A Transatlantic Secular Divide? The Representation of Religion in EU and US Foreign Policy

It is a common—often stereotypical—presumption that Europe is secular and America religious. Differences in international religious freedom and religious engagement policies on both sides of the Atlantic seem to confirm this “cliché.” This article argues that to understand why it has been easier for American supporters to institutionalize these policies than for advocates in the EU, it is important to consider the discursive structures of EU and US foreign policies, which enable and constrain political language and behavior. Based on the analysis of foreign policy documents, produced by the EU and the United States in their relationship with six religiously diverse African and Asian states, the article compares how both international actors represent religion in their foreign affairs. The analysis reveals similarities in the relatively low importance that they attribute to religion and major differences in how they represent the contribution of religion to creating and solving problems in other states. In sum, the foreign policies of both international actors are based on a secular discursive structure, but that of the United States is much more accommodative toward religion, including Islam, than that of the EU.


Introduction
Religion has often been perceived as a dividing line between Europe and the United States.Academic publications on the transatlantic divide are full of anecdotes and examples from both scholarship and public media that purport a division between "secular Europe" and "religious America" (e.g.,Howard 2011, 2-4;Waldschmidt-Nelson 2013, 319f).Whereas most of them refer to domestic politics, occasionally this "cliché" (Berger, Davie, and Fokas 2008, 9) is also borrowed to explain a transatlantic rift in how international relations are understood (Solana 2003).Transatlantic comparisons to assess whether and how such a religious divide becomes apparent in European and American foreign policy are scarce and their findings inconclusive.While some emphasize differences (Hurd, 2007(Hurd, , 2008;;Hampton 2013), others detect more similarities (Mandaville and Silvestri 2015;Wolff 2015).
A first glance at foreign policy approaches that directly address issues of religion seems to confirm the divide.The United States, for example, started much earlier than the European Union (EU) and its member states to promote religious freedom worldwide.Already in 1998, the US Congress adopted the International Religious Freedom Act (IRFA), which made annual reporting on the situation of religious freedom in the world mandatory for the US Department of State.For this purpose, the latter established the Office of International Religious Freedom and appointed an Ambassador-at-Large for International Religious Freedom whose objective is the monitoring of religious persecution and discrimination worldwide. 1dditionally, the US Commission on International Religious Freedom (USCIRF) was established as part of IRFA, "an independent, bipartisan U.S. federal government commission, the first of its kind in the world, dedicated to defending the universal right to freedom of religion or belief abroad." 2 In Europe, by contrast, more than ten years later, a few EU member states established small offices and/or introduced policies tasked with promoting religious freedom abroad; for example, Italy and the Netherlands in 2009and the UK in 2010(Annicchino 2013;Veer 2014;Lindsay 2015).The EU adopted Guidelines on the Promotion and Protection of Freedom of Religion or Belief 3 in 2013, which, among other things, call on EU missions to monitor respect for freedom of religion or belief in third countries.Their reports, though, unlike in the case of the United States, are not public, and so far, "no substantial change in policy practices is discernible" following the adoption of the guidelines (Foret 2017, 19). 4 In 2016, matching the Ambassador-at-Large in the United States, the European Commission appointed a Special Envoy for the promotion of freedom of religion or belief outside the EU, who serves as special adviser to the Commissioner for International Cooperation and Development. 5 In addition, the US Department of State established the Office of Religion and Global Affairs in 2014, which, headed by the Special Representative for Religion and Global Affairs, works to implement the US Strategy on Religious Leader and Faith Community Engagement, provides advice and analysis on foreign policy matters related to religion, and engages in dialogue with religious actors. 6Although similar institutions were established in a few EU member states; for example, France and Germany (Germain 2014;Kreft 2015), and although in both, the EU and the United States, training modules for diplomats on issues of religion were introduced (Bilde 2015; Wolff 2015), such offices remain rare in Europe, and no respective office has been established within the European External Action Service (EEAS), which functions as the EU's main body for the management of its external affairs.
Reasons for the earlier and heightened awareness for issues of religious freedom and religious engagement in US foreign policies compared to Europe may be manifold, including stronger efforts of religious lobbies-although, supporters of these policies have been active on both sides of the Atlantic (Hertzke 2004;Leustean 2011;Bolvin 2013;De Jong 2013;Wolff 2015).However, this article does not focus on the exact causes of the described policy developments.I am more interested in the bigger picture, namely the deeper discursive structures of foreign policy, which enable and constrain political language and behavior.
Based on a quantitative content and qualitative predicate analysis of foreign policy documents, produced in the relationship with six religiously diverse states in Sub-Sahara Africa (Central African Republic, Nigeria, Tanzania), East Asia (Indonesia, the Philippines), and the Middle East (Lebanon), I compare how religion is represented in EU and US foreign policy to explore whether a transatlantic divide exists.My results suggest that the foreign policies of both international actors are based on a secular structure, but that of the EU is less accommodative toward religion than that of the United States, which also includes Islam.I do not claim that these differences represent the main or even the only reason for the differences in when and to what degree both international actors have introduced religious freedom and religious engagement into their foreign policies.But I argue that they help to develop a deeper understanding of why it has been easier for advocates of international religious freedom and religious engagement policies in the United States to institutionalize their concerns than in the EU.
These findings contribute to research on religion in foreign policy, which so far has primarily focused either on the United States or the EU and has primarily been based on survey research and elite interviews, by analyzing foreign policy documents to systematically compare the representation of religion in EU and US foreign policy.Moreover, the findings are useful for furthering mutual understanding on both sides of the Atlantic, which is important for jointly tackling international challenges that might be arising from the growing political role of religion in other parts of the world.
Before I continue, one further explanation is necessary: I focus here on the EU rather than on a single European state to allow for a broader comparison between Europe and the United States.The EU represents many of the interests of its member states abroad.Either negotiated in an intergovernmental process (in case of its foreign and security policy) or produced at the supranational level (in case of development and external trade policies), EU external policies represent the smallest common denominator all member states can agree to, despite their diverse historical experiences and foreign policy interests that they often also pursue independently of the EU.Paraphrasing Smith (2014, 14), the main question, therefore, is whether the accumulation of EU discourses distinguishes the EU from other international actors, such as the United States.But I do not analyze whether and to what extent these EU discourses represent the individual discourse of any member state.
The remainder of the article is structured as follows: The next section discusses the role of secularism and religion in foreign policy and, drawing on the literature on transatlantic issues, identifies domestic sources that potentially lead to differences in the secular structures of EU and US foreign policies.Subsequently, the methodology is explained, which includes the discursive framework, the selection of the included text corpus, and the description of the coding process.Next, the empirical findings are presented, and, finally, their implications for future research are discussed.

