Restrictions on the Religious Practices of Religious Minorities: A Global Survey

bs_bs_banner POLITICAL STUDIES: 2015 VOL 63, 1070–1086 doi: 10.1111/1467-9248.12141 Restrictions on the Religious Practices of Religious Minorities: A Global Survey Jonathan Fox Yasemin Akbaba Bar Ilan University Gettysburg College This study examines sources of religious discrimination with a particular emphasis on comparison of the identityversus structure-related causes. Using a dyad-based analysis, it asks whether and when majorities behave differently in general and whether levels of religious discrimination are different against different minority groups. The Religion and State-Minorities (RASM) dataset, which includes data on 598 minorities in 177 countries covering the years 1990–2008, is used. The results show that specific minority and majority religions tend to have unique patterns of discrimination. Thus religious identity plays an important role in causing religious discrimination. Also, when controlling for other factors, Christian minorities experience the highest levels of discrimination around the world, including in the Christian world other than in Western democracies where Muslims experience the highest levels of discrimination. Keywords: religion; minority; discrimination This study assesses the impact of religious identity of both religious majorities and minorities on religious discrimination against religious minorities. Most previous studies examine only variation in this type of policy based on the majority religion in a state – that is, they use a single score for an entire country and do not assess whether there is any variation in how a country treats its multiple minorities. We posit that the identity of religious minorities is also important. For example, we assess whether Christian majority states treat Muslim and Christian minorities (in this case those belonging to Christian denominations other than that of the majority) differently. Our findings demonstrate that the religious identity of both the majority groups and the minority group matters, with each dyad of majority and minority groups having distinct patterns of discrimination that are consistent over time. The Religion and State-Minorities (RASM) dataset is unique in that it uses religious minorities as the unit of analysis and includes all religious minorities that meet a minimum population threshold.1 As discussed in more detail below, most previous studies of the causes of discrimination, including religious discrimination, focus on structural rather than identity-based causes of discrimination. Most data collections on discrimination use the entire state as their unit of analysis and therefore cannot assess questions of differential treatment within a state. The one notable exception – the Minorities at Risk (MAR) dataset – focuses on ethnic minorities rather than religious discrimination and does not include all relevant religious minorities. Consequently, we focus on the impact of the interaction between majority and minority religious identities on religious discrimination. We use structural factors as controls and rely on previous studies for our list of control variables and discussions of theory regarding structural causes of discrimination. The results show that identity matters. There is © 2014 The Authors. Political Studies © 2014 Political Studies Association RESTRICTIONS ON RELIGIOUS MINORITIES 1071 variation in the overall levels of discrimination between majority groups. Within each majority group there is variation in the level of discrimination against different minority groups. Also, minority groups fare differently under different types of majority groups. The study proceeds as follows. First, we define the term ‘religious discrimination’. Then we review the literature on the role of religious identity as a cause of religious discrimination and with a focus on why the religious identities of both majority and minority groups matter. Next, we discuss research design, variable construction and present the data analysis. The final section outlines our conclusions. The Definition and Significance of Religious Discrimination For the purposes of this study we define ‘religious discrimination’ as state restrictions on the religious practices or institutions of religious minorities that are not placed on the majority religion. Thus, this definition focuses on differential treatment of minorities rather than general repression. Restrictions on all religions in a state are generally violations of civil rights, but because they do not involve unequal treatment they do not meet this definition of ‘discrimination’. This definition is consistent with other definitions (e.g. Gurr, 2000, p. 206). While, from one perspective, religious discrimination can be considered a subset of discrimination against minorities, many consider it particularly important and theoretically distinct. Jonathan Fox (2002; 2004) and Mark Juergensmeyer (1993) argue that only nationalism rivals religion as a cause of ethnic conflict. Perhaps one reason for this is that issues of politics, economics and to a lesser extent culture and even nationalism are often negotiable, while religious issues are often non-negotiable or ‘indivisible’ because many consider it impossible to compromise on a divine truth (Hassner, 2003; 2009; Laustsen and Waever, 2000, p. 719; Svensson, 2007).2 Consequently, the causes of religious discrimination may run deeper than other forms of discrimination.3 While this study looks at the causes of religious discrimination, it is important to note that studies using religion as an independent variable find it to have a significant influence. This is important because it demonstrates that religion is a significant factor in other aspects of domestic politics. This literature links religion to regime type (Fisch, 2002; Midlarsky, 1998; Toft et al., 2011), violence (Basedau et al., 2011; Fox, 2002; 2004; Henne, 2012a; Reynal-Querol, 2002; Roeder, 2003; Rummel, 1997; Satana et al., 2013; Svensson, 2007; Toft et al., 2011; Vanhanen, 1999), terrorism (Ben-Dor and Pedahzur, 2003; Weinberg and Eubanks, 1998) and interstate conflict (Henderson, 1997; Henne, 2012b). Several studies link religion to discrimination but either examine discrimination in general without distinguishing between the targets of that discrimination (Fox, 2008b; Grim and Finke, 2007; 2011; Sarkissian, 2010) or focus on a subset of religious minorities (Fox, 2000; Fox and Sandler, 2003). These studies use the MAR dataset, which includes only some ethnic minorities and does not systematically include all religious minorities. Studies of general discrimination, civil rights violations and conflict similarly tend to use the state rather than a minority group as the unit of analysis (see, e.g. Abouharb and Cingranelli, 2006; Hafner-Burton, 2005; De Soysa and Nordas, 2007) or use the MAR dataset (see, e.g. Gurr, 1993; 2000). The data collection used in this study is the first that includes all relevant minority groups that meet a minimum © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1072 JONATHAN FOX AND YASEMIN AKBABA population threshold. Its 29-component measure of religious discrimination is also more detailed than any previous study.4 Religious Identity and Religious Discrimination This study focuses specifically on the role of religious identity as a cause of religious discrimination and theorizes that the religious identities of both majority and minority groups matter. Of course, identity is not the only cause of religious discrimination. Clearly there exist many religious and secular factors that significantly influence levels of religious discrimination. However, previous studies (Fox, 2007; 2008a; 2008b; Grim and Finke, 