Electoral participation of Muslims in Europe: assessing the impact of institutional and discursive opportunities

(1)

Article

Reference

Electoral participation of Muslims in Europe: assessing the impact of institutional and discursive opportunities

CINALLI, Manlio, GIUGNI, Marco

Abstract

This paper explores the hypothesis that the electoral participation of Muslims varies according to two main types of opportunities, namely, institutional and discursive opportunities (DOs), characterising their country of residence. More specifically, we assess the impact of institutional opportunities (IOs) in terms of civic and cultural dimensions. We then add up the analysis of DOs in their quality of legitimating public debates over Muslims. We conduct our study by relating survey data to standardised contextual indicators of institutional and DOs in four European countries. The results show that both institutional and DOs have an impact, under control of a number of individual characteristics. However, while the civic and cultural dimensions of IOs have an equally important effect, we find that DOs are less crucial for Muslims' electoral participation. A number of broader implications of these findings are discussed.

CINALLI, Manlio, GIUGNI, Marco. Electoral participation of Muslims in Europe: assessing the impact of institutional and discursive opportunities. Journal of Ethnic and Migration Studies , 2016, vol. 42, no. 2, p. 309-324

DOI : 10.1080/1369183X.2015.1102043

Available at:

http://archive-ouverte.unige.ch/unige:79979

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=cjms20

Download by: [Université de Genève] Date: 18 December 2015, At: 02:26

ISSN: 1369-183X (Print) 1469-9451 (Online) Journal homepage: http://www.tandfonline.com/loi/cjms20

Electoral participation of Muslims in Europe:

assessing the impact of institutional and discursive opportunities

Manlio Cinalli & Marco Giugni

To cite this article: Manlio Cinalli & Marco Giugni (2015): Electoral participation of Muslims in Europe: assessing the impact of institutional and discursive opportunities, Journal of Ethnic and Migration Studies, DOI: 10.1080/1369183X.2015.1102043

To link to this article: http://dx.doi.org/10.1080/1369183X.2015.1102043

Published online: 08 Nov 2015.

Submit your article to this journal

Article views: 24

View related articles

View Crossmark data

(3)

Electoral participation of Muslims in Europe: assessing the impact of institutional and discursive opportunities

Manlio Cinalli^aand Marco Giugni^b

aSciences Po Paris, CEVIPOF, Paris, France;^bDepartment of Political Science and International Relations, University of Geneva, Geneva, Switzerland

ABSTRACT

This paper explores the hypothesis that the electoral participation of Muslims varies according to two main types of opportunities, namely, institutional and discursive opportunities (DOs), characterising their country of residence. More speciﬁcally, we assess the impact of institutional opportunities (IOs) in terms of civic and cultural dimensions. We then add up the analysis of DOs in their quality of legitimating public debates over Muslims. We conduct our study by relating survey data to standardised contextual indicators of institutional and DOs in four European countries. The results show that both institutional and DOs have an impact, under control of a number of individual characteristics.

However, while the civic and cultural dimensions of IOs have an equally important effect, we ﬁnd that DOs are less crucial for Muslims’electoral participation. A number of broader implications of theseﬁndings are discussed.

ARTICLE HISTORY Received 25 July 2014 Accepted 25 September 2015 KEYWORDS

Muslims; electoral participation; institutional opportunities; discursive opportunities

Introduction

The idea that the political context sets the parameters within which political participation and mobilisations occur is by now well established in the scholarlyfield of migration and ethnic relations. In this paper, we assess the extent to which this idea holds as well for the electoral participation of European Muslims in their country of residence. In so doing, we engage with a crucial topic for contemporary European politics. The study of voting can help us to appraise potential participatory gaps between European Muslims, on the one hand, and their fellow citizens, on the other. Thisfirst descriptive step of our analysis is important given that many scholars and commentators have been warning for some time about the numerous threats (political apathyfirst of all) endangering contemporary democracy (Gauchet2007; Todd2008).

However, there is a second explanatory step of our analysis that needs to be singled out.

Drawing upon seminal work on political opportunities and the claims making of Muslims (Koopmans and Statham1999; Koopmans et al.2005), we further develop our reﬂection on the impact of the institutionalised political system side by side with discursive contextual aspects (Cinalli and Giugni2013a). In particular, we test the hypothesis that the electoral participation of Muslims varies according to two main types of opportunities which characterise the context of Muslims’ country of residence: institutional opportunities

CONTACT Manlio Cinalli manlio.cinalli@sciencespo.fr

Downloaded by [Université de Genève] at 02:26 18 December 2015

(4)

(henceforth, IOs), and discursive opportunities (henceforth, DOs). In so doing, we aim to further develop the conceptualisation of the contextual determinants of Muslims’voting, while at the same time providing new empiricalﬁndings about Islam and electoral politics.

Our empirical analysis, based on data collected in EURISLAM research project, deals with the following countries: Britain, Germany, Switzerland, and the Netherlands. The selection of these four countries follows from the fact that they are among the major traditional immigration countries and have an important share of Muslims. In line with main scholarship that has contrasted different types of citizenship models, philosophies, regimes, etc. (Brubaker 1992; Favell 1998; Koopmans et al. 2005), this selection also enables us to build up on the opposition between an assimilationist pole on the one hand (Germany and Switzerland), and a multi-culturist pole on the other (Britain and the Netherlands). Our measure of electoral participation relies on a survey conducted in these countries on representative samples of Muslims from four different origins:

former Yugoslavians, Moroccans, Pakistanis, and Turks. The choice of these groups follows the four selected countries by including a major group of Muslim settlement for each country (former Yugoslavians for Switzerland, Moroccans for the Netherlands, Pakistanis for Britain, and Turks for Germany). At the same time, by taking all fours groups in all countries, we can control for the effect of national origin. In this way we can enter the debate over the negative inﬂuence that certain homeland countries and ethnic identities may have on the capacity to integrate into ‘host’ polities (Brimelow 1995; Huntington2004)

Our contextual analysis of IOs draws upon data on a large number of indicators that summarise relevant state policies and legal frameworks that address Muslims in terms of citizenship, cultural difference, and church–state relations (see Statham and Tillie in the introduction). These indicators are also aggregated in a way toﬁt the typical distinction that previous studies have made between the civic dimension, on the one hand, and the cultural dimension, on the other (Koopmans et al.2005). As regards the contextual analysis of DOs, we draw upon data based on claim making so as to consider the framing of citizenship, cultural difference, and church–state relations in the public domain (see Statham and Tillie in the introduction).

