Religious behavior and European veil bans

ABSTRACT

Do societal religious practices affect European policies towards Muslim veils? We argue that public religious behavior has a substantial effect on European countries’ and regions’ decisions regarding whether or not to ban the wearing of the veil in public spaces. Using data from the European Social Survey, we find that countries with higher levels of religious attendance are substantially less likely to enact veil bans than those where religious attendance is less common. We augment these findings with data from Switzerland, where variation across subnational units parallels the patterns witnessed in Europe more broadly: aggregate religious attendance decreases the likelihood of both voting on veil bans and actually enacting them. In environments characterized by a salient secular-religious divide, high levels of religious attendance lead to greater support for the public expression of religion – even for religious outgroups – and this support is often channeled into more accommodating policies towards religious expression.

KEYWORDS:

• Religion and politics
• European politics
• veil bans
• Islam
• Switzerland
• comparative politics

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 URL: https://www.ots.at/presseaussendung/OTS_20190508_OTS0165/fpoe-moelzer-zu-kopftuchverbot-wir-setzen-damit-ein-wesentliches-signal-gegen-den-politischen-islam

2 See Giraudy (2019) for a discussion of the advantages of supplementing cross-national data with evidence from subnational units within a single country.

3 It is also possible that attitudes towards these symbols might depend on whether the target group is a minority (Bilodeau et al. 2018).

4 Of course, it is possible that religiosity (in one form or another) will also lead to greater tolerance towards members of other groups under some circumstances; Doebler (2014) finds this to be the case for several dimensions of personal religiosity (though not all features of religiosity) in Europe.

5 Importantly, as Carol, Helbling, and Michalowski (2015) note, this type of reasoning requires that Christians not view Muslims as a potential replacement in their societies; power relations must be configured such that Muslims could not plausibly deny rights to Christians even if they wished to do so.

6 URL: https://www.opensocietyfoundations.org/fact-sheets/restrictions-muslim-women-s-dress-28-eu-member-states.

7 When necessary, variables are filled in from their most recent nonmissing value; in practice, this typically means simply carrying forward an average value for one year.

8 Studies of European public opinion have demonstrated a relationship between religious attitudes and immigration attitudes (Bansak et al. 2016; Karyotis and Patrikios 2010) and attitudes towards European integration (Hobolt et al. 2011; Azrout and Wojcieszak 2017).

9 As a robustness check, we also test the theory using survival analysis under the framework of a Cox Proportional Hazards model. Results are robust to this specification; see Figure A2 in the Appendix.

10 These estimates are based on model 6 in Table 1, which includes our full set of control variables.

11 The 10th percentile corresponds to 4% of the population attending weekly (Denmark is in this range for most years), while the 90th percentile corresponds to about 32% weekly attendance (such as Italy in recent years).

12 These figures are taken from the Association of Religion Data Archives (http://www.thearda.com/) and carried forward as necessary.

13 These data are taken from United Nations figures.

14 This index of “General Satisfaction” is derived from a factor analysis of three ESS variables: life satisfaction, economic satisfaction, and government satisfaction.

15 As Table 2 demonstrates, using average levels of self-reported happiness leads to similar results.

16 All of the models in Tables 2 and 3 include the set of controls from model 4 in Table 1; the coefficient estimates for these variables are available in the Appendix.

17 We thank an anonymous reviewer for suggesting this possibility to us.

18 Sampling information for the SHP is available at https://forscenter.ch/wp-content/uploads/2019/01/shp_user-guide-w19.pdf.

19 This variable indicates whether or not a canton held a vote on banning the veil.

20 The canton-level measures of foreign-born and Muslim population shares are taken from the Swiss Federal Statistics Bureau.

21 Indeed, there appears to be at least some relationship between Muslim population share and the likelihood of veil bans in both of our datasets; importantly, however, controlling for Muslim population share does not meaningfully affect our findings regarding attendance.