The Differential Impact on Children of Inter- and Intra-Community Violence in Northern Ireland

Abstract

This study explores distinctions in Northern Ireland between inter-community (i.e. sectarian) and intra-community (i.e. nonsectarian) violence and their respective impacts on children, and considers these forms of violence in relation to children’s processes of emotional security about community conflict. Preliminary work was based on focus groups with mothers in Belfast, followed by a quantitative study involving mothers in Derry/Londonderry. Support emerged for a conceptually-based distinction between sectarian and nonsectarian violence and differential prediction of children’s adjustment problems, which was more closely linked with sectarian than nonsectarian community violence. Pertinent to explanatory mechanisms, community violence, especially sectarian, related to mothers’ perceptions of children’s emotional insecurity about community. Findings are discussed in terms of future directions for understanding community violence and child development in cultural context.

The Differential Impact on Children of Inter- and Intra-Community Violence in Northern Ireland

There is increasing concern about the effects on children of exposure to community and political violence, especially in communities with long-standing histories of conflict between ethnic, religious, or cultural groups, such as Northern Ireland. Reflecting this concern, in July 2008, the International Journal of Behavioral Development devoted a special issue to the impact of chronic exposure to war and political violence on children and families (Sagi-Schwartz, Seginer, & Abdeen, 2008).

At the same time, many conceptual and methodological gaps remain in terms of the systematic, process-oriented study of children’s exposure to politically-related conflict and violence and its implications for their well-being and development. From a social ecological perspective, there are multiple types of conflict, such as sectarian and nonsectarian, that may affect children; however, researchers have not made these distinctions necessary to assess their potential differential impact on children. Moreover, from the perspective of research design, theory-driven and process-oriented studies examining effects on children’s regulatory processes offer the potential for further advances in understanding (Cummings & Cummings, 1988; Prinz & Feerick, 2003).

The present study addresses these gaps for understanding the relationship between political violence and children’s development in Northern Ireland in the context of the historical conflict between Protestants and Catholics (known colloquially as the “Troubles”). Specifically, we use ecologically valid measures of children’s exposure to sectarian and nonsectarian antisocial behavior in the community. Furthermore, we consider children’s emotional security about the community as a psychological process related to explanations of the impact of political violence for children in this cultural context.

Effects of Sectarian and Nonsectarian Community Violence on Children

Studies in the United States show that children’s exposure to community violence is related to emotional problems (Singer, Anglin, Song, & Lunghofer, 1995), including depression and stress symptoms (Berton & Stabb, 1996; Gorman-Smith & Tolan, 1998). Links are also found with behavior problems and externalizing disorders, especially aggressiveness (Cooley, Turner, & Beidel, 1995; O’Keefe, 1997), as well as with problems in the cognitive domain (Kuther, 1999) and with poor academic performance in urban elementary school children (Schwartz & Gorman, 2003).

A survey by Smyth and Scott (2000) provided support for the effects of sectarian violence in Northern Ireland on children, including reports of children being physically attacked or verbally abused as well as witnessing violence against others, experiencing school disruption and segregation, and fearing for themselves or their families (see also Fay, Morrissey, Smyth, & Wong, 1999). Both nonsectarian and sectarian forms of community violence pose risk for children, but definitions of, and boundaries between them have been under-explored conceptually and empirically (Trickett, Duran, & Horn, 2003). We define sectarian (or political) antisocial behavior as that which takes place between the two ethnic groups in Northern Ireland (Catholic and Protestant), reflecting negative attitudes and behaviors toward members of the “other” group. By contrast, nonsectarian antisocial behavior refers to “ordinary” crime, that is, antisocial behavior within communities that has no political motive.

Many gaps remain in understanding relations between cultural contexts of ethnic conflict and child development (Feerick & Prinz, 2003). Few studies have examined these issues, including the effects of community violence and, more specifically, exposure to politically-motivated community violence (Mabanglo, 2002). Pathways of influence on children due to politically-motivated community violence and other forms of “ordinary” community violence may be distinct, but this notion has been little examined, with few attempts to discriminate exposure to these different forms of community violence.