2007; 2011) adequately address these issues so we focus on issues of religious identity, which are particularly neglected in the quantitative literature.5 ‘Identity’ is ‘the mechanism through which we locate ourselves in relation to the social world’ (Wilmer, 1997, p. 4). ‘Social identity’ refers to ‘the social categories, attributes or components of the self-concept that are shared with others and therefore define individuals as being similar to others’ (Monroe et al., 2000, p. 421). Sometimes social identity becomes so significant that an out-group is considered an acceptable target of discriminatory behavior (Monroe et al., 2000). We apply this general argument to religious identity and its impact on religious discrimination. Specifically, we ask whether different religious majorities behave differently in general and whether majority groups treat different religious minorities differently. This focus on both majority and minority identity differentiates this study from previous ones and gives it, we argue, a key perspective that previous studies lack. Some studies use a global religious diversity variable that assigns a single score to each state. Some of these studies find that religious diversity has an impact (Reynal-Querol, 2002; Rummel, 1997; Sambanis, 2001; Vanhanen, 1999), while others find it has no impact (Collier et al., 2004; Fearon and Laitin, 2003; Lacina, 2006). There are also studies that focus on other causes of religious discrimination and use the majority religious identity as a control (Fox, 2002; 2008a; 2008b). A strong theoretical base supports the argument that religious identity matters. Religion is generally accepted as a significant source of identity (Breakwell, 1986; Little, 1995; Shupe, 1990), especially due to its power of providing a social, sometimes geographical, cosmological, temporal and metaphysical ‘sense of locatedness’ (Seul, 1999, p. 558). Religious traditions and institutions resist change, thereby bringing stability, predictability and continuity to both individuals and groups (Berger, 1967; Juergensmeyer, 1993; Mol, 1976; Seul, 1999). Religious identity is also strengthened by its ability to convey feelings of security and protections against threats (Abdullah, 1978; Kinvall, 2004). In addition, religious identification is perceived to be a decisive factor in distinguishing political attitudes and preferences (Hayes, 1995; Jelen, 1993; Oldmixion and Hudson, 2008; Wuthnow and Lewis, 2008). Many highlight the role of religious identity in conflict. Perhaps most notoriously, Samuel Huntington (1993) argues that in the post-Cold War world most conflicts will be between different civilizations, which are defined mostly along religious lines. Furthermore, he singles out particular religious identities – primarily Muslims – as more conflict prone.6 However, most empirical studies testing this theory found that Huntington’s © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1073 formulation of religious identity was not the primary influence on conflict – as he predicted – but it often does have an impact (Chiozza, 2002; Fox, 2004; Henderson, 1997; 1998; 2004; 2005; Henderson and Singer, 2000; Henderson and Tucker, 2001; Roeder, 2003; Russett et al., 2000). Yet, most of these studies (as well as most of the studies of religious conflict discussed above), which controlled for the religious identities of the participants, used these variables as controls rather than the focus of the study. Also, other than the study by Fox (2004), which provided a bivariate analysis of the number of conflicts between specific dyads, none of them focused on the interaction between specific minority and majority groups. Thus far the discussion links religious identity to conflict in general. However, the focus of this study is on the interaction between majority and minority religious identity as an influence on religious discrimination – that is, dyad-based religious identity. By ‘dyad’ we mean specific pairings of majority and minority groups. We discuss below several reasons why a specific majority group may treat different minorities differently. While this list is likely not all-encompassing, it is sufficient for demonstrating that the link between dyadic religious identity and discrimination is potentially significant and merits empirical examination. First, many religious traditions have doctrines that mandate different treatment for different categories of minority religions. For example, Islamic doctrine divides nonMuslims into three basic categories. The first is ‘peoples of the book’ – adherents of monotheistic religions such as Jews and Christians. They are considered second in the hierarchy of religions after Islam because they believe in revealed scriptures. Traditionally they are tolerated but subjected to taxes and numerous restrictions. The second category is ‘non-peoples of the book’. While some of these minorities, such as Buddhists and Hindus, are often given some level of tolerance, unlike peoples of the book, there is no entitlement to tolerance. The final category is minorities considered heretical offshoots of Islam, such as the Baha’i. They are often subjected to more severe restrictions (Al-Ghunaimi, 1968; Arzt, 1990; Bassiouni, 1982; Khadduri, 1979; Stahanke, 1999). Judaism similarly differentiates between religions that are monotheist and follow the ‘seven laws of Noah’ and other religions.7 None of this is to say that Islam’s theology is particularly intolerant or that theology determines the policy of any given Muslim majority state. Monotheistic religions by their nature believe in a single truth and are intolerant of those who believe in other truths (Stark, 2001). This is made even more complex because different religious traditions, even related faiths such as the three Abrahamic faiths or even different branches of the same religion, have different interpretations of human rights and different understandings of the way religious minorities should be treated (Bloom et al., 1996; Freeman, 2004; Hoffman, 2003; Martin, 2005). Yet all of these religions have crosscutting elements which support tolerance (Appleby, 2000). Furthermore, while most of the world’s states have a majority religion, the extent to which they follow and support that religion varies considerably. Many states with clear religious majorities – be those majorities Christian, Muslim or other – are essentially secular and not heavily influenced by their religious tradition (Fox, 2008a). This also applies to religious discrimination. For example, some Muslim majority states such as Senegal and Sierra Leone engage in virtually no religious discrimination while © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1074 JONATHAN FOX AND YASEMIN AKBABA others such as Saudi Arabia and Iran engage in significant levels of religious discrimination. Similarly, religious discrimination in Christian majority states varies from virtually nonexistent in states such as Australia, Peru and Namibia to high in states such as Belarus and Bulgaria (Fox, 2008a). Be that as it may, the basic argument presented here is that religious doctrines often prescribe different treatment for different religious minorities. The example described briefly above shows that unlike many religions that mandate blanket intolerance of all other religions, Islam’s complex theology provides for a limited level of tolerance for certain religions. This means that Muslim majority states – to the extent to which they are influenced by this aspect of Islamic doctrine – are more likely to treat different religious minorities differently. A full discussion of all religious doctrines regarding religious minorities for all possible dyads of religious majorities and minorities is beyond the scope