A Transatlantic Perspective on Religion and Secularism in Foreign Policy
Research on international relations and foreign policy analysis long neglected religion due to the deeply embedded assumption amongst most social scientists that the political role of religion would decline over time.This assumption has changed with "the global rise of religion's political influence" since the 1960s and, primarily, the terrorist attacks on September 11, 2001, which "cleared the way for political scientists to turn their analytical attention to this influence" (Philpott 2009, 192).Since then, the literature on religion in foreign policy has grown considerably, particularly with regard to the United States but also the EU.Studies that directly compare EU and US foreign policy with respect to religion, however, are scarce and provide inconclusive answers to the question of whether there is a transatlantic divide in how the EU and the US approach religion in their foreign policies.Some policy studies, for instance, observe a similar "secular bias" among diplomats on both sides of the Atlantic (Mandaville and Silvestri 2015) and a convergence of approaches to engage religious actors (Wolff 2015).Others focus more on differences.Hampton (2013), for example, posits that differences in the social and political importance of religion on both sides of the Atlantic facilitates varied perceptions of threat and Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 security.Her argument goes that, since many-more religious-Americans still believe in the providential blessing of their nation and in fundamental evil that needs to be wiped out, they tend to opt more easily for military force than their-less religious-European counterparts.However, even though providing intriguing insights in the impact of the "religious America-secular Europe" cliché on strategic security cultures, Hampton does not really provide any clues on the question of how the EU and the United States approach and represent religion in their foreign policies.
More helpful in that regard is Hurd's (2007Hurd's ( , 2008) ) identification of two distinct trajectories of secularism effective in European and American foreign policy: a laicist tradition, which is based on a separationist narrative, in which religion is expelled from politics, and a more accommodationist narrative with Christianity and Judaism as unique bases of secular democracy.She argues that, even though each understanding exists on both sides of the Atlantic, the EU tends more toward the laicist trajectory, the United States more toward the Judeo-Christian one.These prevailing narratives, according to Hurd, shape foreign policies.She argues, for example, that "for laicists, Muslim-majority societies can be 'modernized' if, like Turkey, they follow in the footsteps of their secular Western role models and enforce the exclusion of religion from politics" (Hurd 2007, 357).For Judeo-Christian secularists, by contrast, Islam would be incompatible with democracy and any separation between religion and politics, explaining, for instance, the adverse attitude of the United States toward Iran.Even though there may exist more understandings of the secular in international relations than the two identified by Hurd (Philpott 2009, 185), the frequent citation of her approach evidences its wide recognition, and the distinction of two contrasting secularisms is particularly promising for the matter at hand.Therefore, I use her work as a starting point.Primarily, I draw on her understanding of secularism as discursive tradition and political authority to assess whether and how the two distinct secular trajectories identified by her shape how religion is represented in EU and US foreign policy.
Following Hurd (2007Hurd ( , 2008)), I suggest that the foreign policies of the EU and the United States are based on a secular structure because neither the EU nor the United States are (formed of) religious states and they all understand themselves as secular.Secularism, however, does not mean the absence of religion.To the contrary, "multiple secularisms" exist in democratic societies, which reflect different arrangements that states have found to both bound and integrate religion (Bhargava 1998;Stepan 2011;Sheikh and Waever 2012).As Fox (2015, 179) has shown in his extensive study of political secularism worldwide, "true state neutrality on the issue of religion is very much the exception rather than the norm," and this also applies to Europe and North America, where most states legislate religion in one way or another.Therefore, I assume that the secular structures of foreign policies, rather than being void of religion, reflect basic assumptions about religion and politics held on both sides of the Atlantic.Transatlantic differences in these assumptions, I argue, can be detected when considering how religion is represented in foreign policy documents.
Such an analysis is not only useful for the transatlantic comparison.It also addresses the question of which specific form of secularism the EU draws on in its relations with other states.The historical experiences of its member states with respect to religion are diverse.They range from strict separation (e.g., in France) to accommodation and close cooperation (e.g., in Germany) to establishment of religion (e.g., in the United Kingdom and in many Scandinavian countries) (Stepan 2011;Sheikh and Waever 2012).Therefore, the EU chose to remain largely silent on issues of religion until the 1990s, although religious actors and ideas, primarily Catholic, had played an important role in promoting and shaping European integration (Philpott and Shah 2006).Major change came about with the Treaty of Lisbon, which, for the first time, formulated substantive norms on religion (Challand Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 2009; Doe 2009;Mudrov 2016).Even though it leaves national church-state arrangements in the competences of the member states, it prescribes "an open, transparent and regular dialogue" with churches and religious and philosophical nonconfessional organizations (art.17).Willaime (2009), therefore, has pointed out that the secular model of the EU should not be conflated with the French practice of strict separation.Especially since the Treaty of Lisbon came into force in 2009, the EU, according to Willaime, has shared characteristics of a cooperative model of church-state relations, which, based on separation between political and religious systems, recognizes the role of religion in the public sphere (similar Doe 2009, 157).The question remains which of these understandings shape the EU's relationship with other states-its more laicist past, as assumed by Hurd, or its more cooperative present, as pointed out by Willaime.Similar questions appear in the context of American foreign policy.While some claim that religion has an exceptionally prominent role in US foreign policy (e.g., Haynes 2009, 297-300), others identify a lack of engagement with religion in the US Department of State, including the marginalization of the international religious freedom policy structure within the department (Albright 2006;Chaplin and Joustra 2010;Johnston 2011;Farr 2013;Lindsay 2015, 258ff).In addition, there are major differences between administrations that may be reflected in foreign policy documents.For instance, the American presidents "George W. Bush and Barack Obama represent two fundamentally different streams of the American religious tradition" (Devuyst 2010, 36).While the former embodied the narrative considering America to be the "Chosen Nation"-that is, chosen by God-the latter rather typified an opposing conception emphasizing a civil religion, which balances religion and politics and which is more inclusive toward non-Christian faiths (Williams 2013).
To build an analytical framework, the different roles that religion can potentially play in foreign policy need to be considered.Religion can broadly be defined here as consisting of "beliefs, actions and institutions which assume the existence of supernatural entities with powers of action, or impersonal powers or processes possessed of moral purpose" (Bruce 2011, 112). 7These beliefs, actions, and institutions can have diverse impacts on foreign policy; that is, on "the formal policies . . .which affect various military, economic, humanitarian, social, and cultural dimensions of [the] relations with other states and nonstate actors" (Warner and Walker 2011, 114).Drawing on Warner andWalker (2011) andon Sandal andFox (2013, 12-29), who usefully summarize potential impacts of religion on foreign policy and international relations, I identify four potential domestic sources that may mediate how the United States and the EU approach religion in their foreign policies: religiosity of the population including its political leaders, institutional religion-state relations, ideas about the appropriate relationship between religion and politics, and religious tradition.This list of potential domestic sources for the transatlantic secular divide in foreign policy is not exhaustive.But these four factors correspond particularly well with insights from the literature on transatlantic differences concerning religion, which is why I focus on them in the following empirical analysis.