Children’s Emotional Security about Community Violence

The mechanisms by which political tension and sectarian violence relate to children’s well-being and development are even less well-understood. Process-oriented studies of the effects on children are rare, particularly investigations of the psychological factors related to the impact of violence on children. The pertinence of children’s emotional security to understanding the effects of sectarian and nonsectarian violence on children is explored in this study. Emotional security as a construct is rooted in attachment theory, which has historically focused on security in the context of parent-child attachment. This notion has been extended to other family relationships, including the marital subsystem and family wide processes, with a particular emphasis on the importance of emotional security in contexts of conflict (Cummings & Davies, 1996; Davies & Cummings, 1994; Davies, Harold, Goeke-Morey, & Cummings, 2002). Lovell and Cummings (2001) extended the notion of emotional security to include community and political conflict and violence, hypothesizing that these ecological contexts also influenced children’s sense of emotional security.

We hypothesize that children develop a sense of security about their community rooted in their feelings of physical and psychological safety for themselves and their family. In this sense, tension and violence in the community are expected to undermine children’s security in that context. Moreover, although both sectarian and nonsectarian violence are hypothesized as stressors for children, sectarian violence is posited as a more salient stressor due in part to the presumed meaning this violence holds for the identity and well-being of the child and broader family and community systems.

Belfast and Derry/Londonderry

As in much of Northern Ireland, Catholic and Protestant residential areas and facilities in Belfast and Derry/Londonderry are highly segregated. The contested name of Derry (used by Catholics) and Londonderry (common amongst Protestants) highlights the nature of the division. Territorial markings such as wall murals, flags, and curb paintings are salient in both cities. Notably, children are aware of categorical differences between Catholics and Protestants from an early age (Cairns, 1987). Northern Ireland’s youth possess an emotional attachment to their respective social categories, which may be particularly salient when children are under threat or stress (Cairns & Mercer, 1984). Despite the paramilitary ceasefires of 1994 and subsequent shifts towards a power-sharing administration, conflict and violence regularly occur in Belfast and Derry/Londonderry at the interfaces between Catholic and Protestant districts. The minority in given areas modify their travel patterns and use of services and facilities in response to perceptions of fear and threat from others (Shirlow & Murtagh, 2006). The Catholic and Protestant communities understand and interpret changes through disparate political and cultural lenses (Murtagh, 2004), contributing to the bases for continuing violence and conflict between these groups.

Current Study

In the current study, we examine the impact of community conflict on child adjustment, including emotional and conduct problems, via mothers’ perceptions. To that end, we developed an ecologically valid “mother report” measure that distinguishes children’s exposure to sectarian (ethnic/political) and nonsectarian community-level violence in Northern Ireland through a combination of qualitative analysis based on focus groups conducted in Belfast communities and a two-wave quantitative study in Derry/Londonderry.

In addition, we advance children’s security about the community as a possible explanatory mechanism in the impact of political violence on children. We anticipate that insecurity will be especially associated with exposure to sectarian conflict at the community level, and that the resulting insecurity about the community will have negative consequences for children’s adjustment, particularly emotional problems, consistent with emotional security theory (Cummings & Davies, 1996).

Method

Participants and Procedures

Qualitative Study

Focus groups of Protestant or Catholic mothers (n = 33), respectively, took place in community centers in selected working class areas in Belfast. Mothers, ranging in age from 21 to 55 years, discussed their own and their children’s experiences in their community, including how they described their community, perceptions of relations between and within groups, the prevalence of sectarian violence over nonsectarian, and children’s experiences with both. The focus groups took place between January and March 2006. Participants received £30 for their participation.