of this article. This example is intended to demonstrate the fact that religious doctrine can play a factor in differential discrimination against religious minorities.8 Second, the extent or lack of historical presence of religious minorities in a state can influence discrimination. Religious minorities new to a country may be perceived as ‘unorthodox’ or more foreign and thus more of a threat than more established minorities (Little, 1991). Additionally, religions new to the state have had less time to reach an accommodation with the majority. Sometimes states actively single out such new minorities as part of overt policies to protect indigenous culture from outside influences. Conversely, many religious minorities with a long historical presence in these states are considered part of the indigenous culture and are given state support (Fox, 2008a). Third, whether a minority poses a real threat or not, mere perception of threat can trigger religious discrimination. Threat perception is often high when there is a history of conflict between two religions, such as is the case for Catholics, Orthodox Christians and Muslims in the former Yugoslavia or the Hindu minority in Sri Lanka. Also, many states consider ‘sects’ or ‘cults’ such as the Church of Scientology or Sung Yung Moon’s Unification Church as threats to a country’s culture and perhaps to its citizens (Fox, 2008a). A fourth potential source of identity-based religious discrimination is major historical events. These are distinguished from local events in that they are of sufficient magnitude and influence that they can affect inter-religious relations on an international scale. For example, the 9/11 attacks and the war in Iraq have profoundly altered the perception of Islam of many in the West. After 9/11, for many in the West, Islam was a new topic of interest and was perceived by some as associated with terrorism and violence. These events and the global war on terrorism have also altered Muslim perceptions of Christianity and the West. This includes perceptions that Islam was under attack, which fed the ‘defensive’ attitude in many Muslim majority states and the response to protect Islam. The Iraq War is particularly important in this respect because it was the first time since the withdrawal of the colonial powers that a Western power assumed even temporary responsibility for governing a Muslim state (Rabasa et al., 2004; Vamik, 2002). Fifth, discrimination may vary along with the extent of cultural differences between groups (Finke, 2012). Religion is perceived as a significant factor and a source of ‘shared perception’ in the ethnic (communal) group definition supplied by Ted Gurr (1993, p. 3). © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1075 Gurr (2000) argues that a greater salience of ethnocultural identity facilitates discrimination and mobilization for collective action. While accepting that communal identities are multidimensional, he distinguishes some traits, such as race and religion, from others since they are intrinsically more important. Finally, social identity theory (a psychological theory on political and social identity) recognizes the importance of understanding the dynamics of discriminatory intergroup relations. This literature demonstrates that people need to be part of a group that they perceive positively (Halev and Morse, 2003; Tajfel and Turner, 1986, p. 16). Maintaining positive social identity is strongly tied to comparisons between groups (Seul, 1999). People feel better when their group does better than the others (Halev and Morse, 2003). When one’s social identity is unsatisfactory, there are two options for group members: leave the existing group and join another one, or make changes so that the existing group will be perceived positively. Switching religions has high costs (Stark and Finke, 2000) and discrimination becomes a potential tool for realizing the latter option. Studies show that group members feel better about themselves after engaging in discrimination (Lemyre and Smith, 1985; Tajfel and Turner, 1986). Thus, discrimination can be a method to improve a group’s social identity vis-à-vis another group. However, identity differences do not necessarily lead to discrimination and situational factors are important in determining when identity differences lead to discrimination (Monroe et al., 2000, p. 436). Among the situational factors that have been identified as important in leading to group identity and discrimination is religion (Tajfel, 1959). While ideally this study would have included variables for each of these factors, the time- and resource-consuming process of collecting these variables for each minority group is not feasible at this stage. Nevertheless, we use existing variables, when available, to measure these factors. Methodology This study uses the RASM dataset, which includes information on 598 religious minorities in 177 countries coded yearly for the period 1990–2008. This dataset was coded in conjunction with the Religion and State Round 2 (RAS2) dataset.9 Any religious minority which is at least 0.25 per cent of the population or has a population of at least 500,000 (in countries with populations of 200 million or more) are included. These thresholds were determined to be the minimum population thresholds for which sufficient information was consistently available to code the cases.10 Due to the prominence of Christian-Muslim relations, we also coded all Christian minorities in Muslim states and all Muslim minorities in Christian states, regardless of population size. We used the most general aggregation for minorities. For instance, in most cases, all Christians in non-Christian countries are coded together unless different Christian denominations are treated differently. In Christian countries, groups are separated into Catholic, Orthodox and Protestant. In Muslim countries, groups are separated into Sunni, Shi’i and, if applicable, Ahmadis. As explained above, this study’s combination of using a religious minority as the unit of analysis and examining all relevant minorities considering structural and identity related factors is unique. Accordingly, the focus of this study is to examine the issue of how different minorities are treated. While previous studies were able to account for only the © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1076 JONATHAN FOX AND YASEMIN AKBABA religious identity of the majority group in a state and perhaps the extent of religious diversity in that state, this study also examines whether states treat all minorities the same or whether some are singled out for different treatment. The RASM project measures 2911 types of religious discrimination, each coded individually on the following scale: 2 = Significantly restricted; 1 = Slightly restricted; and 0 = Not restricted. When combined, these 29 components create a scale of 0 to 58.12 For the purposes of this measure, ‘religious discrimination’ is defined as restrictions on the religious practices or institutions of minority religions that are not placed on the majority religion. Thus, this definition is based on differential treatment. Limitations placed on all religion in a state, while important, do not imply discrimination under this definition because the minority in question is being treated the same as all other groups. This variable is distinctly religious as compared to similar variables on ethnic conflict such as those developed by Gurr (1993; 2000) for two reasons. First, the unit of analysis is a religious minority. While many religious minorities are also ethnic minorities, this is not the case for all of them. While we have made no systematic analysis of the overlap between the 598 minorities covered in this study and ethnic