INTRODUCTION
The place of minority religious symbols in the public sphere is an increasingly contentious and politicized issue in both Europe and North America. In countries such as Canada, France, Germany, Bosnia-Herzegovina, and Denmark, politicians have proposed or adopted laws to ban minority religious symbols in schools, courts, or the public service. Other laws have also been adopted to prohibit or restrict minority religious symbols in the broader public space, including restrictions on burqas or niqabs in France, Belgium, and municipalities in a number of other countries, and restrictions on the construction of new minarets in Switzerland.
These bans, as well as broader debates around the place of minority religions, have been increasingly portrayed as the product of unique alliances between progressive and conservative forces. In her analysis of debates around the place of Islam in the Netherlands, Liz Fekete (Reference Fekete2006) argues that there is a growing trend among “enlightenment crusaders,” including feminists and gay activists, to support policy prescriptions around integration that are espoused by far right political parties and movements. In his exploration of French debates over the place of Islam in the public sphere, Olivier Roy (Reference Roy2007, xii) asserts that proposals to restrict all religious displays to the private sphere, rooted in what he refers to as “ideological laïcité,” have led the “majority of the secular left to strike an alliance with the Christian Right against Islam.” And according to the philosopher Jocelyn Maclure (Reference Maclure, Rudy, Gervais and Kirkley2011), intellectual debates surrounding the institutional accommodation of minority religious practices in the province of Quebec in Canada have generated an alliance between “liberal neutralists” and “conservative nationalists.” In Maclure's view, whereas “liberal neutralists” view the accommodation of all religions as incompatible with civic values such as gender equality and the separation of church and state, “conservative nationalists” see minority religious accommodation as a threat to the traditions and customs of the majority.
While united in their desire to restrict the visible presence of minority religions in the public sphere, these groups differ when it comes to their broader policy objectives and motivations. Maclure refers to the possibility that groups with distinct motivations are driving opposition to the presence of minority religions in the public sphere as the “strange bedfellow hypothesis.” While “the Christian right” or “conservative nationalists” strive to curb the presence or accommodation of minority religions and to protect symbols associated with the majority, “liberal neutralists,” the “secular left,” or what we refer to in this article as “enlightenment liberals” wish to restrict the visibility of all religions in the public sphere. Moreover, while perceived cultural threats, in-group favoritism and prejudice are presented as driving opposition to minority symbols, secular values and a particular conception of liberalism are said to be at the root of support for blanket restrictions on the presence of religion in the public sphere.
A growing number of scholars have provided evidence of political or intellectual alliances between such strange bedfellows (see also, Lamy Reference Lamy2015; Scott Reference Scott2007), but no study has explicitly explored whether such alliances extend to mass publics. Is mass public support for restrictions on minority religious symbols the product of an alliance between strange bedfellows — that is, one group expressing the desire to ban all religious symbols, including majority ones, and another group expressing the desire to ban only minority religious symbols? And more importantly, if such an alliance exists, do the motivations of mass publics correspond to those of political and intellectual elites? Are they the “enlightenments liberals” and “conservative-nationalists” described by political theorists? This study attempts to answer these questions.
Very few cases offer the opportunity to explore the strange bedfellows hypothesis within a mass public in a natural setting. One recent case is the province of Quebec in Canada, where the visibility of both minority and majority religious symbols in the public sphere were debated simultaneously. In autumn 2013, the Quebec government introduced a legislative proposal known as the “Charter of Values” which would prohibit public sector employees, including doctors, nurses, teachers, and child care workers, from wearing religious symbols such as the hijab, the turban, or the kippah. While the Charter of Values also would have banned wearing “large” crosses, the debate surrounding the proposal focused almost exclusively on minority religious symbols, especially Muslim ones. However, in light of the introduction of the legislative proposal, a number of commentators then also questioned the legitimacy of preserving the large crucifix currently hanging in the National Assembly, the province's legislative body. While many argued that, like other religious symbols, the crucifix should be removed, others contended it should be preserved out of respect for Quebec's cultural heritage. Recognizing this was a potentially divisive issue, the government ultimately opted to delay a final decision on whether or not to remove the crucifix. The crucifix episode, while brief, revealed a division among supporters of the Charter of Values. Debates around the Charter of Values in Quebec suggest that Charter supporters did form a coalition of “strange bedfellows” like Jocelyn Maclure (Reference Maclure, Rudy, Gervais and Kirkley2011) proposes, making it a useful and unique case to study the motivations of mass publics in support for bans on either all religious symbols or only minority religious symbols.
