Graphics 18 (Arbuckle, 2009). Estimates were derived using maximum likelihood estimation and overall model fit was assessed with the χ2/df index, the normative fit index and comparative fit index (TLI, CFI), and the root mean square residual (RMSEA). Good model fit indices are indicated by χ2/df index ≤ 3, a TLI and CFI ≥ .90, and a RMSEA ≤ .08 (Hu & Bentler, 1999).
In the proposed model, past impact of sectarian conflict, relative deprivation, and church attendance were all entered as manifest predictor variables; strength of in-group identity was a latent outcome variable; and, later, social group was a dichotomous moderator. The proposed model was first tested with the full sample of participants. The overall fit for the combined model was adequate, χ2 (34) = 253.17, p<.001, N =692; χ2/df = 7.45; TLI = .90; CFI = .95; RMSEA = .06 (CI: .05, .07). Both relative deprivation (β = −.29, p <.001) and more frequent church attendance (β = .11, p <.05) positively predicted strength of in-group identity. More frequent church attendance related to stronger sense of identity, whereas greater relative deprivation related to weaker sense of identity. In the full model, past impact of sectarian conflict did not significantly relate to current strength of identity. However, these results must be interpreted in light of evidence for differences as a function of group.
To test for Catholic/Protestant group differences in the relations between past impact of sectarian conflict, relative deprivation, church attendance and strength of identity, a multiple group comparison was conducted. First, separate models for Catholics and Protestants were estimated with all parameters left free to vary. This model was then compared to subsequent models in which individual paths were constrained to be equal; significance was computed using the chi-square difference test. For Catholics, two significant paths emerged; greater negative past impact of sectarian conflict (β = .18, p <.05) and more frequent church attendance (β = .16, p <.05) were significantly linked to stronger identity. On the other hand, for Protestants only relative deprivation was significantly related to strength of identity, with more deprivation associated with weaker identity (β = −.43, p <.001). Further supporting the differences between the models, chi-square tests confirmed that each of these paths was significantly different for the two social groups (χ2 diff(1) = 5.43, p <.05 for past impact; χ2 diff(1) = 22.59, p <.05 for relative deprivation; and χ2 diff(1) = 4.13, p <.05 for church attendance). In other words, strength of identity was predicted by distinct phenomenon for Catholics and Protestants in post-accord Northern Ireland. For Catholics, strength of in-group identity was predicted by past impact of sectarian conflict and more frequent current religious attendance, whereas for Protestants, strength of identity with the in-group was linked to judgments of the groups' better access to jobs, standard of living, and political power relative to Catholics.
The current study investigated the underpinnings of strength of in-group identity in Northern Ireland. Understanding the factors that contribute to the strength of identity has implications for mental health and the continuation of tension and violence in this setting of protracted intergroup conflict. As predicted, results suggested different antecedents of strength of in-group identity for Protestant and Catholic women. For Catholic women, strength of identity was positively linked to past negative impact of sectarian conflict and greater religious attendance, whereas for Protestant women, strength of identity was related to more favorable relative comparison to Catholics with regard to employment and political/economic standing.
The evaluation of greater impact of sectarian conflict for self, family, and community in the past related to stronger in-group identity for Catholic women, but not Protestant women when controlling for current relative deprivation and church attendance. Although levels of direct exposure to sectarian violence varied, subjective experience of violence during the Troubles was widespread. Moreover, Catholics have historically been considered the disadvantaged group, including greater experience of discrimination and negative effects. This shared history of the impact of the Troubles for themselves, their families, and their community during the height of sectarian violence and discrimination, coupled with an enduring confidence that their cultural tradition will be preserved, may serve to draw Catholics, the lower status group, together and contribute to a stronger connection to the group, as proposed by SIT
In contrast, current perceptions of comparisons between the social groups was related to strength of in-group identity for Protestant women but not for Catholic women, as expected given the perception of post-accord shifts in relative status for Protestants. Contrary to prediction, the more Protestant women perceived their group to be relatively deprived in comparison to Catholics, the more likely they were to distance themselves from the in-group and experience weaker identity. Stated another way, Protestant woman who were relatively satisfied with regard to the distribution of jobs, standard of living, and political power reported stronger identity to the in-group.