minorities, there are numerous examples of religious minorities that are not ethnically distinct, such as Christians in South Korea and the growing number of Protestants in many of the Catholic majority countries of Latin America. Second, all of the 29 types of discrimination are limitations on religious practices or institutions that specifically target religious minorities. While in theory they could be placed on the institutions of an ethnic minority whose religion is the same as that of the majority, this would only be possible if that minority maintained its own religious institutions. As for practices, very few instances were found of restrictions on religious practices placed exclusively on members of the majority religion who belong to a different ethnic group.13 While it is not possible to go into detail on all 29 types of discrimination, one of the more common types – restrictions on the building, maintaining or repairing of places of worship – provides a good illustration of the religious focus of these variables, the impact of religious discrimination on religious minorities and the specifics of this coding scheme. A ‘significant’ restriction, coded as 2, includes cases where a country does not allow a minority group to build places of worship at all as is the case for countries like Brunei and Saudi Arabia. Other countries where it is extremely difficult but not impossible for minorities to obtain places of worship are also coded as 2. For example, in Kuwait many groups are denied licenses to rent places of worship; and ownership of property by minority religious groups is not allowed. Some groups are allowed to rent private homes for use as places of worship, but few landlords are willing to rent their property.14 ‘Slight’ restrictions, coded as 1, refer to situations where minorities can obtain places of worship, but this ability is limited. For example, Oman allows places of worship on land donated by the Sultan, which is occupied by two Protestant and two Catholic churches. Group worship is not allowed anywhere else and other religions are forced to borrow the existing churches in order to worship.15 This coding also applies to cases where local governments denied permits or approval to build places of worship such as was the case for the building of mosques in Western European countries such as Austria, Denmark, Germany and Italy. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1077 All religious identity groups in this study – both majority and minority groups – are divided into three categories: Christians, Muslims and other. This creates nine potential dyads because the minorities belonging to a different denomination of the same religion as the majority groups are coded as religious minorities. For example, Protestant minorities in Catholic states and Shi’i Muslim minorities in Sunni Muslim states are included in the analysis. The RASM codings have much more specific data than this, including the following categories: Catholic, Protestant, Orthodox Christians, Other Christians, Sunni Muslims, Shi’i Muslims, Other Muslims, Hindu, Buddhist, Jewish, Animist, Zoroastrian, Baha’i, Confucian, Chinese religions, Sikh, and Other. However, we use these three categories because many of the others have too few cases for statistical assumptions to hold and testing more than nine dyads makes the analysis too unwieldy. The first test examines the extent of religious discrimination on a yearly basis between 1990 and 2008. Because a large number of the states in the study became independent after 1990, the statistics for 1990 are for the first year of available data. We then test this relationship using OLS multiple regressions in two separate sets of tests. In the first, we create a dummy variable for each of the nine potential majority-minority dyads. Because one of these nine variables will be coded as 1 it is necessary to exclude at least one of the variables from the regression. Accordingly, we use three models. In each model, three dyads are excluded: model 1 excludes the three dyads with other majorities, model 2 excludes the three dyads with Muslim majorities, and model 3 excludes the three dyads with Christian majorities. The second set of tests looks at Christian, Muslim and other countries separately. Each set of tests has three models, each of which excludes the Christian, Muslim or other minorities. The reasoning behind these two sets of multivariate tests is as follows. The first set of tests examines how specific dyads compare to all other cases. The second set allows us to focus on whether countries with a specific religious identity treat Christian, Muslim and other religious minorities differently. If, as proves to be the case, both sets of tests show that minority religious identities matter, the results will be more robust. Both sets of multivariate tests use the same set of control variables, which are shown by Fox (2008a) to be significant determinants of religious discrimination: Religious Support is from the RAS2 dataset. It measures whether the government supports religion – that is, whether the government enforces religious precepts through law, finances religion or otherwise supports it. There are 51 components that, when combined, create a variable which ranges from 0 to 51.16 This variable is included because states that more strongly support religion are more likely to be intolerant of other religions (Fox, 2008a). Regime is measured using the Polity variable from the Polity Project dataset (Jaggers and Gurr, 1995). This variable ranges from –10 to 10 with −10 being the most autocratic and 10 being the most democratic. This variable is used because overall freedom in the country may influence religious freedom. While many studies also use the Freedom House measure for democracy, it includes civil rights, which makes it unacceptably covariant with the dependent variable used in this study. Regime durability, also taken from the Polity Project, measures how many years a regime has persisted without a change in the Polity measure. We use this variable because stability of the country may influence policies of religious discrimination. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1078 JONATHAN FOX AND YASEMIN AKBABA Minority and majority population size are based on the RAS2 coding of these variables. They are included because, as noted above, many studies found religious demography to be a significant factor in religious conflict and discrimination. The log of population variable controls for population size. States with small populations will have different methods of organizing their government than larger states. This variable is from the World Bank. Economic development is measured by log per capita GDP. This variable is taken from the UN Statistics Division website.17 Higher log per capita GDP indicates more economic development.18 Data Analysis and Discussion The bivariate analysis of levels of religious discrimination between 1990 and 2008, presented in Figure 1, demonstrates that each majority-minority dyad has a distinct pattern of discrimination. These patterns are influenced by both the majority and minority identities. Among majority identities, for the most part, Muslim states discriminate more. Christian and other minorities in Muslim majority states experience levels of discrimination higher than any of the other dyads. Muslim minorities in Muslim majority states experience levels of discrimination higher than all groups in non-Muslim majority states other than Christians in other majority states. Other majority states discriminate consistently more than Christian majority states. Within each majority group, the levels of discrimination against the three categories of minorities are different. However, the