Evaluating the Political Organisation of Muslim Citizens in the Netherlands : The Role of Political Orientation , Education and Multiculturalism

Political participation is an important aspect of the integration of Muslim citizens into western societies. However, Muslims’ formal political participation is often met with resistance from the majority population. In two studies, we investigated among national samples of majority Dutch whether the level of resistance to the democratic political organizations of Muslim citizens is associated with political orientation and level of education. Furthermore, we examined whether these associations are mediated by the endorsement of multiculturalism. Findings from both studies show that the higher educated and politically left-wing individuals are more supportive of Muslim political organisation and that this association is (partly) explained by the endorsement of multiculturalism. Additionally, in Study 2 it is found that attitudes toward cultural tradition and group equality mediate the relations between education and political orientation with multiculturalism.

The rising number of Muslim citizens in Western societies makes the question about their formal political participation increasingly important.There are several ways in which Muslim citizens can have an influence on party politics.One has to do with the "minority electorate" (or 'minority vote') and various studies in different countries have examined the party preference and voting behavior of immigrant-origin groups (e.g., Dancygier & Saunders, 2006;Goldsmith & Holzner, 2015;Voicu & Comşa, 2014).A second one relates to the so-called descriptive representation of minority members in existing political parties due to preferential "co-ethnic voting" based on religious or cultural affinity and minority group-based interests (e.g., Bloemraad, 2013;Schildkraut, 2013;Sobolewska, 2013).
In countries with a multi-party political system there is the third possibility of participating in democratic politics with a religious or ethnic minority political party (group representation).For example, the democratic system in Journal of Social and Political Psychology jspp.psychopen.eu| 2195-3325 the Netherlands -where the current research was conducted -is based on proportional representation which allows various (cultural) groups within the population to be represented in the political institutions of society.Indeed, thirteen political parties are currently represented in Dutch parliament, including small orthodox Christian parties and one-issue parties (e.g.party for the elderly).Thus, the Dutch political system offers the opportunity for Muslim citizens to organize themselves politically, like Christian and one-issue parties.In fact, and similar to for example Spain, Belgium and Denmark, there are and have been several political parties in the Netherlands that participate in politics by explicitly presenting themselves as Islamic, or that claim to draw inspiration from Islam.Some of these political parties have secured seats in municipality councils in the last Dutch local elections (2018; e.g., "NIDA" in Rotterdam, and "Islam Democrats" in The Hague) and in the national parliament in the national election of March 2017 ("DENK").This development of increasing political group representation can be considered an important aspect of Muslim citizens' integration into the host society.It can improve their societal position and enhances the legitimacy of the democratic political system (Bloemraad, 2013;Bloemraad & Schönwälder, 2013;Petrusevska, 2009).However, Muslims' political participation is often met with majority resistance and doubts about their national loyalty and entitlement to be included in the political decision making process (Bloemraad & Schönwälder, 2013;Petrusevska, 2009;Sniderman & Hagendoorn, 2007).This negative attitude may affect the democratic process and can lead to feelings of exclusion and inequality among Muslim citizens who face strong negative sentiments in western societies (e.g., Sniderman & Hagendoorn, 2007;Strabac & Listhaug, 2008).
However, there is little systematic empirical research on the attitudes of majority members toward the democratic political participation of Muslim citizens and how these attitudes can be understood (Hindriks, Verkuyten, & Coenders, 2015;Verkuyten, Hindriks, & Coenders, 2016).In the current research we examine in two studies Dutch majority members' evaluation of the democratic political organization of Muslims in relation to their political orientation, educational level and endorsement of multiculturalism.In Western societies, political and educational polarization represent two separate and important societal cleavages, with people on the political right and the political left, as well as the higher and lower educated, being increasingly apart attitudinally, ideologically and in terms of lifestyle (e.g., Bovens, Dekker, & Tiemeijer, 2014;Layman, Carsey, & Horowitz, 2006).For example, there is consistent and robust evidence that the rejection of immigrants and cultural diversity is stronger among the politically right than among the left (e.g., Hainmueller & Hopkins, 2014;Mudde, 2007;Sears & Henry, 2003), and among the lower compared to the higher educated (e.g., Coenders & Scheepers, 2003;Jenssen & Engesbak, 1994;Meeusen, De Vroome, & Hooghe, 2013).There is no research, however, that examines the concurrent role of political orientation and education in relation to the evaluation of the political organization of Muslim citizens.
Our aim is to examine whether the support for Muslim political organization is associated with majority members' political orientation and educational level.Thus, we want to examine how two important societal cleavages relate to individual attitudes and beliefs.The second aim is to examine whether the endorsement of multiculturalism plays a mediating role in these relations.The ideology of multiculturalism features prominently in Western debates about immigration and cultural diversity (Bloemraad, Korteweg, & Yurdakul, 2008) and the support of multiculturalism has been examined in the social and political psychology (Deaux & Verkuyten, 2014;Rattan & Ambady, 2013), but not in relation to the important question of Muslims' political organisation.The debate about multiculturalism is mainly concerned with identity and cultural differences rather than about the 'loftier goals of political participation' (Joppke, 2010, p. 147).Hence, it is unclear whether majority members' support of multiculturalism is also meaningful for their attitude towards the democratic political organization of Muslim citizens.