Our analysis explores the characteristics of two groups of supporters for a ban on minority religious symbols in Quebec using a survey administered as the debate over the Charter of Values was unfolding: those who supported both the Charter of Values and removing the crucifix from the National Assembly (enlightenment liberals), and those who supported the Charter of Values but opposed the eventual removal of the crucifix (conservative nationalists).
We begin by providing an overview of the context in which the debate over religious symbols unfolded in Quebec, and we briefly highlight similarities and differences with other debates around the world. We then present a discussion of the different theoretical perspectives guiding our analysis to differentiate the motivations of both groups of Charter supporters, stressing the role of liberalism, religiosity, and ethnocentrism. Our findings support the strange bedfellows hypothesis and demonstrate that both groups of Charter supporters had distinct motivations. Hence, while stronger liberal attitudes and low religiosity were key characteristics of Charter supporters who wanted to remove the crucifix, a perceived threat to Quebec culture and generalized prejudice were central characteristics of Charter supporters who wanted to keep the crucifix at the National Assembly. However, negative attitudes specifically toward Muslims also appear to have motivated both groups of Charter supporters.
DEBATES ABOUT RELIGIOUS SYMBOLS IN QUEBEC
Before the 1960s, Catholicism was, like the French language, viewed as central to Quebec's identity. The province was often referred to as the “priest-ridden province”; the Catholic Church played a central role in the daily life of citizens, being largely responsible for education, health, and social services. The conservative Union nationale government's 1936 decision to install a crucifix above the Speaker's chair in the National Assembly is emblematic of the power exercised by the Catholic Church during that period. However, in the 1960s, the province underwent rapid social, cultural, and political transformations — a period referred to as the Quiet Revolution — during which the Quebec state gradually took over many of the Church's responsibilities and the province experienced a gradual, but significant decline in church attendance (Eagle Reference Eagle2011; Meunier and Wilkins-Laflamme Reference Meunier and Wilkins-Laflamme2011).
Until the 1990s, debates over the place of religion focused mostly on the continuing presence of Catholic symbols and practices in public institutions. However, sporadic public discussions about the wearing of headscarves in schools and opening prayer spaces in universities emerged in the province in the 1990s, in part because of the growing number of North African immigrants arriving and settling in Quebec. In the mid-2000s, a series of disputes erupted over rules and policies accommodating religious minorities, including a Quebec soccer federation ban on wearing headscarves during games, requests by Hasidic Jews to have the windows of a YMCA gym tinted to hide women in gym clothes from public view, and a Supreme Court of Canada decision allowing a young Sikh to wear his ceremonial dagger at school. This led to what became known as the “reasonable accommodation crisis.”
In 2007, in order to provide guidelines on how to respond to these accommodation requests, the provincial Liberal government created the Consultation Commission on Accommodation Practices Related to Cultural Differences. The Commission's 2008 report recommended a ban on religious symbols for public servants who must embody “at the highest level the necessary neutrality of the State” and for those who exercise coercive power (Bouchard and Taylor Reference Bouchard and Taylor2008, 151). It also recommended removing the crucifix from the National Assembly. Ultimately, however, very few of the Commission's main proposals were implemented (see, Rocher Reference Rocher, Emongo and Bob2014).
The place of minority religious symbols in the public sphere has been a mainstay of public debates in Quebec ever since. During the 2012 Quebec election, the pro-independence Parti Québécois (PQ) campaigned in part on a promise to ban the wearing of religious symbols for all public employees. The party won the election, but failed to secure a majority government. In September 2013, the PQ minority government proposed the Charter of Values. However, Charter supporters were divided about what should happen to the crucifix at the National Assembly, and Bill 60, the formal legislative proposal that came before the National Assembly in November 2013, did not include any recommendations regarding the crucifix. Hoping to win a majority government, the PQ called an election in March 2014, before the bill was adopted. The election outcome sealed the fate of the Charter of Values and Bill 60: the PQ was defeated and the Liberal Party of Quebec, opposed to the bill, won a majority government.Footnote 1
While public disputes about the place of religion in Quebec have largely been influenced by its history with the Catholic Church, such debates are not unique to the province. As noted in the Introduction, arguments about minority religious symbols abound throughout Europe and North America. But so do debates about majority religious symbols. In Italy, for example, controversies have erupted over the place of crucifixes in polling stations or schools (see, McGoldrick Reference McGoldrick2011), and in the United States there have been legal battles over school prayers and Ten Commandments monuments located in or near government buildings (Finkelman Reference Finkelman2005).