pattern differs from majority group to majority group. In Muslim majority states, the other minorities are subject to the highest levels of discrimination. In Christian and other majority states, it is Christian minorities that experience the highest levels of discrimination. This is particularly interesting given assumptions that Muslims are subject to distinct levels of discrimination in Christian states. Many of these results are statistically significant. Another important finding is that religious discrimination has been increasing over time. Other than for Muslim minorities in Muslim majority states, the levels of religious discrimination against each of the dyads increased during the study period. These results are statistically significant for six of the dyads as well as all cases combined. This confirms pervious results regarding increasing levels of religious discrimination around the world (Fox, 2008a; Grim and Finke, 2011). The dyad-based multivariate analysis, presented in Table 1, confirms the results of the bivariate analysis. In each of the models, two of the dyad variables are statistically significant. As was the case in the bivariate analysis, Christian minorities in Christian and other majority countries were subject to higher levels of religious discrimination and Muslim minorities in Muslim majority countries experienced lower levels of discrimination. Four of the control variables are significant. As expected, states which more strongly support religion discriminate more and democratic states discriminate less. Discrimination is more common in more populous states. Also, larger minorities experience less discrimination. This means that smaller and weaker minorities are more vulnerable to poor treatment by the state. The results for the second set of tests, which provide separate analysis for each majority religion, presented in Table 2, provide similar results. In eight of the nine tests, at least one © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1079 Figure 1: Mean Levels of Religious Discrimination 1990*–2008 12.00 10.00 ChrisƟan-ChrisƟan 8.00 ChrisƟan-Muslim ChrisƟan-Other Muslim-ChrisƟan Muslim-Muslim 6.00 Muslim-Other Other-ChrisƟan Other-Muslim 4.00 Other-Other All Cases 2.00 .00 1990 1993 1996 1999 2002 2005 2008 Notes: Scores for 1990 are for 1990 or the earliest year available. Significance between Christian-Christian and all other groups is <0.05 in 2002–2005, <0.01 in 1997–2001 and <0.001 in 1990–1996. Significance between Christian-Muslim and all other groups is <0.001 in 1990–2008. Significance between Christian-Other and all other groups is <0.001 in 1990–2008. Significance between MuslimChristian and all other groups is <0.001 in 1990–2008. Significance between Muslim-Other and all other groups is <0.001 in 1990–2008. Significance between scores for 1990 and 2008 is <0.05 for Christian-Muslim, Christian-Other, Muslim-Christian and Other-Christian. Significance between scores for 1990 and 2008 is <0.01 for Christian-Christian and Muslim-Other. Significance between scores for 1990 and 2008 is <0.001 for all cases. of the minority identity variables is significant. The polity variable performs as expected in all of the regressions. However, the other significant control variables are only significant in regressions for specific majority groups. Regime durability leads to more religious discrimination in Muslim majority states and larger minorities experience less discrimination in Christian majority states. Religious support, while significant in Christian and Muslim majority states, does not significantly impact on other majority states.19 © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) JONATHAN FOX AND YASEMIN AKBABA 1080 Table 1: Dyad-Based OLS Regressions Predicting Religious Discrimination 2008 Model 1 Dyads (majority-minority) Christian-Christian Christian-Muslim Christian-Other Muslim-Christian Muslim-Muslim Muslim-Other Other-Christian Other-Muslim Other-Other Other controls Religious support Regime (polity) Regime durability Minority population size (%) Majority population Size (%) Log population Log per capita GDP df Adjusted R2 Model 2 Beta Sig. Beta Sig. 0.079 −0.015 −0.064 0.015 −0.167 −0.076 0.050 0.691 0.111 0.686 0.000 0.056 0.165 0.063 0.027 0.001 0.164 0.586 0.441 −0.478 0.017 −0.109 0.007 0.135 0.047 597 0.507 0.000 0.000 0.633 0.001 0.828 0.000 0.219 Model 3 Beta Sig. 0.648 0.000 0.064 0.015 0.614 0.102 0.000 0.000 0.968 0.001 0.226 0.000 0.112 0.132 0.060 0.006 0.000 0.078 0.875 0.016 −0.163 −0.073 0.077 0.015 −0.052 0.420 −0.478 0.022 −0.147 0.019 0.128 0.032 0.000 0.000 0.522 0.000 0.568 0.000 0.412 0.439 −0.467 0.001 −0.106 0.038 0.136 0.060 597 0.496 597 0.504 Perhaps the most interesting result from this table is that controlling for other factors, Christian minorities experience the highest levels of discrimination in all three types of majority state. In Christian majority states, the Christian minority variable is significant in models 1 and 2. In model 3, both Muslim and other minorities experience lower levels of discrimination with Christian minorities as the reference group. In Muslim majority states, Christian minorities have significantly higher levels of discrimination in both models which include the variable. Among other majority groups, Christian minorities suffer from the highest levels of discrimination in model 1. In model 2, none of the identity variables are significant but the Christian minorities are associated with higher levels of discrimination. In model 3, both Muslim and other minorities experience lower levels of discrimination with Christian minorities as the reference group. Conclusions Overall this analysis produces three important results. First, religious discrimination is not monolithic. States treat different religious minorities differently. This is important because most previous studies of the topic (Abouharb and Cingranelli, 2006; Fox, 2008a; Grim and Finke, 2007; 2011) use a global variable to measure all discrimination in a state. Accordingly, these studies miss the fact that state behavior is more complex. This also implies that © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) Christian majority Model 1 Beta Minority Christian Minority Muslim Minority other Religious support Regime (polity) Regime durability Minority population size (%) Majority population size (%) Log population Log per capita GDP df Adjusted R2 Sig. 0.248 0.000 0.064 0.231 0.257 0.000 −0.384 0.000 −0.111 0.075 −0.125 0.015 −0.045 0.419 0.152 0.003 0.093 0.166 338 0.240 Model 2 Beta Sig. Muslim majority Model 3 Beta Sig. 0.181 0.002 −0.070 0.231 0.257 0.000 −0.384 0.000 −0.111 0.075 −0.125 0.015 −0.045 0.419 0.152 0.003 0.093 0.166 338 0.240 −0.174 0.002 −0.261 0.000 0.257 0.000 −0.384 0.000 −0.111 0.075 −0.125 0.015 −0.045 0.419 0.152 0.003 0.093 0.166 338 0.240 Model 1 Beta Sig. 0.139 0.038 −0.125 0.066 0.495 0.000 −0.210 0.008 0.224 0.003 −0.090 0.181 0.023 0.749 −0.057 0.473 −0.011 0.898 135 0.551 Model 2 Beta Sig. 0.286 0.001 0.146 0.066 0.495 0.000 −0.210 0.008 0.224 0.003 −0.090 0.181 0.023 0.749 −0.057 0.473 −0.011 0.898 135 0.551 Other majority Model 3 Beta Sig. Model 1 Beta Sig. Model 2 Beta Sig. 0.214 0.003 0.076 0.372 −0.242 0.001 0.123 0.083 −0.138 0.038 −0.218 0.083 0.495 0.000 0.059 0.436 0.059 0.436 −0.210 0.008 −0.499 0.000 −0.499 0.000 0.224 0.003 0.125 0.129 0.125 0.129 −0.090 0.181 −0.110 0.142 −0.110 0.142 0.023 0.749 0.100 0.249 0.100 0.249 −0.057 0.473 0.314 0.000 0.314 0.000 −0.011 0.898 0.030 0.697 0.030 0.697 135 122 122 0.551 0.492 0.492 Model 3 Beta Sig. −0.067 0.272 −0.230 0.003 0.059 0.436 −0.499 0.000 0.125 0.129 −0.110 0.142 0.100 0.249 0.314 0.000 0.030 0.697 122 0.492 RESTRICTIONS ON RELIGIOUS MINORITIES © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) Table 