Political Orientation
Political orientation can help to explain "why people do what they do, as it organizes their values and beliefs and leads to political behaviour" (Jost, 2006, p. 653).Although the meaning of a left-wing and right-wing political orientation has changed over the years and differs across countries, the motivated social cognition model (Jost, Kruglanski, Glaser, & Sulloway, 2003) argues that there are two psychological aspects distinguishing between left-wing and right-wing people: attitudes towards traditions (versus change) and attitudes towards equality (versus dominance).People on the political right tend to endorse traditionalism and conformity, and tend to justify inequalities between individuals and groups.In contrast, a left-wing orientation is associated with openness to experiences and preferences for greater equality and less group dominance.Multiculturalism and minority political representation imply cultural changes and greater equality for Muslim citizens and this is difficult to reconcile with, for example, social conformity and social dominance beliefs (Duckitt, Wagner, du Plessis, & Birum, 2002).Thus we expected left-wing oriented people to be more likely to support the political organisation of Muslim citizens than right-wing oriented people, and that this is (partly) due to their stronger endorsement of multiculturalism.

Education
Education is an increasingly important and separate dividing line in western societies, including the Netherlands (Bovens et al., 2014).The association between higher education and more positive attitudes towards immigrants and ethnic and religious minority groups is one of the most replicated findings in the social sciences (Jenssen & Engesbak, 1994).This association has been predominantly explained in terms of cognitive abilities and perceived competition (Coenders & Scheepers, 2003;Meeusen et al., 2013).
First, education is associated with cognitive ability and flexibility (Bobo & Licari, 1989;Ohlander, Batalova, & Treas, 2005) making the higher educated better able to understand the importance of basic norms of equality and tolerance underlying the democratic culture (Vogt, 1997).Furthermore, higher levels of cognitive ability make it easier to generalize the principles of tolerance and equality to minority groups (Gaasholt & Togeby, 1995).Through education, people acquire knowledge and information that increases their understanding of the complexity of society and broadens their social perspective.This gives them the ability to try to avoid generalizations such as minority group stereotypes, and increases their understanding of different norms and values, which is associated with higher levels of tolerance (e.g., De Witte, 1999;Stephan & Stephan, 1984).For example, Bobo and Licari (1989) found that cognitive ability is associated with higher levels of political tolerance, while Lipset (1981) and Stouffer (1955) found a positive association between education and support for civil liberties for minorities.
Second, according to ethnic competition theory (Scheepers, Gijsberts, & Coenders, 2002), the higher educated are less likely to face and perceive competition over scarce resources (e.g.housing, jobs) from immigrant and minorities.Furthermore, while the higher educated can benefit from globalization processes, the lower educated with their more vulnerable position are more likely to be negatively affected by increasing globalization with its higher movement and inflow of people from abroad.As a result, the lower educated more strongly feel the need to reduce uncertainties and therefore rely more on their cultural worldview and traditions to provide security (Bekhuis, Lubbers, & Verkuyten, 2014).Thus, for the majority members we expected higher education to be associated with stronger endorsement of the political organization of Muslim citizens through stronger endorsement of multiculturalism.

