Approaching right-wing populism in the context of transnational economic integration: lessons from Luxembourg

ABSTRACT What does right-wing populism look like in Luxembourg, a small European state whose economy heavily relies on strong European integration and foreign and cross-border labor? This article explores this question by looking at the Alternativ Demokratesh Reformpartei (ADR), Luxembourg’s version of a right-wing populist party. It studies the ADR’s discourse during three discursive events: the debate over a burka law (2014–2018), the 2015 Refugee Crisis, and Luxembourg’s 2015 constitutional referendum. Much of the ADR’s discourse is right-wing populist, but with two specificities: the ADR’s embrace of multiculturalism and its use of the issue of the national language to appeal to cultural and economic interests. Both of these specificities reflect into a distinctive concept of the Luxembourgish ‘people’. The Luxembourg case demonstrates the chameleonic nature or populism, or its ability to adapt to its local circumstances, in this instance even adopting features that contradict much of the thinking on right-wing populism.


Introduction
There is an increasing volume of academic literature dedicated to right-wing populist parties in Europe. Much of this literature focuses on right-wing populism in large or mid-sized European countries or in large regions contained within larger nation-states (Akkerman et al., 2017;Kaltwasser et al., 2017;Mudde, 2007;2017b;Mudde & Kaltwasser, 2012;Rydgren, 2007;Wodak et al., 2013;see however De Jonge, 2021). The case of the small, internationally and economically integrated country of Luxembourg, however, is interesting because its socio-economic structure speaks directly to many of the themes that right-wing populists tap into including globalization, mass migration, and the European Union. In contradistinction to other European countries, Luxembourg is often referred to as an immigration success story (Fetzer, 2011), has one of the highest GDP's per capita in the world, and its working class is composed largely of foreign labor while many Luxembourg nationals hold secure, high-paying, public sector jobs (De Jonge, 2021). Because of populism's chameleonic nature (Taggart, 2000;2004), which sees it adopting characteristics particular to its context, it makes sense to ask what does rightwing populism in Luxembourg look like?
The party in Luxemburg with the strongest right-wing populist profile and sustained electoral success is the Alternativ Demokratesch Reformspartei (ADR, Alternative Democratic Reform Party). The party formed in 1987 as the Aktiounskomitee 5/6 Pensioun fir jiddfereen (Action Committee 5/6th -Pensions for Everyone) to campaign on the single-issue of pension reform with the goal of harmonizing private and public sector pension programs; at the time, the public sector pension system was more generous. In 1992 the party changed its name to Aktiounskomitee fir Demokratie a Rentegerechtegkeet (ADR, Action Committee for Democracy and Pension Justice). In 1999 the party received its greatest success with 11.3 per cent of the vote in the general election, although pension reforms in 1998 and 2002, which the ADR strongly influenced despite not being a member of a governing coalition, made the single-issue of pension reform less pressing (Schulze, 2006). The ADR then evolved into a broad issue, established party, adopting its current name in 2006. In the most recent general election in 2018, the party received 8.28 per cent of the vote and gained 4 seats in the 60 seat national parliament.
There is no consensus as to whether the ADR is a right-wing populist party, with some saying yes (Blau, 2005;Zulianello, 2020) 1 and some saying no (Camus, 2017;Poirier, 2012). As Léonie De Jonge (2021) notes, and as this article will discuss, such confusion stems from the fact that a right-wing populist party conforming to conventional understandings of the phenomenon is likely electorally impossible in Luxembourg, leading the ADR to adopt softer stances on important issues compared to other parties, such as the Rassemblement National or the Freiheitliche Partei Österreichs. Nevertheless, it shares many of the same positions and preoccupations as other right-wing populist parties in Europe and is associated with the European parliamentary group European Conservatives and Reformists (ECR), which includes the Law and Justice Party (PiS) from Poland and the Swedish Democrats, two of the more well-known right-wing populist parties in Europe (ADR, 2019). These facts, combined with the lack of a suitable alternative leads to the conclusion that exploring right-wing populism in Luxembourg requires a look at the ADR, a conclusion that De Jonge (2021, p. 11) also makes. Thus, this article focuses on the ADR, discussing the party's particularities as they relate to populism's chameleonic nature.
Recent history has afforded the ADR the opportunity to pronounce itself on topics likely to produce right-wing populist discourse: in 2014, the party began efforts to ban the burka in Luxembourg, in 2015 there was the Refugee Crisis in Europe, and also in 2015, there was a referendum on amending the Luxembourgish constitution to allow foreign residents to vote in national elections. These events touch on core themes related to right-wing populism: Islam, multiculturalism, immigration, European integration and the European Union, national sovereignty, etc. Exploring the discourse, the ADR produced on these topics will thus help assess what right-wing populism in Luxembourg looks like.
