ssoar_reference_gesis-ssoar-22042_study_program_90_search

Property	Value
?:linkContext	famous thesis about the clash of civilizations has generated much debate. Huntington argues that a cultural cleft has emerged between Western and Islamic civilizations -a clash of cultures that is strongly related to religion. Huntington's writings are perhaps best understood as a provocation, and many critics have argued that his thesis cannot be substantiated at macro level (cp. the debate in . Huntington's argument refers to large cultural entities -civilizations -but he also argues that clashes of civilizations might be found inside Western European countries as a consequence of the increase in Muslim immigrants. Huntington claims that Europe is threatened by 'Islamization' and 'Muslims pose the immediate problem to Europe' . According to Huntington, democracy is one of the important issues that divide civilizations. In particular, he notes that Islam is associated with authoritarian rule and Christianity with democracy. However, we still lack studies that analyze a possible clash of civilisations (i.e. sizeable value differences) at the individual level in Western countries (for an exception based on the World Values Survey and consequently with very few respondents from ethnic minorities, see . This article addresses this need, using a relatively unique data set: A Danish survey of values among eight ethnic groups: Ethnic Danes and seven ethnic minorities in which all or some members are Muslims. If Huntington's thesis were correct, we would expect a lower level of support for democratic values among Muslim minorities in Western countries compared to the support shown by the Christian majority population. Huntington's fear is shared by who is concerned that immigration to Western Europe -particularly the increasing number of Muslim immigrants -will result in a struggle between radical Islamism and liberal democracy. Democratic values are fundamental to Western societies. Most European countries have references to democracy and human rights written into their constitutions, and democracy is codified as a principle of the European Union. The preamble of the 2004 draft treaty mentions 'the inviolable and inalienable rights of the human person, freedom, democracy, equality and the rule of law' . Even for supporters of multiculturalism, who argue that societies should accommodate a large variety of values based on cultural differences, the acceptance of parliamentary democracy and freedom of speech and religion are seen as a non-negotiable principles . Scholars like Fukuyama and Huntington fear that immigration may create problems for Western democracies for at least two reasons. Firstly, in general immigrants come from countries that are less democratic than the Western European countries they immigrate to . This observation is documented in The Freedom House Survey that measures a country's level of democracy according to two criteria: Political rights and civil liberties. There is some variation these countries' total score but in general the score is lower than that of any Western European country. Even though we may be able to estimate the character of the regime we have very little reliable evidence about the attitudes in the general population in these countries. Perhaps the support for democracy among the general public is higher than one would expect, judging by the character of regime. At least this result was found by in a study of Arab countries but even so it seems reasonable to argue that in general the general support for democracy in these countries is lower than in the Western world. Secondly, most of the immigrants to Western Europe come from countries that are dominated by Islam. Some critics argue that Islam's relation to democracy is ambiguous, if not dubious. Nasr claims that the 'Islamic view of democracy is not something deeply legitimate, but at best … a tool or tactic that may be useful in gaining the power to build an Islamic state' . By contrast, argues that 'several crucial elements of constitutional democracy and civil society are also moral imperatives of Islam' . The same difference of opinion is found at a public discursive level in Western societies both among politicians and in the general population. This article explores whether the Fukuyama/Huntington thesis about different sets of values among Muslim minorities and the Christian majority population can be substantiated in the case of Denmark. Our analysis is based on a survey among immigrants, descendants, and Danes undertaken in spring, 2006. According to the Fukuyama/Huntington thesis, a comparison of these groups would reflect a higher level of democratic values among Danes than among Muslim minorities. However, the Fukuyama/Huntington thesis does not allow us to determine whether the level of democratic sentiment differs among immigrants and descendants; nor does it allow for differences among ethnic minority groups. That is why a set of alternative hypotheses is needed. An alternative to the religious hypothesis is the assumption that political socialization will change the values of the individual. argues that political socialization is a crucial mechanism for creating political values and skills among citizens. According to , the concept of political socialization had its heyday in the 1980s but has received less attention in recent decades. However, political socialization is becoming increasingly relevant to the development of political involvement and values among immigrants. A general socialization hypothesis suggests that the mere fact of living in a society with democratic institutions will teach individual citizens democratic values. In order to analyse political socialization, we need to differentiate between two groups: Immigrants and descendants. Immigrants (including family-reunited individuals) are people who were born abroad and have come to Denmark either to work or as refugees from their home country. In other words, they have lived only part of their lives in Denmark. They are expected to have a lower level of democratic awareness than native Danes when arriving in Denmark. Based on the political socialization hypothesis, however, we would expect that the longer an immigrant lives in Denmark, the stronger his/her democratic values will become. Descendants are people who have lived in Denmark all their lives. Their socialization has taken place in a society with democratic institutions and a democratic culture. More specifically, they have gone to Danish kindergartens and schools where one of the official goals (formally stated in a law from 1975) is to teach pupils fundamental democratic principles. A comparative analysis has found that Danish schools emphasize 'collaboration and consensus' and concern themselves particularly with 'education for citizenship and democracy, as well as with the academic goals of education . Following the political socialization hypothesis, this emphasis on democracy will have an impact on the values of pupils. For this reason, we would expect descendants to have a more positive attitude towards democracy than their parents (the immigrants). There would only small differences in democratic values between descendants of immigrants and Danes because both groups have lived in a democratic society all their lives. A second alternative hypothesis focuses on the role played by ethnicity rather than religion in determining democratic values. Earlier research has shown that ethnicity has a significant influence on the individual's values . This finding may be interpreted as the result of a number of processes: Often, people from ethnic minorities come from the same region in the country of origin; they have a common destiny (be it as refugees or immigrants), and they are interrelated by family relations, common language and their daily interactions in their host country. To be able to test the hypothesis about the relationship between democratic values and ethnicity (irrespective of religion), the ethnic group needs to reflect a variation of religious faiths. In some ethnic groupings, however, such differences do not exist. Among Turks or Pakistanis more than 95 per cent of the respondents present themselves as Muslims. In such cases, it would be impossible to ascertain whether it is the respondent's ethnicity or his/her religion that has had the greatest influence on his/her democratic values or lack of such. The ethnicity hypothesis can only be tested in ethnic groups with a varied composition of denominations. This limits the investigation to immigrants from West Balkan, where our survey reflects 69 per cent to be Muslims; to immigrants from Iran, where 55 per cent of respondents are Muslims; and to immigrants from Iraq where 82 per cent of respondents are Muslims. To summarize, we operate with the following three hypotheses: • A religious hypothesis that claims that Muslims have fewer democratic values than Christians. • A political socialization hypothesis, which suggests that the longer a person has lived in Denmark the stronger his/her democratic values become. This means that Danes and descendants have more democratic values than immigrants, and that there will be little or no differences between descendants and Danes in terms of democratic mindset. • An ethnicity hypothesis that argues that in ethnic groups with different denominations, ethnicity is a stronger indicator of democratic values than religion. Each of the hypotheses above will be explored in the following. As a precursor to such an analysis, a short introduction to immigration in Denmark appears relevant, describing the data and definitions of the dependent variables. There has always been immigration to Denmark. Until the 1960s, the proportion of ethnic minorities was very small, making Denmark a very homogeneous country in terms of ethnicity. During the late 1960s, the first immigrants from Pakistan, Turkey and the former Yugoslavia came to Denmark as so-called 'guest workers". By the early 1970s, immigration laws prevented more labour immigration, but from the