Quebec is not insulated from disputes in other countries. In fact, French debates have profoundly influenced the discourse around questions of religious symbols in Quebec. According to Koussens and Amiraux (Reference Koussens, Amiraux and Lévesque2014, 55), since the “accommodation crisis,” the idea that the French model of laïcité or, more accurately, a specific interpretation of that model, should be an inspiration for Quebec has spread among politicians, journalists, and intellectuals.Footnote 2 Laïcité in France has taken different meanings over the years. Koussens and Amiraux (Reference Koussens, Amiraux and Lévesque2014, 60) argue that, prior to the 2000s, French laïcité, as embodied in the 1905 legislation on the separation of Church and State, had more in common with the liberalism of Locke than the anti-religiousness of the Enlightenment. However, since then, it has become much more dedicated to the emancipation of individuals from religion and “the regulation of individual expression of beliefs” (Koussens and Amiraux 2014, 67, our translation).
Consequently, the Quebec case can offer lessons that are relevant to other contexts. In the reminder of the article, we explore whether supporters of restrictions on minority religious symbols in Quebec constitute “strange bedfellows” with different motivations behind their desire to ban religious symbols.
STRANGE BEDFELLOWS: ENLIGHTENMENT LIBERALS AND CONSERVATIVE NATIONALISTS
Enlightenment Liberals: Liberal Attitudes, Religiosity, and Atheism
In his study of intellectual debates around religious accommodation in Quebec, Jocelyn Maclure (Reference Maclure, Rudy, Gervais and Kirkley2011) refers to those who view the accommodation of all religions as incompatible with civic values — such as gender equality and the separation of church and state — as “liberal neutralists.” We prefer the term “enlightenment liberals,” which has been used in a number of studies related to debates around cultural diversity and immigration (see, Rostbøll Reference Rostbøll2009; Triadafilopoulos Reference Triadafilopoulos2011; Gustavsson, van der Noll, and Sundberg Reference Gustavsson, van der Noll and Sundberg2016). Arguably, these liberals are not “neutral” about the role of the State with respect to the presence of religious symbols in the public sphere, since they believe that it must actively foster secular values and promote autonomous citizens. Nevertheless, Maclure's profile of this group echoes studies on opposition to religious symbols in mass publics in a number of ways.
First and foremost, attitudes driven by a particular set of underlying values emphasizing the protection and enhancement of individual rights and freedoms, especially those related to gender and morality, are often invoked in discussions about the place of religious symbols in the public sphere (Helbling Reference Helbling2014; Gustavsson, van der Noll, and Sundberg Reference Gustavsson, van der Noll and Sundberg2016). We refer to this constellation of attitudes stressing individual freedom — identified as a distinctive and coherent set by a number of researchers (Rokeach Reference Rokeach1973; Schwartz Reference Schwartz1994; Inglehart and Welzel Reference Inglehart and Welzel2005) — as liberal attitudes. From the perspective of many of those who hold such attitudes, religion is viewed as intrinsically patriarchal and conservative, and religious symbols (both minority and majority ones) seen as symbols of the oppression of women and minority groups, such as the LGBTQ (Lesbian, Gay, Bisexual, Transgender, Queer) community. However, not all Liberals share this perspective.