To answer my research question, I will discuss the concept of right-wing populism and identify its defining characteristics: a vertical exclusion between an inauthentic 'elite' and a pure 'people' combined with a horizontal exclusion between the 'people' and hostile, foreign entities. This definition lays the foundation for the ideational approach to studying right-wing populism outlined by Cas Mudde, which sees populism as a thin-ideology demarcated principally by the aforementioned vertical and horizontal exclusions. I will use this approach and code the ADR's discourse related to the three discursive events mentioned above in MAXQDA. Doing so will enable me to determine how the ADR's discourse fits within the parameters of right-wing populism. I find that much of the ADR's discourse reflects that of other right-wing populist parties, but that there are two unique features: the ADR's recognition and embrace of multiculturalism and the way it uses the issue of the native language, Luxembourgish, to appeal to the economic interests of the electorate. Both of these specificities reflect into a concept of the 'people' that is partly at odds with typical understandings of right-wing populism, and I will discuss them at length.

Literature review
Populism is well-known as a difficult concept to define. There is nevertheless a general consensus that populism, in its essence, retains an opposition between the people and an elite (Taggart, 2000). A minimalist definition along these lines comes from Cas Mudde (2017a, p. 29): populism is an ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, 'the pure people' versus 'the corrupt elite', and which argues that politics should be an expression of the volonté générale (general will) of the people. This definition has the advantage of being applicable to a wide array of populist manifestations, while avoiding the vagueness or limitations that characterize other definitions. Thus, when identifying a populist party or movement, one will look for the trademark distinction between the elites and the people.
Populism exists in both right-wing and left-wing variants, although only the right-wing variety is of interest in this article. A number of authors argue that the salient aspect of right-wing populism is its nationalist dimension (Mudde, 2007, pp. 16-31;Rydgren, 2007;Taguieff, 2015, pp. 26-27;De Cleen, 2017). In this sense, in addition to its stance towards the elite, right-wing populism also contains an 'in/out' exclusion marked by nativism, in which the state 'should be inhabited exclusively by members of the native group ("the nation") and […] non-native elements (persons and ideas) are fundamentally threatening to the homogeneous nation-state' (Mudde, 2007, p. 19). In this sense, right-wing populism contains two exclusions with regards to the 'people': a vertical exclusion towards a multicultural elite and a horizontal exclusion towards foreign entities.
Stemming from their positions regarding the nation, the main themes of right-wing populist parties include national sovereignty, the preservation of the dominant group's culture, and opposition to mass migration and multiculturalism. Right-wing populist parties set themselves in opposition to immigrants they see as incompatible with the national culture (in particular Muslims), national and international multicultural elites who they see as undermining the interests of the nation, and the European Union, which reduces national sovereignty (Akkerman et al., 2017;De Cleen, 2017;Mudde, 2007;Richardson, 2020;Rydgren, 2007;. Two of the most prominent recent markers of national culture that are used as grounds of exclusion are language and religion. Requiring the acquisition of language or the banning of certain religious symbols are part of an effort to underline incompatibility and legitimate measures of migration control (Rheindorf & Wodak, 2020). Right-wing populists will also often speak of the popular will and call for direct democracy or referendums as a way to side-step the goals of elite politicians. Often, however, these moves are aimed at curbing constitutional limits on political power that guarantee constitutional freedoms and minority rights in the pursuit of a nationalist agenda (Mudde, 2007, pp. 155-156;De Jonge, 2016). Right-wing populists will also often complain of a 'political correctness' that brands opponents as xenophobes and racists as a way of silencing the popular will and imposing a multicultural agenda (Betz & Johnson, 2017).
Despite their attachment to national culture and sovereignty, right-wing populists also increasingly work across borders to create what Benjamin De Cleen (2017) calls metapopulism, or populists in different contexts that work together against common international elites, foreign entities, and similar national elites. One form such meta-populism takes is claiming the need to defend Western civilization, a position Rogers Brubaker has termed 'civilizationalism'. This idea sees right-wing populists defending a 'liberal' Europemarked by Enlightenment values, free speech, and rights for marginalized groups such as homosexuals and womenagainst an 'illiberal' Islam (Brubaker, 2017). Benjamin Moffitt (2017) has labeled such positions 'liberal illiberalism', and sees the positioning as a way for right-wing populists to strategically hide their true, electorally infeasible nature, and as an opportunistic way to 'honorably' attack one's enemies.
There are several approaches to studying populism. There is the 'discursive approach' associated with Ernesto Laclau (1977) in which the 'people' is an empty signifier with no particular content that populists use to boost their support. The political-strategic approach concentrates on the relationship between a charismatic leader and the masses (Weyland, 2017). The socio-cultural approach looks at how populist politics flaunts low-brow culture in opposition to high-brow, 'elitist' manners of speaking and acting (Ostiguy, 2017, p. 73). Lastly, the ideational approach views populism as a ''thin' or 'thin-cantered' ideology' that has a limited scope of conceptualization compared to full ideologies such as socialism or liberalism (Mudde, 2017a, p. 30). Populism as such identifies a division between the 'people' and the 'elites' and argues that policies be made in line with the will of the people, while not articulating any specific views on policy issues. In this sense, populism can be said to have general characteristics, but at the same time can also be described as 'chameleonic' in the sense that populist movements, including right-wing variants, take on secondary characteristics tied to their specific contexts, making no two populist movements the same (Taggart, 2000;2004).

Methodology
This article will adopt Mudde's ideational approach to study the ADR's discourse. I am thus looking for the vertical and horizontal boundaries that characterize right-wing populism. In order to find these boundaries, I will examine discourse from the ADR related to three political events in the recent past where such boundaries were likely to become visible: the constitutional referendum of 2015, the 2015 Refugee Crisis, and debates about banning the burka in Luxembourg, which began in 2014. I will complement this analysis by examining the ADR's party programs for the 2018 national general elections and the 2019 European parliamentary elections where the ADR rearticulates relevant positions.