The idea that European Muslims vote in avacuumcan hardly be held in the real world, where actions, exchanges, and discourses of various stakeholders across the public and the policy domain must inevitably concur to inform political behaviour (Cinalli2003,2004).

Here we engage fully with the idea to put Muslims back into their context, so as to assess the extent to which this latter plays a role for explaining their electoral participation.

Through our analysis, however, we also hope to contribute to bridging the gap between scholars working on individual variables at the micro-level and those who deal with contextual determinants at the macro-level. The role of individual characteristics has been acknowledged since the production of seminal works on political participation (Verba and Nie1972; Verba, Nie, and Kim1978), so it needs to integrate the more recent scholarly interest in macro-levelexplanans(Anduiza2002; Franklin2004).

IOs and DOs for the electoral participation of Muslims

Political opportunities have been shown to be of great importance to understand cross- national variations in the levels, forms, and outcomes of social movements and collective

(5)

action (see Kriesi2004 and Meyer 2004 for reviews). In this perspective, scholars have argued that a same grievance may bring about strong contentious dynamics in one country but not in other countries, owing to the lack of political opportunities, for example, in terms of salience of deep cleavages, openness of the political system, strategic alliances, and so forth. More crucially for this article, the opportunity approach has also been applied to thefield of migration and ethnic relations. Of particular interest for us are works focusing on the role of opportunity structures influencing the mobilisation of migrants and ethnic minorities. The main argument in this line of reasoning is that migrants and minorities often lack basic citizenship rights allowing them to exert some leverage on the host society institutions. Hence, the aspects of the political context that affect their mobilisation are to be found not primarily in the overall political system, but rather in the political opportunities that open or close their access to citizenship, including specific migration-related opportunities and opportunities varying across the sub-national level (Koopmans et al.2005).

In spite of its uncontestable merits, however, the concrete application of the concept of political opportunity structure face in our view two main shortcomings and a main challenge. The firs shortcoming lies in the fact that the study of political opportunities has mostly reiterated the classic distinction between individual and collective rights in the field of migration and ethnic relations (Taylor1994; Kymlicka1995) based on the assumption that these two types of rights stand out as independent, often conflicting, dimensions.

Taking a critical stance, in what follows we want to appraise the extent to which the impact of the civic dimension (mostly built on individual rights) goes hand in hand with that of the cultural dimension (mostly built on cultural rights). This interest follows the fact that, in previous studies, we have argued that uneven combinations of individual and collective rights are not that common when looking into the empirical world (Cinalli and Giugni2013a; see also the empty category of‘segregationism’in Koopmans et al.2005, as well as the discussion over the potential oxymoronic nature of French republicanism in Schor2001).

As regards the second shortcoming, this refers to the fact that a systematic treatment of discursive aspects is almost absent in the literature that focuses on contextual determinants of political behaviour. Drawing upon the ‘framing’ scholarship in the ﬁeld of social movement studies (see Snow 2004 and Benford and Snow2000 for reviews), we investigate the role of discourse as well. A number of scholars have argued that policy- makers follow the occurrences in the public domain (Faist1994; Husbands1994), inter- vening in line with dominant anti-immigration and xenophobic rhetoric (Faist 1994;

Thränhardt1995). Yet scholarly research has only recently started to engage with work- able operationalisation of discourse allowing for a systematic cross-national treatment of opportunities (Cinalli and Giugni2013b).

The main challenge is to apply the opportunity approach to the explanation of electoral participation of Muslims at the micro-level, beyond the broader focus on collective mobilisation that is usually kept by scholars of political opportunities. While these latter scholars have sometimes dealt with electoral politics, they have limited their analyses to the meso- level perspective so as to examine the interactions between social movements and election campaigns (Kriesi et al. 1995; McAdam and Tarrow 2010). By focusing on individual Muslims, we can also go beyond an ethnicised understanding that may blur the distinc- tions between citizens, foreigners, and migrants on the simple basis of their common Muslim background.

(6)

Drawing upon scholarship that has emphasised the relevant differences of voting vis-à- vis the broader notion of political participation (Verba, Schlozman, and Brady 1995;

Ramakrishnan2005; Minnite 2009), our analysisﬁts the growing scholarly engagement with electoral politics in theﬁeld of migration and ethnic relations (Fennema and Tillie 1999; Tam Cho 1999; Ramakrishnan and Espenshade 2001; Jang 2009). The main reason to look at the distinctive characteristics of Muslims’ voting is in line with the idea that electoral participation is crucial for governors and policy-makers, who can thus understand better the preferences of the governed (Dagger 1997). Given this crucial function at the core of the whole democratic process, voting is even more important when enabling groups of migrant and various cultural backgrounds to express their preferences and concerns vis-à-vis the autochthonous majority. This is why a large number of democratic countries have granted local voting rights for migrants—or alter- natively advisory boards that give voice to various types of cultural minorities—with the view to maintain the full legitimacy of political institutions.