Liberals are in fact divided on the matter of religious symbols, especially in the public sphere. As argued by Helbling (Reference Helbling2014, 244), the debate goes to the heart of whether liberalism is a substantive way of life (that must be promoted and defended), or is instead a mechanism to reconcile different ways of life. In the literature, these different perspectives are associated, respectively, with enlightenment and reformation liberalisms (Galston Reference Galston1995; Gustavsson Reference Gustavsson and Magone2014). These different conceptions of liberalism give significantly different roles to the state. Enlightenment liberalism sees the state as a defender of liberal values against symbols or practices that are perceived as “illiberal,” and as a promoter of autonomous citizens who are capable of rising above their particularistic identities (Gustavsson, van der Noll, and Sundberg Reference Gustavsson, van der Noll and Sundberg2016: 1725).Footnote 3 In contrast, reformation liberalism is associated with tolerance, diversity, and individual choices in a way that rejects state promotion of any particular definition of the “good life.” Thus, a point of contention between these two liberalisms is what, if anything, the state should do about religious symbols in the public sphere: enlightenment liberalism is inclined to limit the visibility of symbols associated with religious beliefs and practices they see as hindering individual autonomy; reformation liberalism is wary of such restrictions. In a recent empirical test of the impact of endorsing different conceptions of liberalism, Gustavsson, van der Noll, and Sundberg (Reference Gustavsson, van der Noll and Sundberg2016) find that enlightenment liberals are inclined to express negative attitudes toward the Muslim veil (while the opposite is true of reformation liberals).
Are liberal attitudes a motivation to ban religious symbols in Quebec? The answer probably depends on which perspective on liberalism is dominant in the province. While our survey does not allow us to explicitly distinguish between support for enlightenment and reformation liberalism, Charles Taylor (Reference Taylor1994) has argued that liberalism in Quebec takes a different form (“substantive liberalism”) than in the rest of Canada (“procedural liberalism”). Like enlightenment liberalism, Taylor's substantive liberalism is not neutral about what constitutes the good life and is willing to restrict some rights in order to promote collective goals.Footnote 4 Perhaps more importantly, as stated in the previous section, debates around secularism in Quebec have increasingly been influenced by a conception of laïcité that explicitly draw on the anti-religiousness of the Enlightenment (see also, Lamy Reference Lamy2015, 65–70). Accordingly, in Quebec, we expect liberal attitudes concerning individual autonomy to be associated with support for a ban on all religious symbols: that is, we expect those with liberal attitudes to support both the Charter and the removal of the crucifix from the National Assembly.
If enlightenment liberalism is a driving force behind support for a ban on all religious symbols, then we anticipate that ambivalence toward religion, if not an outright absence of any religious beliefs at all, is also associated with favoring such a ban. According to Gustavsson, van der Noll, and Sundberg (Reference Gustavsson, van der Noll and Sundberg2016, 1725), enlightenment liberalism's focus on fostering autonomy “leads enlightenment liberals to disrespect choices that are perceived to be made out of faith or passion, rather than reasoned reflections.” This perspective is similar to the “new atheism,” best represented intellectually by the writings of Dawkins (Reference Dawkins2006), Harris (Reference Harris2006), and Hitchens (Reference Hitchens2007), which opposes the display of religious symbols in public institutions in the name of rationality, science, and often explicitly, the enlightenment tradition. As such, atheism, low religiosity, or secularism might also influence public attitudes toward religious symbols. Indeed, a growing body of literature has shown that they influence a range of political attitudes and behaviors (Beard et al. Reference Beard, Ekelund, Ford, Gaskins and Tollison2013; Clements Reference Clements2013; Ribberink, Achterberg, and Houtman Reference Ribberink, Achterberg and Houtman2013). Some of those studies confirm Tariq Modood's (Reference Modood1994, 72) assertion that the main divide in today's world is not between majority and minority religious groups, but between “(…) those who think religion has a place in a secular public culture and those who don't.” Fetzer and Soper (Reference Fetzer and Soper2003, 250–251) have dubbed this perspective the “solidarity-of-the-religious theory.”Footnote 5 Accordingly, we expect atheists and individuals with low levels of religiosity to support both the Charter and the removal of the crucifix.
Conservative Nationalists: Ethnocentrism, Feelings of Threat and Generalized Prejudice
According to Maclure (Reference Maclure, Rudy, Gervais and Kirkley2011, 142), conservative nationalists, rather than opposing a public place for religion per se, believe that the religion of the majority — Catholicism, in the case of Quebec — “cannot be put on par with the other religions brought to Quebec through immigration.” They distinguish themselves from enlightenment liberals by viewing minority religious accommodation as a threat to the traditions and customs of the majority group. Conservative nationalists' opposition to minority religious symbols may well be rooted in ethnocentrism, which Kinder and Kam (Reference Kinder and Kam2009, 8) — drawing on the seminal work of Sumners (Reference Sumners1906) — define as a pre-disposition to divide the human world into in-groups and out-groups.Footnote 6 Accordingly, “symbols and practices become objects of attachment and pride when they belong to the in-group and object of condescension, disdain, and (in extreme cases) hatred when they belong to out-groups” (Kinder and Kam Reference Kinder and Kam2009, 8).
