Faith in Contention

Abstract

What explains variation in the role of religion in ethnic conflict? Although conflict involving religion is often more violent and longer lasting than other forms of conflict, to date little research has examined the factors explaining the relevance of religion to conflict mobilization. Adopting a rational choice approach, I argue that religion is more likely to be a salient component of conflict when an ethnic group’s religious leaders face local incentives to compete over adherents. I test this approach using a multi-method research design that combines statistical analysis of original time-series data on the salience of religion in conflict with qualitative evidence drawn from the Troubles in Northern Ireland. Both types of analyses support the notion that competition among religious leaders can serve as a precursor to the mobilization of religious sentiment in conflict, which in turn exerts a tremendous influence on the intensity and duration of conflict.

Keywords

religion and politics civil war conflict processes

The last several decades have witnessed a dramatic increase in the use of religious rhetoric and symbolism to promote mobilization in conflict. The percentage of civil wars involving religion more than doubled from the 1960s to the 1990s (Toft, 2007). In the Middle East and North Africa alone, the proportion of armed conflicts involving religious claims increased from 0% in 1975 to 75% in 2011 (Svensson, 2013). Indeed, the number of countries experiencing religious terrorist violence has doubled in the last decade (Pew Research Center, 2014b).

Although the infusion of religion in conflict tends to heighten the intensity and duration of political violence, little is known about the conditions that contribute to the religious character of conflict. Why did conflict in Bosnia take on a religious tone while conflict between Turkish Muslims and Greek Christians in Cyprus remained largely secular? Why did Sinhalese Buddhist monks advocate military force during the Sri Lankan civil war while Hinduism was absent from the rhetoric of the Tamil Tigers?

Adopting a rational choice approach, I posit that religion is more likely to become a salient component of ethnic conflict when multiple religious organizations compete for the support of adherents within an ethnic group. My argument is derived from existing literature suggesting that religious leaders are most politically active when they risk losing adherents to rivals. Applying these insights to the incentives of religious leaders during ethnic conflict, I hypothesize that religious competition within group boundaries has a tremendous influence on the religious character of conflict across group boundaries.

I test this argument by combining statistical analysis of original time-series data on the salience of religion in conflict with qualitative evidence of religious competition during the Troubles in Northern Ireland. Both types of analysis support the notion that religious rhetoric, and by extension the likelihood that conflict will take on the intense and intractable characteristics of religious violence, is strongly influenced by competition among an ethnic group’s religious leaders. I conclude by noting the limitations of this analysis and the implications of these findings for the study of ethnic mobilization and religious conflict.

The Salience of Religion in Ethnic Conflict

Religious conflict is often more violent and longer lasting than other forms of conflict.¹ Rebel groups motivated by religion are often more effective in combat and more resilient to counterinsurgency tactics (Berman, 2011; Berman & Laitin, 2008; Toft & Zhukov, 2015). Because religious conflicts often concern indivisible goals, they are also less likely to be resolved through negotiated settlement (Hassner, 2009; Svensson, 2007; Toft, 2006). These challenges can be seen worldwide, from sustained religious conflict in the Middle East and South Asia to the proliferation of religious terrorist violence across Western Europe and North America.

Despite the destructiveness of religious violence, there have been few attempts to systematically theorize about the salience of religion in conflict or to compare across diverse cases (Feliu & Grasa, 2013). Although some conflicts across religious boundaries are articulated in religious terms, others remain largely secular, and many others contain a mix of religious and secular mobilization. The second Sudanese civil war pitted the “Islamizing” regime in Khartoum against the avowedly secular Southern People’s Liberation Army (Toft, 2007, p. 124). Similarly, although religion served as a critical point of differentiation between Serbs, Croats, and Bosnian Muslims in the Yugoslav Wars, the relevance of religion to conflict mobilization differed considerably across these groups and over time (Sells, 1998). Why is religion central to the mobilizing platform of some groups but absent from that of others, even within the same conflict?

Although few scholars have examined variation in the salience of religion in conflict, considerable research has explored the conditions in which religion serves as a focal point for collective mobilization. This research has coalesced around three categories of explanation: political theology, transnational religious kin, and religious competition.

First, a small body of literature has focused on political theology to explain religion’s involvement in collective mobilization. Philpott (2007) emphasizes the importance of theology to explain the relationship between religion and the state. Henne (2012a) highlights the role of theology in justifying suicide terrorism, whereas Juergensmeyer (2000), Moghadam (2008), and Rapoport (1984) argue that messianic interpretations of religion, often but not exclusively within Islam, contribute to the intractability of conflict.

Although theology undoubtedly plays an important role in religious conflict, the causal relationship between theology and conflict mobilization is difficult to disentangle. In different contexts, the same religious symbols have been interpreted to support radically different political attitudes and behaviors (Appleby, 2000; Stepan, 2000). Theology alone is unlikely to explain why Catholic priests promoted social mobilization in Latin America but not in Northern Ireland. Furthermore, considerable research provides evidence against the notion that Islam is inherently conflict-prone, suggesting instead that radical interpretations of Islam may be endogenous to conflict (Fish, Jensenius, & Michel, 2010; Fox, 2000; Piazza, 2009; Toft, 2007).

A second category of literature highlights the importance of transnational religious kin to explain religion’s role in collective mobilization. Religious groupings are often the largest demographic unit in which individuals claim membership and global religious networks often control considerable mobilizing resources (Grzymala-Busse, 2012; Smith, 1996). By appealing to religion, political entrepreneurs may hope to gain access to these resources (Fox, 2002). Toft (2007) notes the vast resources of the global Muslim community as a potential explanation for the disproportionate presence of Islam in armed conflict. Similarly, Bakke (2013) argues that the Islamization of the Chechen separatist movement can be explained by the need to stimulate support from religious kin in the Middle East.

Although this literature provides a compelling explanation for the relationship between Islam and collective mobilization, this approach fails to explain the role of religion in numerous conflicts beyond the Muslim world. Catholic paramilitaries in Northern Ireland declined to invoke religion despite the considerable mobilizing resources of the global Catholic community (Mitchell, 2006). Similarly, Tamil militants declined to invoke Hinduism during the Sri Lankan civil war despite the rising popularity of Hindu nationalism in neighboring India (Hansen, 1999).

A third category of scholarship explores the role of religion in collective mobilization as a function of competition among religious organizations. Building on the economics of religion, this approach relies on the notion that religious organizations function as firms in a religious marketplace (Finke & Stark, 1992; Iannaccone, 1991, 1992; Stark & Finke, 2000). Adherents are viewed as rational consumers who seek to maximize doctrine, spiritual rewards, and material rewards. Likewise, religious leaders seek to maximize membership and resources for their firms. Where multiple firms compete over adherents, each firm must appeal to the preferences of adherents or risk losing market share. Assuming pluralistic religious preferences across society, religiosity is greatest where religious firms compete for adherents, and adherents are therefore able to find a religious product that satisfies their preferences (Stark & Bainbridge, 1987).

Although developed as an explanation for religiosity, considerable research has applied this approach to the study of political activism. Gill (1998) and Trejo (2009) find that the Catholic Church is more likely to become involved in collective mobilization where it faces competition from Protestant churches. Similarly, Kalyvas (1996) and Warner (2000) find that the emergence of Christian Democratic parties in post-war Europe can be explained in part by the self-interested behavior of the Catholic Church. This approach suggests that religion’s role in collective mobilization is a function of those factors which influence competition among religious firms, namely, a state’s level of religious diversity or a state’s religious regulatory environment.²

This literature provides a compelling explanation for religious political activism in a variety of contexts. However, the standard religious economy model is a poor fit for understanding the role of religion in ethnic conflict. The premise of this model is that competition is greatest where conversion from one religious firm to another is relatively low cost.³ In highly divided societies, however, the theological cost of conversion may pale in comparison with the cost of transcending contentious communal boundaries. As Stark and Finke (2000) note, “in multicultural caste systems where each caste has its own religion, there is much pluralism but no competition” (p. 218). Likewise, in Northern Ireland, communal violence makes religious competition between Protestants and Catholics unlikely despite their relative theological proximity and a nominally “competitive” religious marketplace. If contentious ethnic boundaries render conversion prohibitively costly, the standard religious economy model appears ill suited to explaining variation in the religious character of ethnic conflict.