Religiosity
One way how religion impacts foreign policy refers to the religious worldviews, attitudes, and practices of political leaders and their constituents, which may shape how foreign policies are formulated (Sandal and Fox 2013, 13ff.;Warner and Walker 2011, 123ff).One could, for example, assume that states whose societies are more religious assign more political importance to issues of religion, such as the protection of religious freedom and the engagement with religious actors, than those whose members tend to be less religious.By extension, one could also assume that there are more, and more influential, religious interest groups in more religious societies.8When comparing the prevalence of religious beliefs and practices in the United States with those in Europe, it is safe to say that, overall, Americans are more religious than Europeans, despite some geographical variation on both sides of the Atlantic (Bös and Hebel 2007;Berger et al. 2008, 9ff;Pfaff 2008;Norris and Inglehart 2011, 83ff).Indeed, it is common that US presidents openly refer to their faith and use religious language to legitimize their foreign policies (Berggren and Rae 2006;Inboden 2008;Devuyst 2010), which particularly attracts religious voters.On the special case of the Iraq war, for example, Froese and Mencken (2009, 112) demonstrate that, irrespective of party identity, "Americans who feel that their religious faith should be more influential in political matters have placed their trust in President Bush and, in turn, lent their support to the Iraq War."In Europe, by contrast, invoking religion to legitimize foreign policies is relatively uncommon (Hampton 2013).Consequently, one could expect that religion is also more prevalent in US foreign policy documents than in those issued by the EU.If this were the case, it would reinforce the depiction of a "more religious America" and a "more secular Europe" and the impact of this difference on foreign policy.

The Institutional Relationship between Religion and State
The access of organized religions and religious interest groups and parties, which channel religious values and ideas into foreign policy, is shaped by institutional structures, including those that regulate the religion-state relationship (Haynes 2009, 295;Warner and Walker 2011, 124).Drawing on the religion and politics literature, Warner and Walker (2011, 124), for instance, assume that "all other things being equal, the overt influence of a religion in a state's foreign policy may be lower in a state in which there is high differentiation through a formal, legal separation of religious and state authority" because religious actors and values lack formal access to politics and the state.Considering that in Europe religion and state are in most countries less strictly separated than in the United States (Fox 2015), and Willaime's (2009) already-mentioned claim that the EU shares characteristics of a cooperative model of church-state relations rather than strict separation, one would arrive at a completely different expectation than the one explicated above.One would expect that the "wall of separation" between religion and state that is stipulated in the First Amendment of the US constitution restricts, if not formally then at least discursively, the extent to which the foreign policy administration can focus on issues of religion.Lindsay (2015), for example, in her comparison of religion in US and UK foreign policy detects that, despite the more religious civil culture in the United States, its foreign policy bureaucracy, due to the constitutional precept of separation of religion and state, is more hesitant to engage with issues of religion than the bureaucracy in the UK with its established religion.This could mean for the analysis here that, even though the United States actively promotes international religious freedom and engagement with religious actors, beyond these specific policies religion is as much, or as little, present in foreign policy documents as in those issued by the EU.If this alternative expectation was confirmed by the empirical analysis, it would seriously challenge the "religious America-secular Europe" cliché and suggest that religiosity of the population and political leaders is a less significant, or at least not the only factor, that determines how much importance international actors assign to religion in their foreign policies.