mid-1980s Concepts such as 'ethnicity', 'ethnicity' and 'immigrant' are contentious and difficult to define. In this article, we 'ethnic groups'is used in accordance with ) 'working definition' where ethnie is defined as "a named human community connected to a homeland, possessing common myths of ancestry, shared memories, one or more elements of shared culture, and a measure of solidarity, at least among the elite". Nørregård-Nielsen 2007) has shown that people who say they are Kurds do not differ significantly from other immigrants from the respective country of origin. A similar result was found in a survey among Turkish immigrants in Denmark . A special problem relates to immigrants and refugees from the Balkans since Statistics Denmark only registers the state where the immigrant comes from. The first immigrants came to Denmark in the early 1970s from former Yugoslavia, and in the 1990s, Denmark accepted a number of refugees primarily from Bosnia-Herzegovina. A comparison (not shown here) of the so-called guest workers from ex-Yugoslavia, their reunited family members and the refugees from the Balkan countries in the 1990s shows no significant difference in democratic values among these groups. This may be due to the fact that they come from roughly the same region of ex-Yugoslavia Finally, 'Danes' refer to people who are not immigrants or descendants. It would be more correct to use the term 'ethnic Danes' as many immigrants are in fact Danish citizens, but the term 'Danes' has been chosen for the sake of brevity. Our survey was based on phone interviews with nine ethnic groups: Immigrants and descendants from Pakistan and Turkey, immigrants or refugees from West Balkan, Iraq, Iran and Vietnam, and Danes. Since this article concerns the possible relationship between Islam and democracy, the Vietnamese immigrants have not been included in the subsequent analysis because none of the Vietnamese respondents were Muslims. Among all the other ethnic minorities (except for the Danes), some or all of the respondents have stated that they are Muslims. Among the ethnic minorities, only people who have lived in Denmark for three years or more were included in the survey. The three-year limit was chosen because Danish law grants the right to vote at local elections after three years residency in Denmark. Compared to the general population, the age distribution among the immigrant groups is skewed. Almost all of the respondents were below 50 years of age. Consequently, the survey population was defined as persons aged 18-50. The largest immigrant group was the Turkish immigrants where the population was 22.587; the smallest group of immigrants was the Vietnamese (5.910 persons). Among descendants, only two groups had a size that permitted further study: Turkish descendants amounted to 5.967, and Pakistani descendants numbered 3.268. Respondents were randomly chosen by civil registration number. Interviewers were bilingual and both immigrants and descendants could choose to be interviewed in Danish or in their mother tongue. About 500 persons in each ethnic group were interviewed. The average response rate was 54 per cent, varying between 67 per cent for Danes and 41 per cent for Pakistani descendants. The main reason for the relatively low response rate was the difficulty encountered when having to establish a clear Once contact was established, the number of rejects was relatively small. In the tables, weights are used to compensate for the low response rate. In the statistical analyses, no weights were applied. The number of 'don't know' answers varied from question to question and between ethnic groups. In general, the 'don't know' answers were lowest among Danes. As it is difficult to establish the reasons for such differences, 'don't know' answers have been excluded from the analysis. All data may be obtained from the Danish Data Archive (DDA-21171). Democracy theory tells us that the term 'democracy' has at least two meanings: 'popular sovereignty' and 'the liberty of the individual members of society' . The interpretation of these meanings and their interrelation varies in the different normative democracy theories, but that is of less relevance to the concerns of this article. For our purposes, the two-dimensional structure that these meanings suggest is of greater importance. The first -'popular sovereignty'-will be termed democratic principles. This variable is operationalized in the following question: 'What do you think about each of the following ways of governing Denmark? Having a democratic system' ('Very good', 'Fairly good', 'Fairly bad', 'Bad', 'Don't know'). This question is part of a battery of four questions that has been used by the World Values Survey and the European Values Survey. The other three questions in the battery ask about support for 'Having a strong leader who does not have to bother with parliament and elections", 'Having experts, not government, make decisions according to what they think is best for the country", 'Having the army rule the country". These questions are sometimes combined in an index but in our data we could not find a satisfactory one dimensional structure for all groups. Therefore, the analysis only concerns the 'democratic system' question -the only question that use the word 'democracy". Analyses show that the support for democratic principles is generally high in most Western countries ). The second dimension, democratic rights, is just as important for democracy as free elections. In the present survey, the variable is an index based on two questions: one on religious freedom and one on freedom of speech. Religious freedom has been called 'the first freedom' in the sense that freedom of conscience is both chronologically the first and may be seen as a precondition for other freedoms . It is measured by the following question: 'To what extent do you agree or disagree with the following statement: It is important that all people in Denmark have the possibility of carrying out their religious rituals and wearing religious symbols'('Strongly agree', 'Agree', 'Neither agree nor disagree', 'Disagree', 'Strongly disagree', 'Don't know'). what extent do you agree or disagree that all groups -even groups with extreme attitudes -should have the right to hold meetings and state their cause?'('Strongly agree', 'Agree', 'Neither agree nor disagree', 'Disagree', 'Strongly disagree', 'Don't know'). The two questions are positively correlated for all ethnic groups (γ between 0,12 and 0,36. The γ's are significant except for the Pakistani descendants (cf. Table A.1). The low insignificant correlation in this group is due to the fact that the distribution of responses is very skewed: almost everybody answers 'agrees strongly". For each respondent the answers are added up. The sum varies between 2 and 10. This is recoded as follows: 2=1, 3=2, 4 or 5 = 3, other values = 4. If a question is answered 'don't know', the index is coded 'don't know'. Democratic values are not just related to the formal structural arrangements in society but also to its political culture . Democratic culture has been measured in many ways. In our survey, we decided to develop a measurement for democratic values that was family related. If democratic culture pervades all institutions in society, it would also encompass the private sphere. Democratic culture in the family would mean that the individual members of the family have influence on their own life and are not subjected to any authoritarian family structure. Giddens have suggested a close relation between societal and family democracy. There is only one story to tell about the family today, and that is of democracy..... Democracy in the public sphere involves formal equality, individual rights, public discussion of issues free from violence, and authority which is negotiated rather than given by tradition. The democratized family shares these characteristics. . Family democracy may be studied through the relations between spouses or through the relationship between parents and their children. education". The respondents were asked to grade their answer according to a 10-point scale, where 1 reflected total decision power to the young person, and 10 suggested that parents should decide totally. There is a very high correlation (γ between 0,53 and 0,70) between the two variables for all ethnic groups, cf. . The two variables were entered into an index by adding the values of each variable, and the result subsequently divided into three groups: Index value 2 = the young person has very high influence (code 1), Index value 3-6 = the young person has high influence (code 2). Index value 7-20 = the young person has less influence (code 3). If one question is answered with a 'don't know', the index is coded 'don't know'. Are Muslims less democratic than Christians? The table presents an interesting and quite surprising result. The respective positions of the three groups vary according to which dimension of democratic values we investigate. When it comes to democratic principles -i.e. whether the respondent believes that a democratic system is good thing -the Danes score higher than the two other groups. The Danes' high score on this variable is similar to what has reported from the European Values Survey 1999. The two minority groups' score are somewhat lower but still higher than what is found in many European countries and in the immigrants' homelands (cf. . Consequently, we may conclude that the support for democratic principles is high among Danes and the ethnic minorities. It is thought provoking that the support for democracy is considerably lower when it comes to the second aspect of democracy: Democratic rights. It is perhaps even more surprising that in a country that prides itself on democratic principles, Danes show a low level of acceptance of democratic rights when these rights concern the liberties of minorities. More than twice as many among the ethnic minorities are strongly in favour of democratic rights compared to the Danes. As we will discuss below, this discrepancy is probably