Quantitative Study

The sample for the quantitative study included 130 mothers who were recruited through local community centers in working class areas in Derry/Londonderry, Northern Ireland. Over 90% of participants had lived in Northern Ireland all their lives (M years in NI = 34.4, SD = 8.7 years). The majority reported living in their current ward all their lives (54%); an additional 26% for more than 10 years, 4% for 5–10 years, 10% 1–5 years, 1% for less than 1 year, and 4% failed to provide this information. The sample was evenly split between the two major religious groups, Catholics and Protestants (48% each); the remaining reported being Christian (no denomination; 2%) or having no religious affiliation (2%). Fifty-six percent of mothers were married or living with a partner. Mothers completed the questionnaire with regard to one of their children (n = 67 boys; n = 63 girls) between the ages of 5 and 17 years (M = 12.0 yr, SD = 3.2 yr).

Participants completed the survey as a self-administered battery during the first assessment (June 2006) and as a telephone interview with a research assistant during the second assessment (September–November 2006). Data from the first assessment was used to aid in instrument refinement and to assess test-retest reliability; all substantive analyses were done with wave 2 data. Twenty-six of the mothers who completed the first phase could not be reached for the second phase, reducing the complete sample to n = 104. No significant differences on any of the test or demographic variables were found between mothers who participated in both phases and those who did not. Informed consent was obtained from participants; the project was approved by the committee for the protection of human subjects at the three author-affiliated universities. Participants received £30 for their participation.

Instrument Development: Sectarian and Nonsectarian Antisocial Scales

Community antisocial behavior, particularly that which is politically motivated, is expected to be context-specific. Ecologically valid measurement needs to take into account the ways in which community violence is manifest in the particular social, cultural, or political context of interest. So, we began by developing an ecologically valid measure of children’s exposure to sectarian (political) and nonsectarian community-level violence in Northern Ireland. A first step was qualitative analysis based on focus groups conducted in Belfast communities to explore the distinction between sectarian and nonsectarian violence. We created an initial pool of items that were subsequently developed into initial scales of community violence based on qualitative analyses of mothers’ discussions. The resulting political and community violence scales were developed further in a two-wave quantitative study in Derry/Londonderry, including assessment of the psychometric properties, factor analyses and internal consistency.

Focus Group Analyses

We referred to the transcripts of the focus groups to generate lists of sectarian and nonsectarian antisocial behaviors reported by mothers as occurring in their communities. We noted any act that was mentioned at any point during the discussion, whether in response to a direct question from the facilitator, or mentioned indirectly in the course of describing their community or their lives.

The following is a representative exchange that illustrates several types of sectarian antisocial behaviors, involving mothers from a high intensity Catholic area in Belfast that is separated from a neighboring Protestant community by a high “peace” wall:

M1: And we hadn’t a window the whole summer [2002] and the house was paint-bombed constantly. The windows were constantly put in, constantly being replaced. At that stage there wasn’t really much could be done…I wouldn’t allow my children to put the light on after 7:00 in the living room in case it attracted anybody going by. They knew you lived in the house. This is what the kids were saying: “Mommy they know we’re here. What would putting the light on be?” Until a breezeblock [concrete cinder block] was put through the window on top of my wee girl who was only 13 at the time.[snip]

M2: Not a day or a week has gone by but something comes over that wall. But now last weekend, it was terrible. It was diabolical. The stuff that came over that wall was unbelievable.

M1: It was.

M2: A big heavy duty rock came over my wall the other night. A jack—you know the thing for playing bowls. Snooker balls, golf balls, you name it.

Other exchanges referenced actions such as houses set on fire, children being taunted, people being yelled at from passing cars, vandalism, and politically motivated assaults.