Religious Competition in Ethnic Conflict

However, the absence of competition across religious boundaries does not preclude the possibility of competition within religious boundaries. I argue that religion is more likely to form a salient component of conflict when an ethnic group’s religious leaders have incentives to compete over adherents within ethnic boundaries. This argument brings the logic of the religious economy model down one level of analysis to examine how competition within a single religious community influences the mobilization of religious sentiment in ethnic conflict.

To understand this argument, begin by considering an ethnically divided society in which religious boundaries overlap with ethnic boundaries. To be a member of one ethnic group, one must also profess a particular religious affiliation. To be Sinhalese in Sri Lanka generally entails being a member of the Buddhist faith. To be a Serb in the Balkans is not only to speak Serbian but also generally to identify as a Serbian Orthodox Christian.

Now assume, in keeping with the religious economy model, that religious leaders catering to each group seek to maximize membership and resources for their firms (Gill, 1998; Iannaccone, 1991; Trejo, 2009). Although the high cost of conversion across contentious ethnic boundaries renders inter-group religious competition unlikely, a competitive religious market may operate within the group if adherents have a choice over which denominations or congregations to patronize. In the terminology of Stark and Finke (2000), although conversion across religious boundaries is unlikely, reaffiliation of organizational membership within the group may remain relatively low cost, depending on the structure of the group’s religious organizations.⁴

I define a religious organization as a self-contained, formal civil society organization that produces and distributes religious goods (Gill, 2008, p. 42). Religious organizations vary considerably in their scope and structure, from local and autonomous congregations to regional or national networks of religious bodies (a presbytery) and transnational hierarchical entities (the Catholic Church).

Competition among adherents within an ethnic group can occur only when adherents have a choice in religious organizational membership. When a single organization monopolizes religious allegiance within a group, adherents have little say in organizational membership, and religious leaders therefore have little incentive to cater to the preferences of adherents. Thus, to be a Serb in the Balkans means not only to profess a belief in Orthodox Christianity but also to belong to the Serbian Orthodox Church and, often, to attend services at a particular local parish. Because such religious producers face no competition within group boundaries, they have no structural incentive to engage in activities aimed at increasing membership.

By contrast, where multiple fragmented organizations cater to a single group, adherents within that group have the power to choose their organizational affiliation, and religious leaders must therefore appeal to the preferences of adherents. Although Serbs may have little choice in religious membership, Bosnian Muslims have numerous local mosques to choose from. With multiple religious producers and no standard dictating membership in any given religious organization, Bosnian Muslims enjoy a competitive religious market despite nominal religious homogeneity within the group.

How does within-group religious competition contribute to the religious character of ethnic conflict? To understand this process, it is useful to establish what adherents want from their religious leaders (the demand side) and what religious leaders can offer adherents in conditions of ethnic conflict (the supply side).

On the demand side, the religious economy literature notes that adherents have heterogeneous religious preferences and generally seek to maximize doctrine, spiritual rewards, and material rewards (Stark & Bainbridge, 1987). However, the social and political hostilities associated with ethnic conflict are likely to have a profound effect on religious preferences across the group. Seul (1999) argues that religious leadership becomes more important to individuals experiencing communal violence. Similarly, Buckley (2016) finds that popular demand for religious politics is highest when religion occupies a position of relative weakness in society. Although Buckley does not explicitly consider ethnic conflict, conflict along religious lines is likely to encourage individuals to perceive their religion as weak or under threat. Therefore, although religious preferences are heterogeneous, ethnic conflict is likely to encourage demand for public religious leadership across the group.

On the supply side, religious firms facing competition within group boundaries will respond to this demand by providing a greater supply of public religious leadership than their competitors. In normal conditions, religious firms facing competition have a variety of strategic options, including providing tangible goods to members or lobbying the state for special privileges (Gill, 1998; Hagopian, 2008). Increased demand for public religious leadership during conflict provides firms with two additional strategies.

First, firms may compete by becoming directly involved in politics, either by establishing their own political organizations or by forming alliances with existing political or paramilitary organizations (De Juan, 2008; Kalyvas, 1996; Warner, 2000). The elevated public presence of religious leaders should dramatically increase the likelihood that ethnic conflict is articulated in religious terms. However, direct political activism also risks tainting religious leaders with failure and thus diminishing their authority or alienating their followers (Buckley, 2016; Warner, 2000). For this reason, religious leaders may be wary of directly entering politics.

Second, firms may compete by embracing radical interpretations of theology targeting out-groups. Although most adherents will express an elevated demand for religious leadership during conflict, at least a minority of adherents will also develop a demand for religious chauvinism. Because members of other ethnic groups are, by definition, beyond the reach of the religious firm, we may expect some religious leaders to respond to this demand by espousing radical theologies that vilify members of other groups.⁵ By publicly voicing religious endorsement for conflict against the out-group, this religious outbidding should have a tremendous effect on the religious character of conflict.

Although not all religious firms will respond equally to these market pressures, in the aggregate, we should expect at least a minority of religious leaders to respond to ethnic conflict by becoming highly politically active, either through direct political activism or through articulation of radical religious sentiment targeting out-groups.

However, as noted above, political activism can impose costs on religious leaders. Consequently, religious firms should only become politically active when they face competition within the boundaries of their group, and therefore have incentives to pay the costs of political activism. When a single religious firm dominates a group’s religious market, that firm will face few structural incentives to respond to the tastes of adherents when doing so imposes costs. This logic suggests the following hypothesis:

Hypothesis 1: Religion is more likely to be salient in conflict for ethnic groups with fragmented religious organizations than for ethnic groups with a single religious organization, other factors being equal.

If this hypothesis provides an accurate account of the salience of religion in ethnic conflict, we should expect the religious character of conflict to vary across groups based on the structure of religious organizations. Where fragmented religious firms compete over adherents within a group, the benefits of religious political activism should outweigh the costs, and religion should form a salient component of conflict mobilization. By contrast, where a single religious firm monopolizes religious membership within a group, religious leaders stand to gain little from political activism, and religion should remain peripheral to the conflict. Findings inconsistent with these expectations would undermine the explanatory value of this argument.

Research Design

The goal of this analysis is to test the relationship between the structure of an ethnic group’s religious organizations and the salience of that group’s religious identity in conflict. I accomplish this goal by combining statistical analysis of original time-series data on the salience of religion in conflict with qualitative evidence of religious competition during the Troubles in Northern Ireland. My qualitative evidence draws on original archival research and a sample of 25 interviews with religious and political elites conducted in summer 2015. Below I describe the design of my statistical analysis, including rules regarding the universe of cases and measurement of key variables.

Universe of Cases

I limit the universe of cases in this analysis to religiously distinct ethnic groups experiencing ethnic conflict.⁶ To ensure that the salience of religion can vary within groups, I include only ethnic groups whose members share a religious identity as well as another type of ethnic commonality (language, race, tribe, or nation), both of which distinguish the group from the remainder of the state.⁷ This universe of cases includes 42 ethnic groups in 22 countries.

By focusing on groups distinguished by multiple types of shared identity, I include all groups for which the structural conditions are in place for religion to be a salient component of conflict without rendering this outcome predetermined. This method relies on two assumptions regarding ethnic structure. First, religion is unlikely to be a salient component of group politics unless the group shares a common religious identity. Second, however, we should expect religion to form a central component of group politics during conflict if religion is the only ethnic identity held in common by group members.⁸ By requiring that all groups have multiple forms of shared identity, I hold such demographic considerations constant. Although this selection process omits many groups with salient religious identities, it provides the necessary structural conditions to observe variation in the salience of religion independent of demographic considerations.