related to the general dissociation from Islam found among many Danes. Finally, the data shows a lower support for family democracy among the ethnic minorities that among the Danes. Although the difference is not that large, it is still significant. It may be explained by differences in family structure and family values among Danes and the ethnic minorities. In general, immigrant families are less individualistic, reflecting a more authoritarian culture than we see in Danish families. This initial presentation of data shows that there is no reason to believe that there is a clash of civilizations among Danes and ethnic minorities. On the contrary, in two out of three democracy measures, the differences are fairly small. In the third dimension (democratic rights), the ethnic minorities are more democratic than the Danes. Since the ranking varies for different measures of democratic values, the pattern cannot be explained by reference to a single variable such as the religion of the respondent. Instead, the differences are probably best explained by the framework in which Danes' generally perceive Islam. It is indeed striking that the Danes are the most positive group in relation the democratic system but the most reluctant when it come to accepting the constitutional rights of all citizens in relation to civil liberties. One explanation for this may be found in the framing of the questions and the survey itself. Based on in-depths interviews about democracy among shows that people's perception of concepts and their immediate social context greatly influence their way of answering survey questions. Finding themselves within the framework of a democratic constitution, people are generally positive when asked about relatively abstract democratic ideals. When it comes to specific civil liberties, however, the framing appears to become group-specific; respondents fear that their own liberties will be affected -even infringed -when applied to people of a different faith. According to theories of political socialization, living in a society with democratic institutions and democratic rule will educate the individual in democratic values. Following this hypothesis, the individual's democratic values will become stronger the 1. Immigrants: the longer an immigrant has lived in Denmark, the more positive his/her democratic values will be. Denmark for a shorter period, they are expected to have fewer democratic values than descendants. Since people in these two groups have lived in Denmark all their lives, there will be only minor differences in their democratic values. A socialization hypothesis predicts that immigrants who have stayed in Denmark for a longer period will have stronger democratic values than immigrants who have only recently arrived in Denmark. analysis of world values survey data suggest, however, that the values among immigrants in Europe only change slowly. In contrast study of Mexican Americans found that 'Mexican Americans begin to adopt the American definition of democracy after having resided in the United States for only one year' . Later in the text, however, Camp modifies his conclusion. It appears that it is not the length of stay as such but the immigrant's language proficiency that is the more important determinant of his/her changing values. Camp does not present an interpretation of this conclusion, but his study implies that immigrants who command the language of the host country are more committed to adapting to the ways of their new country and will develop a stronger sense of belonging towards the host country. We used Camp's finding to create a statistical model where the three measures of democratic values are dependent variables. The intervening variables are: The respondent's sense of belonging vis a vis Danish society, the respondent's proficiency in Danish, and the length of the respondent's stay in Denmark. The independent variable is ethnicity. Data was analyzed by so-called 'chain graphical models' using strategies and techniques described by integrating analyses of data with graphical theoretic analyses of Markov graphs. The contact with Danish society is relatively limited and consequently that the probability for political socialization is relatively small. Since descendants have been born in Denmark, the socialization hypothesis would predict that they have a stronger democratic mindset than that of their immigrant counterparts. Descendants have joined Danish schools, kindergartens and have generally had more intimate contact with Danish society. At the same time, however, they have grown up in a family that is different from the average Danish family, implying that their family socialization may conflict with the impact of societal institutions. Only the Pakistani and Turkish minorities can be used for the comparison of immigrants and descendants as these are the only two groups that have lived in Denmark long enough for there to be second generation ethnic minority groups of a sufficient size to undertake analysis. A comparison of immigrants and descendants is complicated by the different age structure of the two groups. There are no descendants over 35 years of age in the data. For this reason, it is only possible to compare descendants and immigrants between the ages of 18 and 34. The scenario reflected by the table is somewhat complicated. Firstly, there are no significant differences between immigrants and descendants in relation to democratic principles, which support the general findings in relation to this variable. Secondly, in relation to family values, we might expect that descendants are more democratic than immigrants, but the table shows no general pattern here. Among the 18-24 year-olds, the Pakistani descendants are more democratic than their immigrant counterparts, but for the Turks, it is the other way around. Among the 25-34 year-olds, there is only a noticeable difference for the Turkish minorities, but here the correlation is reversed: Turkish descendants are more democratic than Turkish immigrants. These differences appear difficult to explain, but it looks as if their family values depend on the specific situation of the ethnic group. Finally, as regards democratic rights there is no difference between descendants and immigrants in the 18-24 year age group. Among the 25-35 year-olds, descendants are, surprisingly enough, less democratic than the immigrants. In fact, it seems as if descendants are becoming more like the Danes (who are very low on democratic rights) than their parents' generation (cf. ). This leads us to a closer analysis of the differences between descendants and Danes of the same age. The best way of testing the denomination and socialization hypotheses is by comparing people of different denomination, who are all socialized in Danish society. We matched the descendants of Muslim faith with Christian Danes of the same age (18-34 years). presents an overview of the distribution of the three dependent democracy variables in these groups, tolerance towards freedom of speech and religious faith is lower than that of the ethnic minorities. When it comes to young people's influence on their own life (choice of education and spouse), the difference between the ethnic groups is much smaller than when discussing more abstract democratic values. There is a significant difference between all the ethnic groups. As the figures in the table reflect, value differences exist in spite of the fact that all three ethnic groups have gone through public education in Denmark. Thus, the hypothesis about formal socialization cannot be supported. The differences must be explained by other factors. One possibility is family socialization. Descendants are not only socialized by social institutions, they also received specific ethnic socialization through their family and network. We lack general information about family socialization in Denmark. However, it would be reasonable to expect that at least some of the descendants have been exposed to relatively strong Islamic values in their families. Some of them may also have attended private Islam-oriented schools but we have no information about this in the data. In the immigrant family, values are shaped both by the country of origin as well as the host society (Foner 1997). In many immigrant families, religion plays a much stronger role than in Danish families, whose lifestyle is generally quite secular. This returns us once more to Huntington's 'clash of civilizations' hypothesis. Is it plausible to assume that devout descendants have less democratic values, whereas religious faith among the Danes has no affect on their democratic values? A multivariate statistical analysis -not shown here because space does not allow it -reflects that for Danes, there is no relationship between religious values and democratic values. Among descendants, however, democratic values are influenced by the religion of the respondent, but there is no difference in democratic values between the Pakistani or Turkish respondents. shows the distribution of democratic values for people with strong religious faith (who report themselves to be 'very religious' or 'somewhat religious') and people who are less devout (who say that they are 'not very religious'). The differences between 'very' and 'less devout' descendants are relatively small, but there is a tendency for the more religious respondents to be less attracted to democratic principles and family democracy and to value religious freedom and freedom of speech higher. This result may be interpreted with reference to 'Stuck between two cultures with which they cannot identify, they find a strong appeal in the universalistic ideology offered by contemporary jihadism' ). While there is no indication in the data that there is a strong fundamentalist Islamic minority in Denmark, Fukuyama's observation may contain some truth, especially for the more religious second generation immigrants for the simple reason that this group is more exposed to the pressures of different value systems. Such an interpretation is supported by an