Ideas about the Appropriate Relationship between Religion and Politics
The relationship between religion and state, however, is not only a legal question.As Hurd (2007Hurd ( , 2008) ) stresses, secularism is a discursive tradition that is nurtured by ideas on where the line between religion and politics should be drawn.Such ideas on what constitutes the appropriate relationship between religion and politics are also incorporated into the secular structure of foreign policies (Warner and Walker 2011, 120).
Several scholars have identified major differences in how the appropriate relationship between religion and politics is understood on both sides of the Atlantic.As already mentioned, much of this literature focuses on the differences between specifically French, rather than European, and US secularism.Stepan (2011), for example, points out that French laicism is based on the "clerically hostile form of freedom of the state from religion," while US secularism is a "clerically friendly form of freedom of religion from the federal state."Accordingly, Kuru (2007, 571) distinguishes two forms of secularism: On the one hand, "assertive secularism," prevailing for instance in France, means "that the state excludes religion from the public sphere and plays an 'assertive' role as the agent of a social engineering project that confines religion to the private realm."On the other hand, "passive secularism . . .requires that the secular state play a 'passive' role in avoiding the establishment of any religion" and is more influential in the United States.Berger et al. (2008, 126), similarly, differentiate a "hard" from a "soft" version of separation predominating in Europe and the United States, respectively.These differences match Hurd's (2008) suggestion of a laicist, or separationist, trajectory of secularism predominating in Europe and a more accommodationist one in the United States.The question remains whether insights gained from arrangements prevailing in France can be extrapolated to the EU level.
The mentioned authors share the assumption that differing understandings of secularism have political implications.One major consequence usually stressed is that the public expression of religiosity in the United States is more protected than in many parts of Europe, where controversies about religious practices and symbols have more often been solved in favor of an ejection of religion from official public spheres (Pew Research Center 2012).Accordingly, religion in the United States is often "seen as a resource (the means by which to resolve secular as well as religious dilemmas); in Europe, it is part of the problem" (Berger et al. 2008, 126).These different views correspond with Philpott's (2007) reasoning about religion's ambivalent role in politics.There are religious actors who support democracy and counter violence and others who do the opposite.The differences discussed in the transatlantic literature suggest that US foreign policy focuses more on the "positive" role that religion can play in other countries, while the EU concentrates more on religion's "negative" role, justifying why it should remain strictly separated from politics.

Religious Tradition
Religious tradition refers to the specific culture that shapes the identity of an international actor, which in turn may affect how this actor defines its foreign policy interests (Sandal and Fox 2013, 28ff.;Warner and Walker 2011, 120).Some of the literature on religion in foreign policy has focused on the impact of differences between denominations, including between different Christian groups; for example, on the general orientation of US foreign policy (Guth 2013), on US policy toward the Middle East (Oldmixon, Rosenson, and Wald 2005;Baumgartner, Fancia, and Morris 2008;Collins et al. 2010), or on European integration (Boomgaarden and Freire 2009;Minkenberg, 2009Minkenberg, , 2012;;Nelsen and Guth 2015).Another question on religious tradition is how Europe's and America's Christian heritage shapes their Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 relations with other faiths, such as Islam, which often is constructed as the main "other" of both the EU and the United States (Hurd, 2007(Hurd, , 2008;;Challand 2009;Smith 2009).To the question of whether there is a transatlantic difference in that regard, one finds no clear response in the literature.Hurd's (2007Hurd's ( , 2008) ) assertion of a prevailing Judeo-Christian trajectory of secularism in US foreign policy suggests a more critical stance toward Islam in US than in EU foreign policy.Others, such as Berger, Davie, and Fokas, argue that due to the predominating separationist notion of secularism in Europe, religion is, as mentioned, primarily seen as a problem and even "more so, whether fairly or not, with reference to Islam" (2008,126).In contrast, the United States' approach of "passive secularism . . .allows for the public visibility of religion" (Kuru 2007, 571), which also includes Islam and therefore implies a more "positive" view of Islam than is prevalent in Europe.
Similar diverging views can be found on Christianity.Hurd's (2008) differentiation between a laicist and a Judeo-Christian trajectory of secularism in EU and US foreign policy, respectively, implies a more positive or protective representation of Christianity and Christians in US foreign policy (whereas, according to the European narrative, religion should be expelled from politics altogether).The existing Christian heritage of the EU, as emphasized for instance by Philpott and Shah (2006), by contrast, suggests more similarities between EU and the United States with respect to Christianity's representation.
Assessing whether the representation of Islam and Christianity is more or less critical in EU than in US foreign policy documents will contribute to addressing these contradictions in the transatlantic literature.