Dependent Variable

The dependent variable of this analysis is the salience of a group’s religious identity in conflict. By the salience of religion in conflict, I refer to the explicit public use of religious rhetoric and symbolism by the political representatives of the group to promote collective mobilization. This rhetoric-based approach focuses on the use of religion to encourage collective action and is ambivalent to whether religious rhetoric is instrumental or sincere. A growing body of literature has demonstrated the utility and precision of this rhetoric-based approach to measuring the religious character of conflict.⁹

To capture this concept, I begin by coding whether each of the political organizations representing an ethnic group uses religious rhetoric to encourage collective action on an annual basis from 1970 through 2012. I identify these organizations based on a series of secondary sources.¹⁰ An organization is considered to represent an ethnic group if it makes some form of ethnic rhetoric central to its political platform (Chandra, 2011). I include political parties, paramilitaries, and protest organizations. For each source and each year, organizations are coded as using religious rhetoric if they (a) include reference to the religious identity of the group in their name, or (b) were identified as making explicitly religious appeals for mobilization by one or more sources.

Based on this information, I generate Religious Salience, a continuous variable measuring the percentage of organizations representing a group in a given year that use religious rhetoric. Higher values on this variable indicate that a larger proportion of the political representatives of a group seek to appeal to their constituents on the basis of religion. This measure reliably captures variation in the extent to which the political representatives of a group frame conflict as religious, both across groups and over the life span of a conflict.

This data generation process identified 495 political organizations over the course of the period from 1970 through 2012. Of these, 160 organizations used religious rhetoric at some point and 335 did not. Figure 1 shows the percentage of organizations using religious rhetoric by year across all groups in the data. These data provide striking confirmation of variation in the salience of religion over time. Across all groups, approximately 26% of organizations in 1970 used religious rhetoric to some extent. By 2006, this number had climbed to 37%. These data are consistent with previous research on the rise of religious conflict over this period (Fox, 2012; Svensson, 2013; Toft, 2007).

Figure 1.

Religious rhetoric over time, 1970 to 2012.

Independent Variable

The purpose of this analysis is to assess the effect of within-group religious competition on the salience of religion in conflict. The majority of research using the religious economy perspective has measured religious competition as a function of religious diversity.¹¹ Such a measure rests on the assumption that professed religious identity can serve as a proxy for religious affiliation, and therefore religious diversity corresponds to fractionalization of the religious market.

Although this assumption is reasonable when considering competition across groups, professed religious identity provides a poor measure of religious competition within groups. As noted above, an ethnic group with near homogeneous religious identity may enjoy a competitive religious market if the group’s religious organizations allow adherents to choose their congregational affiliation. Although the Buddhist community in Sri Lanka is almost uniformly Theravada Buddhist, adherents can choose to worship at countless local Buddhist temples, the leaders of which therefore have some incentive to appeal to the preferences of worshippers (Gombrich, 1988). This type of competitive religious market is not captured by even the most fine-grained measure of religious diversity.

For this reason, the primary independent variable of this study is the structure of an ethnic group’s religious organizations. I capture this concept with a binary variable measuring Religious Structure as either fragmented (1) or monopolistic (0). Fragmented religious organizations are defined by freedom of choice over congregational affiliation and may consist of a multitude of unaffiliated congregations, several hierarchical or networked religious organizations (e.g., a regional presbytery), or a combination of both. By contrast, monopolistic religious organizations are defined by prescribed congregational affiliation and universal nominal membership within the group. Examples include the Serbian Orthodox Church among Serbs in the Balkans or the Catholic Church among Catholics in Northern Ireland.

I code this variable based on secondary sources, including The Blackwell Dictionary of Eastern Christianity (Parry, Brady, Griffith, Melling, & Healey, 2000), The Encyclopedia of Religion (Jones, 2004), The Oxford Dictionary of Islam (Esposito, 2003), and The Oxford Handbook of Global Religions (Juergensmeyer, 2006). For each group, I note the structure of religious organizations and the number of discrete centers of religious authority catering to the group. This variable is time-invariant. Of the 42 ethnic groups included in the data, 23 have fragmented religious organizations and 19 have a monopolistic religious organization.

In robustness checks, I substitute this binary measure with an original Herfindahl index measuring religious fractionalization within each ethnic group according to professed religious identity.¹² Although this measure effectively captures within-group religious diversity according to professed religious identity, this measure provides a suboptimal proxy for the structural conditions favoring religious competition.

Controls

In addition to these variables of interest, I include several control variables that are likely to influence the salience of religion in conflict. I include the theological distance between conflicting groups, relative size of the group, economic development, and political fragmentation of the group. Theologically distant groups may have a lower likelihood of conflict over religious issues (Montalvo & Reynal-Querol, 2005).¹³ Relatively smaller groups may be more likely to use religious rhetoric to compensate for power asymmetries in conflict (Grzymala-Busse, 2012).¹⁴ Less developed economies may be associated with higher levels of religiosity in general, which may translate into greater religious political activism (Inglehart & Welzel, 2005).¹⁵ Finally, political leaders facing competition from rivals may resort to religion to outbid their competitors (Isaacs, 2016; Toft, 2007).¹⁶

I also consider how the relationship between religious structure and the salience of religion is influenced by three additional variables discussed by previous literature. Consideration of these variables is intended to test the robustness of the effect of religious structure, rather than to systematically test these alternative explanations. First, as noted above, research has suggested that religious rhetoric may be more likely when religious claims offer political actors access to a large base of support from transnational religious kin (Bakke, 2013; Toft, 2007).¹⁷ Second, the decline of communism at the end of the Cold War may have encouraged political actors to turn to religion to promote collective action (Huntington, 1996; Juergensmeyer, 1993). Third, as described above, the religious economy approach implies a straightforward relationship between religious rhetoric and a state’s overall religious diversity.¹⁸

In robustness checks, I consider the effect of numerous additional variables, including measures of political opportunity, the legal relationship between religion and the state, and the demographic characteristics of the state. The online appendix contains a description of these additional variables and the results of these robustness checks.

Quantitative Evidence

Hypothesis 1 predicts that religion is more likely to form a salient component of conflict for groups with fragmented religious organizations than for groups with a monopolistic religious organization, other factors being equal. Figure 2 displays the percentage of political organizations using religious rhetoric in each year distinguished by religious structure.

Figure 2.

Religious rhetoric by religious structure, 1970-2012.

Although religious rhetoric varies considerably within each category, religion consistently carries greater salience among groups with fragmented religious organizations than among groups with monopolistic religious organizations. Of the 23 groups with fragmented religious organizations, two thirds (66.7%) experienced some degree of religious rhetoric. Of the 19 groups with a monopolistic religious organization, less than one half (47.4%) experienced any religious rhetoric during this period.

However, Figure 2 provides a poor test of Hypothesis 1. Without considering the additional variables listed above, we have no way of knowing how much of this relationship is caused by religious structure and how much is caused by the coincidence of religious structure with other determinants of religious rhetoric. For this reason, I estimate the effect of religious structure relative to a number of additional factors.

The ideal statistical technique for testing Hypothesis 1 with time-series cross-sectional data would be a fixed-effects regression. By design, however, fixed-effects regressions are unable to estimate the effect of time-invariant variables like Religious Structure. To solve this challenge, I estimate the effect of Religious Structure while controlling for fixed group and year effects using a fixed-effects vector decomposition (FEVD) analysis.¹⁹ In robustness checks, I subject the same data to two alternative methods for estimating time-invariant coefficients: random effects regression and pooled ordinary least squares (OLS) regression. I describe the applicability of these alternative methods and present the results of these robustness checks in the online appendix.