In the case of public opinion toward minority religious symbols, ethno-centrism might operate in one of two ways. First, majority group members might view minority groups as jeopardizing the material well-being, status, and culture of majority group members. According to “group threat theory,” ethnocentrism is the product of competition over finite resources (material and symbolic) between groups in a political community.Footnote 7 More relevant to our study are “symbolic” or “cultural threats” rooted in fears that out-groups are putting the majority's values and culture at risk (Velasco González et al. Reference Velasco González, Verkuyten, Weesie and Poppe2008). A number of scholars have also linked perceived threats to broader insecurities about national identity (Sides and Citrin Reference Sides and Citrin2007; Sniderman and Hagendoorn Reference Sniderman and Hagendoorn2007), similar to Maclure's account of “conservative nationalists.” The group threat perspective suggests that when individuals believe their own (national, cultural, or ethnic) group is at risk, they are more likely to reject “out-groups” and their associated symbols, and to express a strong desire for cultural unity. Indeed, in her study on attitudes toward the veil, Van der Noll (Reference Van der Noll2010) found that perceived threats are associated with support for a ban on the Muslim headscarves. As such, our expectation is that Quebecers who view minorities as constituting a threat to Quebec culture will be Charter supporters opposed to removing the crucifix from the National Assembly. The rationale is that this group of Charter supporters wants to restrict minority religious symbols exclusively in order to protect the status of their own (majority) religious traditions.
Second, majority group identity might generate antipathy toward members of minority groups. Social identity theory posits that in order to evaluate their own members positively, members of a community are prone to view members of other groups in a negative light. Kalkan, Layman, and Uslaner (Reference Kalkan, Layman and Uslaner2009) have referred to such phenomena as a tendency to view all minorities as a “band of others.” More commonly, it is referred to as “generalized prejudice,” defined by Hagendoorn and Sniderman (Reference Hagendoorn and Sniderman2001, 21, quoted in Spruyt and van der Noll, Reference Spruyt and van der Noll2016, 4), as a “consistent tendency to evaluate immigrant groups negatively (…).” A body of literature has found that those who hold negative attitudes toward minorities are less likely to want to extend to these minorities the same rights or privileges as the majority group, or are more likely to oppose policies perceived as benefiting minority groups (Gilens Reference Gilens1995; Sniderman and Hagendoorn Reference Sniderman and Hagendoorn2007). It is perhaps not surprising, then, that Helbling found that xenophobia was associated with support for a ban on the Muslim Headscarf in schools (Helbling Reference Helbling2014). However, we might anticipate people whose negative attitudes toward “out-groups” are rooted in ethnocentrism to have fundamentally different views about the crucifix than enlightenment liberals: for the former, the crucifix is more likely to be seen as a symbol of the majority group. Consequently, if ethnocentrism is a driving force behind support for a ban on minority religious symbols, then the expectation is that those who express antipathy toward ethno-cultural minority groups will support the Charter, but will be opposed to the removal of the crucifix from the National Assembly.
Group-Specific Affect: A Common Bond between Enlightenment Liberals and Conservative Nationalists?
A growing body of literature has focused on the importance of “group-specific affect” in structuring attitudes toward immigration and minority groups (Sides and Gross Reference Sides and Gross2013; Valentino, Brader, and Jardina Reference Valentino, Brader and Jardina2013). Those studies draw on Converse's (Reference Converse2006 [1964], 38) well-known claim that citizens organize their opinions on specific issues around “visible social groupings,” rather than abstract ideological principles. Policy questions thus become a judgment on the groups implicated in the policies.
Since the reasonable accommodation crisis, media coverage in Quebec has overwhelmingly focused on Muslims. They have been primarily presented as the main group challenging the secular nature of Quebec, but also at times as a group who refuse to assimilate and whose culture is deemed incompatible with Quebec culture (see, Antonius Reference Antonius2008; Potvin Reference Potvin2008). In short, Muslims have been criticized either for their religious practices or, more broadly, for their culture. Together, the literature on group-specific affect and the emphasis on Muslims during the Charter of Values debate suggests that negative attitudes toward Muslims in particular could be associated with support either for restricting the wearing of all religious symbols in public institutions, or for restricting only minority religious symbols. Accordingly, we expect Quebecers who hold more negative attitudes toward Muslims in particular to be Charter supporters irrespective of their position on the crucifix.