Although we are more in an explorative than hypothesis-testing mode, we expect that both IOs and DOs will have a significant impact upon the likelihood that Muslims engage in voting activities. Even the specific cultural dimension within the institutional type may have a strong effect for voting if we take in the multiculturalist argument linking group recognition with political integration and full citizenry (Sackman, Peters, and Faist 2003; Modood 2007; Parekh 2008). As regards the shape of the relationship between our independent and dependent variables, some works in thefield of social movements and contentious politics have pointed to a curvilinear relationship between political opportunities and protest: both very closed and very open opportunities constrain protest, while in-between levels encourage the rise of protest actions (Eisinger 1973;

Tilly 1978; Kitschelt1986). Since we are more interested in ascertaining the impact of the openness or closeness of IOs and DOs as such than in discovering patterns of linear or curvilinear relationships, here we assume a linear model, without tackling directly the potential U-curve shape of the relationship between opportunities and electoral participation.

The impact of IOs and DOs will be studied net of certain individual characteristics that the literature has found to affect electoral participation, such as age, gender, and education.

Given our focus on Muslims, we also include additional variables such as religiosity and perceived discrimination. Since the seminal work by Huntington (1996), the common place that Islamic religious beliefs are hardly reconcilable with the basic tenets of Western democracy has also gained increasing scholarly visibility (Gurfinkiel1997; Hun- tington2004). As regards perceived discrimination, some scholars have considered it to be a key variable for analysis, in fact a better predictor than simple socio-economic difficulties that Muslims mayfind within their impoverished and religious-based milieus (Maxwell 2006). We also pay attention to membership in associations. The notion of social integration has been put at the core of a number of analyses that focus on membership in political, social, and cultural associations (Almond and Verba1963; Verba, Schlozman, and Brady1995) as well as social capital (Bourdieu1986; Coleman1988; Putnam1993,2000).

More crucially for this article, scholars of‘ethnic capital’, have argued that diffuse membership in associations can especially foster the political participation of migrants and cultural minorities (Fennema and Tillie 1999; Togeby 2004). Accordingly, our analysis enables us to examine the impact of contextual variables, conceptualised in the two

(7)

main types of IOs and DOs, vis-à-vis a large volume of individual variables referring to the condition of Muslims themselves. At the same time, we can assess the predicting power of these same individual variables while controlling for cross-national differences on key indicators of opportunities.

Data and operationalisation

Our analysis rests on two different data sets generated within the EU-funded project

‘Finding a Place for Islam in Europe’(EURISLAM), a research aimed to investigate cultural patterns and relations of Muslims in several European countries (see Statham and Tillie in the introduction). The individual-level data come from a survey conducted between 2010 and 2011 on random samples of Muslims from four origins (Morocco, Paki- stan, Turkey, and former Yugoslavia). Appendix 1 shows the sample sizes in terms of country and national origins.¹

Our main dependent variable is a measure of electoral participation based on a question asking respondents whether they had voted in the last national elections.²Only respondents eligible to vote in the last national elections were asked this question, so that we could avoid to blur under the ‘no’ responses those who have not voted because they were not allowed to and those who have not done so because of other reasons. The individual-level data also provide the following control variables for our models: age (in years), gender (male as reference category), education (International Standard Classiﬁ- cation of Education 5-point scale), associational membership, religiosity (scale), perceived discrimination, and group. Associational membership has been measured through the following question: ‘Do you join any associations such as sports clubs, religious organiz- ations, labor unions, parents associations, etc. in <Germany/the Netherlands/

Switzerland/the United Kingdom>, where you meet other people?’ Religiosity has been assessed through a scale made of four items: frequency of prayer, frequency of attending a place of worship for religious reasons, self-deﬁnition as a Muslim, and expression or display of religious beliefs (Cronbach’s Alpha for the sample included in the analyses

= .81). As regards perceived discrimination, we rely on the following question: ‘Have you ever experienced hostility or unfair treatment towards you by people in <Germany/

the Netherlands/Switzerland/the United Kingdom?>, Group is a categorical variables aimed to test for differences across the four groups of different origins (former Yugosla- vians, Moroccans, Pakistanis, and Turks, respectively). Finally, in one of the models we control for the country (categorical variable), while in the others the context is measured through our two main types of political opportunities.

Our measures of IOs draw upon extensive data collection that has taken place within the framework of the EURISLAM project (Carol et al.2009; see also Statham and Tillie in the introduction). In particular, we use these data in a way to acknowledge the distinction between the civic and cultural dimensions of IOs for Muslims. The civic dimension includes traditional political indicators on voting rights as well as indicators on the existence, composition, and functioning of Muslim consultative bodies. In addition, we assess the extent to which the principle of anti-discrimination is effectively implemented, dealing with the various spheres that may be covered (education, access to goods and services, employment, social beneﬁts, housing) as well as appraising which types of discrimination count (religion, race and ethnicity, nationality). We also consider the existence and

(8)

functioning of state-established anti-discrimination bodies and their legal mandate. The cultural dimension refers speciﬁcally to the recognition of European Muslims in terms of their religious difference. Accordingly, we have focused on a number of indicators that cover the analysis of state funding of Muslim schools, Islamic teaching in state schools as an alternative to Christian religion, the right of female teachers and students to wear the headscarf in schools, the broadcasting of Islamic religious programmes, as well as the existence of state-paid imams in prisons and the army.

All these indicators have been measured along aﬁve-point scale, on the base of extensive qualitative information for each single indicator. The measures have been computed by attributing a score:−1 corresponds to restrictive policies vis-à-vis Muslims (and therefore closed opportunities) and +1 to liberal policies (and therefore open opportunities), with −0.5, 0, and +0.5 reﬂecting different degrees of intermediate situations. We then summed up the scores of various indicators and calculated averages for each national case so as to use them in the regression analysis. In particular, all scores used in our analyses refer to the situation of policies and arrangements in each country for the year 2002, so as to measure them before the last election referred to in the survey across the four countries (which were held in 2007 in Switzerland, 2009 in Germany, and 2010 in Britain and the Netherlands). The gap between the timing of our scores and the responses to the survey also allows for considering the longue durée that is often needed for the context to have an effect on political behaviour.