Methodology
This study draws on a discursive framework to compare how religion is represented in EU and US foreign policy documents.It complements previous research that has primarily used public opinion and other survey data (e.g., Baumgartner et al. 2008;Boomgaarden and Freire 2009;Froese and Mencken 2009;Guth 2013;Nelsen and Guth 2015;Foret 2017) and elite interviews (Lindsay 2015;Mandaville and Silvestri 2015;Wolff 2015;Foret 2017) to study the role of religion in American and European foreign policy.Studying foreign policy documents has the advantage that it is not limited by preset questions that often make data sets incomparable or by social desirability bias or lack of access to interlocutors.It enables the researcher to lay bare the macrostructures underlying day-to-day foreign policy practices.Previous discursive studies on religion in EU and US foreign policy, however, have focused only on the United States (Berggren and Rae 2006;Devuyst 2010) or on the EU (Larsen 2014) or do not inform the reader about the text corpus used (Hurd 2008).
Discourses can be understood as "structures of signification which construct social realities" (Milliken 1999, 229).They refer to "the space where human beings make sense of the material world, where they attach meaning to the world and where representations of the world become manifest" (Holzscheiter 2014, 144).Discourses do not determine what political actors do, but they constrain their language and political behavior.Therefore, reconstructing discursive structures does not aim at explaining details of foreign policies; for example, the exact causes of why the EU and the United States adopted policies on international religious freedom and religious engagement.Rather, discourse-analytical efforts aspire to identify "significative practices and the knowledge systems underlying them" (Milliken 1999, 236).In the case at hand, this means reconstructing how religion is represented in EU and US foreign policy documents to uncover and compare deeply held attitudes and interpretations.
To make the text corpus comparable, I focus on documents that were produced in the relationship of the EU and the United States with six other states: Nigeria, Tanzania, and the Central African Republic (CAR), as well as Indonesia, the Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 Philippines, and Lebanon.The country selection is primarily used to focus the transatlantic comparison.The analysis does not aim at evaluating the EU's and the United States' roles in these countries.The six countries were selected for three reasons: First, to make sure that religion appears in the documents of the two secular international actors at all, countries were chosen in which religious groups have played a significant political role.Second, to make sure that the analysis can focus on the representations of Islam and Christianity, countries were chosen in which these two are the largest denominations.The six states belong to the religiously most diverse countries in the world.In half of them, Muslims constitute the majority (87.2 percent in Indonesia, 54 percent in Lebanon, and 50 percent in Nigeria); in the other half, Christians represent the largest group (92.5 percent in the Philippines, 61.4 percent in Tanzania, and 50 percent in the Central African Republic).In all, the respective other religion-that is, Christianity or Islam-represents the largest minority. 9This composition is important for the analysis here because it seeks to compare EU and US views on their "own" religion abroad (i.e., Catholic and Protestant Christianity) as well as their main "other" (Islam).This is the main reason why I desisted from including some of the "usual suspects" when thinking about religion and foreign policy, such as Turkey, Iran, Saudi-Arabia, or Israel, or emerging powers, such as China, India, or Russia, because they either are religiously homogenous-Muslim-majority-states or they consist of large groups of neither Muslim nor Catholic/Protestant affiliation.Both would hamper the comparison between representations of "own" and main "other" religion.Third, to allow for some variation I included countries with protracted violent conflicts in relation to religion as well as some which have widely been perceived as peaceful democracies, despite their religious diversity and occasional and/or regional violent incidents (Indonesia, Tanzania).Additionally, to challenge the dominant public image of aggressive Muslims and peaceful Christians, I included a country which saw attacks by Christians against Muslims (CAR).
I retrieved the documents from websites of the EU (EUR-Lex, Council of the European Union, High Representative of the Union for Foreign Affairs and Security Policy, the European Commission's press database, and the European Parliament) and of the United States (White House, Department of State, and US Congress).I first selected all documents between 2001 and 2015 with one of the six country names in the title or which dealt exclusively with one of these countries.I only included documents representing an institutional position to avoid distortion due to individual opinions. 10The time frame allows for the comparison between the time before and after the Treaty of Lisbon came into force in 2009, and between the Bush Jr. and the Obama administrations (the cut-off date was December 31, 2015, to avoid distortions due to the heated election campaigns in 2016).Earlier time periods, unfortunately, could not be included due to a lack of access to a comparable number of US documents from previous administrations. 11In total, this made 1,684 documents (EU: 872, US: 812). 9CIA World Fact Book (https://www.cia.gov/library/publications/the-world-factbook/,accessed: January 7, 2017). 10Therefore, parliamentary questions and responses to them were excluded, as well as statements by journalists or representatives from other states in interviews or joint speeches.Documents simply announcing upcoming events, visits, and press availabilities were not included; documents reporting on these were.Statements by EU delegations, US embassies, and the US Mission to the UN were only included if posted on the mentioned websites of EU and US institutions.Their own websites were excluded for practical reasons because they did not provide access to documents from the entire time frame under investigation.Documents from the European Court of Justice and the European Investment Bank published on the mentioned EU websites were excluded due to a lack of corresponding documents on the American side.Regular reports were also excluded to avoid a distortion of the results due to the extensive reporting system of the United States (e.g., on religious freedom worldwide) that is not matched by a similarly extensive system on the part of the EU.
11 Only the most important US foreign policy documents are archived.After collecting the documents, I first identified, based on a list of thirty-four key words, the documents that mentioned religion for further analysis. 12To compare the quantitative presence of religion in EU and US foreign policy discourse, I calculated for each side the share of documents mentioning religion in relation to the total number of documents.
For the qualitative analysis, I adapted Milliken's discourse analytical method of predicate analysis, which "focuses on the language practices of predication-the verbs, adverbs and adjectives that attach to nouns.Predications of a noun construct the thing(s) named as a particular sort of thing, with particular features and capacities" (Milliken 1999, 232).The "thing" whose representation is reconstructed here is religion.However, in foreign policy documents, religion is not only referred to as a noun but also as an adjective to label other nouns; for example, "religious conflict" or "interfaith dialogue."Therefore, I broadened the focus to include also nouns to which religion is attached as an adjective as they also shed light on how religion's role is interpreted.As can be seen in the following analysis, foreign policy actors and administrations on both sides of the Atlantic hesitate to discuss the political role of religion in length in their documents.In most cases, religion is mentioned in passing.Therefore, predicate analysis lends itself as the most suitable method because it allows for an assessment of how religion is constructed through predication, even though it is not elaborated on in length.
The coding proceeded as follows: Each mention of religion (as noun or as adjective, based on the thirty-four key words) in the documents was coded according to whether predications (corresponding adjectives or nouns) carried a negative or a positive connotation; that is, the mention was coded as "problem" if religion was framed as something contributing to a problem (e.g., to violence or to discrimination), as "resource" if framed as contributing to solving problems (e.g., religious leaders who condemn human rights violations), or as "neutral" if no such tendency could be detected.Afterward, I calculated the number of occurrences of each framing and assessed for each document whether it tended more toward representing religion as "problem" or as "resource" or in a "neutral" manner (if "neutral" descriptions predominated or "problem" and "resource" framings balanced each other out).In a similar manner, I assessed for each document separately how Islam (and related terms, such as Muslim, mosque, etc.) and Christianity (and related terms, such as Christian, church, etc.) were represented-that is, in a rather critical way (e.g., as a perpetrator of violence and discrimination) or in a more positive or protective way, (e.g., as a contributor to peace or as a victim to be protected from violence and discrimination).Again, I ascertained the tendency of the whole document (i.e., more critical, more appreciative, or neutral).The subsequent transatlantic comparisons, thus, used the documents as units of analysis, rather than single sentences or words.
In a nod to debates on transatlantic differences with respect to foreign policy preferences, which emphasize the EU's inclination toward civilian, nonviolent measures and the promotion of human rights and the rule of law in opposition to the United States' focus on military security (Manners 2002;Smith 2014, 204f), I additionally coded the documents with respect to the contexts in which religion was mentioned (e.g., security, human rights, development, democracy/democratization, and others).Afterward, I prepared a ranking for each document to assess whether there are transatlantic differences and similarities with respect to the contexts in which religion is primarily represented.

