Polarising the Christian West: Us and them in far-right parties’ rhetoric

Abstract

Far-right communication often portrays a division between ‘the people’ and perceived antagonists, potentially fuelling intergroup tensions. A pivotal element is the politicisation of religion: in Christian-shaped regions, far-right parties invoke Christianity as part of Western identity – despite their weak religious ties and growing secularisation. This paradox raises questions about religion’s function in far-right discourse. I argue that far-right parties systematically use a conservative reading of Christianity as a demarcation tool – marking not only Islam but also cultural liberalism as incompatible with a Christian-framed ingroup. Using parliamentary data from Europe, I conduct automated text classification to identify religious markers and their co-occurrence with inclusionary–exclusionary rhetoric across countries and over time. Results reveal group-specific framing, distinct patterns of religious invocation compared to other parties, and links of these patterns to events. This study demonstrates how Christian narratives serve as rhetorical resources for boundary construction, challenging the cohesion of even secularised societies.

Introduction

Far-right messages, often built on nativist and populist elements (Harteveld et al., 2022), portray a division between ‘the people’ and perceived antagonists. This boundary-drawing discourse threatens to fuel intergroup tensions, potentially harming democratic processes and societal cohesion. A pivotal element of far-right communication today is the politicisation of religion. In regions with a Christian heritage, far-right parties frame their discourse in religious terms, claiming that the ‘Christian West’ is under threat and must be defended. This strategy enables parties to portray themselves as saviours of alleged native cultural identity, using religion to mobilise voters and promote their right-wing agenda (Marzouki et al., 2016; Mudde, 2019; Zúquete, 2008). In this context, religion fosters nostalgia for the homogeneous nation in some glorified past: one that is perceived as uncorrupted by elites and ‘others’. The invocation of an allegedly lost ideal seems to legitimise the necessity for a battle against those deemed culturally different and their allies.

While some far-right parties are genuinely religious and others do not refer to religion at all, in Western Europe, the dominant pattern is politicising Christianity as a cultural marker, described as “Christianism” in an identitarian sense (Brubaker 2017). Within this framework, Muslims emerge as a primary negative target group for European far-right parties (Brubaker, 2017; Marzouki et al., 2016; Mudde, 2019). At the same time, religion is also politicised through demarcation from cultural liberals by using religiously-rooted traditionalism to legitimise moral claims (Brubaker, 2017; Hennig, 2024; Mudde and Kaltwasser, 2017). However, how far-right parties benefit from invoking Christian narratives remains underexplored. To advance our understanding of their recourse to religion, it is crucial to examine how Christianity functions as a tool for drawing lines of inclusion and exclusion. This is particularly relevant in largely secularised European contexts, where neither the parties nor their electorates are notably religious (e.g. Arzheimer and Carter, 2009; Marcinkiewicz and Dassonneville, 2022; Minkenberg, 2018). Foundational work by Brubaker (2017) and empirical studies on far-right religious communication in Europe (e.g. Schwörer and Romero-Vidal 2020) have taken important first steps in conceptualising and mapping the politicisation of Christianity by the far right. I advance the literature by analysing how Christianity is used as a marker to construct group boundaries in parliamentary discourse, distinguishing inclusionary references to the ingroup and exclusionary references to outgroups.

This study examines far right parties’ use of Christianity by analysing religious language in their political communication. Specifically, it asks: How do far-right parties refer to Christianity in their political rhetoric? Building on the concept of Christianism, I emphasise a conservative reading of Christianity which functions as a unique marker in far-right parties’ discourse. Since far-right parties aim for a homogeneous nation-state, their primary motive in invoking religious elements is likely the creation of clear ingroup-outgroup boundaries. Therefore, I expect far-right parties to make systematic use of religiously-framed appeals to advance their agenda. While the use of religion is unsurprising for parties with a genuine religious platform, it is particularly notable for non-religious far-right parties. Arguing that these parties use Christianity strategically, I expect to find religiously-framed group mentions in both inclusionary and exclusionary forms in their communication. Here, ‘strategically’ refers to the instrumental use of Christianity without stemming from genuine religious commitment. The incentive that far-right parties have for forming their appeals within a conservative Christian framework lies in that this trope functions not only as barrier against Islam but also against cultural liberalism. Muslims and cultural liberals are portrayed as incompatible with the far-right party’s Christian-framed ingroup. This exclusionary rhetoric strengthens established identity frameworks, such as nationalist frames tied to citizenship, which individuals can adopt or relinquish (Brubaker, 1992), as well as so-called Western values, which regularly conflict with far-right traditionalist views.