For the operationalisation of DOs, we rely on another data set which has been built by following a method known as political claims analysis (Koopmans and Statham1999). We have retrieved claims over Islam by any kind of actors. These claims were coded by random sampling articles selected from ﬁve newspapers in each country and covering the period from 1999 to 2008. The choice of newspapers has followed from the need to insure as much as possible a representative and unbiased sample. We have included both quality newspapers and more tabloid-oriented newspapers. Furthermore, we have considered newspapers from different political orientations as well as more‘neutral’newspapers.³Sampling was based on the following keywords: Islam, Muslim, Moslem, mosque, imam, Qur’an (Quran, Qur’ān, Koran, Alcoran, or Al-Qur’ān), headscarf, burqa (burkha, burka, or burqua), and minaret. Since the data collection involved 13 different researchers for the whole coding, we run reliability tests so as to check for intercoder consistency.⁴

Our measure of DOs relies on a speciﬁc variable, namely, the position of claims towards the object. This measure allows us to assess whether claims stand in favour of or against Muslims, which gives us a good grasp of the legitimating public debates over Muslims (see Statham and Tillie in the introduction). All claims whose realisation implies deterioration in the rights or position of Muslims receive code−1, no matter if the reduction is minor or large. The−1 also goes to claims which express a negative attitude with regard to Muslims (both verbal and physical) or a positive attitude with regard to xenophobic and extreme right groups or aims. All claims whose realisation implies an improvement in the rights and position of Muslims (minor or major) receive code +1. This code also goes to claims expressing (verbally or physically) a positive attitude with regard to European Muslims, or a negative attitude with regard to xenophobic and extreme right groups or aims. Neutral or ambivalent claims which are not necessarily related to any deterioration or improvement in Muslims’position or rights and do not express a clear attitude with regard to migrants and minorities or their opponents receive code 0. By averaging the

(9)

scores thus attributed across all claims, we obtain a raw yet helpful indicator of the discursive context in thisﬁeld: the closer the score of this indicator is to−1 the less favourable are the DOs, and vice versa. Given that we want isolate the impact of DOs on the vote of Muslims, claims by the latter have been excluded from the sample so as to reduce risks to blur independent and dependent variables. Appendix 2 shows the raw opportunity scores. These have then been standardised for the analyses by making them range from 0 to 1.

Since we have a binary dependent variable, we test our hypotheses by means of logistic regression. We examine in aﬁrst step the effect of the context in general terms by including country dummies in a logistic regression predicting the electoral participation of Muslims. We then look at the effects of each type of opportunities (distinguishing between the civic and cultural dimensions of the institutional type) by means of three separate models. We cannot include all opportunity variables in the same model owing to their limited variation. Finally, we look at the results in terms of predicted probabilities, ﬁrst by showing the marginal effects of each opportunity variable, then by providing a more detailed look based on predicted probabilities for voting at the observed values of each variable in a graphical way.⁵Appendix 3 describes the variables included in the analyses. The latter were performed with Stata 12.

Findings

As aﬁrst step of our analysis, we show the share of Muslims who have engaged in voting activities—as measured through participation in the last national elections—across the four countries included in our study. As we can see inTable 1, there are important differences across countries. Speciﬁcally, Dutch Muslims are much more likely to vote than their Swiss counterparts, with nearly 8 respondents out of 10 who said that they took part in the last national election. The other two countries lie somewhere in between, with the share of electoral participation being more or less the same in both Britain and Germany.⁶

We do not observe a coherent pattern reflecting those found, for example, by Koop- mans et al. (2005), namely that more open citizenship regimes—both in terms of individual and group rights—favour increases in terms of political participation of migrants. So the connection that has previously been found between IOs in thefield of immigration and ethnic relations politics on the one hand, and the political participation of migrants on the other, does not seem to apply to the electoral participation of European Muslims, or it does so only to some extent. Dutch Muslims, for example, vote more often than Swiss Muslims, but also more often than British Muslims (although Britain and the Netherlands have similarly favourable opportunities). In addition, thesefindings are crucial as they indicate that electoral participation may follow a distinct pattern compared to other forms of political participation at the meso-level by organised publics. The fact that voting activities of

Table 1.Electoral participation of Muslims across countries (percentages).

Germany Switzerland Britain The Netherlands

Has voted in last national election 65.08 49.64 63.65 78.25

Has not voted in last national election 34.92 50.36 36.35 21.75

Total 100% 100% 100% 100%

N 524 411 762 708

(10)

Muslim citizens in Britain are as low as they are in Germany (and in fact even lower) is thus in contrast with the strong presence of Muslim collective actors and their claims in the British public sphere (Cinalli and Giugni2013a).

Once we have ascertained the distribution of our dependent variable at the aggregate level, the next step of our analysis consists in assessing whether the context plays a role by looking at the effect of the country of residence of Muslims. To do so, we ran a logistic regression with the country variable among the covariates, along with the individual-level controls. The results, displayed inTable 2, obviously reﬂect the distributions shown in the previous table and clearly attest to the important impact of the context for the electoral participation of Muslims, just as this has been found for voters in general (Franklin 1996). Speciﬁcally, Muslims have more chances to vote if they live in one of the three countries other than Switzerland (the reference category). The effect is particularly strong for the Netherlands: the odds of voting for a Dutch Muslim are about three and a half times as high as for a Swiss Muslim.

Thisfirst regression model gives us the opportunity to discuss the role of the control variables. Not surprisingly, age and education have a statistically significant effect on voting. These are variables—especially education—that are typically strong predictors of political participation, including electoral participation. Yet the effects are not very strong, especially so in the case of age. Being a member of associations such as sports clubs, religious organisations, labour unions, and other associations is also a strong predictor of participation, thus confirming the importance of associational involvement for political engagement. The effect of this variable is stronger, suggesting that, in the case of Muslims, relational resources are more important than personal ones for getting involved in politics.⁷We alsofind a significant effect of the group variable, attesting to the fact that the national origin matters. However, this holds only for the former Yugosla- vians, who are less likely to vote than Turks (the reference category). In contrast, sex,

Table 2.Effect of country on electoral participation of Muslims (odds ratios).