Religion in EU and US Foreign Policy Documents
In the theoretical considerations above, I identified a series of rival expectations on the relative prominence and representation of religion in general, and of Islam and Christianity specifically, in EU and US foreign policy documents.In the following section, I present the findings of the analysis on each of these expectations, to assess whether there is evidence for a transatlantic divide or not.

Prominence of Religion
One expectation that followed from the differing prevalence of religiosity on both sides of the Atlantic was that issues of religion were more prominent in US than in EU foreign policy documents.When considering religion-state arrangements, by contrast, one would expect that the representation of religion is quite restricted in US foreign policy and, therefore, not more prominent than in EU foreign policy.At first glance, the analysis of the documents confirms the expectation that religion is more present in US foreign policy documents than in those of the EU (by a bit more than six percentage points) (table 1, first two columns, first row).
This difference even increases to almost nine percentage points when considering that foreign policy decision-making is a preserve of executive bodies (21.02 percent of their documents contain references to religion in the United States, compared to 12.18 percent in the EU).However, the appearances of religion are often very brief.If one removes all documents where religion is only mentioned in passing (i.e., once or twice in just one sentence), the difference shrinks considerably to 6.58 percent for the EU and 10.51 percent for the United States.This suggests that the attention for issues of religion within executive bodies is relatively low and does not differ considerably on both sides of the Atlantic, whereas the attention for issues of religion is primarily driven by the legislatures, particularly in the EU (58.82 percent of the European Parliament's documents and 29.41 percent of US congressional documents mention religion).
The comparison of different time periods in the EU and the United States (table 1, last four columns) confirms the assumption that religion overall has become more prominent in foreign policy discourses after the Treaty of Lisbon and that it was more prominent under the Bush Jr. administration than under Obama.
Attention for issues of religion, in addition, is driven by specific US and EU interests, as well as by dynamics in the target countries.In case of Indonesia, for instance, US diplomats used their good relations with the country to emphasize their good relations with Islam or, in other words, to controvert their allegedly negative view of Islam after 9/11.Many documents stress that Indonesia is the largest Muslim democracy in the world, demonstrating that Islam and democracy are no contradiction in terms.EU-Indonesia relations, by contrast, are based primarily on trade, in which religion has not been considered an important element.In Nigeria, by Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 contrast, EU institutions focused a lot on prominent human rights cases that the United States were less vocal about, such as the case of Amina Lawal who was sentenced to death by stoning for adultery by Sharia courts in a northern state of Nigeria. 13With the rise of Boko Haram, the foci of the United States and the EU converged, but Nigeria remained a country where the EU is especially aware of issues of religion.In CAR, clashes between Muslim fighters and Christian militias led to the displacement of large parts of the civil population as well as acts of reprisal against the Muslim minority, explaining why attention for issues of religion in CAR jumped up in 2013.Similar cycles of attention can be related, in Lebanon, to the assassination of Prime Minister Rafik Hariri in 2005 and the subsequent Cedar Revolution and to Israeli airstrikes following a series of rocket attacks and raids into Israeli territory by Hezbollah in 2006, as well as, in the Philippines, to highs and lows of the conflict between the government and the Moro Islamic Liberation Front between 2000 and the preliminary peace agreement in 2012.
In conclusion, the numbers suggest that religion is not a top priority either of the EU nor of the US foreign policy administration.Even though there is some difference, one can hardly argue that religion is a lot more present in US than in EU foreign policy documents.Even quite the opposite is the case if one looks separately at the legislatures.This result challenges the assumption that higher levels of societal religiosity translate into higher levels of attention for religion in foreign policy.Religious actors may be able to push some of their specific projects onto the foreign policy agenda, as the evangelical movement in the United States did with the International Religious Freedom Act and a few other acts (Hertzke 2004;Haynes 2009, 299); and more overtly religious authorities, such as George W. Bush, may give more utterance to faith than less overtly religious authorities (see also Devuyst 2010).Overall, however, foreign policy administrations on both sides of the Atlantic seem to be similarly hesitant to actively engage with religion, confirming the "secular bias" under which both operate, previously emphasized by Mandaville and Silvestri (2015, 3).