Table 1 displays the results of five alternatively specified regressions estimating the effect of Religious Structure on the salience of religion in conflict. Across all models, I control for theological distance, group proportion, gross domestic product (GDP) per capita, and political fragmentation. Models 2 through 5 also control for the lagged dependent variable (LDV) to account for the non-independence of observations over time.²⁰ Model 3 considers the effect of a group’s transnational religious kin, Model 4 controls for the effect of the Cold War, and Model 5 controls for a state’s overall religious diversity.

Table 1.

Regression Models Explaining Religious Salience.

	Model 1	Model 2	Model 3	Model 4	Model 5
Religious structure	26.96**	8.372**	12.00**	7.924**	7.130**
	(8.027)	(2.582)	(4.077)	(2.690)	(2.399)
Theological distance	4.166	−5.329	−11.39*	−5.007	−4.812
	(10.47)	(4.627)	(5.785)	(4.676)	(4.535)
Group proportion	28.32	−12.56	−17.01	−12.07	−12.56
	(29.13)	(13.35)	(15.01)	(13.47)	(13.08)
ln(GDP per capita)	−1.971	−0.786	−4.237*	−2.301*	−0.786
	(2.840)	(1.061)	(1.731)	(1.121)	(0.994)
Political fragmentation _{t − 1}	1.794	0.366	0.003	0.129	0.366
	(1.142)	(0.419)	(0.485)	(0.450)	(0.399)
ln(religious kin)			7.744*
			(3.169)
Cold War				−4.783
				(2.771)
State religious diversity					−1.684**
					(0.612)
Religious salience _{t − 1}		0.750**	0.691**	0.738**	0.750**
		(0.038)	(0.048)	(0.039)	(0.037)
Constant	4.226	17.35	−99.93	31.22*	23.26*
	(26.97)	(11.96)	(54.75)	(12.94)	(11.42)
R ²	0.707	0.868	0.877	0.870	0.868
Observations	662	661	563	661	661

Standard errors in parentheses. GDP = gross domestic product.

p < .05. **p < .01.

The findings presented in Table 1 confirm that the structural conditions favoring religious competition within an ethnic group are positively associated with the salience of religion in conflict. The significant positive coefficient on Religious Structure in Model 1 indicates that a move from monopolistic religious organizations to fragmented religious organizations corresponds to a roughly 27% increase in the proportion of the group’s political organizations using religious rhetoric in conflict. Indeed, Religious Structure is consistently the most powerful single predictor for Religious Salience across all five models presented in Table 1.

This relationship holds while controlling for theological distance, group proportion, GDP per capita, and political fragmentation. As anticipated, GDP per capita appears to have a negative effect on religious rhetoric and political fragmentation appears to have a positive effect, though these relationships come short of statistical significance in most models. These findings are not significantly different when controlling for the LDV (Religious Salience _t _{− 1}).

Models 3 through 5 offer further evidence for Hypothesis 1. Religious Structure continues to have a significant effect on the salience of religion when controlling for the size of a group’s transnational religious kin, the Cold War, and a state’s overall religious diversity. As anticipated, the positive coefficient on Religious Kin in Model 3 suggests that transnational religious kin have a positive effect on religious rhetoric. Similarly, Model 4 indicates that the Cold War is associated with lower levels of religious rhetoric, although this relationship falls short of statistical significance. Contrary to the expectations of the religious economy approach, the negative coefficient on State Religious Diversity in Model 5 suggests that higher levels of religious diversity at the state level may be negatively related to the salience of religion in conflict.

These findings provide compelling evidence that the structural conditions encouraging religious competition within a group are strongly associated with the proliferation of religious rhetoric in conflict. By extension, this analysis suggests that the intense and intractable qualities of religious violence have their origins, in part, in the behavior of local religious leaders concerned with the viability of their organizations. These findings are replicated in numerous robustness checks, including subjecting the data to a random effects regression and a pooled-OLS regression, substituting Religious Structure with a continuous measure of within-group religious fractionalization, considering alternative lag periods, and including numerous additional control variables.²¹

Qualitative Evidence

Although I provide robust statistical evidence for Hypothesis 1, cross-national statistical analysis can offer only limited insight into the behavior of religious leaders in conflict. To better explore these findings, I explore the behavior of Protestant and Catholic religious leaders during the Troubles in Northern Ireland. My evidence draws on original archival research and a sample of 25 interviews with religious and political elites conducted in summer 2015. This case study is intended to demonstrate the empirical validity of my statistical findings, and is not intended to serve as an independent test of my argument.

The conflict between Northern Ireland’s Protestants and Catholics, commonly referred to as the Troubles, lasted from 1968 to 1998 and resulted in approximately 3,500 deaths (Coogan, 2002). This conflict provides an ideal demonstration of my argument for two reasons. First, in addition to religion, the two communities in Northern Ireland are distinguished by national identity. To be Protestant generally means to consider oneself British, whereas to be Catholic is generally to identify as Irish. Due to this overlap, conflict involving either group may be articulated in national or religious terms.

Second, this conflict displays internal variation in all relevant variables. At the height of the Troubles in the 1970s, Protestant political leaders frequently emphasized the religious character of the conflict while their Catholic counterparts relied almost exclusively on secular nationalist rhetoric.²² Elite interviews reinforced this variation. Although Protestant leaders consistently described the conflict as religious, Catholic religious and political elites stressed the secular nature of the conflict. In the words of one Catholic priest, “it was both a religious and a political conflict—religious for Protestants but political for Catholics” (G. Reynolds, personal communication, August 12, 2015).

Regarding religious structure, the Protestant community in Northern Ireland is fragmented into three major denominations (the Presbyterian Church in Ireland, the Church of Ireland, and the Methodist Church in Ireland) and dozens of smaller denominations. By contrast, the Catholic Church enjoys a monopoly over the Catholic community. This variation has a considerable effect on religious practice. As one former leader of the Irish Presbyterian Church explained, “if you’re Protestant, you go to the church you like the most. If you’re Catholic, you go to the nearest church because you have to” (J. Dunlop, personal communication, August 19, 2015).

As I demonstrate below, this structural variation had a considerable influence on the behavior of religious leaders during the Troubles, which in turn shaped the rhetoric adopted by major parties to the conflict.

Protestant Competition and Religious Mobilization

Given the fragmented structure of Protestant religious organizations, I expect Protestant religious leaders to respond to ethnic conflict by becoming increasingly politically active. Evidence suggests that the Protestant religious market in Northern Ireland was characterized by considerable interdenominational competition even before the start of the Troubles. In 1953, the Methodist Church declared a “year of evangelism,” and pointed to the need to outdo the Church of Ireland through a program of active proselytism (“Evangelistic Drive in 1953,” 1952, p. 9). The Presbyterian Church followed suit in 1959, adopting a program designed to expand their market share among Protestants (Presbyterian Church in Ireland, 1964). The most intense competition revolved around the relatively small Free Presbyterian Church of Ulster (FPCU) led by firebrand preacher Ian Paisley. Although the FPCU was an insurgent force in the Protestant religious market from its foundation in the early 1950s, religious competition remained a relatively apolitical process during this period (Moloney, 2008).

As anticipated, however, Protestant religious competition took on an increasingly political character with the rise of communal tensions in the late 1950s. Especially after the Irish Republican Army’s (IRA) Border Campaign of 1956-1962, a growing share of Protestants expressed a desire for public religious leadership.²³ In 1956, FPCU leaders played a central role in establishing Ulster Protestant Action (UPA; 1965), a political organization with the goal of protecting the “Protestant faith and heritage” of Northern Ireland. Under the banner of the UPA, the FPCU was able to expand its ranks through considerable activism and public protest.

Although the anti-Catholic tone of Free Presbyterian mobilization has been well documented, less attention has been paid to the FPCU’s vocal criticism of other Protestant denominations (Bruce, 2009). Building on a growing pool of disaffected Protestants, FPCU protests and publications increasingly sought to characterize other Protestant churches as weak and conciliatory to the Catholic Church.²⁴ This behavior provides striking evidence of religious competition through public activism. As one former Presbyterian minister explained, the Free Presbyterians “were out to win as many Protestants over to their cause as they could . . . There was a kind of warfare going on” (A. Kennedy, personal communication, August 29, 2015).