Age 1.039 (.005)***

Male 1.075 (.103)

Education 1.234 (.062)***

Religiosity 1.005 (.070)

Generation (ref.:ﬁrst)

In-between 1.103 (.143)

Second 1.121 (.172)

Associational membership 1.515 (.161)***

Group (ref.: Turks)

Former Yugoslavians 0.690 (.096)**

Moroccans 0.940 (.123)

Pakistanis 1.037 (.143)

Country (ref.: Switzerland)

Britain 1.859 (.258)***

The Netherlands 3.629 (.527)***

Germany 1.831 (.261)***

Constant 0.118 (.037)***

Log pseudolikelihood −1365.376

N 2321

Notes: Robust standard errors between parentheses.

*p≤.05.

**p≤.01.

***p≤.001.

(11)

religiosity, generation (second and in-between as opposed toﬁrst generation), as well as the variables pertaining to Moroccans and Pakistanis, are not statistically signiﬁcant.

The absence of an effect for religiosity is worth noting. Religiosity might well be associated with certain cultural traits of Muslims or even certain political attitudes, but apparently it does not affect their electoral participation. Contrary to the commonsensical idea, often also conveyed by the media, the political behaviour of Muslims in Europe does not seem to be dictated by their religious beliefs or practices.

Table 3gives the results of three logistic regressions predicting electoral participation as a function of all of opportunity variables discussed earlier, plus the controls. Here we focus on the contextual variables, as the controls have already been discussed.⁸

The results clearly attest to the relevance of the context for explaining the electoral participation of Muslims. The odds of having voted in the last national elections are more than twice higher for Muslims living in a context characterised by open IOs than for those in a closed context, both in terms of the civic dimension (Model 1) and the cultural dimension (Model 2). While IOs have a positive impact on the likelihood that Muslims engage in voting activities, DOs do not seem to matter as much (Model 3).

The latterfinding does not match with previous research that has stressed the importance of discourse to explain variations of political claims making in thefield of Islam (Cinalli and Giugni2013a). Although we have too little information to explain this mismatch, we can suggest in a speculative manner two main research directions that may be explored in the future. On the one hand, the mismatch may be built into the specific public nature of DOs. Since these latter are made of discourses and symbolic rhetoric that prevail in the public domain, they have their stronger impact upon the actors that most likely access the public domain through their claims making, namely organisations, groups, and other corporate actors. On the other hand, the modest impact of DOs on Muslims’

Table 3.Effect of three opportunity variables on electoral participation of Muslims (odds ratios).

Model 1 Model 2 Model 3

Opportunity variables

IOs (civic dimension) 2.077 (.227)*** – –

IOs (cultural dimension) – 2.020 (.206)*** –

DOs – – 1.454 (.180)**

Controls

Age 1.040 (.005)*** 1.040 (.005)*** 1.042 (.005)***

Male 1.066 (.101) 1.067 (.102) 1.043 (.098)

Education 1.251 (.063)*** 1.249 (.063)*** 1.256 (.063)***

Religiosity 1.032 (.071) 1.030 (.071) 1.048 (.071)

Generation (ref.:ﬁrst)

In-between 1.175 (.150) 1.167 (.150) 1.244 (.158)

Second 1.154 (.174) 1.131 (.171) 1.199 (.179)

Associational membership 1.713 (.174)*** 1.728 (.175)*** 1.746 (.177)***

Group (ref.: Turks)

Former Yugoslavians 0.714 (.098)* 0.715 (.099)* 0.708 (.096)*

Moroccans 1.054 (.135) 1.057 (.135) 1.048 (.134)

Pakistanis 0.985 (.132) 0.980 (.131) 0.968 (.128)

Constant 0.129 (.040)*** 0.134 (.042)*** 0.142 (.046)***

Log pseudolikelihood −1385.367 −1383.566 −1402.207

N 2321 2321 2321

Notes: All opportunity indicators have been standardised on a 0–1 scale. Robust standard errors between parentheses.

*p≤.05.

**p≤.01.

***p≤.001.

(12)

electoral participation could be a good news for those who look with anxiety at the electoral integration of Islam in Europe. That is, DOs may have an important impact only upon certain forms of the action repertoire of Muslims, but these latter would be less sen- sitive to sudden changes of the most unstable component of opportunities—discourse—

when engaging as full citizens with the fundamentals of democracy. Future research, in this case, may test whether DOs are especially crucial for explaining political behaviour of other less integrated individuals and groups.

Odds ratios can hardly be compared across models, although in this case the three opportunity variables have been standardised as 0–1 scales and all other variables in the models are the same. Therefore, we go a step further by showing predicted probabilities, which can be compared and are easier to interpret.Table 4shows the marginal effects of the three opportunity variables, keeping all other variables at their means.⁹ The most important effect is observed for the civic dimension of IOs: going from the lowest to the highest value of the score leads to a 16% increase in the predicted probabilities of voting. The effect of the cultural dimension is nearly equally strong, with a change of 15%. Finally, here we can see that the effect of DOs, while being positive, is half the size of those of IOs, corresponding to 8%. In brief, all three opportunity variables increase the likelihood of Muslims engaging in voting activities, but the impact is much stronger for the institutional type than for the discursive type.

Figure 1gives us a more detailed picture by showing the predicted probabilities of electoral participation for each opportunity variable at their observed values, based again on estimates from the logistic models shown earlier. Again, the other variables are held at their means. We can see in the three graphs that the predicted probabilities increase when we go from a more closed to a more open context, and this holds across all opportunity variables. The stronger effect of IOs as compared to DOs, however, is reﬂected in the higher slope of the lines in the related graphs. This corresponds to the greater marginal effects of both dimensions of IOs observed earlier (which were the changes in predicted probabilities when moving from 0 to 1 on the opportunity variables).