The focus groups also provided examples of the nonsectarian antisocial behavior occurring in Northern Ireland. For example, in response to the facilitator’s question asking whether, on a daily basis, they were concerned with what happens within their community in terms of (nonsectarian) antisocial behavior, mothers agreed they were threatened by teenagers within their own communities. One replied, “‥‥Years ago you wouldn’t have heard anything about rapes happening in the area or people breaking into houses.” Another replied, “Drugs as well” to which a fellow mother added, “Yeah, there is the drugs culture as well [snip]. In our day you never heard of cocaine or heroin. Everybody smoked the odd joint and that was about it. Now it’s everybody taking coke and all.” Other excerpts from the groups included reference to problems with drunkenness, fighting, “joyriding”, child “cheekiness” (impudence), knifings, murder, and assaults that were not politically/ethnically motivated.

Measure creation and refinement

Based on a list of behaviors inspired by the focus groups, we created initial versions of the Sectarian Antisocial Behavior Scale (SAB) and Nonsectarian Antisocial Behavior Scale (NAB). Using these scales, mothers in the first wave of the quantitative study reported the frequency of children’s exposure to sectarian and nonsectarian antisocial behavior, respectively. Changes were made to the instruments before wave 2 testing based on factor analyses, further consideration of relevant conceptualizations (e.g., Cummings & Davies, 2002), and reports by experimenters about the use of the instruments.

Exploratory factor analysis of the revised scale, using principle axis factor extraction and an oblique (Promax) rotation, indicated a four-factor solution. The factor loadings are presented in Table 1. The first factor reflected sectarian antisocial behavior and contained all 11 sectarian behaviors. The other factors reflected different facets of nonsectarian antisocial behavior, including substance use, theft, and violent crimes. Together the solution accounted for 71% of the total variance. The eigenvalues for all four factors were > 1.0. The internal consistency (α) for the sectarian antisocial behavior scale was .94, and α = .68 for a total nonsectarian antisocial behavior scale. The α for each scale exceeded or approached the .70 standard of acceptability (Nunnally, 1978). SAB and NAB scales were significantly correlated, r (104) = .47, p < .001, but also reflected considerable unique variance as indicated by the moderate correlation. Accordingly, the sectarian scale contained items 1–12, and the nonsectarian scale contained items 13–19 (see Table 1). These scales were used for all remaining analyses.

Table 1.

Factor loadings for the revised Antisocial Behavior Scales

	Factor 1	Factor 2	Factor 3	Factor 4
3. Someone threatened by people from the other community	.99
5. Someone shouted at from cars by people from the other community	.98
10. Children taunted by people from the other community, including verbal communication, text messaging, instant messaging or other forms of communication	.89
4. Someone chased on the street by people from the other community	.88
2. Name calling by people from the other community	.83
9. Blast bombs or petrol bombs exploded by the other community	.80
7. Houses or churches paint-bombed by the other community	.71
12. Children from the other community allowed to get away with crime and misbehaviour	.71
6. Stones or other objects thrown over walls	.67
8. Windows put in by the other community	.52
1. Someone beaten up by people from the other community	.42
11. Deaths or serious injuries from violent or destructive acts by the other community	.42
15. Fighting in or outside of bars		.90
13. Drunkenness		.62
14. Drugs being sold or used		.56
17. Robberies/muggings				.68
16. Home break ins				.62
18. Murders			.96
19. Stabbings			.82

Instrument Development: Children’s Emotional Security about the Community

Initial instrument and instrument revisions

An initial version of the Security in the Community Scale (SIC) was created by adapting items from existing emotional security scales about family relations and was supported by focus group responses. For example, mothers described their families being restricted in their movement and being “sort of terrified” everyday walking down the street. For this instrument, mothers in the quantitative study reported on the extent to which statements were 0 (not at all like my child) to 4 (a whole lot like my child).

The preliminary version of the security measure administered in Wave 1 contained 11 items. Exploratory factor analysis failed to produce a conceptually meaningful solution, which led us to revise the instrument based on conceptual considerations for the second wave of data collection. A subsequent exploratory factor analysis of a revised 10- item scale in Wave 2 indicated a three factor solution. Together the solution accounted for 62% of the total variance. Applying the Kaiser’s eigen rule (Tabachnick & Fidell, 2001), the eigenvalues for all factors were > 1. The factor loadings are presented in Table 2. The first factor tapped children’s fear and vigilance; the second factor feelings about security about where children lived, and the final factor security about the areas dominated by the other group. These scales combine into an overall security score with α =.69. The overall scale, reflected in the items in Table 2, is used in the analyses.