As expected, other Protestant denominations responded to the FPCU’s challenge by drawing themselves further into the public sphere. The success of the UPA ushered in a proliferation of Protestant pressure groups with the backing of rival denominations, including the Evangelical Protestant Society and the Ulster Orange and Protestant Committee (Bryan, 2000). When the FPCU expanded into the rural village of Rasharkin, the local Presbyterian minister sought to undercut their growth by replicating their style of public proselytism. “It was very much demand driven,” the minister explained in an interview, recalling his decision to print pamphlets denouncing the Free Presbyterians (A. Kennedy, personal communication, August 29, 2015). Similar processes of local competition and public activism were replicated across Northern Ireland. The tone of the time is well captured by a headline from The Belfast Telegraph asking, “Should Publicity Be Used to Sell Religion?” (1959, p. 8).

By the middle of the 1960s, Protestant religious competition began to have a clear effect on relations with Catholics. In June 1966, FPCU leader Ian Paisley led a protest march from his Free Presbyterian congregation to the headquarters of the Irish Presbyterian Church in central Belfast. En route, the crowd of approximately 1,000 protesters marched through the predominantly Catholic Cromac Square neighborhood. Violent confrontations between Free Presbyterians and Catholics marked one of the first instances of sectarian violence in Northern Ireland for decades (O’Callaghan & O’Donnell, 2006).

The following Sunday, attendance at Paisley’s Free Presbyterian service was standing room only, with a large crowd overflowing into the surrounding streets. “With the help of God and the Protestants of Ulster,” the FPCU leader declared at a sermon several weeks later, “the day is coming when I will be in the House of Commons” (“Paisley Says He’ll Be MP at Stormont,” 1966, p. 1). Later that year, Paisley announced the formation of the Protestant Unionist Party (PUP) and the FPCU began regularly running clergy for political office (Bruce, 2009).

In response to this escalation of religious competition, clergy from several rival denominations also became politically active. In 1973, Methodist minister Robert Bradford ran for a seat in the Northern Ireland Assembly. By his own account, his political career was motivated by religious conviction and a desire to expand the evangelical wing of Methodism (Bradford, 1984, p. 81). Mimicking the rhetoric of his Free Presbyterian counterparts, Bradford’s campaign emphasized the liberalism of other Protestant denominations and the threat posed by the Catholic Church while encouraging voters to join his particular denomination.

In 1975, Presbyterian minister W. Martin Smyth (1965) entered the political arena following the success of a series of pamphlets calling for the defense of traditional Protestantism against the “tremendous growth of heretical sects,” including the FPCU (p. 9). Although Smyth used his political exposure to argue for a mainstream style of Presbyterianism, his rhetoric was equally informed by ongoing communal violence. “There must be no compromise between soldiers of Christ and their enemies,” Smyth declared in a pamphlet recounting the history of Protestantism in the United Kingdom (Smyth, 1960, p. 13).

This proliferation of religious political activism continued throughout the Troubles. In 1982, a by-election pitted Smyth against FPCU minister William McCrea, with one Catholic periodical derisively predicting “divine intervention” (1982, p. 3) in the poll. Indeed, from the early 1970s through the late 1990s, nearly 20% of Protestant Members of Parliament from Northern Ireland were clergymen (Bruce, 1998).

As anticipated by the religious competition approach, this rapid expansion of public religious activism had two effects on the Protestant social landscape. First, Protestant clergy were generally rewarded for their activism. In 1966, the year of the Cromac Square riots, the East Belfast FPCU congregation outgrew its facility so rapidly it was forced to rent a public auditorium to conduct church services. Over the next 3 years, the FPCU nearly doubled its size and annual revenue while many of its leaders sought political office (Bruce, 2009). By contrast, membership in the Church of Ireland, the least politically active of the Protestant denominations, declined sharply during this period (Brewer, 2004).

Second, Protestant religious competition had a direct influence on the mobilizing rhetoric of Protestant paramilitary organizations. Both the Ulster Defence Association (UDA) and the Ulster Volunteer Force (UVF), prominent Protestant paramilitaries, gained support in the early 1970s by espousing rhetoric identical to that used by Protestant clergy in their bids for religious influence. In 1973, the UDA’s Ulster Militant published a series of articles lamenting the betrayal of Protestantism and denouncing Irish nationalists as “papists” intent on imposing Vatican control over the Province (UDA, 1973, p. 6). The July 1976 edition of the UVF’s (1976) Combat stressed that Northern Ireland “is the most Protestant part of the United Kingdom . . . .We cannot now compromise with ‘Political Popery’” (p. 9).

Although there is little indication that religious leaders directly engaged in violence, evidence does indicate that the rhetoric of Protestant clergy provided religious justification for paramilitary violence and encouraged paramilitary recruitment. Reflecting on his pathway to violence, one former militant claimed he was “sorry [he] ever heard of that man Paisley or decided to follow him” (Boulton, 1973, p. 54).

The Catholic Church and Religious Monopoly

In contrast to the religious significance of the Troubles among Protestants, the religious component of Catholic communal identity remained largely peripheral to the conflict. As noted above, the Catholic Church enjoys a monopoly over religious allegiance among Northern Ireland’s Catholics. Facing little competition over adherents, I expect such an organization to be unwilling to become involved in costly political activism during conflict.

As anticipated, the Church generally avoided politics throughout the Troubles. In the rare instance that the Church did become involved in politics, it generally did so to protect its financial interests in the realm of education (McGrath, 2000). As one priest noted, “there was an emphasis on protecting the Church as an institution” (G. Donegan, personal communication, August 21, 2015). Indeed, the costs of entering politics weighed heavily on Church leaders throughout the Catholic hierarchy. One survey conducted in 1986 found that just 4% of Catholic priests in Northern Ireland were willing to give specific political advice to their parishioners (McElroy, 1991). One priest argued at the time that clergy intervention in the conflict “could make an unacceptable situation even worse” (Wilson, 1983, p. 25). In the absence of competition over adherents, the Church had little incentive to pay these costs.

Importantly, Catholic clergy remained out of the political fray despite three factors that might have favored religious activism had they faced competition over adherents. First, Catholic adherents, like their Protestant counterparts, expressed a clear demand for public religious leadership during the conflict. By one account, unwillingness to condemn the British state was “damaging the Catholic Church’s credibility in the eyes of many of its people” (“The Churches in N. Ireland,” 1980, p. 23). Especially in parishes with considerable paramilitary support, clergy were aware that their unwillingness to take a stand in the conflict alienated them from the local community (Brewer, Higgins, & Teeney, 2011). In North Belfast, protests followed one priest’s condemnation of the IRA in the early years of the Troubles (McElroy, 1991).

Second, many Catholic priests privately supported the goals of the Irish nationalist movement. A survey conducted in 1986 found that 91% of priests in Northern Ireland supported a united Ireland (McElroy, 1991). Archbishop Tomas O Faich, leader of the Irish Catholic Church in the 1980s, was well known for his personal support for Irish unification. Asked about his political influence, however, the Archbishop stressed that he “would not dream of using his ecclesiastical position to promote nationalism” (McElroy, 1991, p. 57).

Third, the political representation of the Catholic community was relatively open in the early stages of the Troubles. The once dominant Nationalist Party witnessed an erosion of support among Catholics in the late 1960s, and the IRA splintered into the Official IRA and the Provisional IRA in 1969 (Coogan, 2002). This opening encouraged the mobilization of new political actors seeking to represent the Catholic community, including the Social Democratic and Labour Party (SDLP; Murray & Tonge, 2005). If Catholic clergy were waiting to enter the political arena, the early 1970s provided the ideal opportunity to seek out new political allies.