A number of interpretations are possible. We had assumed that because boundaries between groups in Northern Ireland are generally considered to be largely entrenched and impermeable (Hargie, Dickson, Mallet, & Stringer, 2008), individuals would not be able to distance themselves and instead would pull together in an effort to preserve their groups' status, resulting in stronger identity (Wohl, Branscombe, & Reysen, 2010). It may be that boundaries between Protestant and Catholic are rigid in this context, but that diversity is present within the dominant group. Indeed, Protestant are more likely than Catholics to choose a range of identities such as Loyalist, Nationalist, British, or Northern Irish (Muldoon et al., 2007). The current results are in line with work in other parts of the world with regard to individuals feeling weaker connection to their in-group when they consider social change to be negative or when their group is judged negatively by outsiders. For example, following the dismantlement of the Soviet Union, researchers in Russia found that those who considered social change to be rapid and negative reported greater relative deprivation, which was associated with poorer evaluations of the in-group (de la Sablonniere, Tougas, & Lortie-Lussier, 2009). The results can be interpreted in terms of either relative deprivation or relative gratification, both of which have been shown to produce an in-group bias (Guimond & Dambrun, 2002; Dambrun, Taylor, McDonald, Crush, & Meot, 2006). It is important to note that, as Guimond and Dambrun contend, the processes underlying relative deprivation and relative gratification may be distinct and differentially related to social attitudes. Alternatively, it may be the case that the results with this relatively young and exclusively female sample reflect a more heavily gendered aspect of political Loyalism. Perhaps for Protestant mothers who reported lower relative deprivation, satisfaction with group status and resources signals a more promising future for their children and family and thus strengthens their feelings of loyalism.
In addition, church attendance was a significant predictor of strength of identity for Catholic women, although not for Protestant women. Religion plays a role in defining group boundaries in the Northern Irish conflict (Trew, 2004; Whyte, 1990). Some see the role of religion simply as a marker for economic and political interests, or at most a concept in the defining of social identity in Northern Ireland (Muldoon, 2000). Our data suggests that church attendance indeed plays a role in strength of identity, at least for Catholic women in this context. It may be that for Catholic women, religious and social identity are more integrally connected than for Protestant women, perhaps due in part to more frequent attendance at local services or a world view predicated on the principle of communion with one another. Whether other dimensions of religiosity beyond church attendance relate in similar ways to social identity is a promising topic for future research.
The study is not without limitation. The models examined in these analyses conceptualize social identity as the product of adversity posed by political violence history, perceived position and access relative to the other group, and religious attendance, while it is possible that identity colored their memories or characterizations of these constructs. Common method variance associated with self-report questionnaires may have biased the findings as well. Using a single item to assess church attendance may have reduced the reliability of the assessment. Future research should also investigate the antecedents of in-group identity for youth, and the connections between parents' and children's attitudes (Muldoon, McLaughlin, & Trew, 2007). Notably, the data was collected between 2006-2008 and the political and economic context in Northern Ireland has continued to evolve. However, given the long history of the conflict and the slow pace of change in attitudes and intergroup relations, it is reasonable to expect that the pattern of results still applies today.
The study focuses exclusively on women, a group traditionally underrepresented in research on these issues in Northern Ireland (McEvoy, 2009). Notably, analysis of related variables in the 2012 Northern Ireland Life and Times Survey (ARK, 2012) reveal no gender differences (e.g., intergroup attitudes and behaviors related to contact, respect, and understanding of culture and traditions). It is conceivable, then, that these results may extend to men as well. Furthermore, the sample was recruited in socially-deprived neighborhoods in Belfast, areas that experienced a significant amount of violence and strife during the Troubles. Generalizability to findings across Northern Ireland broadly may be limited.
The current study contributes to our understanding of the multidimensional antecedents of strength of identity in the context of sectarian strife and social change. For the historically disadvantaged group on the “rise,” past violence continued to be a particularly salient predictor of strength of identity, whereas for the historically advantaged group on the “decline,” maintaining a sense of relative power was important. By identifying the roots of strength of identity, the results have implications for future expectations and the long-term prospects for peace and reconciliation in areas of historic intergroup conflict (Leach & Williams, 1999). For example, intervention programs aimed at altering identification as a means of influencing intergroup relations would be enhanced by nuanced understanding of relative predictors (Hammack, 2009). Although the current sample was drawn from Northern Ireland, the processes studied may generalize to other regions of the world experiencing shifts in group status in protracted conflicts.
This research was support by NICHD grant 046933-05, We would like to thank the many women in Northern Ireland who participated in the project.
Marcie C. Goeke-Morey, Catholic University of America Ed Cairns, University of Ulster.
Laura K. Taylor, University of North Carolina at Greensboro.
Christine E. Merrilees, University of Notre Dame Peter Shirlow, Queens University, Belfast.
E. Mark Cummings, University of Notre Dame.