2: OLS Regressions for Religious Discrimination 2008, Separated by Majority Religion 1081 1082 JONATHAN FOX AND YASEMIN AKBABA the use of religious demography variables’ impact on conflict (Collier et al., 2004; Fearon and Laitin, 2003; Lacina, 2006; Reynal-Querol, 2002; Rummel, 1997; Sambanis, 2001; Vanhanen, 1999) likely misses a critical element. Different levels of religious discrimination imply a wider range of differential treatment between minorities, which likely has a significant impact on a conflict’s dynamics. Thus, studies that involve religious minorities that do not have data specific to each relevant minority are missing a critical element of the equation. This study confirms the findings of previous research that the religion of the majority group also matters (Fox, 2008a; Grim and Finke, 2007; 2011). Even when controlling for other factors including minority religious identity, each majority group engaged in significantly different levels of religious discrimination. However, this does not mean that this discrimination is uniform among majority groups. For example, among Muslim majority states, religious discrimination is particularly high in the Middle East and the Muslim majority states of Asia that were never part of the Soviet Union. However, a number of Muslim majority states in Africa engage in no religious discrimination at all. These include Burkina Faso, Gambia, Niger, Senegal and Sierra Leone. Similarly, religious discrimination in Christian states varies according to world region, with states in the former Soviet bloc discriminating the most. Interestingly, the lowest levels of religious discrimination in Christian majority states are not found in Western democracies. Rather, Latin American states, on average, have the lowest levels of religious discrimination among Christian majority states. It is also crucial to emphasize that 82.5 per cent of states engage in religious discrimination against at least one type of minority. The majority of all types of states discriminate. This is true whether they are divided by majority religion, region or democracy. Discrimination is the rule, not the exception. While this study focuses on what influences levels of discrimination, it is important not to forget its overall ubiquity. The second important result is that Christian minorities are the most at risk for discrimination across the globe. This is true regardless of religious majority. Even Christian majority states tend to treat Christians belonging to other denominations worse than they do other minority religions. A total of 13 of the 15 most serious cases of ChristianChristian discrimination are in the former Soviet bloc. These cases involve primarily non-indigenous Christian groups, especially American evangelical groups but also indigenous Christian groups such as schismatic Orthodox Christians. Among Christian majority states, this result is mostly consistent across world region. It is true of the former Soviet bloc, Asia, Africa and Latin America, but not Western democracies where Muslims experience higher levels of religious discrimination than Christian minorities. Finally, these results confirm that religious discrimination is rising. While previous studies have shown this rise, this study shows that it applies consistently across majority-minority dyads. This rise takes place as part of a more general rise in state involvement in religion, which also includes increases in state support for religion and the regulation of all religion.20 This demonstrates that the trend of rising religious discrimination is not limited to certain types of states or certain religious minorities. Rather, it is a general rise across the board. While religious discrimination is the dependent variable in this study, it has wider implications for any study using religious discrimination as either a dependent or © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1083 independent variable. Religious discrimination is not applied equally to all groups in a state and studies that do not account for this are missing a critical element of this dynamic. For example, in the context of ethnic conflict, Gurr (1993; 2000) demonstrates that discrimination is a key cause of conflict but that this discrimination is group-specific. The state-level variables that most others use to measure religious discrimination (e.g. Abouharb and Cingranelli, 2006; Fox, 2007; 2008a; Grim and Finke, 2011) cannot accomplish this, but the group-level variables presented in this study can. Accordingly, future research should include developing and integrating group-level variables into research designs. (Accepted: 4 February 2014) (Published online: 9 June 2014) About the Authors Jonathan Fox is Professor of Political Studies at Bar Ilan University in Ramat Gan, Israel; a Senior Research Associate at the Begin-Sadat Center for Strategic Studies; and Director of the Religion and State project (http:// www.religionandstate.org). His research focuses on various aspects of religion and politics. His recent books include An Introduction to Religion and Politics: Theory and Practice (Routledge, 2013) and Religion in International Relations Theory: Interactions and Possibilities (Routledge, 2013) (co-authored with Nukhet Sandal). Jonathan Fox, Department of Political Studies, Bar Ilan University, Ramat Gan, Israel; email: jonathan.fox@biu.ac.il Yasemin Akbaba is an Associate Professor at Gettysburg College. Her research focuses on mobilization of ethnic and religious groups and the effects of religious discrimination on ethnic and religious conflict. Her publications (single authored and co-authored) have appeared in Journal of Peace Research, Nationalism and Ethnic Politics, Ethnopolitics, Civil Wars, International Interactions, Politics and Religion and Politics, Religion and Ideology. She received Gettysburg College’s Thompson Award for distinguished teaching. Yasemin Akbaba, Gettysburg College, 300 North Washington Street, Glatfelter Hall, Campus Box 406, Gettysburg, PA 17325-1400, USA; email: yakbaba@gettysburg.edu Notes 1 The threshold is discussed in detail in the methodology section of this study. 2 This relates to a larger literature on divisible and indivisible conflicts, which argues ‘a good or issue is perceived as indivisible if it is perfectly cohesive, has unambiguous boundaries and cannot be substituted or exchanged for another good or issue. All sacred places fulfill these three conditions “Indivisible conflicts constitute” situations in which ... no compromise settlement is mutually preferable to conflict’ (Hassner, 2003, pp. 8–9). See also Fearon (1995). The case study literature also highlights the intractability of religious conflict. See e.g. De Silva (1986) and Little (1995). For a discussion of the survey-based literature on religion and tolerance, see Eisenstein (2008). 3 Several empirical studies also examine the correlation between religious freedom and democracy. Sarkissian (2012) argues that repressing religious institutions inhibits civil society and a potential source of democratic opposition. Driessen (2010) finds that many democratic governments regulate religion. Birnir and Satana (2013) demonstrate that religion influences democratic coalition politics. 4 A previous version of this dataset exists that includes a shorter time period (1990–2002) and only 24 components (Akbaba and Fox, 2011). 5 For a discussion of the general causes of discrimination, see Davenport (2007) and Park (1987). 6 Huntington’s clash of civilizations theory is criticized for a variety of reasons. From a specifically ‘identity’ perspective, it is blamed for having a ‘solitarist’ approach to human identity that focuses only on civilization (or religion) as a source of identity (Sen, 2008). While this study focuses on religious identity, we do not deny that there are other important sources for collective identities. 