interesting finding in our survey: When asked whether they feel that they have become more religious during the last three years, 37 per cent of the descendants answer in the affirmative, compared to 26 per cent of the immigrants from the same ethnic groups, and 9 per cent of the Danes. Another viable explanation is found in the fact that the descendants may feel that their religion is being discriminated against by the native majority; this leads them to put more emphasis on their right to religious freedom and freedom of speech. Generally, it is difficult to distinguish between ethnicity and religion because most of the immigrants only belong to one denomination: Islam. As has been shown above, the respondent's degree of devoutness has an influence on his/her democratic values, but in the following we shall explore whether differences in democratic values are higher within ethnic groups or within religious groups. If a religious explanation were to be The statistical tests show that there are no significant differences in democratic values among the denominational groups in each ethnicity group. This strongly supports the ethnicity hypothesis -i.e. that the destiny and living conditions of a certain group are the more important determinants of a democratic mindset. However, if an ethnicity hypothesis should be supported, it would also require significant differences among the Muslims in all ethnic groups. This is precisely what we see in . For each variable, there are significant differences among the ethnic groups. Thus, there is no basis for arguing that Muslim respondents are less democratic than non-Muslims respondents. In all three cases, Muslims from Iraq have the strongest democratic values, followed by Muslims from Iran and Muslims from West Balkan. This result is difficult to interpret but might be explained by the history of the three countries or perhaps with reference to the different types of Islam that characterize the groups (e.g. Sunni, Shia). Unfortunately, no data exists that may support such a conclusion. Besides being of interest in their own right, studies of ethnic minorities tell us a lot about the host society . Often, the conclusions drawn by such studies have been relatively unclear because researchers only study the ethnic minorities and not the majority. In this respect, the present study is an exception because it includes data on the ethnic majority (the Danes) as well as data on a number of ethnic minorities. Our study undertaken has used three variables to measure democratic values: The support for democratic principles, democratic rights and family democracy. The study concludes that the support for democratic rights is high among the ethnic minorities as well as the majority population. Even though minorities show less support for democratic principles, they still score higher than the populations of many West European countries. As regards family democracy, there appears to be a similar pattern as for democratic principles (but we have no information about the distribution of family values in other European countries). The most important difference is found in the support for democratic rights -the freedom of speech and freedom of religion. Here, the Danish majority population is much less democratic than the ethnic minorities. Thus, democratic values do not give rise to a clash of civilizations, as Huntington would have it -at least not when we consider immigrants in Western Societies. Contrary to what Huntington's thesis suggests, Christian Danes are less democratic than Muslim immigrants when it comes to granting democratic rights to all members of society. Generally, democratic values do not differ between immigrants and descendants of the same ethnic group, and a change in values does not appear to depend on the length of stay in the host society. These results emphasize the role of ethnic belonging rather than the denomination or socio-economic factors. For instance, we found that among ethnic minorities with a mixed religious composition, the differences between Muslims are greater than the in-minority differences. These results support our general findings: Ethnicity is more important for democratic values than denomination. Ethnicity is important because of the common history and destiny of the various ethnic groups. This 'ethnic habitus' means that democratic values are based on a common history and fate and reproduced within ethnic groups primarily through social interaction and a shared language within the group. The importance of ethnicity might not come as a surprise to sociologists. As noted by , immigrant populations arrive in Europe and their experiences vary for specific historical reasons. The Muslim headscarf, for instance, is an acceptable part of public life in Britain, but not in France. In Denmark, the headscarf is quite common in public life but also a highly contested symbol. The difference in acceptance of foreign symbols and customs need to be understood in the context of the individual country's history. Denmark has a history of a strong nation-religion relations, and has been dominated by a single denomination. It would be interesting to study whether other countries sharing the same characteristics also appear to be less willing to accept the democratic rights of all citizens. χ 2 = 39,732, df = 6; p= 0,000 χ 2 = 106,816, df= 6; p = 0,000 χ 2 = 36,591, df = 4; p= 0,000. Note: The differences between the high and low religiosity categories are significant. Democratic principles γ = -0,21 p = 0,009; democratic rights γ = 0,17 p = 0,008; family democracy = -0,18, p = 0,006. Note: The differences between the national groups are significant. Democratic principles χ 2 = 41,018, df=6, p=0,000; democratic rights χ 2 = 99,702, df=6, p=0,000; family democracy χ = 14,867, df=4, p=0,005. Creating a Virtuous Society: Immigration and Australia's Policies of Multiculturalism JØRGEN GOUL 2002 Danskernes holdning til indvandrere Learning Democracy in Mexico and the United States Mexican Studies/Estudios Mexicanos How People Think, Reason, and Feel about Rights and Liberties American Journal of Political Science The Sociology of Religion, London: Sage FLEDELIUS, CARSTEN 2006 Personal communication FONER, NANCY 1997 'The Immigrant Family: Cultural Legacies and Cultural Changes Interethnic Migration Review Citizens and the State Support for the Democratic System Identity, Immigration, and Liberal Democracy Journal of Democracy Etniske gruppers vaerdier, København: Ministeriet for flygtninge, indvandrere og integration HUNTINGTON, SAMUEL P. 1993 'The Clash of Civilizations? The Third Way: The Renewal of Social Democracy The Clash of Civilizations? The Debate _1997 The Clash of Civilizations and the Remaking of the World Order Universal Values and Muslim Democracy Journal of Democracy Human Values and Beliefs -a Cross-Cultural Source Book Based on the 1999-2002 Values Surveys Tørre Tal Om Troen Critical Citizens. Global Support for Democratic Governance Mapping Political Support in the 1990s: A Global Analysis Computerized exploratory Screening of Large-Dimensional Contingency Tables Compstat 1986. Proceedings in Computational Statistics. 7 th Symposium User Guide to DIGRAM. A Program for Discrete Graphical Modeling An Informal Introduction to Graphical Modelling Mental Health Service Evaluation Udlaendingeservice NASR, VALI 2005 'The Rise of Muslim Democracy Udlaendingestyrelsen ___ 2006b Tal og fakta -Udlaendinges tilknytning til arbejdsmarkedet og uddannelsesystemet Constants and Contexts in Pupil Experience of Learning and Schooling: Comparing Learners in England, France and Denmark Comparative Education Muslim Immigrants in Western Europe: Persisting Value Differentiation or Value Adaptation Worrisome Signs, Modest Shifts Journal of Democracy Basic Books SAPIRO, VIRGINIA 2004 'Not Your Parent's Political Socialization: Introduction for a New Generation Annual Review for Political Science Demokrati Danskernes saerpraeg Do Islamic Orientations Influence Attitudes towards Democracy in the Arab World? Citizens and the State Support for Democratic Values ADDRESS: Department of Sociology, 5 Østerfarimagsgade, DK-1014 Copenhagen K, Denmark. Email: pg@soc Note: Ellipses show significant difference between descendents and immigrants in the same ethnic group (χ 2 , p<0,05) when controlled for age F o r P e e r R e v i e w O n l y Democracy and Denomination: Democratic values among Muslim minorities and the majority population in Denmark Peter Gundelach ethnic@surrey.ac.uk F o r P e e r R e v i e w O n l y Democracy and Denomination: Democratic values among Muslim minorities and the majority population in Denmark GROBID - A machine learning software for extracting information from scholarly documents Ethnic and Racial Studies Immigrants, Islam, survey data, national identity, Denmark, Democracy Immigrants Islam survey data national identity, Denmark, democracy Based on a survey of ethnic minorities (immigrants as well as descendants) and the ethnic majority, this article analyses the distribution and explanation of democratic values in Denmark. Democratic values are measured according to three criteria: Democratic principles, democratic rights and family democracy. With respect to democratic rights, the analysis shows that Muslim immigrants and descendants score slightly lower on democratic principles and family democracy; however, they have more democratic values than the majority of the populations in Western Europe. Democratic values do not differ between immigrants and descendants of the same ethnic group, and differences in values do not depend on immigrants' length of stay in the host society. Finally, among ethnic minorities of mixed religious composition, the differences between the Muslim parts of these groups are larger than the differences between ethnic groups. Thus, differences in democratic values are explained by ethnicity, not religion. Huntington's (1993 Huntington's ( , 1997 famous thesis about the clash of civilizations has generated much debate. Huntington argues that a cultural cleft has emerged