The additional information here, as compared to the previous results, is that we can see how the predictive margins of electoral participation change when moving from one value of the contextual variables to the other. Moreover, we can see whether the observed changes in the predictive margins are signiﬁcant or not. This varies depending on the speciﬁc contextual variable at hand. The greatest change on the civic dimension of IOs occurs between the second and the third opportunity scores. This can partly be explained by the fact that the largest variation in the opportunity score occurs between these two

Table 4.Marginal effects of opportunity variables on electoral participation of Muslims.

Average marginal effect

95% conﬁdence interval lower bound

95% conﬁdence interval upper bound

IOs (civic dimension) .16*** .11 .21

IOs (cultural dimension)

.15*** .11 .20

DOs .8*** .3 .14

Notes: Conditional marginal effects based on logistic regressions shown inTable 3. Variables are set at their means.

*p≤.05.

**p≤.01.

***p≤.001

(13)

Figure 1.Predicted probabilities of electoral participation of Muslims at observed opportunity scores.

Notes: Predictive margins with 95% conﬁdence intervals based on logistic regressions shown inTable 3.

Other variables are set at their means.

(14)

cases, but also stems from the strong effect that this contextual dimension has on electoral participation. This change, moreover, is statistically significant, as we can see in the confidence intervals which do not overlap. In contrast, the changes observed between thefirst and the second, as well as between the third and the fourth opportunity scores, are not significant, as can be seen in the overlapping confidence intervals. The same patterns holds for the other cultural dimension of IOs, both concerning where the greatest change occurs and where the significant differences are. Finally, as we now know, DOs display somewhat a different pattern. Here the changes in the predictive margins are more homogeneous across the four opportunity scores owing to their more regular dis- tance between each other. Most importantly, only the difference between the more closed and the more open context are statistically significant, meaning that in order for DOs to matter one needs to move from a very closed to a very open context.

Conclusion

In this paper, we have built upon recent scholarly work in the field of Islam with the specific aim to appraise the impact of certain features of the context on the electoral participation of Muslims in Europe. The main challenge for us has been to implement a research design that has included contextual variables pertaining to institutional and discursive types of opportunities. Overall, our analysis has shown that both types of opportunities matter. Yet, we have found that DOs are less crucial for Muslims’ electoral participation than IOs. Thisfinding is in contrast with what we found in previous research on Muslims’ claims making in the public space (Cinalli and Giugni 2013a). We have argued that this is a good news for those who believe in the integration of Muslims as full citizens, since it implies that they are likely to ignore sudden changes of discourse

—such as those that are typically pushed forward by anti-migrant and right-wing actors

—when going to the polls.

At the same time, we have emphasised that there are still crucial differences across countries when considering performances in terms of voting. So Dutch Muslims are much more likely to vote than their Swiss counterparts, while British and German Muslims lie somewhere in between. Looking at the micro-level variables, age and education are significant, yet with not very strong effects. Other individual variables are not statistically significant. By contrast, associational membership is a good predictor of participation, thus showing that, at least according to our analysis, relational resources are more important than individual ones for the electoral participation of Muslims. National origins do not show a significant effect, with the exception of former Yugoslavians.

We can stress a number of broader implications of ourﬁndings. We have shown that more knowledge is needed about the different channels of political access for European Muslims, and for the whole European citizenry more broadly. In particular, given that DOs seem to play a very different role for electoral and non-electoral participation, scholars are called to go beyond an approach that looks at the same determinants to explain voting and contentious action (but see Pattie, Seyd, and Whiteley 2003; Teorell, Sum, and Tobiasen 2007; Stolle and Hooghe 2011). We have also shown that one cannot assume a strong continuity between Muslim migrants on the one hand, and the European citizens of Muslim faith that spring out of them on the other, particularly when focusing on issues of political behaviour. So the explanation of voting of European Muslims may be

(15)

out of reach for scholars using only conceptual and operational tools drawn from the migrationﬁeld.

Lastly, our results can enhance a speciﬁc theoretical debate about political participation that deserves further space for discussion. We have shown that the probabilities of voting increase when moving from a more closed to a more open situation of opportunities.

Focusing in particular on IOs, the probabilities of voting have the strongest growth for low levels of opportunity scores. This indicates that, in line with early curvilinear concep- tualisations of political opportunities (Eisinger 1973; Tilly1978; Kitschelt 1986), small openings of IOs for European Muslims may already be sufﬁcient to stimulate their political participation through voting. This is an evidence that even a little policy intervention can go a long way to promote a more inclusive and participatory type of democracy.

Disclosure statement

No potential conﬂict of interest was reported by the authors.

Notes

1. Respondents were left the choice between the language of the host country and their native language.

2. The focus on elections allows us to take into account the same comparable voting events in each country, with no consideration of speciﬁc institutional features, such as, for example, the presence of direct democracy (voting on popular initiatives and referenda) in Switzerland.

3. The following newspapers have been used as a source for the coding:De Volkskrant,Trouw, NRC Handelsblad, De Telegraaf, and Het Paroolin the Netherlands; Neue Zürcher Zeitung, Blick,Tagesanzeiger,Le Matin, andLe Tempsin Switzerland;Bild,Süddeutsche Zeitung,Frank- furter Allgemeine Zeitung,Welt, andTagesspiegelin Germany;Daily Mail,Daily Mirror,The Guardian,The Sun, andThe Timesin Britain;Het Laatste Nieuws,Le Soir,Gazet Van Antwer- pen,La Dernière Heure, andDe Standaardin Belgium.

4. These tests have yielded a strong consistency both with regard to the selection of claims and their description. The Cronbach alpha for selection bias (computed on a sample of 15 articles) is .91. The Cronbach alpha for description bias (computed on a sample of 4 articles) is, respectively, .97, .98, .98, and .98, for an average of 0.98 (see Cinalli and Giugni2013bfor further meth- odological remarks).