Religion: Problem or Resource?
Another expectation identified above considered differing ideas about the appropriate relationship between religion and politics on both sides of the Atlantic, which would result in religion being represented primarily as a problem in EU foreign policy and more often as a resource in US documents.The analysis, indeed, confirms this expectation.
As mentioned, references to religion in EU and US foreign policy documents are usually very brief.There is hardly any explicit discussion of the role of religion in the six included countries.However, the predicates and nouns associated with religion (or related terms) give some indication of how the role of religion is interpreted.The EU as well as the United States, when representing religion as a problem, primarily do so in the context of violence: tensions and attacks where violent conflict persists, and terrorism and separatism in countries where terrorist and separatist groups are active.Both also stress problems of religious intolerance, extremism, fundamentalism, and radicalization.From time to time they both underline, though, that religion cannot, and should not, be blamed for violence and terrorism.When representing religion as a resource, both primarily highlight the role and potential of religious leaders and groups in promoting peace, human rights, and interfaith dialogue and cooperation and mention religion in the context of reconciliation, respect, and moderation.Furthermore, they sometimes emphasize religious diversity as a strength that encourages tolerance and peaceful coexistence.Such  (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) predominantly framing religion as a problem or a resource (percent).
"negative" and "positive" representations of religion can be found in both EU and US foreign policy documents.However, in EU documents, those with a negative connotation prevail, whereas in US documents those carrying a positive connotation can be found much more often (figure 1).When looking at the six country cases separately, it is the existence of violence that drives up critical references to religion.This becomes particularly clear when comparing the cases of Tanzania and Nigeria.Tanzania and Nigeria belong to the religiously most diverse countries worldwide (with about equal shares of Muslims and Christians), but Tanzania is a rather peaceful democracy, despite occasional incidents of religious intolerance, and Nigeria is a country torn by protracted conflict and violence.This is reflected in the documents, with Tanzania being the only of the six countries where even the EU issued more documents with more "positive" references to religion than critical ones (50 percent versus 33 percent), and Nigeria the only country in which even the United States issued more critical documents (35.90 percent) than appreciative ones (33.33 percent).
Another transatlantic difference is that the EU and the United States, even though both predominantly frame religion as a security issue, then diverge with respect to the other contexts.For the EU, religion is primarily a security issue (62.31 percent of its documents frame religion predominantly in the context of security), secondly a human rights issue (34.62 percent), and only marginally an issue of democratization (8.46 percent) and development (6.15 percent).For the United States as well, religion is mainly a security issue (53.14 percent), secondly though an issue of democracy (20.57percent) and development (16.00 percent), and only marginally a human rights issue (6.29 percent). 14The latter supports the contention that the United States' international religious freedom policy-that is, a policy that is explicitly framed as a human rights policy-lacks practical influence in American foreign-policy making.
In sum, the EU represents religion primarily as a problem for security-for instance, as a contributing factor in violent conflicts and terrorism-and as a problem for human rights, especially of religious minorities (such as in Indonesia) and of women and girls, and in the context of harsh punishment (both particularly in Nigeria) and freedom of expression (emphasized primarily in the case of Lebanon).The United States, by contrast, represents religion more often as a resource for dialogue and reconciliation, democratization, civil society and community development, the provision of education and public health (especially HIV/Aids prevention15 ), and as inspiration for humanitarianism, for example, in the context of disaster relief.These findings confirm that differing understandings of the appropriate relationship between religion and politics on both sides of the Atlantic have a bearing on how the EU and the United States represent religion in their foreign policies.They also suggest that the EU, though consisting of member states with different religionstate arrangements and historical experiences, is guided in its foreign policies by an understanding of religion as primarily fueling conflict and violence when mingling with politics-a notion that has not lost ground since the Treaty of Lisbon came into force, even though it advocates a more cooperative relationship with religion.

Representations of Islam and Christianity
Finally, I identified several rival expectations on how the representation of Islam and Christianity in EU and US documents might differ.What the analysis reveals is that EU documents are more critical of Islam than US documents, while US documents are a bit more protective of Christianity than those of the EU.
First of all, however, it is notable that especially the EU, but also the United States, largely refrain from explicitly referring to any specific faith in their foreign policy documents.More than half of all documents of the EU that refer to religion do not mention terms related to Islam (51.54 percent) or to Christianity (73.08 percent) at all; almost as many in the case of the United States (45.14 percent and 69.71 percent, respectively).One can only speculate that this has to do with the inclination of diplomats to avoid accusations of bias toward one religion or another.With respect to the rest of the documents, 24.62 percent of all EU documents analyzed here include more critical than positive references to Islam, while in the case of the United States this concerns only 10.29 percent (table 2, first column for EU and United States each).
As can be seen in table 2, EU and US discourses have become less critical of Islam and more protective of Christianity over time.This may be a reflection of the increasing politicization of Islam in international relations, rendering Western actors more cautious in their public utterances about Islam in an attempt to prevent or reconcile conflict.But the frequent attacks on Christians in many parts of the world have also augmented concerns in the West about their security and protection, explaining the uptick in protective references to Christianity in the analyzed documents.
In the United States, however, this is also a reflection of President Bush's more critical stance toward Islam.Foreign policy documents have become less critical under Obama, with just 5.71 percent of documents representing Islam primarily in a critical light, as opposed to 17.14 percent under Bush Jr.Even though Bush tried to counter accusations of being hostile toward Islam ("I hear the language about the United States, that United States is anti-Islam.It's just not true.We view Islam as a religion of peace" [ Bush 2008]), he also coined terms like "Islamo-fascism" (Bush 2006), which, in linking Islam with fascism, has a very negative connotation.
There is quite some variation in the EU's and the United States' notions of Islam according to country, which primarily reflects domestic developments: both, for instance, are most critical of Islam in the documents on Nigeria and the Philippines, where Islamist groups have been responsible for terrorism and violent insurgencies, and more positive in the cases of CAR (with a Muslim minority under attack) and Indonesia (as Muslim democracy).The United States, in addition, is quite critical of Islam in Lebanon, where Hezbollah operates-even more so than the EU-and positive in the case of Tanzania, a relatively peaceful country where a large part of the population is Muslim.This suggests that the domestic security situation has more of an impact on the representation of different faith groups than the general attitude toward a specific group.