between Western and Islamic civilizations -a clash of cultures that is strongly related to religion. Huntington's writings are perhaps best understood as a provocation, and many critics have argued that his thesis cannot be substantiated at macro level (cp. the debate in Huntington 1996) . Huntington's argument refers to large cultural entities -civilizations -but he also argues that clashes of civilizations might be found inside Western European countries as a consequence of the increase in Muslim immigrants. Huntington claims that Europe is threatened by 'Islamization' (Huntington 1996, p. 204) and 'Muslims pose the immediate problem to Europe' (Huntington 1996, p. 204) . According to Huntington, democracy is one of the important issues that divide civilizations. In particular, he notes that Islam is associated with authoritarian rule and Christianity with democracy. However, we still lack studies that analyze a possible clash of civilisations (i.e. sizeable value differences) at the individual level in Western countries (for an exception based on the World Values Survey and consequently with very few respondents from ethnic minorities, see Pettersson 2007) . This article addresses this need, using a relatively unique data set: A Danish survey of values among eight ethnic groups: Ethnic Danes and seven ethnic minorities in which all or some members are Muslims. If Huntington's thesis were correct, we would expect a lower level of support for democratic values among Muslim minorities in Western countries compared to the support shown by the Christian majority population. Huntington's fear is shared by Fukuyama (2006) who is concerned that immigration to Western Europe -particularly the increasing number of Muslim immigrants -will result in a struggle between radical Islamism and liberal democracy. Democratic values are fundamental to Western societies. Most European countries have references to democracy and human rights written into their constitutions, and democracy is codified as a principle of the European Union. The preamble of the 2004 draft treaty mentions 'the inviolable and inalienable rights of the human person, freedom, democracy, equality and the rule of law' (Eur-Lex, vol. 47, December 16, 2004, p. 3) . Even for supporters of multiculturalism, who argue that societies should accommodate a large variety of values based on cultural differences, the acceptance of parliamentary democracy and freedom of speech and religion are seen as a non-negotiable principles (Borowski 2000) . Scholars like Fukuyama and Huntington fear that immigration may create problems for Western democracies for at least two reasons. Firstly, in general immigrants come from countries that are less democratic than the Western European countries they immigrate to (Puddington and Piano 2005) . This observation is documented in The Freedom House Survey that measures a country's level of democracy according to two criteria: Political rights and civil liberties. There is some variation these countries' total score but in general the score is lower than that of any Western European country. Even though we may be able to estimate the character of the regime we have very little reliable evidence about the attitudes in the general population in these countries. Perhaps the support for democracy among the general public is higher than one would expect, judging by the character of regime. At least this result was found by Tessler (2003) in a study of Arab countries but even so it seems reasonable to argue that in general the general support for democracy in these countries is lower than in the Western world. Secondly, most of the immigrants to Western Europe come from countries that are dominated by Islam. Some critics argue that Islam's relation to democracy is ambiguous, if not dubious. Nasr claims that the 'Islamic view of democracy is not something deeply legitimate, but at best … a tool or tactic that may be useful in gaining the power to build an Islamic state' (Nasr 2005, p. 13) . By contrast, Ibrahim (2006) argues that 'several crucial elements of constitutional democracy and civil society are also moral imperatives of Islam' (Ibrahim 2006, p. 7) . The same difference of opinion is found at a public discursive level in Western societies both among politicians and in the general population. This article explores whether the Fukuyama/Huntington thesis about different sets of values among Muslim minorities and the Christian majority population can be substantiated in the case of Denmark. Our analysis is based on a survey among immigrants, descendants, and Danes undertaken in spring, 2006. According to the Fukuyama/Huntington thesis, a comparison of these groups would reflect a higher level of democratic values among Danes than among Muslim minorities. However, the Fukuyama/Huntington thesis does not allow us to determine whether the level of democratic sentiment differs among immigrants and descendants; nor does it allow for differences among ethnic minority groups. That is why a set of alternative hypotheses is needed. An alternative to the religious hypothesis is the assumption that political socialization will change the values of the individual. Almond and Verba's famous study, The Civic Culture (1963) argues that political socialization is a crucial mechanism for creating political values and skills among citizens. According to Sapiro (2004) , the concept of political socialization had its heyday in the 1980s but has received less attention in recent decades. However, political socialization is becoming increasingly relevant to the development of political involvement and values among immigrants. A general socialization hypothesis suggests that the mere fact of living in a society with democratic institutions will teach individual citizens democratic values. Alternative hypotheses: Socialization and nation In order to analyse political socialization, we need to differentiate between two groups: Immigrants and descendants. Immigrants (including family-reunited individuals) are people who were born abroad and have come to Denmark either to work or as refugees from their home country. In other words, they have lived only part of their lives in Denmark. They are expected to have a lower level of democratic awareness than native Danes when arriving in Denmark. Based on the political socialization hypothesis, however, we would expect that the longer an immigrant lives in Denmark, the stronger his/her democratic values will become. Descendants are people who have lived in Denmark all their lives. Their socialization has taken place in a society with democratic institutions and a democratic culture. More specifically, they have gone to Danish kindergartens and schools where one of the official goals (formally stated in a law from 1975) is to teach pupils fundamental democratic principles. A comparative analysis has found that Danish schools emphasize 'collaboration and consensus' and concern themselves particularly with 'education for citizenship and democracy, as well as with the academic goals of education ' (Osborn 2001, p. 274) . Following the political socialization hypothesis, this emphasis on democracy will have an impact on the values of pupils. For this reason, we would expect descendants to have a more positive attitude towards democracy than their parents (the immigrants). There would only small differences in democratic values between descendants of immigrants and Danes because both groups have lived in a democratic society all their lives. A second alternative hypothesis focuses on the role played by ethnicity rather than religion in determining democratic values. Earlier research has shown that ethnicity has a significant influence on the individual's values (Gundelach and Nørregård-Nielsen 2007) . This finding may be interpreted as the result of a number of processes: Often, people from ethnic minorities come from the same region in the country of origin; they have a common destiny (be it as refugees or immigrants), and they are interrelated by family relations, common language and their daily interactions in their host country. To be able to test the hypothesis about the relationship between democratic values and ethnicity (irrespective of religion), the ethnic group needs to reflect a variation of religious faiths. In some ethnic groupings, however, such differences do not exist. Among Turks or Pakistanis more than 95 per cent of the respondents present themselves as Muslims. In such cases, it would be impossible to ascertain whether it is the respondent's ethnicity or his/her religion that has had the greatest influence on his/her democratic values or lack of such. The ethnicity hypothesis can only be tested in ethnic groups with a varied composition of denominations. This limits the investigation to immigrants from West Balkan, where our survey reflects 69 per cent to be Muslims; to immigrants from Iran, where 55 per cent of respondents are Muslims; and to immigrants from Iraq where 82 per cent of respondents are Muslims. To summarize, we operate with the following three hypotheses: • A religious hypothesis that claims that Muslims have fewer democratic values than Christians. • A political socialization hypothesis, which suggests that the longer a person has lived in Denmark the stronger his/her democratic values become. This means that Danes and descendants have more democratic values than immigrants, and that there will be little or no differences between descendants and Danes in terms of democratic mindset. • An ethnicity hypothesis that argues that in ethnic groups with different denominations, ethnicity is a stronger indicator of democratic values than religion. Each of the hypotheses above will be explored in the following. As a precursor to such an analysis, a short introduction to immigration in Denmark appears relevant, describing the data and definitions of the dependent variables. Immigration to Denmark There has always been immigration to Denmark. Until the 1960s, the proportion of ethnic minorities was very small, making Denmark a very homogeneous