Table 2.

Factor loading for the revised Security in the Community Scale

	Factor 1	Factor 2	Factor 3
6. My child has at times been unable to sleep at night because of violence in our area.	.93
7. Sometimes my child feels that something very bad is going to happen in our community.	.86
3. My child stays in because of the threat posed by the other community.	.68
2. My child feels threatened by people approaching from the other community.	.63
4. My child is being hindered from going to doctor, cinema, dentist, or other places because of threat of violence by the other community.	.61
8. When my child thinks about the problems in our community, he/she feels that things will work out in the end.		.72
10. My child feels safe when walking through our local areas during the marching season.		.69
5. My child’s experiences in this community have provided a secure foundation for meeting other people and going places.		.60
1. My child feels intimidated by murals, flags, and curb paintings.			.83
9. My child feels safe when walking through an area dominated by the opposite religion during the day.			−.60

Measures

In the quantitative study, mothers completed a brief demographic questionnaire (e.g., mother age and marital status, child age and gender, length of time living in Northern Ireland, religious community) as well as their opinion of whether relations between Protestants and Catholics are better than they were 5 years before, worse, or about the same.

Exposure to political and community violence

The Sectarian Antisocial Behavior Scale (SAB) and Nonsectarian Antisocial Behavior Scale (NAB) are advanced for the first time in this study (see above). Mothers reported the frequency of children’s exposure to sectarian and nonsectarian antisocial behavior, respectively, in their community in the previous three months on a five-point scale: 0 (not in the last 3 months), 1 (once in the last 3 months), 2 (every month), 3 (every week), or 4 (every day).

Security in the community

The Security in the Community Scale (SIC) is also advanced for the first time in this study. Mothers reported the extent to which statements were 0 (not at all like my child) to 4 (a whole lot like my child).

Children’s adjustment

Mothers completed the Strengths and Difficulties Questionnaire (SDQ: Goodman, 1997) as an assessment of their child’s adjustment. Respondents report whether 25 behaviors are 0 (not true), 1 (somewhat true), or 2 (certainly true) with regard to the target child over the past six months. Psychometric properties are well established (see Goodman & Scott, 1999). Within our sample, the alphas for the SDQ scales at phase 2 were as follows: total difficulties (α=.71), emotional symptoms (.57), conduct problems (.59), hyperactivity/inattention (.72), prosocial behavior (.62). The peer problems subscale was dropped from analyses due to its low reliability in the sample (alpha = .22).

Results

Children’s Reported Exposure to Community Violence

Every mother in the sample indicated that their child was aware of at least one sectarian antisocial act occurring weekly in their community. Scores were M = 6.67, SD = 9.45, for the sectarian and M = 3.53, SD = 3.56, for the nonsectarian scales. T-tests showed that Protestant children had greater sectarian antisocial behavior exposure than Catholic children (M = 9.09, SD = 10.28 and M = 3.05, SD = 5.18, respectively), t (83.7) = −3.77, p < .001, with no differences in exposure to total nonsectarian antisocial behavior (M = 3.26, SD = 3.69 and M = 4.07, SD = 3.56, respectively), t (95) = 1.09, p = .281. No gender differences were found.

Discriminant validity was supported in that mothers’ predictions that relations between Protestants and Catholics would be worse five years in the future than they are today were related to their reports of sectarian violence exposure (b = 0.02, t = −2.51, p = .014) but not nonsectarian violence exposure (b = −0.002, t = −0.08, p = .937).