Using data from ParlEE V3 (Sylvester and Greene, 2023), V4 (Sylvester et al., 2024), and ParlSpeech V2 (Rauh and Schwalbach, 2020), I analyse speeches by far-right politicians across four Christian-cultural contexts, which provide theoretically meaningful variation in the socio-political environment in which these parties operate today and allow an examination of whether Brubaker’s (2017) patterns hold across cases. Employing zero-shot classification with GPT-4o, I identify religious markers and their co-occurrence with inclusionary-exclusionary rhetoric. Results demonstrate that across countries and over time, far-right discourse features religiously-framed group mentions significantly more often than that of other parties, with exclusionary mentions dominating. This supports the argument that far-right parties strategically use Christian narratives for ingroup-outgroup constructions, highlighting (1) religion’s role in shaping how different groups are addressed, (2) distinct patterns in how far-right parties invoke religiously-framed group mentions compared to other parties, and (3) connections between these patterns and specific topics and events. Mapping the presence and patterns of such mentions also provides a complementary quantitative test of prior work on the politicisation of Christianity in (Western) Europe. More broadly, appeals embedded in a conservative Christian framework may extend beyond discourse. By legitimising exclusionary rhetoric through cultural or moral justifications, they can reinforce group-based antagonism, even among the non-religious, and thereby undermine cohesion and threaten the pluralist foundations of liberal democracy.

Literature review

Far-right ¹ actors are ideologically rooted in nativism, which merges nationalism with xenophobia, and holds ‘that states should be inhabited exclusively by members of the native group (“the nation”) and that non-native elements (persons and ideas) are fundamentally threatening to the homogeneous nation-state’ (Mudde, 2007: 37). A maximal definition adds authoritarianism and, usually, populism. Authoritarianism narrowly implies a rigid law-and-order focus (Mudde, 2007). Populism, in the widely used ideational approach (Hawkins and Kaltwasser, 2018; Kaltwasser and Taggart, 2025), views society as divided between a virtuous ‘people’ and a corrupt ‘elite’ and insists that politics should reflect the people’s general will (Canovan, 1999; Mudde, 2004).

Far-right parties build their political communication on nativist and populist narratives, contrasting ‘natives’ with ‘non-natives’ or ‘the people’ with ‘the elite’ to draw boundaries of inclusion and exclusion (Harteveld et al., 2022). Both populist categories are generally shaped by context, host ideology and political power of respective actors (Mudde and Kaltwasser, 2013); in the European context, the literature underscores the close link between nativism and populism, as nativism helps define ‘the people’ (Brubaker, 2020; De Cleen and Stavrakakis, 2017), while anti-elitism often portrays elites as enabling crime through permissive migration policies, or even pushing conspiratorial claims that ‘progressive’ elites orchestrate migration (Mudde, 2019).

The nativist logic revolves around drawing group boundaries primarily based on cultural traits. In this context, the literature suggests that the politicisation of Christianity appears pivotal to contemporary far-right narratives. In (Western) European countries, far-right parties employ Christianity selectively rather than advancing a substantive religious agenda (Brubaker 2017; Minkenberg 2018: 387). Thus, religion is, primarily, a cultural marker to further define both who belongs to their ingroup and who must be excluded (Marzouki et al., 2016). As such, “Christianism”, comprises facets that are leveraged with varying emphasis by far-right parties depending on political and religious context, following a broadly similar concept across the region (Brubaker 2017).

When it comes to far-right parties’ religious communication, the antagonism between the Christian West and the predominantly Muslim ‘Other’ is central. In Christian-shaped regions, far-right parties portray the Christian heritage as fundamental to Western identity, framing migration as a threat to the alleged native culture (Marzouki et al., 2016; Minkenberg, 2018; Mudde, 2019). Muslims are depicted as a cultural threat, while Christian narratives and symbols serve as identity markers to reject Islam and associated groups as outsiders (Brubaker, 2017; Cremer, 2023; Marzouki et al., 2016). In its ‘antithetical opposition to Islam’ (Brubaker, 2017: 1194; Hennig and Hidalgo, 2021) “Christianism” also intersects with the far right’s paradoxical adoption of liberal rhetoric, including what Brubaker (2017: 1193) calls ‘illiberal invocations of liberalism’, insofar as it demands the selective secularisation and liberalisation of religion itself. Secularism, selectively adopted by the far right, mitigates tensions with the state’s commitment to neutrality by advocating for restrictions on the public visibility of (Muslim) religion, while also serving to delegitimise explicitly Christian arguments in civic or political initiatives, such as advocacy for refugees. Liberal principles, selectively invoked, reduce conflict with the state’s commitment to religious freedom, by characterising ‘our way of life’ as culturally incompatible with the perceived backwardness of Islam (Brubaker, 2017). ² Furthermore, the far right employs the concept of “Judaeo-Christian” roots within its Christianism in a rhetorical attempt to craft an exclusive yet ostensibly tolerant identity that seeks to obscure its antisemitic traits and extremist characteristics, while excluding Muslims (Hafez 2014).