country in terms of ethnicity. During the late 1960s, the first immigrants from Pakistan, Turkey and the former Yugoslavia came to Denmark as so-called 'guest workers". By the early 1970s, immigration laws prevented more labour immigration, but from the mid-1980s Definitions Concepts such as 'ethnicity', 'ethnicity' and 'immigrant' are contentious and difficult to define. In this article, we 'ethnic groups'is used in accordance with Smith's (2001, p. 13 ) 'working definition' where ethnie is defined as "a named human community connected to a homeland, possessing common myths of ancestry, shared memories, one or more elements of shared culture, and a measure of solidarity, at least among the elite". Nørregård-Nielsen 2007) has shown that people who say they are Kurds do not differ significantly from other immigrants from the respective country of origin. A similar result was found in a survey among Turkish immigrants in Denmark (Necef 1996) . A special problem relates to immigrants and refugees from the Balkans since Statistics Denmark only registers the state where the immigrant comes from. The first immigrants came to Denmark in the early 1970s from former Yugoslavia, and in the 1990s, Denmark accepted a number of refugees primarily from Bosnia-Herzegovina. A comparison (not shown here) of the so-called guest workers from ex-Yugoslavia, their reunited family members and the refugees from the Balkan countries in the 1990s shows no significant difference in democratic values among these groups. This may be due to the fact that they come from roughly the same region of ex-Yugoslavia Finally, 'Danes' refer to people who are not immigrants or descendants. It would be more correct to use the term 'ethnic Danes' as many immigrants are in fact Danish citizens, but the term 'Danes' has been chosen for the sake of brevity. Data Our survey was based on phone interviews with nine ethnic groups: Immigrants and descendants from Pakistan and Turkey, immigrants or refugees from West Balkan, Iraq, Iran and Vietnam, and Danes. Since this article concerns the possible relationship between Islam and democracy, the Vietnamese immigrants have not been included in the subsequent analysis because none of the Vietnamese respondents were Muslims. Among all the other ethnic minorities (except for the Danes), some or all of the respondents have stated that they are Muslims. Among the ethnic minorities, only people who have lived in Denmark for three years or more were included in the survey. The three-year limit was chosen because Danish law grants the right to vote at local elections after three years residency in Denmark. Compared to the general population, the age distribution among the immigrant groups is skewed. Almost all of the respondents were below 50 years of age. Consequently, the survey population was defined as persons aged 18-50. The largest immigrant group was the Turkish immigrants where the population was 22.587; the smallest group of immigrants was the Vietnamese (5.910 persons). Among descendants, only two groups had a size that permitted further study: Turkish descendants amounted to 5.967, and Pakistani descendants numbered 3.268. Respondents were randomly chosen by civil registration number. Interviewers were bilingual and both immigrants and descendants could choose to be interviewed in Danish or in their mother tongue. About 500 persons in each ethnic group were interviewed. The average response rate was 54 per cent, varying between 67 per cent for Danes and 41 per cent for Pakistani descendants. The main reason for the relatively low response rate was the difficulty encountered when having to establish a clear Once contact was established, the number of rejects was relatively small. In the tables, weights are used to compensate for the low response rate. In the statistical analyses, no weights were applied. The number of 'don't know' answers varied from question to question and between ethnic groups. In general, the 'don't know' answers were lowest among Danes. As it is difficult to establish the reasons for such differences, 'don't know' answers have been excluded from the analysis. All data may be obtained from the Danish Data Archive (DDA-21171). The dependent variables: Democratic values Democracy theory tells us that the term 'democracy' has at least two meanings: 'popular sovereignty' and 'the liberty of the individual members of society' (Thomassen 1995) . The interpretation of these meanings and their interrelation varies in the different normative democracy theories, but that is of less relevance to the concerns of this article. For our purposes, the two-dimensional structure that these meanings suggest is of greater importance. The first -'popular sovereignty'-will be termed democratic principles. This variable is operationalized in the following question: 'What do you think about each of the following ways of governing Denmark? Having a democratic system' ('Very good', 'Fairly good', 'Fairly bad', 'Bad', 'Don't know'). This question is part of a battery of four questions that has been used by the World Values Survey and the European Values Survey. The other three questions in the battery ask about support for 'Having a strong leader who does not have to bother with parliament and elections", 'Having experts, not government, make decisions according to what they think is best for the country", 'Having the army rule the country". These questions are sometimes combined in an index but in our data we could not find a satisfactory one dimensional structure for all groups. Therefore, the analysis only concerns the 'democratic system' question -the only question that use the word 'democracy". Analyses show that the support for democratic principles is generally high in most Western countries (Fuchs, Guidorossi and Svensson 1995; Klingemann 1999 ). The second dimension, democratic rights, is just as important for democracy as free elections. In the present survey, the variable is an index based on two questions: one on religious freedom and one on freedom of speech. Religious freedom has been called 'the first freedom' in the sense that freedom of conscience is both chronologically the first and may be seen as a precondition for other freedoms (Casanova 1994, p. 40) . It is measured by the following question: 'To what extent do you agree or disagree with the following statement: It is important that all people in Denmark have the possibility of carrying out their religious rituals and wearing religious symbols'('Strongly agree', 'Agree', 'Neither agree nor disagree', 'Disagree', 'Strongly disagree', 'Don't know'). Freedom of speech and association is measured by the following question: 'To what extent do you agree or disagree that all groups -even groups with extreme attitudes -should have the right to hold meetings and state their cause?'('Strongly agree', 'Agree', 'Neither agree nor disagree', 'Disagree', 'Strongly disagree', 'Don't know'). The two questions are positively correlated for all ethnic groups (γ between 0,12 and 0,36. The γ's are significant except for the Pakistani descendants (cf. Table A.1). The low insignificant correlation in this group is due to the fact that the distribution of responses is very skewed: almost everybody answers 'agrees strongly". For each respondent the answers are added up. The sum varies between 2 and 10. This is recoded as follows: 2=1, 3=2, 4 or 5 = 3, other values = 4. If a question is answered 'don't know', the index is coded 'don't know'. Democratic values are not just related to the formal structural arrangements in society but also to its political culture (Almond and Verba, 1963) . Democratic culture has been measured in many ways. In our survey, we decided to develop a measurement for democratic values that was family related. If democratic culture pervades all institutions in society, it would also encompass the private sphere. Democratic culture in the family would mean that the individual members of the family have influence on their own life and are not subjected to any authoritarian family structure. Giddens have suggested a close relation between societal and family democracy. There is only one story to tell about the family today, and that is of democracy..... Democracy in the public sphere involves formal equality, individual rights, public discussion of issues free from violence, and authority which is negotiated rather than given by tradition. The democratized family shares these characteristics. (Giddens 1998, p. 93) . Family democracy may be studied through the relations between spouses or through the relationship between parents and their children. education". The respondents were asked to grade their answer according to a 10-point scale, where 1 reflected total decision power to the young person, and 10 suggested that parents should decide totally. There is a very high correlation (γ between 0,53 and 0,70) between the two variables for all ethnic groups, cf. Table A1 . The two variables were entered into an index by adding the values of each variable, and the result subsequently divided into three groups: Index value 2 = the young person has very high influence (code 1), Index value 3-6 = the young person has high influence (code 2). Index value 7-20 = the young person has less influence (code 3). If one question is answered with a 'don't know', the index is coded 'don't know'. Are Muslims less democratic than Christians? Table 1 reflects The table presents an interesting and quite surprising result. The respective positions of the three groups vary according to which dimension of democratic values we investigate. When it comes to democratic principles -i.e. whether the respondent believes that a democratic system is good thing -the Danes score higher than the two other groups. The Danes' high score on this variable is similar to what Svensson (2004) has reported from the European Values Survey 1999. The two minority groups' score are somewhat lower but still higher than what is found in many European countries and in the immigrants' homelands (cf. Inglehart et al 2004) . Consequently, we may