Sectarian and Nonsectarian Community Violence as Predictors of Reported Child Adjustment

The means and standard deviations for the adjustment scales were: emotional symptoms (M = 3.05, SD = 2.15), conduct problems (M = 1.97, SD = 1.84), hyperactivity/inattention (M = 4.08, SD = 2.70), prosocial behavior (M = 9.02, SD = 1.53), and total difficulties (M = 11.36, SD = 5.38).

Exposure to SAB predicted greater total difficulties, r (104) = .19, p = .050, emotional symptoms, r (104) = .28, p = .004, and conduct problems, r (104) = .23, p = .019, as well as less prosocial behavior in children, r (104) = −.27, p = .007. Exposure to NAB predicted greater total difficulties, r (104) = .26, p = .009, emotional symptoms, r (104) = .22, p = .025, and conduct problems, r (104) = .21, p = .031. These relations remained significant in regression analyses controlling for religious group membership, and child age and gender. Moreover, these variables did not moderate relations between antisocial behavior exposures and adjustment.

Next, a series of multiple regressions were run for each adjustment outcome while entering both SAB and NAB simultaneously in the model. SAB predicted greater emotional symptoms (B = 0.05, SE B = 0.02, p = .034) and lower prosocial behavior (B = −0.04, SE B = 0.02, p = .016), over and above NAB. The NAB was not a significant predictor in the context of the SAB. This pattern of relationships remained when analyses controlled for religious group, child age, and child gender.

Community Antisocial Behavior as a Predictor of Children’s Reported Insecurity about the Community

Both sectarian and nonsectarian antisocial exposure predicted mothers’ reports of their children’s insecurity about the community, r (104) = .65, p < .001 and r (104) = .45, p < .001, respectively. They both remained significant predictors of insecurity when child age, child gender, and religious group were entered into the model. When both SAB and NAB were entered simultaneously into a regression model, sectarian antisocial behavior emerged as the only significant predictor of insecurity, B = 0.04, SE B = 0.01, p < .001. The mean score for the SIC was 1.11 (SD = 0.71), with no differences as a function of religious group membership or child gender.

Children’s Reported Emotional Security about the Community as a Predictor of their Reported Adjustment

Children’s reported emotional insecurity about the community in turn predicted greater emotional symptoms, r (104) = .19, p = .05, and, when controlling for religious group, and child age and gender, more total symptoms, B = 1.63, SE B = 0.83, p = .05. No moderators of relations between children’s emotional security about the community and adjustment outcomes were found.

Discussion

The current study begins the process of filling gaps in our understanding of the impact of community level antisocial behavior and ethnically/politically motivated antisocial behavior on children in Northern Ireland. Mothers’ reports confirmed that, despite ceasefires and peace accords, children in the region continue to be exposed to significant levels of sectarian antisocial behavior and violence. Further, mothers in two different cities in Northern Ireland appear to be able to make this distinction. Moreover, based on these reports, children’s exposure to community violence, particularly sectarian antisocial behavior, appears to undermine emotional security, and would also appear to be related to adjustment.

Consistent with prediction, children’s sectarian and nonsectarian violence exposure predicted their perceived adjustment. Based on data from their mothers, sectarian and nonsectarian violence exposure were significant predictors of children’s total difficulties, emotional symptoms, conduct problems, and lower prosocial behavior (sectarian antisocial behavior only), even after controlling for demographic variables of religious group membership and child age and gender. These findings, we believe, help to underscore the significance of violence in the political and community context for children. The relation of nonsectarian violence exposure to children’s adjustment is consistent with results of exposure to community violence in studies in the United States (Martinez & Richters, 1993). When entered simultaneously into multiple regression analyses, sectarian violence, but not nonsectarian violence, predicted greater emotional problems in children. These results support the notion that these two forms of violence provide distinct pathways of influence, and warrant being measured separately.