Empirical research shows that far-right parties across Europe emphasise religious references more than other parties (Schwörer and Fernández-García, 2020; Schwörer and Romero-Vidal, 2020). They focus negatively on Islam and Muslims (Rosenberg, 2023; Schwörer and Romero-Vidal, 2020), with this trend increasing over time (Schwörer and Fernández-García, 2020). Manifestos show the most negative evaluations of Islam correspond to the most positive evaluations of Christianity, reflecting their narrative as defenders of the ‘Christian West’ (Schwörer and Romero-Vidal, 2020). Rosenberg (2023) further argues that far-right religious rhetoric responds mainly to Muslim population growth rather than Christian decline.

Beyond opposing Islam, far-right parties also politicise Christianity in contrast to cultural liberalism (Brubaker, 2017; Hennig, 2024; Mudde and Kaltwasser, 2017). ³ Cultural liberalism, understood as ‘stressing the emancipatory power of individual rights’, explicitly extends to guaranteeing rights regarding ethnic, sexual, and gender identity (Laruelle, 2022: 309). Hennig (2024) argues that this demarcation serves as another approach for advancing illiberal politics. Central is traditionalism rooted in religion as a source of moral legitimacy (Marzouki et al., 2016). Far-right parties invoke Christianity in opposition to LGBT+ rights, gender sensitive policies, and abortion. Yet, systematic investigation of far-right rhetoric along this boundary has been far missing from the literature. Notably, Brubaker (2017) and Hennig (2024) argue that Christianity is incorporated to appeal to (more conservative) Christians, presenting it as a ‘substantive religious message’ (2017). At first glance, this claim appears plausible, as far-right parties share common ground with Christian traditionalism with respect to family concepts, gender roles, and reproduction (Akkerman, 2015: 38).

However, the embedding of Christianity in the political communication of far-right parties remains striking for two reasons. First, religiously rooted far-right parties are rare and predominantly observed in Eastern Europe. Unlike Christian Democratic parties, parties such as the Alternative for Germany (AfD) and the Freedom Party of Austria lack roots in Christian principles and do not prioritise religious issues (Minkenberg, 2018: 377). Second, religion is of little importance to their electorate, as survey-based research finds that voter’s religiousness predominantly plays only a minor role in support for far-right parties (e.g. Arzheimer and Carter, 2009; Marcinkiewicz and Dassonneville, 2022). More recently, Arzheimer (2025) finds religiosity and Islamophobia are substantively uncorrelated at the individual level.

In sum, the contrast between limited substantive links to religion and the prominence of religious references raises questions about Christianity’s strategic function in far-right rhetoric. This study addresses this puzzle by examining how Christianity is employed as a discursive tool for drawing lines of inclusion and exclusion in largely secularised contemporary Europe.

Theory

How do far-right parties refer to Christianity in their political rhetoric? I expect far-right parties to make systematic use of religious appeals to create clear ingroup-outgroup boundaries. With theoretical reference to the group construction of social identity (Tajfel, 1982; Tajfel and Turner, 1986), I argue that Christianity, as conservatively construed by these parties, is a unique way to distinguish between ‘us’ and ‘them’ in terms of social categorisation: unique in that it facilitates a distinction from both Islam and cultural liberalism.

In line with this, far-right parties use Christianity as a marker for shaping what can be described as a ‘Christian identity’. This broad, culturally embedded framework facilitates ingroup affiliation, as it does not require formal denomination, active religious practice, or deeply internalised convictions. What makes this particularly effective, I argue, is the accessibility of Christian identity as low-threshold yet powerful tool for fostering identification. By politicising Christianity in this way, far-right parties create an easy point of connection for their ingroup while simultaneously reinforcing exclusionary narratives. Christian traditions, customs, and moral norms provide a rich, premade set of links that far-right parties can use to strengthen group identity. This provides an advantage, as it allows them to mobilise around an existing framework rather than having to construct their own from scratch.

Although European far-right parties are, as mentioned above, not rooted in Christian principles, their reluctance to engage with substantive religious messages may also partly reflect the limited resonance such appeals would have in largely secularised (Western) European societies (Peker 2022). ⁴ Nevertheless, as Roy (2013) notes, Christianity continues to shape cultural identity in these contexts. Whether through intergenerational transmission within families, religious education at school, or festivities, such as public holidays, Christian customs, such as ceremonies, celebrations, and moral values, continue to shape societal life and norms (see also Vlachová and Hamplová, 2023). Moreover, the church remains a societal and moral authority, even in its weakened and declining form. ⁵ As a result, growing up in these societies often involves some degree of Christian socialisation, making the appropriation of Christianity a useful tool for mobilisation.