5. Marginal effects and predicted probabilities are based on the regression models and have variables set at their means. Since we have separate samples for each country which have then be pooled in a same data set, we estimate robust standard errors, hence allowing to relax the assumption that error terms are identically distributed (homoskedasticity). Ideally, we would have estimated cluster-robust standard errors so as to also relax the assumption that error terms are independently distributed (serial correlation). However, with only four clusters (countries) the treatment might be worse than the illness and we therefore avoid that. As a result, we pay more attention to the magnitude of the effects than to the signiﬁcance level, as the standard errors in our regression models might be underestimated, leading to too small con- ﬁdence intervals.

6. Theseﬁgures are quite high. As is known, questions about voting are subject to responses dictated by social responsibility as people tend to overestimate their own participation so as to covey the image of good citizens. In our case, however, this is not a problem a far as we assume that this effect is constant across respondents.

7. It should be stressed that the relationship between associational membership and voting is subject to a strong risk of reverse causality. One is therefore only observing a correlation

(16)

between these two variables rather than causality. In our case, however, this is not so proble- matic as we are only using associational membership as a control variable.

8. For the reasons mentioned earlier, beyond statistical signiﬁcance, we are especially interested in the magnitude of the effects (as measured through odd ratios, marginal effects, and predicted probabilities).

9. The observed values here refer to the standardised opportunity scores.

References

Almond, Gabriel A., and Sidney Verba. 1963.The Civic Culture. Princeton: Princeton University Press.

Anduiza, Eva. 2002.“Individual Characteristics, Institutional Incentives and Electoral Abstention in Western Europe.”European Journal of Political Research41 (5): 643–673.

Benford, Robert D., and David A. Snow. 2000. “Framing Processes and Social Movements: An Overview and Assessment.”Annual Review of Sociology26: 611–639.

Bourdieu, Pierre. 1986. “The Forms of Capital.”In Handbook of Theory and Research for the Sociology of Education, edited by J. G. Richardson, 241–258. New York: Greenwood.

Brimelow, P. 1995.Alien Nation: Common Sense about America’s Immigration Disaster. New York:

Random House.

Brubaker, Rogers. 1992. Citizenship and Nationhood in France and Germany. Cambridge, MA:

Harvard University Press.

Carol, Sarah, Manlio Cinalli, Zuhal Kavacik, Ruud Koopmans, and Lucy G. Maas. 2009. Final Integrated Report Work Package 1.

Cinalli, Manlio. 2003. “Socio-politically Polarized Contexts, Urban Mobilization and the Environmental Movement: A Comparative Study of Two Campaigns of Protest in Northern Ireland.”International Journal of Urban and Regional Research27 (1): 158–177.

Cinalli, Manlio. 2004.Horizontal Networks vs. Vertical Networks in Multi-organisational Alliances:

A Comparative Study of the Unemployment and Asylum Issue-Fields in Britain. EurPolCom Working Papers 8: 1–25. Leeds: University of Leeds.

Cinalli, Manlio, and Marco Giugni. 2013a. “Political Opportunities, Citizenship Models and Political Claim-Making over Islam.”Ethnicities13 (2): 147–164.

Cinalli, Manlio, and Marco Giugni. 2013b.“Public Discourses about Muslims and Islam in Europe.”

Ethnicities13 (2): 131–146.

Coleman, James S. 1988.“Social Capital in the Creation of Human Capital.”American Journal of Sociology94: S95–S120.

Dagger, Richard. 1997. Civic Virtues: Rights, Citizenship, and Republican Liberalism. Oxford:

Oxford University Press.

Eisinger, Peter K. 1973.“The Conditions of Protest Behavior in American Cities.”The American Political Science Review67 (1): 11–28.

Faist, Thomas. 1994.“How to Deﬁne a Foreigner? The Symbolic Politics of Immigration in German Partisan Discourse, 1978–1992.”West European Politics17 (2): 50–71.

Favell, Adrian. 1998.Philosophies of Integration: Immigration and the Idea of Citizenship in France and Britain. Basingstoke: Palgrave.

Fennema, Meindert, and Jean Tillie. 1999. “Political Participation and Political Trust in Amsterdam. Civic Communities and Ethnic Networks.” Journal of Ethnic and Migration Studies25 (4): 703–726.

Franklin, Mark N. 1996.“Electoral Participation.”InComparing Democracies: Elections and Voting in Global Perspective, edited by Lawrence Leduc, Richard G. Niemi, and Pippa Norris, 216–235.

Thousand Oaks, CA: Sage.

Franklin, Mark N. 2004.Voter Turnout and the Dynamics of Electoral Competition in Established Democracies since 1945. Cambridge: Cambridge University Press.

Gauchet, Marcel. 2007.La Démocratie d’une crise à l’autre. Paris: Cécile Defaut.

Gurﬁnkiel, Michel. 1997. “Islam in France: The French Way of Life Is in Danger.”Middle East Quarterly4 (1): 19–29.

(17)

Huntington, Samuel P. 1996.The Clash of Civilizations and the Remaking of the World Order.New York: Simon & Schuster.

Huntington, Samuel P. 2004. Who Are We? The Challenges to America’s National Identity.

New York: Simon & Schuster.

Husbands, Charles T. 1994. “Crises of National Identity as the ‘New Moral Panics’: Political Agenda-Setting about Deﬁnitions of Nationhood.”New Community20 (2): 191–206.

Jang, Seung-Jin. 2009.“Get Out on Behalf of Your Group: Electoral Participation of Latinos and Asian Americans.”Political Behavior31 (4): 511–535.

Kitschelt, Herbert. 1986. “Political Opportunity Structures and Political Protest: Anti-nuclear Movements in Four Democracies.”British Journal of Political Science16 (1): 57–85.

Koopmans, Ruud, and Paul Statham. 1999.“Political Claims Analysis: Integrating Protest Event and Political Discourse Approaches.”Mobilization4 (2): 203–221.

Koopmans, Ruud, Paul Statham, Marco Giugni, and Florence Passy. 2005.Contested Citizenship:

Immigration and Cultural Diversity in Europe. Minneapolis: University of Minnesota Press.