Multiculturalism
We examined the mediating role of multiculturalism as a diversity ideology because multiculturalism is considered a hierarchy attenuating ideology (Sidanius & Pratto, 1999).Multiculturalism emphasizes the recognition of cultural group differences together with the importance of full and equal participation of all cultural groups in society (Berry, 2006;Guimond, de la Sablonnière, & Nugier, 2014).Thus, from the perspective of majority members multiculturalism is about cultural openness and change (rather than societal traditions) and the promotion of group equality (rather than group dominance).This means that the ideological content of multiculturalism corresponds with the core attitude dimensions of traditionalism (vs.change) and equality (vs.dominance) underlying political orientation and education.Research has found a positive link between education and multiculturalism, and between left-wing political orientation and multiculturalism (e.g., Breugelmans & Van de Vijver, 2004;Breugelmans, Van de Vijver, & Schalk-Soekar, 2009;Hooghe & De Vroome, 2015).Furthermore, support for multiculturalism has been found to be associated with the endorsement of minority rights (e.g., Levin et al., 2012;Verkuyten, 2006).Therefore, we expected that majority members who more strongly endorse multiculturalism will support the political organisation of Muslim citizens more, and that multiculturalism will mediate the association between political orientation and education with support of Muslim political organization.

To Summarize
We expected higher educated and politically left-wing oriented individuals to be more likely to support Muslims' democratic political organisation, and that their higher level of support can be explained by their stronger endorsement of multiculturalism.Two studies were conducted to test these expectations.Study 1 used a relatively small national sample of Dutch majority respondents and in Study 2 we tried to replicate the findings in Study 1 among a larger, representative sample.Thus, Study 2 responds to the ongoing debate about replications in social psychology and political science (King, 1995;Simons, 2014), and the argument that reproducibility is critical for scientific evidence.In addition, Study 2 tested whether the association between political orientation and education, on the one hand, and endorsement of multiculturalism, on the other, can indeed be explained by attitudes towards cultural tradition (versus change) and equality (versus dominance) as two underlying dimensions (Jost et al., 2003).This allows us to develop a further understanding of the psychological reasons why majority members endorse multiculturalism.

Survey and Participants
Data for the first study were obtained from an online survey on 'Societal developments in the Netherlands', carried out by TNS/NIPO on behalf of Utrecht University.
i TNS/NIPO is one of the main (political) survey agencies in the Netherlands that maintains a large database of self-identified ethnic Dutch individuals.This online panel is not based on self-selection and is representative for gender, age, education and region.For each research a random sample of the panel is invited by email to participate.For the current research the response rate was 51% which is normal for survey research in the Netherlands (see Stoop, 2005).The sample contained 173 self-identified Dutch majority respondents who themselves and their parents were born in the Netherlands.
ii Subsequently, three persons were dropped as they had not answered questions on their level of education.Almost half of the sample (N = 170) was female (48.8%), and the age of the respondents ranged from 19 to 81 years (M = 51.83,SD = 16.38).On a question of religious belonging, none of the respondents indicated that they were Muslims.

Measurements
Endorsement of multiculturalism and support for Muslims' political organisation were measured on 7-point Likert scales (1 = 'totally disagree' to 7 = 'totally agree'.Support for Muslims' political organisation was assessed by three items taken from previous research (Verkuyten, 2017): 'Islam has to have its own voice in political issues, just like other religions', 'It is important for Muslims that an Islamic political party will be established in the Netherlands', and 'Muslims have to cooperate more in order to get more influence in Dutch society'.Support for multiculturalism was measured with eight items that have been previously used in the Netherlands (Verkuyten, 2006; see Appendix).
Education was based on the highest completed level (e.g., Meeusen et al., 2013).Following the Dutch education system and previous studies in the Netherlands (e.g., Tolsma, de Graaf, & Quillian, 2009) seven levels of educational attainment were distinguished: (1) primary education (2.9%), (2) low levels of secondary education (8.8%), (3) intermediate levels of secondary education (12.9%), (4) lower levels of tertiary education (32.9%), ( 5) upper levels of secondary education and propaedeutic exam of (applied) university (12.9%), (6) Bachelor's degree from (applied) university (25.9%), and (7) Master's degree from university (3.5%).The distinction between these levels of achieved education is comparable to the international ISCED-measure that is used, for example, in the European Social Survey.Similar to other research in the Netherlands (e.g., De Graaf, De Graaf, & Kraaykamp, 2000; Van Tubergen & Van de Werfhorst, 2007), education was treated in the analysis as a continuous variable which allows to investigate the difference between lower and higher educated participants.
Scores were recoded such that a higher score indicated a more left-wing orientation.Previous studies have shown that the measure has good construct validity (see Jost, 2006) but because it is a single item a correction was made for possible measurement error (Wang & Wang, 2012).Based on multiple item-measures of political orientation (e.g., Fibert & Ressler, 1998;Kemmelmeier, 1997), a reliability of .50 for the single item was assumed.iv In the analyses age and gender were taken into account.Further, majority members in the lower strata of society tend to perceive more ethnic minority threats (e.g., Scheepers et al., 2002).Therefore, two indicators of socioeconomic status, household income and employment status (1 = employed, 0 = unemployed/other), were controlled for.To measure household income, participants were asked to indicate their annual gross household income by choosing one of 27 categories (ranging from < €4.600 (1) and €4.600 -€6.300 (2) to €284.500 -€310.700( 26) and > €310.700 ( 27)).