First, establishing a connection with Christianity elucidates Islam as a foreign religion. By pointing to the Christian heritage of their countries, far-right parties portray Christianity as a cultural characteristic of their nation. They often even link ‘being a Christian’ with ‘belonging to a nation’. In contrast, Muslims in Europe are strongly associated with migration from Muslim-dominated regions. As religion, particularly Islam, is closely linked to ethnicity, it often functions as a socially acceptable proxy for ethnic difference, enabling far-right actors to signal exclusion of ethnic groups without explicitly invoking racist categories. Consequently, hostility based on alleged cultural incompatibility has become widespread, making Islam a particularly easy target and a distinctive opponent to a Christian ingroup.

Second, given its exclusionary nature (see Allport, 1966, on the ambivalent character of religion), the far-right parties’ conservative understanding of Christianity stands in contrast to the precepts of Christian ethics, such as compassion and love thy neighbour, and is far from a tolerant and inclusive set of values. Rather, Christian traditionalism serves as set of norms and values and as source of moral legitimacy (Marzouki et al., 2016) for maintaining the boundary with culturally-liberal principles. Through this understanding, Christianity contrasts with liberalisation efforts for equality and diversity, including the participation of marginalised groups such as migrants or LGBT+ individuals, and women. It also stands in contrast to initiatives and agendas by (more) ‘progressive’ forces, including political measures and legislation. Notably, it partly opposes positions of the churches themselves, most prominently regarding migration, including their engagement in sea rescue and initiatives such as church asylum, as well as their internal processes and reform movements. With this (almost illiberal) rejection of equality and diversity, references to Christianity are reintroduced into political and public debates, even in otherwise secularised societies. The so-called Christian identity positions the ingroup in opposition not only to the ‘otherness’ of Islam and its associated groups but also to culturally-liberal groups, as both are perceived as undermining Christian civilisation and culture. This rhetorical approach is, thus, more selective than its supposed equivalents. It may facilitate specific group oppositions beyond, for example, national frameworks based on citizenship, which can be chosen or renounced individually (Brubaker, 1992), or so-called ‘Western values’, which often conflict with the far right’s portrayal of moral conservatism.

In this way, the intergroup structuring provides a starting point for group-based emotional attachments. The categorisation of who represents the ingroup and who represents the outgroup, enables far-right parties to identify who belongs to their own group and who does not. Accordingly, they can address their ingroup with identity-forming claims that strengthen group boundaries and encourage individuals to assume the group identity. By using religious connotations to address an ingroup that is not necessarily religious or formally affiliated, far-right parties politicise Christianity to differentiate their allegedly Christian ingroup (‘us’) from members of the outgroup (‘them’), regardless of actual affiliation and religiousness itself. I refer to this type of usage as strategic: Christianity is employed instrumentally as a tool for boundary drawing. Far-right parties can pursue this ingroup-outgroup opposition to promote and foster these devaluing patterns by appealing to the tendency toward (group) comparison (Tajfel and Turner, 1986).

Building on this, I anticipate that they use inclusionary and exclusionary religiously-framed group mentions in their communication. First, far-right parties making inclusionary Christian claims aim to strengthen the identification with the ingroup, which may not be as congruent as it is portrayed. This is likely to be particularly true for those who may have initially considered themselves less Christian, regardless of whether they are expressing dislike for the outgroup.

H1

Far-right parties use inclusionary religiously-framed group mentions to a greater extent than competing political parties.

Second, when demarcated from Christian identity, the outgroups are placed in opposition to the ingroup. To this end, religious connotations exclude and potentially devalue the outgroups to varying degrees that range from exclusionary or negative attributions to derogatory language in the context of constructed threat scenarios.

H2

Far-right parties use exclusionary religiously-framed group mentions to a greater extent than competing political parties.

H3a

Far-right parties use exclusionary religiously-framed group mentions to address a Muslim-related outgroup.

H3b

Far-right parties use exclusionary religiously-framed group mentions to address a culturally-liberal outgroup.

Echoing recent research (Rosenberg, 2023; Schwörer and Romero-Vidal, 2020), far-right parties effectively use exclusionary religious rhetoric, as their Christian-labelled ingroup is initially not coherent or stable. It appears easier to unite the ingroup under the ‘Christian label’ when religious markers target an outgroup rather than just claiming themselves as Christians. Thus, I expect far-right religious framing to be predominantly outgroup-focused, serving as a dog whistle for exclusionary policies.

H4

Far-right parties use exclusionary religiously-framed group mentions to a greater extent than inclusionary religiously-framed group mentions.

Based on prior work (Rosenberg, 2023; Schwörer and Fernández-García, 2020), I further expect that how far-right parties use religious language is shaped by both topic-specific and time-related conditions. Periods such as 2015 and its aftermath, marked by migration from Muslim-majority regions, brought migration to the forefront of public discourse (e.g. Eberl et al., 2018; Greussing and Boomgaarden, 2017), potentially making anti-Muslim frames advantageous for far-right parties. Conversely, legislative negotiations on LGBT+ rights or abortion might prompt increased religiously-framed exclusion of cultural liberals via moral conservatism.