These events take place on both European and national scales, allowing for a wider contextualization of the ADR as well as a look at the specificity of the Luxembourgish context. To better understand the discursive events in question, it is necessary to briefly provide their context.
On 25 February 2015, the Luxembourgish Parliament passed a law declaring a referendum for changing the constitution to take place on 7 June 2015. There were three proposed changes: lowering the voting age to 16, setting a term limit of 10 years for those serving as elected officials, and granting the right to vote in national election to foreigners residing in Luxembourg for more than 10 years and having previously participated in a communal or European election. The referendum was put forth by the governing coalition, composed of the Green party, the Luxembourg Socialist Workers' Party (LSAP), and the liberal Democratic Party (DP), referred to in media and by other parties as the 'Gambia' coalition because the parties' colors (green, red, blue) are found in that country's flag. During the campaign, virtually all elected parties, members of the business community, labor unions, and newspapers supported the extension of voting rights to foreign residents. The powerful labor union of public functionaries, the Confédération générale de la fonction publique (CGFP), was the only labor union to oppose the measure. The center-right Christian Social People's Party (CSV), historically the largest party in Luxembourg, also opposed the measure, but did so somewhat equivocally, calling on voters primarily to make an informed choice. The ADR was the only elected party to campaign for an unequivocal 'No' to all three ballot initiatives (De Jonge, 2016;Kies, 2019). In the end, 78 per cent of Luxembourgers voted against extending the vote to foreign residents, while 81 per cent rejected lowering the voting age, and 70 per cent rejected limiting mandates to 10 years.
The 2015 Refugee Crisis was a European-wide event that saw millions of migrants coming to Europe to flee war and poverty. The migrant flow hit a peak in August and September 2015, prompting German Chancellor Angela Merkel to declare on September 4 that it would open its borders and allow the migrants free entry into Germany. Countries such as Germany and Sweden subsequently took in high rates of migrants. The EU had floated the idea of mandatory refugee resettlement for member countries, but countries such as Poland and Hungary, who refused to take refugees throughout the crisis, successfully opposed the motion. In 2015 Luxembourg received 2447 asylum applications, a significant increase from 2014 when there were 1091 applications (Direction de l'immigration, 2017, p. 2). The ADR was the only established party in Luxembourg to raise questions and criticize the handling of the Refugee Crisis.
In 2011 France and Belgium passed laws making it illegal to wear face veils in public spaces. The laws effectively banned the burka, the full-body covering seen in Muslim societies, and the nikab, in which only the eyes are visible. In July 2014, the ADR proposed a similar law in Luxembourg, presented by the party and referred to in the general public as the 'burka law'. The law had little support beyond the ADR, and no action was taken through 2015. However, in late 2015 during the Refugee Crisis, the CSV introduced their own bill that would also ban the burka in public places. The law did not have support beyond the CSV and the ADR and there was no movement on the issue until 26 April 2018. On that date, the Luxembourgish parliament passed a law sponsored by the ruling 'Gambia' coalition banning clothing that conceals the face (i.e. the burka) in certain public spaces, such as schools, hospitals, or courts. The law passed 32-28, with opposition coming from the CSV and ADR, both of whom advocated for the banning of face veiling in all public spaces (Schumacher, 2018).
For analysis, I selected the texts that the ADR promoted on its website, which does not constitute the totality of produced discourse but, due to the process of self-selection, can be considered representative. The texts include parliamentary speeches, party press releases, open letters, blog entries, campaign flyers, and other assorted party pamphlets, as well as the party programs for the 2018 Luxembourg general election and the 2019 European Parliament election. In total, there were 26 documents (9-Refugee Crisis, 9-referendum, 8-burka debate) ranging in length, resulting in 296 coded segments. I read the texts in their original language of Luxembourgish. I entered the texts in MAXQDA and coded along the vertical and horizontal axes mentioned above, looking for any mentions, both positive and negative, of the people, elites, and foreign entities.
My discourse analysis is thus corpus based, taking into account a group of texts produced by an institutional actor related to particular discursive events. The units of analysis are not single words, but segments of the textranging from a few words to a sequence of sentencesthat express distinct and complete ideas related to specific topics. By focusing on these segments within a corpus of texts, I am able to identify an ideological structure 'beyond the sentence' (Upton & Cohen, 2009, p. 587). In this instance, using a corpusbased approach with segment analysis will allow me to identify the presence of a rightwing populist ideological structure. My discourse analysis is thus qualitative in nature, with the coding statistics helping to summarize and provide an overall feel of the texts. Using Mudde's ideational approach, I will be able to concentrate on the core concepts employed in right-wing populist discourse and identify the ADR's ideological structure in these regards with the goal of identifying how this ideological structure adapts to local conditions. Accordingly, I coded the segments in four paired categories: an inclusive vs. exclusionary/nativist notion of the 'people' and a positive vs. negative/antagonistic notion of the 'elites'. I coded as inclusive statements that refer to the 'people' as welcoming to foreign entities, and coded as exclusionary/nativist statements that indicated the presence of a closed or bounded community marked by elements of nativism; I coded as positive statements that refer to 'elites' in a positive way, and as negative those that portray 'elites' in a negative light or an antagonistic position with the 'people'. Within this framework, I coded the texts inductively and applied a code when the segment referenced a particular actor.