Measurement Model
Table 1 shows the model fit statistics of the full measurement models.An overall confirmatory factor analysis including multiculturalism and political organisation of Muslim citizens was fitted by maximum likelihood.This model fitted the data poorly (Model 1.0) and modification indices suggested that the multiculturalism item on broadcasting of television programs for and from immigrant minorities (see Appendix) cross-loaded on the measure of support for Muslims' political organisation.Therefore this item was dropped.Note.SCF = Scaling Correction Factor for the χ 2 -test of model fit using MLR as the estimator.
The modification indices further indicated that two multiculturalism items should be allowed to covary.This resulted in a model that fitted the data well (Model 1.2).In this final measurement model, support for multiculturalism was measured with seven items that had a high composite reliability (ρ = .857;Kline, 2012).Support for Muslims' political organisation was measured with three items and this scale also was reliable (ρ = .810).

Plan of Analysis
A structural model was estimated using Structural Equation Modelling (SEM) in Mplus (version 7.3;Muthén & Muthén, 2014) with maximum likelihood with robust standard errors (MLR) as the estimator.
v To investigate the mediation hypothesis, the direct effects from education and political orientation on support for political organisation of Muslim citizens were modelled, as well as the indirect effects through endorsement of multiculturalism.To account for missing values, political orientation and household income were endogenized, assuming bivariate normality conditional on the control variables in the model (Enders, 2010). vi

Descriptive Findings
Table 2 shows the descriptive findings for the main variables.A set of Wald's tests showed that the estimated mean of endorsement of multiculturalism did not significantly differ from the neutral mid-point of the scale (Wald χ 2 (1) = 0.885, p = .347),while it was below the mid-point on support for Muslim political organisation (Wald χ 2 (1) = 96.093,p < .001).This latter finding indicates that on average respondents were rather negative towards the democratic political organization of Muslim citizens.All correlations were in the expected directions.Both education and political orientation were positively correlated with endorsement of multiculturalism and support for Muslims' political organisation.Multiculturalism and political organisation were also strongly correlated.Furthermore, education and political orientation were not associated, supporting the proposition that both are separate dividing lines in Dutch society (Bovens et al., 2014).

Support for Muslims' Political Organisation
Figure 1 shows the findings for the hypothesized relations between education, political orientation, endorsement of multiculturalism and support for Muslims' political organisation, while controlling for gender, age, employment status and household income (model fit is reported in Table 1, Model 1.4).As expected, higher educated individuals and politically left-wing oriented individuals were more likely to endorse multiculturalism as a diversity ideology.In turn, endorsement of multiculturalism was positively related to support for the political organisation of Muslim citizens.The indirect effects of political orientation (β = .361,p < .01)and education (β = .109,p < .05),and the non-significant direct effects on support for Muslims' political organisation indicate that multiculturalism fully mediated the relations between education and political orientation, on the one hand, and support for Muslims' political organisation, on the other (see Figure 1).In total, the model explains 39.2% of the variance of support for multiculturalism, and 44.0% of the variance of support for Muslims' political organisation.In summary, Study 1 shows that majority members with a higher level of education or a more left-wing political orientation were more supportive of the political organisation of Muslim citizens, and this higher support can be explained by their stronger endorsement of multiculturalism.