conclude that the support for democratic principles is high among Danes and the ethnic minorities. It is thought provoking that the support for democracy is considerably lower when it comes to the second aspect of democracy: Democratic rights. It is perhaps even more surprising that in a country that prides itself on democratic principles, Danes show a low level of acceptance of democratic rights when these rights concern the liberties of minorities. More than twice as many among the ethnic minorities are strongly in favour of democratic rights compared to the Danes. As we will discuss below, this discrepancy is probably related to the general dissociation from Islam found among many Danes. Finally, the data shows a lower support for family democracy among the ethnic minorities that among the Danes. Although the difference is not that large, it is still significant. It may be explained by differences in family structure and family values among Danes and the ethnic minorities. In general, immigrant families are less individualistic, reflecting a more authoritarian culture than we see in Danish families. This initial presentation of data shows that there is no reason to believe that there is a clash of civilizations among Danes and ethnic minorities. On the contrary, in two out of three democracy measures, the differences are fairly small. In the third dimension (democratic rights), the ethnic minorities are more democratic than the Danes. Since the ranking varies for different measures of democratic values, the pattern cannot be explained by reference to a single variable such as the religion of the respondent. Instead, the differences are probably best explained by the framework in which Danes' generally perceive Islam. It is indeed striking that the Danes are the most positive group in relation the democratic system but the most reluctant when it come to accepting the constitutional rights of all citizens in relation to civil liberties. One explanation for this may be found in the framing of the questions and the survey itself. Based on in-depths interviews about democracy among Americans, Chong (1993) shows that people's perception of concepts and their immediate social context greatly influence their way of answering survey questions. Finding themselves within the framework of a democratic constitution, people are generally positive when asked about relatively abstract democratic ideals. When it comes to specific civil liberties, however, the framing appears to become group-specific; respondents fear that their own liberties will be affected -even infringed -when applied to people of a different faith. Does living in Denmark breed democratic values? According to theories of political socialization, living in a society with democratic institutions and democratic rule will educate the individual in democratic values. Following this hypothesis, the individual's democratic values will become stronger the 1. Immigrants: the longer an immigrant has lived in Denmark, the more positive his/her democratic values will be. Comparing immigrants and descendants: Since immigrants have lived in Denmark for a shorter period, they are expected to have fewer democratic values than descendants. Comparing descendants and Danes: Since people in these two groups have lived in Denmark all their lives, there will be only minor differences in their democratic values. Immigrants: Length of stay and democratic values A socialization hypothesis predicts that immigrants who have stayed in Denmark for a longer period will have stronger democratic values than immigrants who have only recently arrived in Denmark. Pettersson's (2007) analysis of world values survey data suggest, however, that the values among immigrants in Europe only change slowly. In contrast Camp's (2003) study of Mexican Americans found that 'Mexican Americans begin to adopt the American definition of democracy after having resided in the United States for only one year' (Camp 2003, p. 14) . Later in the text, however, Camp modifies his conclusion. It appears that it is not the length of stay as such but the immigrant's language proficiency that is the more important determinant of his/her changing values. Camp does not present an interpretation of this conclusion, but his study implies that immigrants who command the language of the host country are more committed to adapting to the ways of their new country and will develop a stronger sense of belonging towards the host country. We used Camp's finding to create a statistical model where the three measures of democratic values are dependent variables. The intervening variables are: The respondent's sense of belonging vis a vis Danish society, the respondent's proficiency in Danish, and the length of the respondent's stay in Denmark. The independent variable is ethnicity. Data was analyzed by so-called 'chain graphical models' (Lauritzen 1996) using strategies and techniques described by Kreiner (1996) integrating analyses of data with graphical theoretic analyses of Markov graphs. The contact with Danish society is relatively limited and consequently that the probability for political socialization is relatively small. Immigrants and descendants Since descendants have been born in Denmark, the socialization hypothesis would predict that they have a stronger democratic mindset than that of their immigrant counterparts. Descendants have joined Danish schools, kindergartens and have generally had more intimate contact with Danish society. At the same time, however, they have grown up in a family that is different from the average Danish family, implying that their family socialization may conflict with the impact of societal institutions. Only the Pakistani and Turkish minorities can be used for the comparison of immigrants and descendants as these are the only two groups that have lived in Denmark long enough for there to be second generation ethnic minority groups of a sufficient size to undertake analysis. A comparison of immigrants and descendants is complicated by the different age structure of the two groups. There are no descendants over 35 years of age in the data. For this reason, it is only possible to compare descendants and immigrants between the ages of 18 and 34. The scenario reflected by the table is somewhat complicated. Firstly, there are no significant differences between immigrants and descendants in relation to democratic principles, which support the general findings in relation to this variable. Secondly, in relation to family values, we might expect that descendants are more democratic than immigrants, but the table shows no general pattern here. Among the 18-24 year-olds, the Pakistani descendants are more democratic than their immigrant counterparts, but for the Turks, it is the other way around. Among the 25-34 year-olds, there is only a noticeable difference for the Turkish minorities, but here the correlation is reversed: Turkish descendants are more democratic than Turkish immigrants. These differences appear difficult to explain, but it looks as if their family values depend on the specific situation of the ethnic group. Finally, as regards democratic rights there is no difference between descendants and immigrants in the 18-24 year age group. Among the 25-35 year-olds, descendants are, surprisingly enough, less democratic than the immigrants. In fact, it seems as if descendants are becoming more like the Danes (who are very low on democratic rights) than their parents' generation (cf. Table 1 ). This leads us to a closer analysis of the differences between descendants and Danes of the same age. Danes and descendants The best way of testing the denomination and socialization hypotheses is by comparing people of different denomination, who are all socialized in Danish society. We matched the descendants of Muslim faith with Christian Danes of the same age (18-34 years). Table 3 presents an overview of the distribution of the three dependent democracy variables in these groups, tolerance towards freedom of speech and religious faith is lower than that of the ethnic minorities. When it comes to young people's influence on their own life (choice of education and spouse), the difference between the ethnic groups is much smaller than when discussing more abstract democratic values. There is a significant difference between all the ethnic groups. As the figures in the table reflect, value differences exist in spite of the fact that all three ethnic groups have gone through public education in Denmark. Thus, the hypothesis about formal socialization cannot be supported. The differences must be explained by other factors. One possibility is family socialization. Descendants are not only socialized by social institutions, they also received specific ethnic socialization through their family and network. We lack general information about family socialization in Denmark. However, it would be reasonable to expect that at least some of the descendants have been exposed to relatively strong Islamic values in their families. Some of them may also have attended private Islam-oriented schools but we have no information about this in the data. In the immigrant family, values are shaped both by the country of origin as well as the host society (Foner 1997). In many immigrant families, religion plays a much stronger role than in Danish families, whose lifestyle is generally quite secular. This returns us once more to Huntington's 'clash of civilizations' hypothesis. Is it plausible to assume that devout descendants have less democratic values, whereas religious faith among the Danes has no affect on their democratic values? A multivariate statistical analysis -not shown here because space does not allow it -reflects that for Danes, there is no relationship between religious values and democratic values. Among descendants, however, democratic values are influenced by the religion of the respondent, but there is no difference in democratic values between the Pakistani or Turkish respondents. Figure 2 shows the distribution of democratic values for people