However, not one of these new political entities framed Catholic identity in religious terms. Although the Protestant political landscape became increasingly religious, the SDLP championed an explicitly secular nationalism (Murray, 1998). One founding SDLP member ultimately left the party after the media took to calling it the Catholic SDLP. “Things like that imply closeness to the Catholic religion,” he explained, “I wanted no religion near politics” (Murray, 1998, p. 63). Catholic political and paramilitary publications from this period regularly sought to undermine any characterization of the conflict as religious. In September 1976, an article in Sinn Féin’s Republican News noted that the conflict “is not a religious war . . . it is a war of liberation against British Imperialistic dominance” (“A Review of the Situation,” 1976, p. 8).

This process provides compelling evidence that the lack of competitive pressure on the Catholic Church allowed it to overlook the preferences of adherents in favor of maintaining a favorable relationship with the British state. As a result, the Church declined to become involved in the politics of the Troubles and the political organizations representing Catholics consistently characterized the conflict as a secular nationalist struggle rather than a religious war.

Conclusion

The goal of this analysis has been to explore the conditions favoring the salience of religion in ethnic conflict. I present a model treating religious organizations as rational firms and arguing that competition among these firms contributes to the religious character of conflict. I test this approach using quantitative analysis of original time-series data on the salience of religion in conflict and qualitative evidence drawn from the Troubles in Northern Ireland.

I find that religion is more likely to play a salient role in conflict when a group’s religious organizations are fragmented, and religious leaders therefore face incentives to compete over adherents. Alternatively, religion is more likely to remain peripheral to conflict when a single religious organization monopolizes religious membership within the group and religious leaders therefore lack incentives to become involved in the costly politics of conflict.

This argument provides a compelling explanation for the behavior of religious leaders during the Troubles in Northern Ireland. Facing competition over adherents, Protestant religious leaders were rewarded for increasingly radical political activism. By contrast, because Catholic clergy felt little incentive to appeal to the preferences of adherents, no religious political movement emerged among Catholics. Although Protestant paramilitary organizations framed the conflict in religious terms, their Catholic counterparts distanced themselves from Catholicism.

These findings provide compelling evidence that the salience of religion in ethnic conflict, and by extension the likelihood that conflict will take on the intense and intractable characteristics of religious violence, can be traced to the structural incentives of an ethnic group’s religious leaders. In short, religious competition within group boundaries has a tremendous effect on the religious character of conflict across group boundaries.

However, it is important to understand the limitations of this analysis. First, I focus on groups characterized by overlapping ethnic cleavages. Although this focus is critical for observing variation in the salience of religion, this analysis omits many conflicts in which religion is a highly salient factor. However, religious organizations are among the most influential and resource-rich civil society organizations worldwide (Smith, 1996). Although the particular demographic structure examined in this analysis is ideal for observing the effects of religious competition, we should expect competition among religious firms to have similar effects in a wide variety of demographic contexts.

Second, I rely on a binary measure of religious structure to capture the conditions favoring within-group religious competition. As noted above, within-group religious competition should occur whenever adherents have a choice in organizational membership. For this reason, this binary measure provides a strong proxy for the conditions permitting within-group religious competition. However, this measure remains an approximation and says little about the scope or intensity of this process. Future research should devise more precise measures of within-group competition, including market fractionalization based on firms, rather than identity.

Finally, while I focus on competition among religious organizations, this process may provide an explanation for the salience of other forms of collective identity. Abstracted away from religion, this argument relies on the dual notions that civil society leaders are rational actors who seek to maximize resources for their firms and that conflict encourages consumers to prefer those leaders who take a greater role in politics. When organizations are strongly associated with a specific category of collective identity, we may expect competition among these organizations to contribute to the salience of that identity in conflict.

However, while not unique to religion, two factors suggest that this type of competition may be especially prevalent among religious organizations. First, religion is more often associated with authoritative, resource-rich civil society organizations than other forms of collective identity (Laitin, 2000; Smith, 1996). Second, religious membership is more likely to be exclusive than membership in other types of organizations. Although these qualities are a common feature of religion, it is possible that linguistic, racial, national, or tribal organizations may hold these qualities as well. Future research should explore the applicability of this model to the salience of other forms of collective identity in conflict.

Footnotes

Acknowledgements

The author would like to thank Eva Bellin, Steven Burg, Richard English, Anthony Gill, Christopher Hale, Laurence Iannaccone, Jason Klocek, Yasser Kureshi, Victoria McGroary, David Patel, Jared Rubin, Steven Pfaff, and Carolyn Warner, as well as the editors and anonymous reviewers of Comparative Political Studies. All remaining errors are the responsibility of the author.

Author’s Note

Earlier versions of this article were presented at the 2015 American Political Science Association conference, the 2015 International Studies Association conference, and the 2015 Graduate Student Workshop of the Institute for the Study of Religion, Economics and Society at Chapman University.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work has been supported by grants from the Tobin Project and the Research Circle on Democracy and Cultural Pluralism at Brandeis University.

Notes

Author Biography

Matthew Isaacs is a doctoral candidate in political science at Brandeis University. His research focuses on topics related to religious and ethnic conflict, including the politics of religious organizations, peacebuilding, and quantitative approaches to religion.

References

Achen

(2001). Why lagged dependent variables can suppress the explanatory power of other independent variables (Working paper). Retrieved from https://www.princeton.edu/csdp/events/Achen121201/achen.pdf

Appleby

R. S.

(2000). The ambivalence of the sacred: Religion, violence, and reconciliation. Lanham, MD: Rowman & Littlefield.

Asal

Pate

Wilkenfeld

(2008). Minorities at risk organizational behavior data and codebook, Version 9/2008. Retrieved from http://www.cidcm.umd.edu/mar/data.asp

Bakke

K. M.

(2013). Copying and learning from outsiders? Assessing diffusion from transnational insurgents in the Chechen Wars. In Checkel

J. T.

(Ed.), Transnational dynamics of civil war (pp. 31-62). New York, NY: Cambridge University Press.

Banks

A. S.

Muller

T. C.

Overstreet

W. R.

Isacoff

J. F.

(Eds.). (2014). Political handbook of the world. Thousand Oaks, CA: CQ Press.

Basedau

Koos

(2015). When do religious leaders support faith-based violence? Evidence from a survey poll in South Sudan. Political Research Quarterly, 68, 760-772.

Basedau

Pfeiffer

Vüllers

(2016). Bad religion? Religion, collective action, and the onset of armed conflict in developing countries. Journal of Conflict Resolution, 60, 226-255.

Beck

Katz

J. N.

(2011). Modeling dynamics in time-series cross-section political economy data. Annual Review of Political Science, 14, 331-352.

Bercovitch

DeRouen

(2005). Managing ethnic civil wars: Assessing the determinants of successful mediation. Civil Wars, 7, 98-116.

10.

Berman

(2011). Radical, religious, and violent: The new economics of terrorism. Cambridge, MA: MIT Press.

11.

Berman

Laitin

D. D.

(2008). Religion, terrorism and public goods: Testing the club model. Journal of Public Economics, 92, 1942-1967.

12.

Boulton

(1973). The UVF, 1966-1973: An anatomy of loyalist rebellion. Norfolk, UK: Torc Books.

13.

Bradford

(1984). A sword bathed in heaven: The life, faith and cruel death of the Rev. Robert Bradford B.Th. M. P. Hanta, UK: Pickering Publishing.

14.

Brewer

(2004). Continuity and change in contemporary Ulster Protestantism. The Sociological Review, 52, 265-283.

15.

Brewer

Higgins

G. I.

Teeney

(2011). Religion, civil society, & peace in Northern Ireland. New York, NY: Oxford University Press.

16.

Brown

James

(2015). Religious characteristics of states dataset. Retrieved from http://www.thearda.com/Archive/Files/Descriptions/BROWN.asp

17.

Bruce

(1998). Conservative Protestant politics. New York, NY: Oxford University Press.

18.

Bruce

(2009). Paisley: Religion and politics in Northern Ireland. New York, NY: Oxford University Press.