Study 2
Considering the ongoing debate about the importance of replications (e.g., King, 1995;Simons, 2014), the first aim of Study 2 was to try to replicate these findings using a larger, representative sample.Additionally, in Study 2 we further investigated the link between political orientation and education, on the one hand, and the endorsement of multiculturalism, on the other.Attitudes towards cultural traditions (versus change) and about equality (versus dominance) are considered to be the two core attitudinal dimensions of people's political orientation (Jost et al., 2003), and are also related to level of education (Meeusen et al., 2013).Furthermore, cultural recognition and the promotion of equality are considered defining aspects of multiculturalism (Berry, 2006;Guimond et al., 2014).This means that the link between political orientation and education with multiculturalism can be expected to run through the attitudes towards tradition and (in)equality.This was tested in Study 2.

Method Survey and Participants
Similar to Study 1, data was collected among a representative sample (for gender, age, education and region) of self-identified Dutch majority members that was drawn from the TNS/NIPO-panel.Respondents who participated in Study 1 were not invited to participate in Study 2. The response rate was 57% and in total, there were 810 participants.For eight of these participants at least one of their parents was born abroad and therefore were not considered in the analysis.Furthermore, 11 participants were dropped because they had a missing score on education.The final sample consisted of 791 respondents.In total, 50.1% was female, and the age of the participants ranged between 18 and 87 years (M = 50.62,SD = 17.18).Again, none of the respondents indicated that they were Muslims.
Cultural traditionalism was assessed by three items (7-point scales; see Appendix).The attitude towards societal equality was measured with four items (7-point scales; see Appendix) that were derived from the social dominance orientation equality subscale (SDO-E; Ho et al., 2012).

Measurement Model
A measurement model including latent factors for multiculturalism, traditionalism, equality and Muslim political organisation was fitted in Mplus, using maximum likelihood with robust standard errors (MLR) as the estimation method.This model fitted the data reasonably well (Table 1, Model 2.0).However, results suggested that freeing the error covariance between two items of traditionalism (one addressing the protection of Dutch norms and values, and one addressing the maintenance of Dutch culture and traditions) improves the fit.Including this modification resulted in a significantly better model with a good fit (Table 1, Model 2.2), which was considered the final model.
In this model, traditionalism had a composite reliability of ρ = .811,and the composite measure for equality was also reliable (ρ = .849).Endorsement of multiculturalism had a composite reliability of ρ = .782,and the reliability of support for Muslims' political organisation was ρ = .853.

Plan of Analysis
We again used Structural Equation Modelling in Mplus, estimated by MLR for testing the structural model.As in Study 1, gender and age, household income and employment status were control variables.Household income was again endogenized assuming bivariate normality conditional on the control variables in the model to account for missing values (25.8%;Enders, 2010).To investigate whether the findings in the two Studies are similar despite the slightly different measurement of multiculturalism, we first fitted a structural model similar to Study 1 (Figure 1).Subsequently, the path model including traditionalism and equality as additional mediators was estimated.Note.Unstandardized coefficients presented, with standard errors between parentheses; original direct effects in italics.Age, gender, household income and employment status were included as control variables.Gender (female) was negatively related to support for Muslims' political organisation (B = -.278,p < .01),while age was positively related to it (B = .007,p < .05).Both age and employment status were positively related to traditionalism (B = .011,p < .001for age, B = .169,p < .01 for employment status).Household income was negatively related to socio-political equality (B = -.022,p < .05).*p < .05,one-tailed.**p < .01,one-tailed.***p < .001,one-tailed.
Figure 2 further shows that traditionalism and equality are two attitude dimensions through which education and political orientation were associated with the endorsement of multiculturalism.The indirect effects on multiculturalism were significant (for political orientation, β = .261,p < .001,and for education β = .075,p < .001),but there also were direct effects on multiculturalism.Furthermore, traditionalism and equality had a direct effect on support of Muslim political organization, independently of multiculturalism.These findings indicate that traditionalism and equality played a broader ideological role in the investigated associations.In total, the model explained 55.2% of the variance of support for Muslim political organisation, while 43.7% of the variance in the endorsement of multiculturalism was explained.ix