Conclusions
Studying foreign policy documents is a promising way to identify and compare discursive structures that enable and constrain foreign policies.The comparison of how religion is represented in EU and US foreign policy documents reveals similarities and differences.On the one hand, the occurrence of religion, despite slight differences, is rather similar, and rather scarce when just considering executive bodies, which are more influential in foreign policy making than-the more attentive (to religion)-legislatures.On the other hand, EU and US foreign policy documents differ profoundly in how they represent religion.While the EU perceives religion primarily as a problem, especially for security and human rights, the United States portrays religion much more often as a resource for security, democratization, and development.Similarly, the document analysis reveals that the EU is more critical toward Islam than the United States.
Without claiming that these differences in the discursive structures constitute the main explanation for foreign policy differences on both sides of the Atlantic, they still help developing a deeper understanding for the question of why it is much more difficult for supporters of religious freedom and engagement policies in the EU to find a sympathetic ear than in the United States.If dominant discourses suggest that religion is a problem rather than a resource, it is much more likely that demands for promoting the freedom of religious communities and for engaging with them are turned down.
These findings, furthermore, allow a more nuanced view on the "religious America-secular Europe" cliché and its effects on foreign policy.Unlike sometimes assumed, transatlantic differences with respect to religion do not determine how much importance EU and US foreign policy authorities attach to religion, but they shape how religion is represented.Thus, the findings challenge the assumption that more prevalence of religiosity in the United States would directly translate into more attention for issues of religion in US foreign policy, although more overtly religious actors can leave their mark in foreign policy documents, as illustrated by the comparison between the Bush and the Obama administration.Yet, the document analysis rather supports observations of a secular bias among Western diplomats similar on both sides of the Atlantic, in the United States further reinforced by the constitutional precept of strict separation of religion and state (Lindsay 2015), which Mandaville and Silvestri (2015, 3) interpret as "major impediment to making sense of religion in the world today and for integrating greater attention to religion in the foreign policy process."At the same time, the findings support reasonings about transatlantic differences, such as those put forward by Hurd (2008), that Downloaded from https://academic.oup.com/fpa/advance-article-abstract/doi/10.1093/fpa/orz013/5475144 by Aston University user on 07 May 2019 imply that US foreign policy is nurtured by accommodationist ideas and EU foreign policy based on separationist notions.Despite the diversity of religion-state relations in EU member states, the representation of religion in EU foreign policy documents indeed reveals separationist ideas: After all, religion is perceived primarily as contributing to conflict and strife if mingling with politics.This implies the idea that it would be better to separate religion from politics if security and human rights are to be protected.With respect to Islam, though, my findings challenge Hurd's (2007Hurd's ( , 2008) ) assertion that the Judeo-Christian basis of American secularism would render foreign policy particularly hostile toward Islam in other countries.They rather support views that American secularism is more accommodative toward all religions (e.g., Kuru 2007), including Islam, which is also reflected in foreign policy.
On a more general note, the foreign policies of secular international actors can be characterized by differing representations of religion.These interpretations are nurtured by domestic sources, such as norms and attitudes deeply held by foreign policy bureaucrats as well as legal and discursive precepts on the appropriate relationship between religion and politics.In addition, as the analysis also shows, attention for issues of religion and the representation of religion in general, and Islam in particular, are further mediated by dynamics within target states.
More research is needed to assess the reliability of the findings of this study.The comparison could be expanded by including, for instance, foreign policy documents on states with additional religious majorities (e.g., Israel, India, or Myanmar), on some usual, religiously less pluralistic, "suspects" (e.g., Iran, Saudi-Arabia, Turkey), or some major powers (e.g.,China or Russia) to corroborate or modify the findings.A comparison of pre-and post-9/11, not possible in this study for practical reasons, would be another valuable addition.The most interesting question for future research, however, would be to explore more in-depth how exactly the identified different secular structures and representations of religion on both sides of the Atlantic shape the implementation of foreign policies by the EU and the United States, and how these structures interact with other factors, such as the strategies of religious lobbies or the individual religiosity of policy-makers.Such an analysis, however, would go beyond studying discourse and include other structural and agency factors to further expand knowledge about the role of religion in American and European foreign policy.

Figure 1 .
Figure1.Proportions of documents (2001Proportions of documents ( -2015) )  predominantly framing religion as a problem or a resource (percent).

Table 1 .
Proportions of documents that include references to religion (percent)

Table 2 .
Proportions of documents with mainly critical or positive/protective references to Islam and Christianity (percent) ("neutral" or "absent" not displayed)