In making these distinctions, one must take into consideration that in a right-wing populist context, the notion of the 'people' has an ambiguous stance and can refer to a wide range of actors (Biancalana & Mazzoleni, 2020). Of note, like its German counterpart Volk, the Luxembourgish word for the 'people', Vollek, can refer to either an ethnos, a bounded ethnocultural nation that shares a common history, language, and culture, or demos, the group of citizens that constitute a democratic society (ADR, 2015a). Another Luxembourgish term for 'people' is Leit, but this term is devoid of ethnic or political significance and can refer to 'people' as a general abstract category. Noting these distinctions, I created the following codes as a way to identify a reference to the 'people': Leit, society (Gesellschaft, or the idea of living together), members of the Luxembourgish nation (Luxembourgers), popular will, citizens (Bierger), Vollek, language, and the nation/nation-state. Important to note is that language has an ambiguous status as both the means of integration for newcomers, but also as grounds for the exclusion of foreign elements (Rheindorf & Wodak, 2020); accordingly, I often coded language as simultaneously inclusive and exclusionary depending on the context. In line with the ADR's meta-populism, I also include references to other European right-wing populist parties and 'Western Civilization' as a code for the 'people'. As other indicators of a boundary with the 'people' I also included in my codes 'immigrants' (those in the country legally and not seeking asylum), 'refugees/migrants' (those seeking asylum), and 'Islam'.
For elites, I identified actors on three levels: national, European, and international. The national level includes the Gambia coalition, the CSV, a 'vague' elite where specific actors are not mentioned, the media, and an assortment of 'miscellaneous' actors (churches, labor unions, and other actors). The European level includes the European Union, Germany, mainstream politicians in France, Belgium, and Austria, the media, and an unnamed 'vague' elite. The international level includes the United Nations and 'vague' elites.

Results
To achieve a higher degree of nuance in the discourse, I have analysed each discursive event separately. Doing so allows for a more precise identification of actors and the evaluations the ADR attributed to them.

The 2015 referendum
The discourse during the 2015 referendum was marked by a clear vertical antagonism, with 25 coded negative segments -23 of national elites and 2 of European elitesand no positive mentions of elites. The ADR criticized the Luxembourgish media (1 segment) and 'miscellaneous' civil institutions (churches, labor unions etc.) (1 segment) for supporting the measure and warned that supporting the measure would lead to more power for the EU (2 segments). Four segments referred to 'vague' elites present in Luxembourg's ambient discourse, who, for example, compared a 'No' position with South-African Apartheid (Kirsch, 2015). Seventeen of the coded segments referred to the ruling Gambia coalition, which the ADR criticized among other things for denouncing its opponents as xenophobic or disregarding constitutional procedures in implementing the referendum and of pursuing the changes as a way to guarantee their own electoral relevance in the future (ADR, 2015a, p. 2015). 2 In his parliamentary speech after the referendum, Gast Gibéryen also accused the LSAP of being part of a condescending and out of touch socio-economic 'elite'. He called on the coalition to resign and stated: But power-hungry as this government is, it sticks to its guns! The trust of the citizens in them has eroded! From June 7th on, it will no longer be the government for the people (Vollek), but the government against the people (Vollek)!' (Gibéryen, 2015, p. 568) 3 Such statements are the sine qua non of populist rhetoric and demonstrate a clear vertical opposition between the authentic 'people' and an out of touch elite (Mudde, 2017a). The ADR also routinely engaged in rhetoric that evoked the political correctness of their opponents and that resulted in unwarranted accusations of racism and xenophobia.
The ADR's conception of people was more nuanced: 57 coded segments were exclusionary/nativist and 29 coded segments were inclusive. The ADR recognized Luxembourg as a nation (15 segments), or Vollek (2 segments), with a distinct history, values, culture, and language that expresses its political will through the state (ADR, 2015a, p. 8). Extending the right to vote to foreign residents (immigrants, 27 segments) would threaten the Luxembourgish nation in different ways: foreign residents would represent their respective national interests over those of citizens (ADR, 2015a, p. 11); different religious communities (i.e. Islam, 2 segments) would form communitarian voting blocks and make demands contradictory to Luxembourgish culture (ADR, 2015a, p. 15); Luxembourgers (3 segments) would risk becoming a minority voting block in their own country (ADR, 2015a, p. 10). This last point plays on the theme of demographics and national sovereignty, but unlike in other contexts, where the demographic threat comes from Islam (Turner, 2019), the threatening population in Luxembourg is mostly European in origin (Statec, 2020a). Language (8 segments) was also an important theme for both economic and cultural reasons: foreigners would eventually be allowed to enter the public service sector without learning Luxembourgish (ADR, 2015a, p. 15), and reflecting language as a cultural barrier (Rheindorf & Wodak, 2020), those who do not learn Luxembourgish, do not celebrate national holidays, or know Luxembourgish history remain foreigners (Auslänner) and should not vote (Kirsch, 2015).