General Discussion
Political participation is an important aspect of the integration of Muslim minorities into western societies, but it is often met with resistance from the majority (e.g., Bloemraad, 2013;Verkuyten, 2017).In two studies, we investigated whether the level of support for the democratic political organisation of Muslims can be explained by two important socio-cultural cleavages, namely political orientation and education (Bovens et al., 2014).Moreover, it was examined whether the level of support for Muslims' political organisation is a result of more general beliefs about multiculturalism, and whether these beliefs mediate the relation between political orientation and education with the level of support.
The findings in both studies indicate that Dutch majority members tend to resist Muslim formal political organisation and that the level of resistance was related to education and political orientation.The non-significant (Study 1) and weak (Study 2) correlation between education and political orientation and their independent associations with support for Muslims' political organisation, indicate that both are indeed two separate societal cleavages.
Those with a higher level of education were found to be more supportive of the political organisation of Muslim citizens than lower-educated people, and the same applied to politically more left-wing oriented individuals versus those with a right-wing political orientation.Additionally, political orientation appears to be a more important dimension for the level of support than education.A likely reason is that, compared to education, political orientation as a construct is more closely related to changes in societal politics resulting from the formal political participation of immigrant-origin groups.Furthermore, in the analyses two indicators of socio-economic status (household income and employment status) were controlled for which might have reduced the effect of education.
Multiculturalism is considered a hierarchy-attenuating ideology in which both the acceptance of cultural diversity and the importance of equality is emphasized (Berry, 2006;Guimond et al., 2014;Sidanius & Pratto, 1999).Therefore, we tested whether the endorsement of multiculturalism mediates the associations between political orientation and education with the support for Muslim political organisation.In Study 1, the findings indicate that multiculturalism fully mediated the associations, and in Study 2 the mediation was partial.Thus, the endorsement of multiculturalism explained why higher-educated individuals and left-wing individuals were more supportive of Muslims' political organisation.The difference in mediation between the two studies might be because Study 2 had more statistical power due to the larger sample size.Next to that, the operationalization of multiculturalism slightly differed between both studies.In Study 1, ideological abstract (e.g., maintenance of minorities' identity) and more concrete (media, education) statements were used to measure the endorsement of multiculturalism, while in Study 2 only abstract statements were used (Yogeeswaran & Dasgupta, 2014; see Appendix).Conceptually, the concrete statements are closer to the political organisation of Muslim citizens which may explain why the measure including those items (in Study 1) played a more important mediating role than the more abstract ideological measure (in Study 2).
In Study 2 we further examined whether attitudes towards cultural tradition (vs.change) and equality (vs.dominance) explained why political orientation and education were positively related to multiculturalism endorsement.This test was based on the motivated cognition model of political orientation (Jost, 2006) and research on education, as well as the argument that multiculturalism favours cultural openness and group equality (Berry, 2006;Guimond et al., 2014).In agreement with this reasoning, it was found that the higher educated and left-wing individuals were more in favour of equality and less traditional, and in turn more strongly endorsed multiculturalism.Thus, support was found for the cognitive ability mechanisms of education (Coenders & Scheepers, 2003;Meeusen et al., 2013), and for political ideology as motivated social cognition (Jost, 2006).However, the attitudes towards cultural tradition and equality did not fully mediate the associations and also had a direct statistical effect on the support of Muslim political organization.This indicates that other mechanisms also explain why the higher educated and left-wing oriented individuals demonstrated stronger endorsement of multiculturalism and stronger support for the political organisation of Muslim citizens.For example, the higher educated, or left-wing people, might perceive less realistic and symbolic threats and therefore endorse multiculturalism more strongly (Hainmueller & Hopkins, 2014).And attitudes towards cultural tradition and equality might be associated with the support of Muslim political organization through the endorsement of other cultural diversity ideologies, such as colourblindness or assimilation (Guimond et al., 2014).Future studies could examine these possible mechanisms.