Data and methods

To assess religious rhetoric in the communication of far-right parties, I use an LLM for text classification. LLMs efficiently handle complex language tasks using vast pre-trained data. Building on this approach, I aim to further develop the detection of group mentions, following Licht and Sczepanski’s (2024) study on social groups in political rhetoric. To add qualitative insight, I manually examine a random selection of 30% of the detected mentions, stratified by country and year.

Data

Employing data from ParlEE V3 (Sylvester and Greene, 2023), V4 (Sylvester et al., 2024), and ParlSpeech V2 (Rauh and Schwalbach, 2020), I analyse speeches by politicians from far-right and other parties in the national parliaments of four European countries. Parliamentary speech, as a conduit between political actors and citizens, constitutes a valuable source for studying inclusionary-exclusionary rhetoric that invokes religious markers. Unlike party manifestos or press releases, which primarily convey party messaging, parliamentary debates capture how parties perform their positions in a public, interactive, and often confrontational setting. Plenary sessions feature direct exchanges between parties, including critiques of political opponents or societal groups, allowing one to observe processes of symbolic boundary-drawing. Although speeches are formally addressed to fellow parliamentarians, their content is widely disseminated through media coverage. This public visibility (Tresch, 2009) and citizen relevance (Proksch and Slapin, 2012) make debates a central stage for strategic messaging: parties can signal their messages to a broad audience, and voters are likely to receive and interpret these cues. In the context of this study, migration-related debates or legislative negotiations on LGBT+ rights are a likely arena where religious markers may be used to emphasise “us” versus “them”. In contrast to manifestos, which are tied to election cycles (Sagarzazu and Klüver, 2017), plenary debates unfold in response to ongoing events, allowing one to trace how speakers dynamically reference current developments when framing groups in religious terms.

Building on Brubaker’s insight, that far-right parties invoke Christianity in different ways globally, while displaying overarching patterns in the (Western) European context, I select cases to examine whether these patterns hold across Europe’s main Christian traditions: Catholic (Italy), Protestant (Sweden), mixed (Germany), and Orthodox (Greece) ⁶ . Furthermore, these cases present various degrees of secularisation, state-church relations, and party system characteristics, including the presence (or absence) of politically-relevant Christian-democratic parties and varying numbers of far-right parties. Each context features a politically-relevant far-right party (Rooduijn et al., 2024) that politicises Christian heritage despite predominantly non-religious electorates. ⁷ However, the focus is not on denomination per se, but on two aspects: religion’s inherent traditionalism, which underpins moral conservatism, and the distinction from Islam, which is often implicitly contrasted with Christianity. For details on the final dataset, see Appendix A1.

Pre-processing

Each corpus, comprising speeches from all parties in one country, is split into sentences. A single sequence may contain more than one religious marker. Sentences were translated into English using the EasyNMT package. Open-source machine translation models are reliable, reproducible, and cost-effective tools for multilingual text analysis (Licht et al., 2024). Automated translation, thus, offers a pragmatic and robust process (Maier et al., 2021). The accuracy is manually reviewed by a native speaker on a random subsample. Given the large amount of data and the rarity of religious references in party communication, I pre-filter the data using a zero-shot classifier. ⁸ Of 5,093,626 total sentences, 2% contain at least one religious reference, ⁹ which is consistent with earlier findings (e.g. Schwörer and Fernández-García, 2020). While the share is relatively small, the vast overall amount of speech data still allows for a meaningful investigation of far-right parties’ recourse to Christianity. The final dataset includes 190,443 sentences.

Classification task

To conduct the classification task, I employ GPT-4o by OpenAI, which outperforms other LLMs in identifying both explicit and implicit group mentions and their inclusionary or exclusionary framing (see Appendix B1 for model comparison). Choosing a closed proprietary model over open-source alternatives improves performance but reduces transparency and replicability (Barrie et al., 2024) and increases costs. To mitigate these limitations, I assess robustness by reproducing the classification task on a subsample and evaluating consistency (see validation subchapter). The task proceeds in two steps. First, the model extracts text segments that explicitly or implicitly refer to religiously-framed groups or related characteristics. These mentions are defined broadly, as the focus is on religious framing as external attribution rather than on inherently religious groups. Second, the model classifies the type of each mention using the following categories: Christian ingroup mention, Muslim outgroup mention, culturally-liberal outgroup mention, and other. It then outputs whether a statement contains the religiously-framed group mention or not (full prompt in Appendix B2). This reveals which groups far-right parties mention using religious markers, how and how often.

A Christian ingroup mention includes inclusionary references to Christian tradition, symbols, values, faith, church representatives, and communities.

I know that our school has a thousand problems (…), but this battle is a cultural battle in defence of our identity, our values and our Christian principles (05.12.2018, Lega).

A Muslim outgroup mention includes exclusionary references to Muslims and Islamic tradition, symbols, beliefs and practices, entities and communities.