However, the ADR was also careful to portray the Luxembourgish nation (13 segments) as welcoming to immigrants (12 segments) who made positive contributions and could be integrated, especially through language (3 segments). In an election pamphlet, for example, Jeff Engelen (who became a parliament deputy in 2018) explained his decision to become a Luxembourg citizen and is identified as a 'born Dutchman' (gebiertegen Hollänner) (ADR, 2015b). The ADR also celebrated the fact that Luxembourg was host to so many immigrants without the presence of racist and xenophobic politics (ADR, 2015b). It also referenced the successful integration of successive waves of immigrants beginning in the nineteenth century and continuing through the twentieth and into the twenty-first, including Italians, Portuguese, Yugoslavians (refugees, 1 segment), and most recently from Eastern Europe. What is more, the ADR recognized Luxembourg as a multicultural society, noted that 45% of residents in Luxembourg were foreign and that the economy had over 160,000 cross-border workers, and celebrated Luxembourg's remarkable tolerance of foreigners (ADR, 2015a, p. 16). Such statements are surprising from a party within the right-wing populist family. However, it is worth noting that most of the immigrants the ADR viewed positively were European and those that integrated into the country did so by, among other things, learning Luxembourgish, while Islam, as a foreign communitarian religious group, remains as a subtly referenced outsider that threatens the Luxembourgish community.

The 2015 Refugee Crisis
During the Refugee Crisis, there were 42 coded segments critical of elites at all levels, and 4 positively coded segments for elites at European and international levels. The positive segments reaffirmed the ADR's general support for the free movement of people within Europe (Schengen Accords) and expressed support for the United Nations' efforts at aiding refugees. Otherwise, the ADR was critical of Germany and Angela Merkel (4 segments) and of the EU (9 segments): they held the EU responsible for uncontrolled immigration, rejected refugee quotas, and supported the idea of national sovereignty. The ADR was also critical of the UN (4 segments) and in particular, rejected the UN Migration Pact (ADR, 2019). There were also references to vague elites at all levels (7 segments: 3 national, 3 European, 1 international) and the media on a European level (1 segment): these entities pursued an agenda of mass migration/open borders or thwarted open discussions of the effects of the Refugee Crisis. Domestically the ADR was critical of the ruling DP-LSAP-Green coalition (17 segments): the coalition mismanaged the crisis, and its political correctness was both a threat to free speech and prevented it from recognizing and adequately dealing with the security threat posed by radical Islam. Playing on the themes of border control and political correctness, the ADR's comments place it within a meta-populism that has common international enemies (Merkel, the EU, the UN) and opposes similar national elites engaged for example in politically correct politics (De Cleen, 2017).
There were 56 coded segments for an exclusionary/nativist notion of people and 21 coded segments for an inclusive notion of the people. The vast majority of the inclusive segments noted that Luxembourg (nation-state, 7 segments) had a moral responsibility to help refugees and that refugees (9 segments) had a right to seek asylum. The ADR also thanked refugee helpers (1 segment) and suggested giving special priority to persecuted Christians and Yezidis (1 segment) (ADR, 2015c, p. 3). Language (2 segments) was again an important means of integration. In a parliamentary speech Fernand Kartheiser made it clear that he wanted migrants that stayed in Luxembourg to be full participants in society (1 segment), stating: 'We don't want any parallel societies! […] we don't want to exclude, we want to hear them participate, and that means through the Luxembourgish language!' (Kartheiser, 2015a, p. 10) 4 At the same time, the ADR also presented refugees and other migrants as problematic (28 segments): the ADR often spoke against economic migrants posing as refugees, and spoke of 'asylum-tourism' (Kartheiser, 2015a, p. 9) or 'asylum shopping' (ADR, 2015c, p. 5); they warned against the threat and consequences of continued mass migration; they evoked the cultural incompatibility of migrants; by prioritizing Christian and Yezidi refugees they implied that Muslims were the true persecutors and not deserving of asylum; and in the wake of the terror attack on the Bataclan in Paris on 13 November 2015, Fernand Kartheiser evoked the security threat posed by a 'radicalized minority' (radikaliséierter Minoritéit), that is, Islam (7 segments), stating that he did not want a 'Molenbeek à la luxembourgeoise' (Kartheiser, 2015b, p. 95). Further engaging in civilizationalism and meta-populism, the ADR also defended other right-wing populist parties (4 segments), particularly those in East Europe who opposed the EU's handling of the crisis. It also defended the nation-state's (6 segments) sovereign ability to decide immigration policy, in particular in line with the integration capacity of the (ideally Christian) refugees (ADR, 2015c, p. 3). The remaining coded segments include Society (2), Leit (1), Popular Will (2), and Citizens (3). These codes played on similar themes, such as the limits of Luxembourgish society's integration ability or the importance for people (Leit) to speak freely without being accused of populism. Language (3 segments) was also mentioned as a requirement for integration, underlining its potential as grounds of inclusion and exclusion.