DATA AND METHODOLOGY
To test the strange bedfellows hypothesis, the study relies on data from an online survey conducted in January and February 2014 as the debate over the Charter of Values was unfolding. The sample is composed of 1000 respondents from Quebec.Footnote 8 Respondents were asked whether they supported or opposed the Charter of Values, and whether they agreed or disagreed with the following statement: “The crucifix should be removed from the National Assembly.”Footnote 9 The data indicate that 49% of respondents were opposed to the Charter of Values, 27% were supportive of the Charter but opposed to the removal of the crucifix, and 24% were supportive of the Charter and in favor of removing the crucifix.Footnote 10 For brevity, we will now refer to these two groups as “pro-Charter and pro-crucifix” and “pro-Charter but anti-crucifix,” respectively.
Details on the construction of our dependent and independent variables are reported in Appendix A, but three measures — generalized prejudice, antipathy toward people of Muslim faith, and antipathy toward people of Jewish faith, require further description. Generalized prejudice is measured by a seven-item index of feelings toward minority groups, where a score of 10 means strong positive feelings toward all groups and a score of 0 means strong antipathy. To test for the relationship between Charter support and antipathy toward people of Muslim faith in particular, we use a variable measuring the gap between feelings toward Muslims and feelings toward all minority groups. This variable ranges from -10 to 10; it takes a positive value when feelings toward Muslims are more positive than for all minority groups, and takes a negative value when feelings toward Muslims are more negative than for all minority groups. We also examine how distinctive feelings toward Jews are related to support for the Charter of Values.Footnote 11 Including measures of feelings toward Muslims and Jews allows us to examine whether antipathy toward religious minorities in general (rather than Muslims specifically) motivated Charter supporters.
Table 1 reports the descriptive characteristics for all three groups of respondents. While the group “pro-Charter but anti-crucifix” appears to hold stronger liberal attitudes and to be more likely to have no religion or to be less religious, the group “pro-Charter and pro-crucifix” appears to hold stronger generalized prejudice and to be more likely to perceive immigration as a threat to Quebec culture. Also consistent with our hypothesis, the data indicate that both groups appear to hold more negative feelings toward Muslims than Charter opponents.
Characteristics of charter of values supporters
Source: Provincial Diversity Project. Note: Entries report within-group characteristics.
Because there is likely considerable overlap among some of these attitudes and characteristics, Table 2 reports correlations between our independent variables. Most significantly, it shows that those who express liberal attitudes also tend to be less religious (r = −0.46), and perhaps not surprisingly, those who view immigration as a cultural threat exhibit more negative feelings toward minority groups (r = −0.43). For this reason, a multivariate strategy is most appropriate for our analysis.
Relationships between explanatory variables (correlation coefficients)
Source: Provincial Diversity Project. ***p < 0.001; **p < 0.01; *p < 0.05.
TESTING THE STRANGE BEDFELLOWS HYPOTHESIS: MULTIVARIATE ANALYSIS
In order to confirm the characteristics of each group of Charter supporters, we performed a multinomial logit analysis, in which respondents opposed to the Charter of Values are the reference category. Table 3 reports the unstandardized B coefficients of the analysis, but in the discussion which follows we translate the log-odds derived from the multinomial logit results into predicted probabilities to facilitate the interpretation of effects. When predicted probabilities are calculated, all other independent variables in the model are held constant at their means. In addition to the variables discussed in the section above, the analysis also controls for a number of socio-demographic variables, including age, gender, employment, education, and language spoken at home. Since the PQ proposed the Charter of Values, we also control for PQ partisanship.
Characteristics of Charter of Values supporters (multivariate analysis)
Entries are multinomial logit unstandardized B coefficients.
Source: Provincial Diversity Project. ***p < 0.001; **p < 0.01; *p < 0.05; cp < 0.10.