19.

Bryan

(2000). Orange parades: The politics of ritual, tradition and control. London, England: Pluto Press.

20.

Buckley

(2016). Demanding the divine? Explaining cross-national support for clerical control of politics. Comparative Political Studies, 49, 357-390.

21.

Carey

S. C.

Mitchell

N. J.

Lowe

(2013). States, the security sector, and the monopoly of violence: A new database on pro-government militias. Journal of Peace Research, 50, 249-258.

22.

Central Intelligence Agency. (2012). World Factbook. Washington, DC: Author. Retrieved from https://www.cia.gov/library/publications/resources/the-world-factbook/

23.

Chandra

(2004). Why ethnic parties succeed: Patronage and ethnic head counts in India. New York, NY: Cambridge University Press.

24.

Chandra

(2006). What is ethnic identity and does it matter? Annual Review of Political Science, 9, 397-424.

25.

Chandra

(2011). What is an ethnic party? Party Politics, 17, 151-169.

26.

Chandra

Boulet

(2012). A combinatorial language for thinking about ethnic identity. In Chandra

(Ed.), Constructivist theories of ethnic politics (pp. 179-227). New York, NY: Oxford University Press.

27.

Chaves

Cann

D. E.

(1992). Regulation, pluralism, and religious market structure: Explaining religion’s vitality. Rationality and Society, 4, 272-290.

28.

Chaves

Schraeder

P. J.

Sprindys

(1994). State regulation of religion and Muslim religious vitality in the industrialized West. The Journal of Politics, 56, 1087-1097.

29.

Chesnut

R. A.

(2003). Competitive spirits: Latin America’s new religious economy. New York, NY: Oxford University Press.

30.

Chiozza

(2002). Is there a clash of civilizations? Evidence from patterns of international conflict involvement, 1946-97. Journal of Peace Research, 39, 711-734.

31.

The churches in N. Ireland. (1980, Spring). Belfast Bulletin, pp. 1-51.

32.

Coogan

T. P.

(2002). The Troubles: Ireland’s ordeal and the search for peace. New York, NY: Palgrave Macmillan.

33.

De Juan

(2008). A pact with the devil? Elite alliances as bases of violent religious conflict. Studies in Conflict & Terrorism, 31, 1120-1135.

34.

DeVotta

(2005). From ethnic outbidding to ethnic conflict: The institutional bases for Sri Lanka’s separatist war. Nations and Nationalism, 11, 141-159.

35.

Divine intervention. (1982, March 4). An Phoblacht, p. 3.

36.

Ellingsen

(2000). Colorful community or ethnic witches’ brew? Multiethnicity and domestic conflict during and after the Cold War. Journal of Conflict Resolution, 44, 228-249.

37.

Esposito

(Ed.). (2003). The Oxford dictionary of Islam. New York, NY: Oxford University Press.

38.

Evangelistic drive in 1953. (1952, October 22). The Belfast Telegraph, p. 14.

39.

Europa Productions. (1970). Europa world year book. New York, NY: Routledge.

40.

Fearon

J. D.

Laitin

D. D.

(1996). Explaining interethnic cooperation. American Political Science Review, 90, 715-735.

41.

Feliu

Grasa

(2013). Armed conflicts and religious factors: The need for synthesized conceptual frameworks and new empirical analyses—The case of the MENA region. Civil Wars, 15, 431-453.

42.

Finke

(1990). Religious deregulation: Origins and consequences. Journal of Church and State, 32, 609-626.

43.

Finke

Guest

Stark

(1996). Mobilizing local religious markets: Religious pluralism in the Empire State, 1855 to 1865. American Sociological Review, 61, 203-218.

44.

Finke

Stark

(1988). Religious economies and sacred canopies: Religious mobilization in American cities, 1906. American Sociological Review, 53, 41-49.

45.

Finke

Stark

(1992). The churching of America, 1776-1990: Winners and losers in our religious economy. New Brunswick, NJ: Rutgers University Press.

46.

Fish

M. S.

Jensenius

F. R.

Michel

K. E.

(2010). Islam and large-scale political violence: Is there a connection? Comparative Political Studies, 43, 1327-1362.

47.

Fox

(2000). Is Islam more conflict prone than other religions? A cross-sectional study of ethnoreligious conflict. Nationalism and Ethnic Politics, 6(2), 1-24.

48.

Fox

(2002). Ethnoreligious conflict in the late twentieth century: A general theory. Lanham, MD: Lexington Books.

49.

Fox

(2012). The religious wave: Religion and domestic conflict from 1960 to 2009. Civil Wars, 14, 141-158.

50.

Gill

(1998). Rendering unto Caesar: The Catholic Church and the state in Latin America. Chicago, IL: University of Chicago Press.

51.

Gill

(1999). Government regulation, social anomie and protestant growth in Latin America: A cross-national analysis. Rationality and Society, 11, 287-316.

52.

Gill

(2008). The political origins of religious liberty. New York, NY: Cambridge University Press.

53.

Gombrich

R. F.

(1988). Theravada Buddhism: A social history from ancient Benares to modern Colombo. New York, NY: Routledge.

54.

Grim

B. J.

Finke

(2007). Religious persecution in cross-national context: Clashing civilizations or regulated religious economies? American Sociological Review, 72, 633-658.

55.

Grim

B. J.

Finke

(2011). The price of freedom denied: Religious persecution and conflict in the 21st century. New York, NY: Cambridge University Press.

56.

Grzymala-Busse

(2012). Why comparative politics should take religion (more) seriously. Annual Review of Political Science, 15, 421-442.

57.

Hagopian

(2008). Latin American Catholicism in an age of religious and political pluralism: A framework for analysis. Comparative Politics, 40, 149-168.

58.

Hale

C. W.

(2015). Religious institutions and civic engagement: A test of religion’s impact on political activism in Mexico. Comparative Politics, 47, 211-230.

59.

Hansen

T. B.

(1999). The saffron wave: Democracy and Hindu nationalism in modern India. Princeton, NJ: Princeton University Press.

60.

Hassner

(2009). War on sacred grounds. Ithaca, NY: Cornell University Press.

61.

Henderson

E. A.

(1997). Culture or contiguity? Ethnic conflict, the similarity of states, and the onset of war, 1820-1989. Journal of Conflict Resolution, 41, 649-668.

62.

Henderson

E. A.

Tucker

(2001). Clear and present strangers: The clash of civilizations and international conflict. International Studies Quarterly, 45, 317-338.

63.

Henne

P. S.

(2012a). The ancient fire: Religion and suicide terrorism. Terrorism and Political Violence, 24, 38-60.

64.

Henne

P. S.

(2012b). The two swords: Religion–state connections and interstate disputes. Journal of Peace Research, 49, 753-768.

65.

Horowitz

M. C.

(2009). Long time going: Religion and the duration of crusading. International Security, 34, 169-193.

66.

Huntington

S. P.

(1996). The clash of civilizations and the remaking of world order. New York, NY: Simon & Schuster.

67.

Iannaccone

L. R.

(1990). Religious participation: A human capital approach. Journal for the Scientific Study of Religion, 29, 297-314.

68.

Iannaccone

L. R.

(1991). The consequences of religious market structure: Adam Smith and the economics of religion. Rationality and Society, 3, 156-177.

69.

Iannaccone

L. R.

(1992). Religious markets and the economics of religion. Social Compass, 39, 121-131.

70.

Inglehart

Welzel

(2005). Modernization, cultural change, and democracy: The human development sequence. New York, NY: Cambridge University Press.

71.

Isaacs

(2016). Sacred violence or strategic faith? Disentangling the relationship between religion and violence in armed conflict. Journal of Peace Research, 53, 211-225.

72.

Jones

(Ed.). (2004). Encyclopedia of religion. Detroit, MI: Macmillan Reference.

73.

Juergensmeyer

(1993). The new Cold War? Religious nationalism confronts the secular state. Berkeley: University of California Press.

74.