Debate over the burka law
During discussions over the burka law, there were 26 negatively coded segments, all of national elites, and 7 positive segments on a European level only. The positive segments came in reference to politicians in France, Belgium, and Austria (4 segments), who passed laws against the burka, and to the European Court of Human Rights (3 segments), who ruled favorably on the French and Belgian initiatives. The ADR attacked the Gambia coalition (15 segments) mostly for not tabling its 2014 burka bill for parliamentary debate. They accused the ruling coalition of political correctness and denying the democratic process, or the popular will. Of note, the ADR accused the Green and LSAP parties of hypocrisy, since these parties are typically known for their defence of women's rights (ADR, 2015d). Fernand Kartheiser made the same point about self-proclaimed feminist groups (Miscellaneous, 1 segment) who opposed the burka law (Kartheiser, 2018). Such comments are emblematic of right-wing populists' liberal illiberalism (Moffitt, 2017). The ADR also criticized the CSV (10 segments) for tabling essentially the same bill the ADR proposed in 2014, except doing so during Refugee Crisis. The ADR also engaged in strongly populist rhetoric, and after the burka issue was again ignored in November 2015, and likely also in reference to the referendum and in the context of the Refugee Crisis, the party accused the ruling coalition of a 'culture war [Kulturkrich] […] against its own people [Vollek]!' (ADR, 2015e). This last comment echoes the one Gast Gibéryen made after the referendum vote, except this time with an explicit mention of culture. The statement thus contains both the vertical and horizontal exclusions that typify right-wing populism (De Cleen, 2017;Mudde, 2017a;Rydgren, 2017).
For the ADR's concept of people, 23 coded segments were exclusionary/nativist and 3 were inclusive. In its discourse, the party spoke about the cultural incompatibility between Islam (11 segments) or (Islamic) immigrants (1 segment) and the Luxembourgish nation (1 segment) and society (3 segments), and referred to the will of the people through the concept of Vollek (2 segments), citizens (1 segment) and popular will (1 segment). A statement by Fernand Kartheiser during parliamentary debate over the passage of the Green-LSAP-DP coalition's burka law in 2018 summed up many of the ADR's positions: And we do not want to succumb, step by step, to demands that are foreign to our culture, that do not correspond to our Judeo-Christian civilization, our humanistic heritage. We are not giving up our civilization and the values that make up our culture, to succumb to the pressure of various radical elements here. We want to preserve society as it is, as the CSV said, the social contract and the culture that we have, that remains. And people who come here should align with that, with our values, and not the other way around! (Kartheiser, 2018, p. 494) 5 Like many right-wing populists engaging in civilizationalism (Brubaker, 2017), Kartheiser mentions Western civilization (3 segments) as opposed to foreign entities, such as radical Islam, and he appeals to Luxembourgish culture and values as grounds for exclusion. Indeed, the desire to ban the burka itself can be construed as a tactic to stigmatize Muslim migrants and legitimize border control (Rheindorf & Wodak, 2020). Nonetheless, in this context even, the ADR recognized Luxembourgish society (3 segments) as open and multicultural. In the same speech above, Kartheiser recognized the need for immigrants to integrate 'into a society that is multicultural, in the sense that many different cultures are found here' (Kartheiser, 2018, p. 494). 6 As he argued, such integration would prevent parallel societies and segregation from forming. As indicated above, however, there are clear limits on demands that are too foreign (Islamic) for Luxembourgish traditions.

Playing on right-wing populist themes
In the ADR's discourse, there is a clear vertical separation between the people and the elites, although the horizontal separation between people and foreign entities is more nuanced. The ADR's discourse plays on many of the themes of right-wing populists, although in different ways depending on the debate. The party was the most rightwing populist during the burka debate, where they located Luxembourg as part of a broader Judeo-Christian West and claimed to defend women's rights against an illiberal Islam. Their portrayal of the Luxembourgish nation as threatened by outsiders as well as their reference to political correctness as an inhibitor of popular will was also a common feature of the party's discourse in all three instances. In all three instances, such rhetoric can be construed as fitting the right-wing populist playbook in which the popular will is a way of protecting the Luxembourgish nation from outsiders and legitimizing the restriction of minority rights (Betz & Johnson, 2017;Mudde, 2007).
There are, however, two unique aspects of the ADR's positions. One is the ADR's recognition of Luxembourg as a multicultural society. Interestingly, the ADR recognizes and celebrates this multicultural society, including Luxembourg's ability to integrate successive waves of immigrants, as uniquely Luxembourgish, even as a defining part of the Luxembourgish nation itself (although with the exception of Islam). A second unique feature is the ADR's use of the issue of language. Like other right-wing populist parties, the ADR uses language as a marker of foreignness and to protect the nation. But the party also uses it to appeal to the economic interests of Luxembourgers. Such was the case during the referendum where the ADR raised the prospect of the Luxembourgish language losing its status in the public service sector, allowing non-Luxembourgers to compete with Luxembourgers in this domain. Both of these specificities translate into unique features within the ADR's concept of the 'people' that I will discuss more fully in the following section.