The analysis presented in Table 3 indicates that partisanship was a factor in shaping opinions about the Charter of Values, aside from any other considerations that might have influenced opinions about the Charter. PQ supporters were significantly more likely to support the Charter, but were split over the matter of the crucifix. Hence, the probability of being pro-Charter and pro-crucifix increases by 19 points (from 21% to 40%) when respondents identify with the PQ, and the probability of being pro-Charter and anti-crucifix increases by 25 points (from 14% to 39%). Moreover, Francophones were more likely to be Charter supporters even when controlling for all other variables (13 points more likely than non-Francophones to be pro-Charter and pro-Crucifix, and 22 points more likely to be pro-Charter and anti-crucifix).
Beyond partisanship and language, the analysis lends support for the strange bedfellows hypothesis. First, the analysis indicates that those holding stronger liberal attitudes are more likely to support the Charter and the removal of the crucifix. The probability of being pro-Charter and anti-crucifix is 18 points higher among respondents with a high score (9.0) on the scale of liberal attitudes in comparison to someone with a low score (5.0) — 31% vs. 13%.Footnote 12 Liberal attitudes are not associated with being pro-Charter and pro-crucifix
Second, as expected, atheists and those who do not belong to any religion were more likely to support both the Charter and the removal of the crucifix. The predicted probabilities indicate that the likelihood of being pro-Charter and anti-crucifix is 19 points higher (35% vs. 16%) among people who are atheists or without religion compared to those with a religious affiliation other than Catholicism.Footnote 13 It is noteworthy that the probability of being pro-Charter and pro-crucifix is 14 points higher (37% vs. 23%) among Catholic respondents than among those with other religious affiliations. Moreover, people for whom religion is of great importance, regardless of their religious denomination, are 19 points less likely to be pro-Charter and anti-crucifix than those for whom religion is not important at all in their lives (10% vs. 29%).
Third, and also consistent with our expectations, a perceived threat to Quebec culture and greater antipathy toward minority groups are associated with supporting the Charter, but only among those who are opposed to the removal of the crucifix. The predicted probability of being pro-Charter and pro-crucifix is 15 points higher (37% vs. 22%) when respondents express the view that immigration represents a threat to Quebec culture as compared to when they do not hold this opinion, and is 15 points higher (36% vs. 21%) when respondents' feelings toward minority groups are quite negative (4.1) as opposed to quite positive (8.4).Footnote 14
Finally, beyond the effects of their general feelings toward all minority groups, respondents who have distinct antipathy toward Muslims are more likely to be Charter supporters of either type. The probability of being pro-Charter and anti-crucifix is six points higher when respondents express strong distinctive antipathy toward Muslims (−3.4) in comparison to when their score for Muslims is the same for other minority groups (0), and the probability of being pro-Charter and pro-crucifix is seven points higher.Footnote 15 Not only is greater antipathy toward Muslims a significant characteristic of both groups of Charter supporters, but it is also equally salient for both groups. However, there is no conclusive evidence of a relationship between distinct antipathy toward Jews and Charter support. Those with distinct antipathy toward Jews are not significantly more likely to be pro-Charter and anti-crucifix. They are more likely to be pro-Charter and pro-crucifix, but the relationship is modest and significant only at the 0.10-level (p value = 0.086).Footnote 16
These results lend empirical support to the strange bedfellows hypothesis at the mass public level, and suggest two distinct and mutually exclusive narratives for supporting the Charter of Values. While liberal attitudes, atheism, and low religiosity distinguish Charter supporters who wanted to remove the crucifix (the group we call enlightenment liberals) from Charter opponents, a perceived threat to Quebec culture and generalized prejudice distinguish Charter supporters against the removal of the crucifix (the group we call conservative nationalists) from Charter opponents. Moreover, the analysis indicates that both groups of Charter supporters express greater antipathy toward Muslims than Charter opponents. This result might be explained by the fact that debates around the Charter of Values were primarily focused on the Muslim veil; whether both groups of Charter supporters' greater antipathy toward Muslims has the same sources is beyond the scope of this study.Footnote 17
CONCLUSION
Legislative proposals to ban minority religious symbols in the public sphere have multiplied across many Western democracies. What motivates mass publics to ask for such proposals or to support them? A group of scholars looking at political and intellectual elites argues that attempts to limit the visibility of minority religious symbols in the public sphere is the product of unique alliances between conservative and progressive political forces. From their perspective, erstwhile political adversaries have joined forces to restrict the presence of minority religious symbols in the public sphere. In this study, we investigated whether this “strange bedfellows” hypothesis extends to the mass publics, and most importantly, we examined whether motivations of mass publics correspond to “enlightenment liberal” and “conservative-nationalist” ideas described by political theorists (Maclure Reference Maclure, Rudy, Gervais and Kirkley2011).