Inner peace cannot withstand any further immigration of this specific kind: no archaic clan structures, no culturally alien and, in part, unlawful guiding principles of Islam, including the threat of terrorism (18.01.2018, AfD).

A culturally-liberal outgroup mention includes Christian-framed references to groups, individuals, or viewpoints that are supposedly not sufficiently committed to Christian culture, its traditional symbols and values. Persons or ideas associated with a culturally-liberal agenda such as ‘progressive’ forces are marked as outsiders using a religious marker.

Greece is a Christian country, it is a country of Orthodox people and it will remain so whether the Prime Minister and the rest of you promote multicultural politics, whether you like it or not (13.11.2019, LAOS).

The fourth category, other, includes religiously-framed group mentions that do not fit the categories above.

Validation

To assess the consistency of the classification task, I calculate Krippendorff’s Alpha as a measure of the model’s reliability over three iterations. With values ranging from 0.91 to 0.97 across categories, the classification shows sufficient consistency, ensuring a stable and reproducible process. In the validation process, I assess the model’s performance against human judgement, the gold standard for textual data analysis (Nelson et al., 2018; Song, 2020). To validate and compare the results, a random sample comprising 1% of the corpus was drawn, with stratification based on country. The selected sentences were manually annotated by the author, following the classification prompt (Appendix B2). Table 1 reports the basic performance metrics.

Table 1.

Performance metrics.

	F1	Precision	Recall
Christian ingroup mention	0.78	0.77	0.79
Muslim outgroup mention	0.83	0.79	0.88
Culturally-liberal outgroup mention	0.67	0.57	0.80

Good recall and moderate precision performances for Christian ingroup mentions and Muslim outgroup mentions yield robust F1 scores of 0.78 and 0.83, respectively. This good performance is likely attributable, in part, to the broad definition of group mentions adopted, which encompasses a wide range of group-related characteristics. By contrast, the F1 score is notably lower for religiously-framed mentions of the culturally-liberal outgroup – a not unexpected outcome, considering the high complexity of this task. In particular, the weak precision indicates that I most likely overestimate the degree to which this outgroup appears in the parties’ discourse. These findings should therefore be considered in light of this limitation.

Qualitative examination

I complement my analysis with a qualitative perspective, aiming to gain deeper insights, capture further rhetorical nuances, and identify patterns linked to specific events. I analyse a subset comprising 30% of all speech segments that contain at least one religious group mention, as classified by the automated task. Each segment is coded by the author according to a predefined scheme combining both deductive and inductive elements (Appendix B3).

Results

Hypothesis 1 predicted a higher level of inclusionary religiously-framed group mentions in the communication of far-right parties compared to other parties. This hypothesis is confirmed. When referring to religion, far-right parties refer to the Christian-framed ingroup in 8.8% of their references, while other parties do so in 6.2% (Figure 1).

Figure 1.

Religiously-framed group mentions by far-right parties compared to other parties when referring to religion in plenary debates, 2009–2019 (or party entry).

Hypothesis 2 predicted a higher level of exclusionary religiously-framed group mentions in the communication of far-right parties compared to other parties. This hypothesis is confirmed: When referring to religion, far-right parties refer to outgroups in exclusionary religiously-framed ways in 17.3% of the references, while other parties do so in 3.8%.

Hypotheses 3a and 3b proposed that exclusionary religiously-framed group mentions by far-right parties address both Muslims and cultural liberals. Muslims are referenced in 16.3% of religious references, providing clear support for H3a. In contrast, mentions of cultural-liberals are extremely rare (0.98%). ¹⁰ Although such references are present in all cases, the prevalence is very low, providing limited support for H3b. The stark contrast with the much higher frequency of Muslim outgroup mentions highlights clear differences in the salience of these outgroups. Thus, this finding necessitates a cautious interpretation, particularly given cross-country differences (Appendix C1).

Hypothesis 4 predicted a higher level of exclusionary religiously-framed group mentions than inclusionary ones. This is supported. Exclusionary mentions (17.3%) occur almost twice as often as inclusionary ones (8.8%). Muslim-related mentions dominate, aligning with previous findings (Rosenberg, 2023; Schwörer and Romero-Vidal, 2020). This gap is widest in Germany, where, along with Sweden, Muslim mentions far exceed Christian ones. Notably, in both countries, religiously-framed mentions towards cultural liberals are not substantive.

Unlike far-right parties, other parties tend to emphasise Christian over Muslim mentions. In Germany and Sweden, where Christian Democrats are politically relevant, CDU/CSU and Kristdemokraterna (KD) focus on the Christian-framed ingroup, while AfD and Sweden Democrats (SD) predominantly refer to the Muslim-related outgroup (see Appendix D1 for religious group mentions by party family, classified following Döring and Manow, 2024). Interestingly, in Greece, where secularisation is comparatively low, religious rhetoric is rare outside far-right discourse.