Limitations
The present research has some limitations.First, the measure used to assess political orientation is well-established in political psychology (Jost, 2006) but might not fully capture the political cleavages that characterize many western societies (Middendorp, 1992).It has been argued, for example, that the left-right distinction can refer to economic or cultural issues and that these two dimensions do not have to correspond.For instance, the far-left can endorse equality together with cultural traditionalism, while the far-right can also emphasize the importance of cultural tradition but together with dominance (Knutsen, 1995).
Second, the cross-sectional nature of the data means that no conclusions can be drawn about whether someone's political orientation causes multiculturalism endorsement and support for Muslims' political organisation.Yet, the model that we tested was theoretically derived and it is not likely that educational level and political orientation are consequences of beliefs about cultural diversity and support of Muslim' political rights.Furthermore, it is likely that general cultural diversity beliefs highlighting the broad goals of group recognition and equality (multiculturalism) predict the specific ways in which these goals can be achieved (support of political organization).More abstract or general ideological beliefs tend to determine more specific, concrete attitudes (Maio, Olson, Bernard, & Luke, 2006; see also Yogeeswaran & Dasgupta, 2014).Yet, future research should examine these associations further by using longitudinal data.
Third, we focused on Muslims' political participation in terms of formal organization.Future studies could examine attitudes towards other forms of Muslims political engagement and participation, such as voting, starting a petition, running for office, and online activism.It is possible that some forms of political participation are more difficult to accept by majority members than others, which could lead to different findings.Fourth, the context in which this research was conducted raises questions about the generalizability of the findings, both to other countries and to other minority groups.Although other western countries face similar intercultural tensions, particularly in relation to Muslim minorities, the Dutch political system and its history of multicultural policies might make it somewhat specific.Therefore, it would be useful to examine the support of the political organization of minority groups more generally in other countries and among other groups, including non-Muslim immigrant and minority groups.

Conclusion
With the rising number of Muslim citizens in many western societies the question about their political participation is increasingly important.Politics is about power and raises critical questions about social influence and societal change.Furthermore, political participation has important symbolic and normative implications for the political incorporation of minority groups and the legitimacy of the democratic system.By focusing on majority members' support for the democratic political organization of Muslim citizens, this research makes a novel contribution to the literature on attitudes towards immigrant-origin groups.In two studies using national samples, we showed that Dutch majority members are rather resistant towards the democratic political organisation of Muslim citizens.Furthermore, education and political orientation represent two separate and important societal cleavages, attitudinally and ideologically (Bovens et al., 2014;Layman et al., 2006).In line with this, we showed that political orientation and education were independently associated with support for the political organization of Muslims and that this is mainly due to the endorsement of the ideology of multiculturalism.

Notes
i) The data used in Study 1 and 2 contained various other constructs, and parts of these data have been analysed in other papers (e.g.Hindriks et al., 2015;Verkuyten & Martinovic, 2015;Verkuyten, Martinovic, Smeekes, & Kros, 2016).However, the current theoretical focus and analyses are novel and the findings of the two studies have not been published previously.
ii) Further information on the data collection and sampling procedure can be derived from www.tns-nipo.com.The name of the survey agency has recently been changed to Kantar-TNS.
iii) None of the respondents had followed no education.As a sensitivity check, education was transformed into years of education (per category: (1) eight years, (2) and (3) twelve years, (4) and (5) fourteen years, (6) seventeen years, (7) eighteen years), and this yielded similar results.iv) By fixing the variance and the intercept of the measure.A sensitivity analysis imposing other levels of reliability (.4 and .6)showed that results are robust.v) Some of the items used for the second study were found to have a somewhat skewed distribution.For the sake of consistency, it was therefore decided to use MLR as the estimator in both studies.vi) On political orientation, 17.9% of the respondents had a missing value, while 24.9% had a missing value on household income.vii) We also tested whether political orientation and education had non-linear effects on multiculturalism and Muslim political organization, and this was not the case.Furthermore, there also were no significant interaction effects between political orientation and education in predicting multiculturalism and support for Muslim political organization.viii) To account for single item measurement error of political orientation, a reliability of .5 was assumed by fixing the variance and intercept.Assuming different levels of reliability (.3, .4 and .6)did not affect the results.Replacing education in levels by education in years also did not affect the results.ix) Similar to Study 1 we also tested for non-linear effects of political orientation and education as well as their statistical interaction.There were only two small quadratic effects: of education on attitude towards traditionalism and of political orientation on multiculturalism.In addition there only was one significant interaction effect on traditionalism. vii

Table 1
Model Fit Statistics of Measurement Models and Path Models From Study 1 and Study 2

Table 2
Descriptive Findings and Correlations of the Main Variables From Studies 1 (First Four Rows), and 2 (Last Six Rows)Note.Descriptive statistics and correlations are derived from Mplus, using effect coding as identification method.