Our findings are consistent with such an interpretation. Supporters of the Charter of Values in Quebec appeared to form a coalition of strange bedfellows, two groups with different policy objectives and motivations matching the expected profiles of “enlightenment liberals” and “conservative nationalists.” In terms of policy objectives, the former group distinguishes itself by asking for a ban on all religious symbols, whereas the latter group wants to restrict only minority symbols. We find that prejudice and a perceived threat to Quebec culture are relevant motivations only for those who want to ban minority religious symbols exclusively. In sharp contrast, those who want to ban all religious symbols appear motivated by principles such as secularism and a particular set of liberal attitudes.
The application of the strange bedfellows hypothesis to the domain of public opinion sheds an important light on the structure of debates over the place of minority religious symbols in the public sphere among the mass publics. Existing research on the subject has consistently shown that perceived threats and prejudices are key motivations for asking for bans on minority religious symbols (Saroglou et al. Reference Saroglou, Lamkaddem, Van Pachterbeke and Bruxant2009; Van der Noll Reference Van der Noll2010); the evidence with respect to the role of principles has been far less conclusive (Helbling Reference Helbling2014). Our study does not dispute the significance of perceived threats and prejudices; the evidence is unequivocal that antipathy toward ethnic minority groups — even groups with no explicit link to any particular religion — and a sense that immigration represents a threat to the majority culture, are key motivators for a significant group of supporters of bans on minority religious symbols. However, our study demonstrates that in order to have a more complete understanding of the motivations of mass publics, we must also consider people's views about bans on majority religious symbols. By taking into account views on both minority and majority religious symbols, our study suggests that previous research may well have underestimated the role of principles such as secularism and liberalism in shaping opinions about religious symbols in the public sphere.
Our analysis also raises questions about the foundations of Charter supporters' attitudes toward Muslims in particular. Quite importantly, we find that both enlightenment liberals and conservative nationalists express greater antipathy toward Muslims, thus supporting the claim that aside from the effects of generalized prejudice, people's views about public policies are also partially structured by their attitudes toward Muslims. It is indeed possible that enlightenment liberals' antipathy toward Muslims has different origins than the antipathy expressed by conservative nationalists. However, an exploration of this possibility is beyond the scope of this study.
Can our Quebec findings be generalized to other national contexts? We believe they might help us to understand the dynamics of public opinion regarding the place of minority religious symbols in other contexts, especially in Europe. Although Quebec is a small sub-national political jurisdiction in North America, in many ways it has much in common with European states when it comes to issues surrounding the place of religion in the public sphere. First, church attendance levels in Quebec are among the lowest in North America and are more similar to levels found in some European countries (Bibby Reference Bibby2008). Second, the relatively small population of Quebec and its minority status within North America offer useful comparisons to many European countries. Cultural and linguistic anxieties play a key role in the Quebec public sphere, more so than in the rest of North America (Bouchard Reference Bouchard2012). Indeed, our findings strongly echo those of Van der Noll and Saroglou (Reference Van der Noll and Saroglou2015) on funding for religious education in Germany, in which the authors distinguish between those who oppose public funding for all religious schools and those who only oppose the public financing of Muslim schools.
Beyond the question of attitudes toward minority religious symbols, our study also offers broader insights into contemporary political transformations and debates. More specifically, it helps make sense of the blurring of traditional lines of political division, as far-right political parties have increasingly co-opted themes and discourses once associated with left political parties, such as the defense of liberal values and gender equality (Akkerman and Hagelund Reference Akkerman and Hagelund2007; Mayer Reference Mayer2013). As such, our study of the Quebec case confirms that not only has opposition to minority religious symbols and practices reshaped political coalitions at an elite level, but also that such new coalitions can reshape traditional political alliances within mass publics.
APPENDIX A: CONSTRUCTION OF VARIABLES
APPENDIX B: FACTOR ANALYSIS (INDEX OF GENERALIZED PREJUDICE)