7 For details, see Broyde (1997). 8 The argument that theology or doctrine can influence policy is a common one. See e.g. Fox (2002), Juergensmeyer (1993) and Toft et al. (2011). 9 For more on the RAS2 dataset, see http://www.religionandstate.org 10 An examination of the treatment of smaller minorities would likely provide further enlightenment, but locating and coding all such minorities consistently is impractical. A selective examination would likely disproportionally include minorities that are politically active or suffer from high levels of discrimination, which would result in selection bias. 11 A full listing of these variables is available at: http://www.religionandstate.org 12 For a reliability analysis of these variables, cross-correlations and a discussion of alternative versions of creating this index, see Fox (2011). 13 The RAS project codes these types of restrictions using a separate variable that measures restrictions on the majority religion, so such practices are not included in the religious discrimination variable used in this study. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1084 JONATHAN FOX AND YASEMIN AKBABA 14 US State Department Report on Religious Freedom, 2006. Available online at: http://www.state.gov/j/drl/rls/irf/ 15 US State Department Report on Religious Freedom, 2002, 2003, 2004, 2007, 2008, 2009. Available online at: http:// www.state.gov/j/drl/rls/irf/ 16 For a more detailed description of official religious support and religious legislation, see Fox (2008a; 2013a). 17 http://www.unstats.org/unsd/default.htm 18 While many studies use other measures such as the UN’s Human Development Index to measure economic development, we chose to use log per capita GDP because the Human Development Index measure was missing for several of the states included in this study. 19 A robustness analysis controlling for world region instead of majority religion provides similar results. 20 For a more detailed discussion of this rise and its causes, see Fox (2007; 2008a; 2013b). References Abdullah, T. (1978) ‘Identity Maintenance and Identity Crisis in Minangkabau’, in H. Mol (ed.), Identity and Religion: International Cross Country Approaches. Beverly Hills, CA: Sage, pp. 151–67. Abouharb, M. R. and Cingranelli, D. L. (2006) ‘The Human Rights Effect of World Bank Structural Adjustment, 1980–2001’, International Studies Quarterly, 50, 233–62. Akbaba, Y. and Fox, J. (2011) ‘The Religion and State-Minorities Dataset’, Journal of Peace Research, 48 (6), 807–16. Al Ghunaimi, M. T. (1968) The Muslim Conception of International Law and the Western Approach. Leiden: Martinus Nijhoff. Appleby, R. S. (2000) The Ambivalence of the Sacred: Religion, Violence and Reconciliation. Lanham, MD: Rowman & Littlefield. Arzt, D. E. (1990) ‘The Application of International Human Rights Law in Islamic States’, Human Rights Quarterly, 12 (2), 202–30. Basedau, M. G. S., Vüllers, J. and Wegenast, T. (2011) ‘Do Religious Factors Impact Armed Conflict? Empirical Evidence from Sub-Saharan Africa’, Terrorism and Political Violence, 23 (5), 752–79. Bassiouni, M. C. (1982) ‘Sources of Islamic Law and the Protection of Human Rights in the Islamic Criminal Justice System’, in M. C. Bassiouni (ed.), The Islamic Criminal Justice System. London and New York: Oceana Publications, pp. 3–53. Ben-Dor, G. and Pedahzur, A. (2003) ‘The Uniqueness of Islamic Fundamentalism and the Fourth Wave of International Terrorism’, Totalitarian Movements and Political Religions, 4 (3), 71–90. Berger, P. L. (1967) The Sacred Canopy. Garden City, NY: Anchor. Birnir, J. K. and Satana, N. S. (2013) ‘Religion and Coalition Politics’, Comparative Political Studies, 46 (1), 3–30. Bloom, I., Martin, J. P. and Proudfoot, W. L. (eds) (1996) Religious Diversity and Human Rights. New York: Columbia University Press. Breakwell, G. (1986) Coping with Threatened Identities. London: Methuen. Broyde, M. J. (1997) ‘The Obligation of Jews to Seek Observance of Noahide Laws by Gentiles: A Theoretical Review’, in D. Shatz, C. I. Waxman and N. J. Diament (eds), Tikkun Olam: Social Responsibility in Jewish Thought and Law. Northvale, NJ: Jason Aronson, pp. 103–44. Chiozza, G. (2002) ‘Is There a Clash of Civilizations? Evidence from Patterns of International Conflict Involvement, 1946–97’, Journal of Peace Research, 39 (6), 711–34. Collier, P., Hoeffler, A. and Söderbom, M. (2004) ‘On the Duration of Civil War’, Journal of Peace Research, 41 (3), 253–73. Davenport, C. (2007) ‘State Repression and Political Order’, Annual Review of Political Science, 10, 1–23. De Silva, K. M. (1986) ‘Religion, Nationalism and the State’, USF Monographs in Religion and Public Policy. No. 1. Tampa, FL: University of South Florida. De Soysa, I. and Nordas, R. (2007) ‘Islam’s Bloody Innards: Religion and Political Terror, 1980–2000’, International Studies Quarterly, 41 (4), 927–43. Driessen, M. D. P. (2010) ‘Religion, State and Democracy: Analyzing Two Dimensions of Church-State Arrangements’, Politics and Religion, 3 (1), 55–80. Eisenstein, M. A. (2008) Religion and the Politics of Tolerance: How Christianity Builds Democracy. Waco, TX: Baylor University Press. Fearon, J. D. (1995) ‘Rationalist Explanations for War’, International Organization, 49 (3), 379–414. Fearon, J. D. and Laitin, D. D. (2003) ‘Ethnicity, Insurgency, and Civil War’, American Political Science Review, 97 (1), 75–90. Finke, R. (2012) ‘Origins and Consequences of Religious Restrictions: A Global Overview’. Presidential address, Annual Meeting of the Association for the Sociology of Religion. Fisch, M. S. (2002) ‘Islam and Authoritarianism’, World Politics, 55 (1), 4–37. Fox, J. (2000) ‘Religious Causes of Discrimination against Ethno-Religious Minorities’, International Studies Quarterly, 44, 423–50. Fox, J. (2002) Ethnoreligious Conflict in the Late Twentieth Century: A General Theory. Lanham, MD: Lexington Books. Fox, J. (2004) Religion, Civilization and Civil War: 1945 through the New Millennium. Lanham, MD: Lexington Books. Fox, J. (2007) ‘Religious Discrimination: A World Survey’, Journal of International Affairs, 61 (1), 47–67. Fox, J. (2008a) A World Survey of Religion and the State. New York: Cambridge University Press. Fox, J. (2008b) ‘State Religious Exclusivity and Human Rights’, Political Studies, 56 (4), 928–48. Fox, J. (2011) ‘Building Composite Measures of Religion and State’, Interdisciplinary Journal of Research on Religion, 7 (8), 1–39. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) RESTRICTIONS ON RELIGIOUS MINORITIES 1085 Fox, J. (2013a) An Introduction to Religion and Politics: Theory and Practice. London: Routledge. Fox, J. (2013b) ‘Is It Really God’s Century? An Evaluation of Religious Support and Discrimination from 1990 to 2008’, Politics and Religion, 7, 4–27. doi:10.1017. Fox, J. and Sandler, S. (2003) ‘Regime Types and Discrimination against Ethno-Religious Minorities: A Cross-Sectional Analysis of the Autocracy-Democracy Continuum’, Political Studies, 51 (3), 469–89. Freeman, M. (2004) ‘The Problem of Secularism in Human Rights Theory’, Human Rights Quarterly, 26, 375–400. Grim, B. J. and Finke, R. (2007) ‘Religious Persecution in Cross-National Context: Clashing Civilizations or Regulating Religious Economies’, American Sociological Review, 72, 633–58. Grim, B. J. and Finke, R. (2011) The Price of Freedom Denied. New York: Cambridge University Press. Gurr, T. R. (1993) Minorities at Risk. Washington, DC: United States Institute of Peace Press. Gurr, T. R. (2000) Peoples versus States: Minorities at Risk in the New Century. Washington, DC: United States Institute of Peace