Kriesi, Hanspeter. 2004.“Political Context and Opportunity.”InThe Blackwell Companion to Social Movements, edited by David A. Snow, Sarah Soule, and Hanspeter Kriesi, 67–90. Oxford:

Blackwell.

Kriesi, Hanspeter, Ruud Koopmans, Jan Willem Duyvendak, and Marco Giugni. 1995.New Social Movements in Western Europe. Minneapolis: University of Minnesota Press.

Kymlicka, Will. 1995.Multicultural Citizenship: A Theory of Liberal Rights. Oxford: Clarendon.

Maxwell, Rahsaan. 2006.“Muslims, South Asians and the British Mainstream: A National Identity Crisis?”West European Politics29 (4): 736–756.

McAdam, Doug, and Sidney Tarrow. 2010.“Ballots and Barricades: On the Reciprocal Relationship Between Elections and Social Movements.”Perspectives on Politics8 (2): 529–542.

Meyer, David S. 2004.“Protest and Political Opportunities.”Annual Review of Sociology30: 125–145.

Minnite, Lorraine C. 2009. “Lost in Translations? A Critical Reappraisal of the Concept of Immigrant Political Incorporation.” In Bringing Outsiders In Transatlantic Perspectives on Immigrant Political Incorporation, edited by J. L. Hochschild and J. Mollenkopf, 48–59.

Ithaca, NY: Cornell University Press.

Modood, Tariq. 2007.Multiculturalism. Cambridge: Polity Press.

Parekh, Bhikhu. 2008.A New Politics of Identity: Political Principles for an Interdependent World.

Basingstoke: Palgrave.

Pattie, Charles, Patrick Seyd, and Paul Whiteley. 2003. “Citizenship and Civic Engagement:

Attitudes and Behaviour in Britain.”Political Studies51 (3): 443–468.

Putnam, Robert. 1993.Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ:

Princeton University Press.

Putnam, Robert. 2000. Bowling Alone: The Collapse and Revival of American Community.

New York: Simon & Schuster.

Ramakrishnan, S. Karthick. 2005.Democracy in Immigrant America: Changing Demographics and Political Participation. Redwood City, CA: Stanford University Press.

Ramakrishnan, S. Karthick, and Thomas J. Espenshade. 2001. “Immigrant Incorporation and Political Participation in the United States.”International Migration Review35 (3): 870–909.

Sackman, Rosemarie, Bernhard Peters, and Thomas Faist, eds. 2003. Identity and Integration:

Migrants in Western Europe. Farnham: Ashgate.

Schor, Naomi. 2001.“The Crisis of French Universalism.”Yale French Studies100: 43–64.

Snow, David A. 2004. “Framing Processes, Ideology, and Discursive Fields.” In The Blackwell Companion to Social Movements, edited by David A. Snow, Sarah Soule, and Hanspeter Kriesi, 380–412. Oxford: Blackwell.

Stolle, Dietlind, and Marc Hooghe. 2011.“Shifting Inequalities: Patterns of Exclusion and Inclusion in Emerging Forms of Political Participation.”European Societies13 (1): 119–142.

Tam Cho, Wendy K. 1999.“Naturalization, Socialization, Participation: Immigrants and (Non-) Voting.”The Journal of Politics61 (4): 1140–1155.

Taylor, Charles. 1994.Multiculturalism. Princeton, NJ: Princeton University Press.

(18)

Teorell, Jan, Paul Sum, and Mette Tobiasen. 2007. “Participation and Political Equality: An Assessment of Large-Scale Democracy.” In Citizenship and Involvement In European Democracies: A Comparative Analysis, edited by Jan Van Deth, José R. Montero, and Anders Westholm, 384–414. London: Routledge.

Thränhardt, Dietrich. 1995.“The Political Uses of Xenophobia in England, France and Germany.” Party Politics1 (3): 321–343.

Tilly, Charles. 1978.From Mobilization to Revolution. Reading, MA: Addison-Wesley.

Todd, Emmanuel. 2008.Après la démocratie. Paris: Folio.

Togeby, Lise. 2004.“It Depends…How Organisational Participation Affects Political Participation and Social Trust among Second-Generation Immigrants in Denmark.”Journal of Ethnic and Migration Studies30 (3): 509–528.

Verba, Sidney, and Norman H. Nie. 1972.Participation in America: Political Democracy and Social Equality. New York: Harper & Row.

Verba, Sidney, Norman. H. Nie, and Jae-on Kim. 1978. Participation and Political Equality. A Seven-Nation Comparison. New York: Cambridge University Press.

Verba, Sidney, Kay L. Schlozman, and Henry E. Brady. 1995.Voice and Equality. Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press.

Appendix 1. Sample sizes

Britain Germany The Netherlands Switzerland Total

Former Yugoslavians 200 255 151 250 856

Moroccans 200 256 250 182 888

Pakistani 350 162 152 150 814

Turks 350 355 250 281 1236

Total 1100 1028 803 863 3794

Appendix 2. Raw opportunity scores

Britain Germany The Netherlands Switzerland

IOs (civic dimension) .12 −.23 .21 −.16

IOs (cultural dimension) .60 −.22 .74 −.19

DOs .14 −.37 .00 −.16

Appendix 3. Descriptives

N Mean SD Min. Max.

Electoral participation 2405 0.659 0.474 0 1

Age 3767 38.226 13.162 17 79

Male 3794 0.536 0.499 0 1

Education 3653 2.754 1.004 0 4

Religiosity 3770 0.174 0.746 −1.375 1.850

In-between generation 3793 0.270 0.444 0 1

Second generation 3793 0.231 0.421 0 1

Associational membership 3771 0.309 0.462 0 1

IOs (civic dimension) 3794 0.480 0.416 0 1

IOs (cultural dimension) 3794 0.465 0.453 0 1

DOs 3794 0.538 0.390 0 1