Juergensmeyer

(2000). Terror in the mind of God: The global rise of religious violence. Berkeley: University of California Press.

75.

Juergensmeyer

(Ed.). (2006). The Oxford handbook of global religions. New York, NY: Oxford University Press.

76.

Kalyvas

S. N.

(1996). The rise of Christian democracy in Europe. Ithaca, NY: Cornell University Press.

77.

Kurzman

(1998). Organizational opportunity and social movement mobilization: A comparative analysis of four religious movements. Mobilization: An International Journal, 3(1), 23-49.

78.

Laitin

D. D.

(2000). Language, conflict and violence: The straw that strengthens the camel’s back. European Journal of Sociology, 41, 97-137.

79.

Maoz

Henderson

E. A.

(2013). The world religion dataset, 1945-2010: Logic, estimates, and trends. International Interactions, 39, 265-291.

80.

McElroy

(1991). The Catholic Church and the Northern Ireland crisis, 1968-1986. Dublin, Ireland: Gill & Macmillan.

81.

McGarry

O’Leary

(1995). Explaining Northern Ireland. Malden, MA: Wiley-Blackwell.

82.

McGrath

(2000). The Catholic Church and Catholic schools in Northern Ireland. Dublin, Ireland: Irish Academic Press.

83.

Mitchell

(2006). Religion, identity and politics in Northern Ireland: Boundaries of belonging and belief. Burlington, VT: Ashgate.

84.

Moghadam

(2008). Motives for martyrdom: Al-Qaida, Salafi jihad, and the spread of suicide attacks. International Security, 33(3), 46-78.

85.

Moloney

(2008). Paisley: From demagogue to democrat? Dublin, Ireland: Poolbeg Press.

86.

Montalvo

J. G.

Reynal-Querol

(2005). Ethnic polarization, potential conflict, and civil wars. American Economic Review, 95, 796-816.

87.

Murray

(1998). John Hume and the SDLP: Impact and survival in Northern Ireland. Dublin: Irish Academic Press.

88.

Murray

Tonge

(2005). Sinn Fein and the SDLP: From alienation to participation. Dublin, Ireland: The O’Brien Press.

89.

National Consortium for the Study of Terrorism and Responses to Terrorism. (2012). Global terrorism database. Retrieved from http://www.start.umd.edu/gtd

90.

O’Callaghan

O’Donnell

(2006). The Northern Ireland government, the “Paisleyite Movement” and Ulster Unionism in 1966. Irish Political Studies, 21, 203-222.

91.

Paisley says he’ll be MP at Stormont. (1966, July 20). The Belfast Telegraph, p. 1.

92.

Parry

Brady

Griffith

S. H.

Melling

D. J.

Healey

(2000). The Blackwell dictionary of Eastern Christianity. Malden, MA: Wiley-Blackwell.

93.

Pearce

(2005). Religious rage: A quantitative analysis of the intensity of religious conflicts. Terrorism and Political Violence, 17, 333-352.

94.

Pew Research Center. (2014a). Global religious diversity (Religion and Public Life Project). Retrieved from http://www.pewforum.org/2014/04/04/global-religious-diversity/

95.

Pew Research Center. (2014b). Religious hostilities reach six-year high (Religion and Public Life Project). Retrieved from http://www.pewforum.org/2014/01/14/religious-hostilities-reach-six-year-high/

96.

Pfaff

Gill

A. J.

(2006). Will a million Muslims march? Muslim interest organizations and political integration in Europe. Comparative Political Studies, 39, 803-828.

97.

Philpott

(2007). Explaining the political ambivalence of religion. American Political Science Review, 101, 505-525.

98.

Piazza

J. A.

(2009). Is Islamist terrorism more dangerous? An empirical study of group ideology, organization, and goal structure. Terrorism and Political Violence, 21, 62-88.

99.

Plümper

Troeger

V. E.

(2007). Efficient estimation of time-invariant and rarely changing variables in finite sample panel analyses with unit fixed effects. Political Analysis, 15, 124-139.

100.

Posner

(2005). Institutions and ethnic politics in Africa. New York, NY: Cambridge University Press.

101.

Presbyterian Church in Ireland. (1964). Annual report. Belfast, UK: Author.

102.

Rabushka

Shepsle

K. A.

(1972). Politics in plural societies: A theory of democratic instability. New York, NY: Macmillan.

103.

Rapoport

D. C.

(1984). Fear and trembling: Terrorism in three religious traditions. American Political Science Review, 78, 658-677.

104.

A review of the situation: Past, present and future. (1976, September 14). Republican News, pp. 2-8.

105.

Roeder

P. G.

(2003). Clash of civilizations and escalation of domestic ethnopolitical conflicts. Comparative Political Studies, 36, 509-540.

106.

Ruane

Todd

(1996). The dynamics of conflict in Northern Ireland. New York, NY: Cambridge University Press.

107.

Sells

M. A.

(1998). The bridge betrayed: Religion and genocide in Bosnia. Berkeley: University of California Press.

108.

Seul

J. R.

(1999). “Ours is the way of God”: Religion, identity, and intergroup conflict. Journal of Peace Research, 36, 553-569.

109.

Should publicity be used to sell religion? (1959, June 3). The Belfast Telegraph, p. 8.

110.

Smith

(1996). Correcting a curious neglect, or bringing religion back in. In Smith

(Ed.), Disruptive religion: The force of faith in social movement activism (pp. 1-25). New York, NY: Routledge.

111.

Smyth

W. M.

(1960). The message of the banners. Belfast, UK: The Evangelical Protestant Society.

112.

Smyth

W. M.

(1965). Why Presbyterian? A booklet of biblical and historical reasons. Rathfriland, UK: Outlook Press.

113.

Stark

Bainbridge

W. S.

(1987). A theory of religion. New York, NY: Lang Publishing.

114.

Stark

Finke

(2000). Acts of faith: Explaining the human side of religion. Berkeley: University of California Press.

115.

Stepan

A. C.

(2000). Religion, democracy, and the “twin tolerations.” Journal of Democracy, 11(4), 37-57.

116.

Svensson

(2007). Fighting with faith: Religion and conflict resolution in civil wars. Journal of Conflict Resolution, 51, 930-949.

117.

Svensson

(2012). Ending holy wars: Religion and conflict resolution in civil wars. St Lucia, Australia: University of Queensland Press.

118.

Svensson

(2013). One God, many wars: Religious dimensions of armed conflict in the Middle East and North Africa. Civil Wars, 15, 411-430.

119.

Toft

M. D.

(2006). Issue indivisibility and time horizons as rationalist explanations for war. Security Studies, 15, 34-69.

120.

Toft

M. D.

(2007). Getting religion? The puzzling case of Islam and civil war. International Security, 31(4), 97-131.

121.

Toft

M. D.

Zhukov

Y. M.

(2015). Islamists and nationalists: Rebel motivation and counterinsurgency in Russia’s North Caucasus. American Political Science Review, 109, 222-238.

122.

Trejo

(2009). Religious competition and ethnic mobilization in Latin America: Why the Catholic Church promotes indigenous movements in Mexico. American Political Science Review, 103, 323-342.

123.

Tusicisny

(2004). Civilizational conflicts: More frequent, longer, and bloodier? Journal of Peace Research, 41, 485-498.

124.

Ulster Defence Association. (1973, January 20). The priest as an all-rounder. Ulster Militant, pp. 6-7.

125.

Ulster Protestant Action. (1965). UPA constitution. Belfast, UK: n.p.

126.

Ulster Volunteer Force. (1976, May). A message to Ulster Protestants. Combat, p. 9.

127.

Warner

C. M.

(2000). Confessions of an interest group. Princeton, NJ: Princeton University Press.

128.

Wilson

(1983, July 5). Confidentiality and the clergy. Iris, pp. 25-27.

129.

World Bank. (2013). World development indicators. Retrieved from http://data.worldbank.org/data-catalog/world-development-indicators