The notion of distinguishing inter- and intra-group community conflict may also be relevant to communities in the United States and elsewhere (Cummings, Goeke-Morey, Schermerhorn, Merrilees, & Cairns, in press). The concept may apply most readily to communities divided by race or ethnicity, but may also extend to communities divided on the basis of income, immigrant status, political or religious ideology, or sexual orientation. Furthermore, the presence of homeless individuals or gangs may foster a sense of “us” vs. “them” among residents of a community. In these communities, notions of inter- and intra- community stressors or conflict may be relevant to the understanding of the well-being for members of communities, and possibly serve as a basis for conceptualizing more effective community intervention programs.

One possible mechanism we hypothesize to account for the impact of community violence, particularly sectarian antisocial behavior, is emotional security (Lovell & Cummings, 2001). In this study we extend the test of emotional security theory to the extra-familial context of the community, addressing gaps in theory-driven and process-oriented research on the effects of political and community violence on children. Our preliminary results show that insecurity is related to both nonsectarian and sectarian violence, even after controlling for demographic variables including group membership. Further, our results suggest that children’s security about community seems to be especially undermined by sectarian violence exposure. This insecurity, in turn, places children at risk for greater adjustment problems, particularly emotional symptoms. This finding is consistent with theoretical conceptualizations and empirical work, suggesting that emotional security in the family context is closely related to children’s experiences of anxiety and fear. The construct of emotional security in the community should continue to be refined and tested empirically. For example, analyses in this study suggest that violence predicts insecurity, which in turn predicts adjustment problems, but longitudinal analyses are needed to rigorously test emotional security as a generative mechanism accounting for the influence of community violence on children’s adjustment. Moreover, the emotional security construct holds promise for advancing understanding of children’s development in the context of communities worldwide, particularly high risk communities fraught with crime and violence.

A major contribution, we believe, is the initiation of the development of instruments designed to assess children’s exposure to sectarian and nonsectarian community-level violence as it is currently being experienced in Northern Ireland. Ecologically sound measurement is essential for understanding the context of ethnic conflict and its potential influence on children’s functioning. Culturally-distinct forms of sectarian antisocial behavior are likely to vary in different cultural contexts of political discord. Any specific instruments may be of limited utility to researchers examining ethnic conflict in other regions of the world. It will therefore be important to develop culturally-appropriate measures to maximize the utility of these constructs. Also, it is likely that sectarian antisocial behaviors within the same region may vary across time as well, making scale development an ongoing process.

A number of limitations of the current study warrant consideration. The most important of these is that reports were made exclusively by mothers. This was done of necessity given the exploratory nature of our work and the associated ethical issues. Obtaining children’s or others’ reports of conflict exposure and adjustment would further enhance the validity of the measures and lessen the concerns of common method variance. Also, the sample in the current project was one of convenience. A larger, more representative sample of Catholics and Protestants matched for socioeconomic status and related variables would strengthen our ability to compare the experiences of children in these two communities.

Notwithstanding these limitations, we would argue that a notable contribution towards future work is our demonstration that sectarian and nonsectarian forms of community antisocial behavior can be distinguished, at least by mothers. The potential importance of this distinction is illustrated, we believe, by the fact that sectarian antisocial behavior was especially predictive of child adjustment, in turn raising questions as to how and why these particular events are more challenging for children. Moreover, emotional security was supported as a possible explanatory mechanism for effects on child adjustment. The present study we would therefore argue thus helps advance conceptualizations for new directions in evaluating the effects of community violence on children.

Finally, we would of course wish to point out that in order to understand fully the impact of political and community violence on children, we must consider not only processes within the child, such as emotional security and coping abilities, but also interpersonal factors such as family processes (Gibson, 1989). To date only scant research takes into account multiple contexts of children’s exposure to conflict and violence, including governmental and societal influences, on the one hand, and family processes, on the other. With regard to family, supportive and harmonious family environments and parents’ concern and control may mediate for children the stress of community conflict. Conversely, community conflict and violence may contribute to marital and family conflict and parents’ own adjustment problems (e.g., alcohol abuse), magnifying the risks for children in these contexts. Research combining all these elements is badly needed if a complete understanding of the impact of political conflict on children is to be advanced.