Discussion
To understand the specificities of the ADR it is important to understand the socio-economic structures that define life in Luxembourg and that are unique to it. Primary among these structures is Luxembourg's political and economic integration with neighboring and foreign countries and the structure of Luxembourg's labor market, particularly with regards to the location of foreigners and Luxembourgers within it. 7 Luxembourg's status as a small country, centrally located on the European continent, has led it historically to be a central player in political and economic transnational cooperation. For example, through much of the nineteenth century and until 1919, the independent state of Luxembourg was part of a customs union with other German provinces (Harmsen & Högenauer, 2020). After World War II, Luxembourg was at the heart of the creation of the European Union. They participated in the creation of the European Coal and Steel Community (ECSC) in 1951, signed the Treaty of Rome in 1957 that established the European Economic Community (EEC), and signed the Treaty of Maastricht in 1992, creating the European Union. Luxembourg City is furthermore host of several European institutions, including the European Court of Justice and the European Investment Bank (CVCE, 2021). More recently, the country has become a center of international financial investment, with over 4.5 trillion euros placed in investment funds in Luxembourg in 2019, representing an important economic boon for the country (Lalieu, 2019). Luxembourg has also succeeded in attracting multinational companies, such as PricewaterhouseCoopers, Deloitte, and Amazon, which offer high-paying jobs and are among the country's largest employers (Statec, 2020d).
Luxembourg's integration in the EU and its transnational cooperation with bordering states also greatly impacts its population structure and the structure of its labor force. An extremely high percentage of Luxembourg's resident population is foreign. Of the 626,100 people living in Luxembourg, 47 per cent (296,500) hold foreign nationality (Statec, 2020a). In addition to these foreign residents, around 43 per cent of Luxembourg's labor force, of a total of 479,868, are cross-border workers, meaning they live in bordering countries and commute to Luxembourg to work (Statec, 2020b). In total, only 26.7 per cent of the Luxembourg labor force is made of Luxembourgers, and only in the sectors of public administration (91 per cent) and education (77 per cent), both of which are part of the public service sector, do Luxembourgers make up a majority of the labor force (Statec, 2020c). Foreign residents and cross-border workers obviously make meaningful contributions to the well-being of the country on a scale that no other European country can replicate. These foreign residents furthermore are overwhelmingly of European origin (84 per cent) (Statec, 2020a), while the number of Muslims in Luxembourg remains low compared to other European countries (2.3 per cent in 2012) (Pew Research, 2012); there is thus less ethnic or racial strife in Luxembourg than other countries with high rates of non-European immigration.
In addition, Luxembourgers are economically very well off, in particular in comparison to foreign workers. A study based on statistics from 2018 shows that 80 per cent of Luxembourgish workers occupy white-collar jobs versus 57 per cent of foreign residents and 64 per cent of cross-border workers. At the same time, 65 per cent of Luxembourgers worked in high-paying job sectors, versus 42 per cent of foreign residents and 37 per cent of cross-border workers. Foreign residents and border workers are also vastly overrepresented in lower-paying blue-collar job sectors, such as construction, horeca (hotel, restaurants, cafés), industry, commerce, and service jobs (Statec, 2020c). Also of note, the two sectors where Luxembourgers dominate the labor force, education and public administration, are also respectively the second and fourth highest-paying job sectors in Luxembourg (Statec, 2020e). 8 These socio-economic structural facts, unique to Luxembourg, form the local context in which the ADR expresses itself. The presence of foreign nationals living and working in Luxembourg is a well-accepted daily fact of life and generates massive economic benefits for the country. Luxembourg has made a conscious decision to make these aspects a fundamental element of their country. The ADR is well aware of this situation, as evidenced by their affirmative statements towards multiculturalism and integration, and modulates its discourse accordingly. However, there are important nuances concerning this point. The Luxembourgish nation itself is not presented as multicultural, but more so the Luxembourgish territory. This territory includes the foreign residents and cross-border workers, who Luxembourgers tolerate and give space to live their cultural identity. Hence, these foreigners are somehow outside the Luxembourgish nation. At the same time, the ADR presents the mostly European foreign residents and cross-border workers that make up the multicultural society as compatible with Luxembourgish culture and thus within the limits of the tolerance that defines the Luxembourgish nation. The limits of this tolerance are found when the ADR contrasts these groups to a radical and culturally incompatible Islam that remains excluded from both the Luxembourgish nation and its multicultural society. In this way, Luxembourg's multicultural society is somehow both excluded and included as a defining part of a Luxembourgish 'people'.
One might be tempted to see the ADR's rhetoric and positions as part of a move by rightwing populists to present a more respectable face to mask more radical views, a process Ruth Wodak (2013) has described as 'Haiderization'. However, unlike the Rassemblement National (formerly the Front National) or the Freiheitliche Partei Österreichs, the ADR did not begin as a radical right-wing party with a strong racial component, but rather as a single-issue party campaigning on pension reform. As the party evolved, it adopted more positions to address a full spectrum of issues. Their views on immigration, multiculturalism, and integration have not changed substantially over time and have never included an overt rejection of immigrants in a racialist way (Poirier, 2012). Rather the ADR seems to genuinely embrace its beliefs, which reflect the reality of life in the small Duchy.