Press. Hafner-Burton, E. M. (2005) ‘Right or Robust? The Sensitive Nature of Repression to Globalization’, Journal of Peace Research, 42 (6), 679–98. Halev, J. S. and Morse, E. T. (2003) ‘National Identity and Self-Esteem’, Perspectives on Politics, 1, 515–32. Hassner, R. E. (2003) ‘ “To Halve and to Hold”: Conflict over Sacred Space and the Problem of Indivisibility’, Security Studies, 12 (4), 1–33. Hassner, R. E. (2009) War on Sacred Grounds. Ithaca, NY: Cornell University Press. Hayes, B. C. (1995) ‘The Impact of Religious Identification on Political Attitudes: An International Comparison’, Sociology of Religion, 56 (2), 177–94. Henderson, E. A. (1997) ‘Culture or Contiguity: Ethnic Conflict, the Similarity States and the Onset of War, 1920–1989’, Journal of Conflict Resolution, 41 (5), 649–68. Henderson, E. A. (1998) ‘The Democratic Peace through the Lens of Culture, 1820–1989’, International Studies Quarterly, 42 (3), 461–84. Henderson, E. A. (2004) ‘Mistaken Identity: Testing the Clash of Civilizations Thesis in the Light of Democratic Peace Claims’, British Journal of Political Science, 34, 539–63. Henderson, E. A. (2005) ‘Not Letting the Evidence Get in the Way of Assumptions: Testing the Clash of Civilizations with More Data’, International Politics, 42 (4), 458–69. Henderson, E. A. and Singer, J. D. (2000) ‘Civil War in the Post-Colonial World, 1946–92’, Journal of Peace Research, 37 (3), 275–99. Henderson, E. A. and Tucker, R. (2001) ‘Clear and Present Strangers: The Clash of Civilizations and International Conflict’, International Studies Quarterly, 45 (2), 317–38. Henne, P. S. (2012a) ‘The Ancient Fire: Religion and Suicide Terrorism’, Terrorism and Political Violence, 24 (1), 38–60. Henne, P. S. (2012b) ‘The Two Swords: Religion-State Connections and Interstate Disputes’, Journal of Peace Research, 49 (6), 753–68. Hoffman, V. J. (2003) ‘Contending Legitimacies: Muslim Perspectives’, in E. Kolodziej (ed.), A Force Profonde. Philadelphia, PA: University of Pennsylvania Press, pp. 45–68. Huntington, S. P. (1993) ‘The Clash of Civilizations?’, Foreign Affairs, 72 (3), 22–49. Jaggers, K. and Gurr, T. R. (1995) ‘Tracking Democracy’s Third Wave with the Polity III Data’, Journal of Peace Research, 32 (4), 469–82. Jelen, T. G. (1993) ‘The Political Consequences of Religious Group Attitudes’, Journal of Politics, 55 (1), 178–90. Juergensmeyer, M. (1993) The New Cold War? Religious Nationalism Confronts the Secular State. Berkeley, CA: University of California Press. Khadduri, M. (1979) War and Peace in the Law of Islam. Baltimore, MD: Johns Hopkins University Press. Kinvall, C. (2004) ‘Globalization and Religious Nationalism: Self, Identity and the Search for Ontological Security’, Political Psychology, 25, 741–67. Lacina, B. (2006) ‘Explaining the Severity of Civil Wars’, Journal of Conflict Resolution, 50 (2), 276–89. Laustsen, C. B. and Waever, O. (2000) ‘In Defense of Religion: Sacred Referent Objects for Secularization’, Millennium, 29 (3), 705–39. Lemyre, L. and Smith, P. M. (1985) ‘Intergroup Discrimination and Self Esteem in the Minimal Group Paradigm’, Journal of Personality and Social Psychology, 49, 660–70. Little, D. (1991) Ukraine: The Legacy of Intolerance. Washington, DC: United States Institute of Peace Press. Little, D. (1995) ‘Belief, Ethnicity and Nationalism’, Nationalism and Ethnic Politics, 1 (2), 284–301. Martin, J. P. (2005) ‘The Three Monotheistic World Religions and International Human Rights’, Journal of Social Issues, 61 (4), 827–45. Midlarsky, M. I. (1998) ‘Democracy and Islam: Implications for Civilizational Conflict and the Democratic Peace’, International Studies Quarterly, 42 (3), 458–511. Mol, H. (1976) Identity and the Sacred: A Sketch for a New Social-Scientific Theory of Religion. Oxford: Basil Blackwell. Monroe, K. R., Hankin, J. and Van Vechten, R. B. (2000) ‘The Psychological Foundations of Identity Politics’, Annual Review of Political Science, 3, 419–47. Oldmixion, E. and Hudson, W. (2008) ‘When Church Teachings and Policy Commitments Collide: Perspectives on Catholics in the US House of Representatives’, Politics and Religion, 1 (1), 113–35. Park, H. S. (1987) ‘Correlates of Human Rights: Global Tendencies’, Human Rights Quarterly, 9, 405–13. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5) 1086 JONATHAN FOX AND YASEMIN AKBABA Rabasa, A. M., Benard, C., Chalk, P., Fair, C. C., Karasik, T., Lal, R., Lesser, I. and Thaler, D. (2004) The Muslim World after 9/11. Santa Monica, CA: RAND Corporation. Reynal-Querol, M. (2002) ‘Ethnicity, Political Systems and Civil Wars’, Journal of Conflict Resolution, 46 (1), 29–54. Roeder, P. G. (2003) ‘Clash of Civilizations and Escalation of Domestic Ethnopolitical Conflicts’, Comparative Political Studies, 36 (5), 509–40. Rummel, R. J. (1997) ‘Is Collective Violence Correlated with Social Pluralism?’, Journal of Peace Research, 34 (2), 163–75. Russett, B., Oneal, J. R. and Cox, M. (2000) ‘Clash of Civilizations, or Realism and Liberalism Déjà Vu? Some Evidence’, Journal of Peace Research, 37 (5), 583–608. Sambanis, N. (2001) ‘Do Ethnic and Nonethnic Civil Wars Have the Same Causes?’, Journal of Conflict Resolution, 45 (3), 259–82. Sarkissian, A. (2010) ‘Religious Reestablishment in Post-Communist Polities’, Journal of Church and State, 51 (3), 472–501. Sarkissian, A. (2012) ‘Religious Regulation and the Muslim Democracy Gap’, Politics and Religion, 5 (3), 501–27. Satana, N. S., Inman, M. and Birnir, J. K. (2013) ‘Religion, Government Coalitions and Terrorism’, Terrorism and Political Violence, 25 (1), 29–52. Sen, A. (2008) ‘Violence, Identity and Poverty’, Journal of Peace Research, 45 (1), 5–15. Seul, R. J. (1999) ‘ “Ours is the Way of God”: Religion, Identity and Intergroup Conflict’, Journal of Peace Research, 36 (5), 553–69. Shupe, A. (1990) ‘The Stubborn Persistence of Religion in the Global Arena’, in E. Sahliyeh (ed.), Religious Resurgence and Politics in the Contemporary World. Albany, NY: State University of New York Press, pp. 17–26. Stahanke, T. (1999) ‘Proselytism and the Freedom to Change Religion in International Human Rights Law’, Brigham Young University Law Review, 1, 251–350. Stark, R. (2001) One True God: Historical Consequences of Monotheism. Princeton, NJ: Princeton University Press. Stark, R. and Finke, R. (2000) Acts of Faith: Explaining the Human Side of Religion. Berkeley, CA: University of California Press. Svensson, I. (2007) ‘Fighting with Faith: Religion and Conflict Resolution in Civil Wars’, Journal of Conflict Resolution, 51 (6), 930–49. Tajfel, H. (1959) ‘Quantitative Judgment in Social Perception’, British Journal of Psychology, 50, 16–29. Tajfel, H. and Turner, J. C. (1986) ‘The Social Identity Theory of Intergroup Behavior’, in S. Worchel and W. G. Austin (eds), Psychology of Intergroup Relations, second edition. Chicago, IL: Nelson-Hall, pp. 7–24. Toft, M. D., Philpott, D. and Timothy, S. (2011) Shah God’s Century: Resurgent Religion and Global Politics. New York: W. W. Norton. Vamik, V. (2002) ‘September 11 and Societal Regression’, Group Analysis, 35, 456–82. Vanhanen, T. (1999) ‘Domestic Ethnic Conflict and Ethnic Nepotism: A Comparative Analysis’, Journal of Peace Research, 36 (1), 55–73. Weinberg, L. and Eubanks, W. (1998) ‘Terrorism and Democracy: What Recent Events Disclose’, Terrorism and Political Violence, 10 (1), 108–18. Wilmer, F. (1997) ‘Identity, Culture and Historicity: The Social Construction of Ethnicity in the Balkans’, World Affairs, 160 (1), 3–16. Wuthnow, R. and Lewis, V. (2008) ‘Religion and Altruistic US Foreign Policy Goals: Evidence from a National Survey’, Journal for the Scientific Study of Religion, 47 (2), 191–209. © 2014 The Authors. Political Studies © 2014 Political Studies Association POLITICAL STUDIES: 2015, 63(5)