Qualitative insights

Manual examination clarifies the rhetorical patterns. Across countries, mentions are frequently embedded in a narrative of Christianity under threat and far-right actors as its defenders. Yet, the links to Christianity vary depending on the country, which may partly also reflect party-level differences, in addition to country-level factors. In Germany, inclusionary references emphasise cultural continuity, invoking a ‘Judaeo-Christian foundation’ as the basis of native civilisation. SD links Christianity and national heritage culturally, likely due to the country’s high secularisation: ‘Christian cultural heritage is […] something that all citizens, whether religious or not, can identify with and value.’ (SD, 06.04.2011). In contrast, Greek far-right discourse, especially Golden Dawn (XA) but also Independent Greeks (ANEL), LAOS, and EL, ties the ingroup identity to Greek Orthodoxy, blending religion with concepts of nation, family, and moral traditionalism. To a lesser extent, Italian far-right rhetoric frames the ingroup as the Italian people, portraying them as a Christian-rooted value community that emphasises family and the protection of life.

Muslims, and to a much lesser extent, cultural liberals, are frequently depicted as threats. Boundaries between the Muslim faith and its fundamentalist interpretations often become blurred. In particular, Greek and Italian far-right parties, frequently conflate Muslims with Islamist fundamentalism. In Sweden, additionally, the emphasis on Christian persecution in Muslim-majority countries is used to stoke national fears: ‘When Islamists fire at the Red Cross in combat zones, there is a similarity to our no-go areas, where ambulances are attacked’ (19.05.2016, SD). In Germany, the framing focuses on Muslims’ alleged backwardness and antisemitism as incompatible with the liberal West or Christian culture. Greek and Italian discourses place greater emphasis on religion itself. Noteworthy is the simultaneous emphasis on Christian, institutionally embedded, legacy and on secularism when they accuse ‘new Italians [as] not suited to positively confronting European secular Christian culture’ (13.10.2015, FDI).

Threat narratives regarding cultural liberals often accuse them of being ‘too liberal’, especially when aligned with the left or in national or supranational official positions. They are frequently depicted as failing or unwilling to act, or, in conspiratorial terms, for example, as actively promoting migration. Religiously-framed opposition to supposed cultural liberals frequently combines these claims with anti-elite rhetoric: ‘European identity (…) forged by the Christian roots that the Europe of banks and bureaucrats has denied and disowned’ (19.03.2019, FDI). In Greece and Italy, such discourse occasionally links to moral traditionalism, portraying advocates of inclusive policies as ‘absurd’, ‘ideological’, and as undermining Christian family values: ‘This is a real metaphorical slap in the face to the hypocrisy of political correctness and unwanted cultural submission, and may this slap (…) reach all those who, driven by ideology, (…) seek to educate and raise our children by denying them respect for our values, our traditions and our Christian principles’ (05.12.2018, Lega). Debates on LGBT+ rights, such as same-sex partnerships and transgender rights, feature similar rhetoric: ‘I do not want and will not contribute with my vote to anything that harms Christian principles and attempts to alter the form of the family, which was and is dad, mum, and children and not parent number 1 and parent number 2’ (ANEL, 08.05.2018). In Greece, exclusionary rhetoric towards cultural liberals also targets secular reforms, including religious education, criticising ‘progressive’ actors and initiatives.

Insights from analysis over time

Figure 2 shows spikes in religiously framed group mentions during specific periods and distinguishes between far-right parties and other parties, showing trends across countries; country-specific trends are reported in Appendix D2. Early peaks in far-right religious references reflect far-right discourse in Sweden and Italy. Complementing these quantitative trends with qualitative analysis reveals they often correspond to particular debates or events, although patterns vary slightly by country. Christian mentions peak around 2011. In Italy, higher levels (see Appendix D2) follow the 2010 attack on a Coptic church, with ‘progressive’ Christianity used to justify restrictive migration policies. Similarly in Sweden, which accepted many Syrian refugees from 2011 and granted permanent residency in 2013, SD, newly in parliament, uses inclusionary Christian rhetoric to pressure the government: ‘If we do not do this now, I fear that in a few years' time we will be standing here discussing a ban on Lucia processions (…) or the removal of the yellow cross from the Swedish flag’ (06.04.2011).

Figure 2.

Religiously-framed group mentions by far-right parties compared to other parties when referring to religion in plenary debates over time. Note: The figure distinguishes between far-right parties and other parties, showing trends across countries. The far-right line reflects the discourse of far-right parties in countries where they are represented at the given time.