The party's position on the Luxembourgish language is also unique. Language has a multifaceted status in the ADR's discourse, as both the means for integration and exclusion, and as a way to address cultural and economic concerns. To be sure, the importance of language to Luxembourgish identity and daily life is paramount in the eyes of Luxembourgers, as polls and surveys have repeatedly demonstrated (Kies, 2019). However, looking at the demographic breakdown of Luxembourg's economy, Luxembourgers are very well off and dominate two of the best-paying sectors in the economy: public administration and education. In total, 16% of Luxembourgers work in education, and 12.4% work in public administration, representing a sizeable portion of the Luxembourgish work force and electorate (Statec, 2020c). Importantly, entrance into these sectors requires a high level of fluency in Luxembourgish. The ADR's defence of Luxembourgish is thus also a means to defend the economic interests of Luxembourgers who work, or would work, in the public service sector. The ADR made this a point during the 2015 referendum, and in the party program for the 2018 national election proposed making a C2 level speaking ability of Luxembourgish, essentially native fluency, a requirement for obtaining a public service sector job (ADR, 2018, p. 14). The economic aspect of the language interest is real, and it is thus perhaps no surprise that during the 2015 referendum, the only labor union to oppose the extension of the vote to foreign residents was the powerful CGFP.
Interestingly, this defence of language demonstrates a conceptualization of the Luxembourgish 'people' that is somewhat at odds with current understandings of populism. Populism is predicated on an opposition between the 'people' and the 'elite', and this distinction implies an important socio-economic rift between the middle and/or working class and upper-class elites who represent their own interests to the detriment of other groups. Looking at the structure of Luxembourg's economy, however, it is Luxembourgers, who work overwhelmingly white-collar jobs that often have language requirements, that constitute a sort of economic elite as compared to foreign residents and border workers who dominate the blue-collar sectors. The ADR's defence of language as well as its embrace of Luxembourg's multicultural landscape, which economically benefits native Luxembourgers, can thus been seen as them appealing to the interests of a 'people' that in an important way forms a privileged 'elite'.

Conclusion
When taking into consideration the chameleonic nature of populism (Taggart, 2000;2004), one can better understand the specificities of the ADR, namely their stances towards multiculturalism and their use of the issue of language. These specificities give them a conceptualization of 'people' that is atypical when compared to other right-wing populist parties. Both of these specificities are rooted primarily in the structure of Luxembourg's internationally integrated political system and economy, which attracts large foreign companies and foreign capital and heavily relies on the labor of foreign residents and cross-border workers. These dynamics make it impossible for a right-wing populist party characterized by strong nationalism and overt xenophobia from having any electoral success (see also De Jonge, 2021). While unique to Luxembourg, these results also point to the potential for comparison. Such comparisons are possible with other small EU states such as Cyprus or Malta, small non-EU states such as Andorra, or with the canton of Geneva, Switzerland, which like Luxembourg, is at the heart of international integration efforts, has a high percentage of foreign residents, and is heavily dependent on cross-border workers. Such comparative work might unveil commonalities that allow for theory building for right-wing populism in small European states that challenges how the phenomenon is understood in larger states or that highlights how it operates in larger states (i.e. might the limits imposed by the size of small states, which require political and economic integration with neighboring states, be a common factor preventing a strong right-wing populist party from succeeding in certain states?); or perhaps such work would reveal the degree to which each case remains unique (for similar discussions see Thorhallsson, 2018;Thorhallsson & Wivel, 2006).
Notwithstanding its specificities, the ADR retains features that are broadly shared by other right-wing populists, particularly their antagonistic positions towards Islam and multicultural, politically correct 'elites' at the EU and national levels. Such positions allow them to participate in a meta-populism with other right-wing populist parties in Europe and make common causes with the PiS or Victor Orban. The ADR's participation in a broader meta-populism, combined with their (for most right-wing populist parties) atypical views on multiculturalism and integration, is likely a reason why the party has been difficult for researchers to classify. Nevertheless, taking the chameleonic nature of populism in mind, one can classify the ADR as right-wing populism à la luxembourgeoise, with all the caveats and nuances that implies. Notes 1. These affirmations come with caveats. Lucien Blau (2005, p. 89) describes the ADR as a weak form of right-wing populism, while Mattia Zulianello (2020) classifies the ADR as right-wing populism, but of a national-conservative as opposed to a radical right variety.
2. See also Kies' (2019) explanation on this point. As they relate, there was widespread anger over the way in which the coalition introduced the constitutional amendments. The coalition did not have the required 2/3 vote in parliament in favor of the constitutional amendment, and sought to use the referendum to pressure other parties to accept the reforms. A second, legally binding referendum would be held only after the 2/3 parliamentary majority would be obtained in this manner. . 'D'Integratioun ass an eisen Aen och an enger Gesellschaft, déi multikulturell ass, an deem Sënn, datt vill verschidde Kulture sech hei erëmfannen, mä Multikulti heescht net, datt mer egal wat mussen toleréieren.' My translation. 7. De Jonge (2021, pp. 92-100) discusses many of the same socio-economic structures in Luxembourg and their influence on right-wing populism. However, her discussion focuses on identifying the reasons for the failure or success of right-wing populism in Luxembourg, whereas my discussion explores how these socio-economic factors influence the ADR's populist conception of the 'people.' 8. For reference, the brute annual salary for those in education is 98,835 euros per year and for those in public administration it is 86,582 euros. These salaries are much higher than the average brute annual salary in horeca (33,796 euros), construction (42,750 euros), and retail (46,766 euros). In health care, where Luxembourgers are also well represented, at 44 per cent of the labor force, the average brute annual salary is 68,303 euros (Statec, 2020c.