Muslim outgroup mentions also peak in Sweden that time. In Sweden (2010) and Germany (2017), far-right parliamentary entry coincides with surges in anti-Muslim rhetoric (see Appendix D2), mainly linked to migration debates, as the qualitative examination shows. In Sweden, Christian and Muslim mentions move in opposite directions over time: between 2010 and 2013, Christian mentions increase while Muslim mentions decline; after 2013, Muslim mentions rise as Christian mentions become less frequent. In Greece, Christian mentions slightly increase in 2015 and 2016 (Figure D2), linked to reforms proposing a value-neutral subject instead of religious education, reflecting broader state-church tensions. In 2018, as part of the new government, Italy’s Lega contrasts its vision of a Christian Europe with the EU, opposing supranationalism, cultural liberalism, and secularism: ‘It is no coincidence that for this EU, everything that defines us is a problem, everything that makes up our identity. We are not allowed to call ourselves Italian, […] Christian, […] man or woman, […] mother or father.’ Notably, Italian far-right references to the Pope decline after 2013, likely due to Francis’ comparatively more liberal stance on migration compared to Benedict XVI.

Nevertheless, broader patterns emerge across countries. Muslim outgroup mentions rise around 2016 (Figure 2), increasing in Italy from 2014 (Figure D2). As the qualitative analysis demonstrates, this rise often links to Daesh’s caliphate declaration and increasing refugee arrivals via the Mediterranean – turning (Muslim) migration into a key issue for Lega under Salvini. The subsequent upward trend (Figure 2), centers around Islamist attacks in Europe, such as the Paris attacks in January and November 2015 and the perception that they represent a dual threat: terrorism and migration. This pattern appears across countries, although less pronounced in Greece, where mentions relate more to migration, Islam’s presence, and tensions with Turkey. Notably, the 2015 migration does not evoke a second peak in Muslim mentions across countries.

Exclusionary mentions of cultural liberals are rare (Figure 2) and often tied to national debates, explaining the absence of broader trends. Greek and Italian far-right parties mention them slightly more often (Figure D2). In Greece, transgender rights legislation persists as a topic throughout the period, while Italy’s 2013 spike links to LGBT + rights debates. Similar claims appear in 2016 amid the recognition of same-sex partnerships, which, for instance, are framed as violating ‘the sanctity of marriage’. Furthermore, Christian holidays are used to frame ingroup and outgroup claims, portraying traditions and customs as threatened.

Conclusion

This study analysed how far-right parties use religious markers in their parliamentary discourse, combining automated text classification and manual examination. Extending existing literature, I argued that far-right parties systematically invoke Christianity, yet do so through a conservative lens as a distinct marker. Comparing their rhetoric to that of other parties in parliamentary debates, results show that far-right parties use both inclusionary (H1) and exclusionary (H2) group mentions framed in religious terms more frequently than competitors. Beyond a Muslim-related outgroup (H3a), they mention a culturally-liberal outgroup very rarely (H3b), favouring exclusionary over inclusionary rhetoric (H4).

By systematically mapping religiously-framed group mentions across countries and over time, this study provides a complementary test to existing work on far-right parties’ politicisation of religion. It uncovers the instrumental use of religious framing, empirically demonstrating how it shapes references to different groups, how these patterns differ from other parties’ discourse, and how it is associated with particular topics and events. However, given the very low share of religiously-framed references to culturally-liberal outgroups, it remains unclear whether voters perceive these narratives as relevant in parliamentary debates. Such mentions could be more prominent in other arenas, such as transnational forums, campaigns, or far-right media, rather than in national parliaments.

While my analysis provides important insights into far-right religious rhetoric, it has limitations that future research should address. First, the topics in which group mentions appear and their evolution over time, warrant further systematic investigation. Second, since Muslim mentions are often linked to migration, it would be valuable to assess whether their frequency shifts when migration is less prevalent in plenary debates. Third, including other sources, such as campaign materials or social media, would enhance understanding of far-right religious rhetoric, since parliamentary speech is shaped by institutional mechanisms that constrain participation. Fourth, although the cases capture relevant variation, a larger sample would yield more robust evidence on general trends. Finally, while the findings support the idea that far-right parties strategically use religious appeals to construct opposing groups, the implications of this strategy require further investigation beyond what text-as-data approaches can capture.

Overall, this study contributes new insights into the role of religion in far-right communication, showing how recourse to Christianity shapes ingroup-outgroup construction. Methodologically, it adds to the detection of group mentions using LLMs. More broadly, religiously-framed appeals may have consequences beyond rhetoric, potentially strengthening social boundaries even among non-religious audiences. By casting exclusionary narratives in cultural or moral terms, such rhetoric can heighten ingroup attachment and outgroup hostility. The strategic use of religious framing to emphasise group distinctions could, thus, foster affective divides with implications for cohesion and the resilience of liberal societies.

Supplemental material

Supplemental material - Polarising the christian west: Us and them in far-right parties’ rhetoric

Supplemental material for Polarising the christian west: Us and them in far-right parties’ rhetoric by Lucienne Engelhardt in Party Politics.