with strong religious faith (who report themselves to be 'very religious' or 'somewhat religious') and people who are less devout (who say that they are 'not very religious'). The differences between 'very' and 'less devout' descendants are relatively small, but there is a tendency for the more religious respondents to be less attracted to democratic principles and family democracy and to value religious freedom and freedom of speech higher. This result may be interpreted with reference to Fukuyama's (2006) 'Stuck between two cultures with which they cannot identify, they find a strong appeal in the universalistic ideology offered by contemporary jihadism' (Fukuyama 2006, p. 11 ). While there is no indication in the data that there is a strong fundamentalist Islamic minority in Denmark, Fukuyama's observation may contain some truth, especially for the more religious second generation immigrants for the simple reason that this group is more exposed to the pressures of different value systems. Such an interpretation is supported by an interesting finding in our survey: When asked whether they feel that they have become more religious during the last three years, 37 per cent of the descendants answer in the affirmative, compared to 26 per cent of the immigrants from the same ethnic groups, and 9 per cent of the Danes. Another viable explanation is found in the fact that the descendants may feel that their religion is being discriminated against by the native majority; this leads them to put more emphasis on their right to religious freedom and freedom of speech. Is religion more important than ethnicity in relation to democratic values? Generally, it is difficult to distinguish between ethnicity and religion because most of the immigrants only belong to one denomination: Islam. As has been shown above, the respondent's degree of devoutness has an influence on his/her democratic values, but in the following we shall explore whether differences in democratic values are higher within ethnic groups or within religious groups. If a religious explanation were to be The statistical tests show that there are no significant differences in democratic values among the denominational groups in each ethnicity group. This strongly supports the ethnicity hypothesis -i.e. that the destiny and living conditions of a certain group are the more important determinants of a democratic mindset. However, if an ethnicity hypothesis should be supported, it would also require significant differences among the Muslims in all ethnic groups. This is precisely what we see in Figure 3 . For each variable, there are significant differences among the ethnic groups. Thus, there is no basis for arguing that Muslim respondents are less democratic than non-Muslims respondents. In all three cases, Muslims from Iraq have the strongest democratic values, followed by Muslims from Iran and Muslims from West Balkan. This result is difficult to interpret but might be explained by the history of the three countries or perhaps with reference to the different types of Islam that characterize the groups (e.g. Sunni, Shia). Unfortunately, no data exists that may support such a conclusion. Conclusion Besides being of interest in their own right, studies of ethnic minorities tell us a lot about the host society (Davie 2007) . Often, the conclusions drawn by such studies have been relatively unclear because researchers only study the ethnic minorities and not the majority. In this respect, the present study is an exception because it includes data on the ethnic majority (the Danes) as well as data on a number of ethnic minorities. Our study undertaken has used three variables to measure democratic values: The support for democratic principles, democratic rights and family democracy. The study concludes that the support for democratic rights is high among the ethnic minorities as well as the majority population. Even though minorities show less support for democratic principles, they still score higher than the populations of many West European countries. As regards family democracy, there appears to be a similar pattern as for democratic principles (but we have no information about the distribution of family values in other European countries). The most important difference is found in the support for democratic rights -the freedom of speech and freedom of religion. Here, the Danish majority population is much less democratic than the ethnic minorities. Thus, democratic values do not give rise to a clash of civilizations, as Huntington would have it -at least not when we consider immigrants in Western Societies. Contrary to what Huntington's thesis suggests, Christian Danes are less democratic than Muslim immigrants when it comes to granting democratic rights to all members of society. Generally, democratic values do not differ between immigrants and descendants of the same ethnic group, and a change in values does not appear to depend on the length of stay in the host society. These results emphasize the role of ethnic belonging rather than the denomination or socio-economic factors. For instance, we found that among ethnic minorities with a mixed religious composition, the differences between Muslims are greater than the in-minority differences. These results support our general findings: Ethnicity is more important for democratic values than denomination. Ethnicity is important because of the common history and destiny of the various ethnic groups. This 'ethnic habitus' means that democratic values are based on a common history and fate and reproduced within ethnic groups primarily through social interaction and a shared language within the group. The importance of ethnicity might not come as a surprise to sociologists. As noted by Davie (2007) , immigrant populations arrive in Europe and their experiences vary for specific historical reasons. The Muslim headscarf, for instance, is an acceptable part of public life in Britain, but not in France. In Denmark, the headscarf is quite common in public life but also a highly contested symbol. The difference in acceptance of foreign symbols and customs need to be understood in the context of the individual country's history. Denmark has a history of a strong nation-religion relations, and has been dominated by a single denomination. It would be interesting to study whether other countries sharing the same characteristics also appear to be less willing to accept the democratic rights of all citizens. χ 2 = 39,732, df = 6; p= 0,000 χ 2 = 106,816, df= 6; p = 0,000 χ 2 = 36,591, df = 4; p= 0,000. Note: The differences between the high and low religiosity categories are significant. Democratic principles γ = -0,21 p = 0,009; democratic rights γ = 0,17 p = 0,008; family democracy = -0,18, p = 0,006. Note: The differences between the national groups are significant. Democratic principles χ 2 = 41,018, df=6, p=0,000; democratic rights χ 2 = 99,702, df=6, p=0,000; family democracy χ = 14,867, df=4, p=0,005. onwards, Denmark has accepted a number of refugees from countries such as Iraq, Iran, Somalia and Vietnam. According to the Danish Ministry of Integration (2006a), a total of six per cent of the Danish population are immigrants or descendants from non-Western countries. It is not easy to estimate the number of Muslims in Denmark because the authorities do not register religious denomination. Based on a careful analysis of a number of sources, Jacobsen (2007) estimates that the number of Muslims in Denmark ranged around 207.000 in 2006. This corresponds to 3.8 per cent of the Danish population. The structural societal position of minorities is clearly less favourable than that of native Danes. The rate of employment as well as the level of education and income is considerably lower among immigrants -and among the descendants (Ministry of Integration 2006b). 7 The concepts of immigrant and descendant are regrettably limited to the technical definitions provided by Statistics Denmark since these definitions constitute the foundation for our data collection. According to Statistics Denmark an immigrant is defined as a person born abroad whose parents are both (or one of them if there is no available information on the other parent) foreign citizens or were both born abroad. If there is no available information on either of the parents and the person was born abroad, the person is also defined as an immigrant. This means that the category immigrant also includes refugees. Statistics Denmark (2008) defines a descendant as a person born in Denmark whose parents (or one of them if there is no information on the other parent) are either immigrants or descendants with foreign citizenship. If there is no available information on either of the parents and the person in question is a foreign citizen, the person is also defined as a descendant. Ethnic group refers to the country of origin as registered by Statistics Denmark. If a descendants has parents with different ethnicities, the descendants is assigned the ethnicity of his/her mother (for details, see Statistics Denmark 2008). This assigned ethnicity does not necessarily correspond with the respondent's perceived ethnic belonging. However, the data contain no information about the respondent's self-reported ethnic origin -with one important exception: Respondents were asked if they were of Kurdish origin. A statistical analysis (Gundelach and ( Fledelius 2006). For the purposes of this article, it seems acceptable to consider respondents from West Balkan as a single group. respondent's civil registration number and his/her phone number. For the purpose of this article, values concerning the husband-wife relationship are less useful because marital relationships are heavily infused with gender roles issues. The parent-child relationship, by contrast, is less contaminated when it comes to democratic values. The index is based on the following question: 'How much influence should a young person/parents have on the young person's a) choice of partner or b) choice of an initial overview of democratic values in the Christian majority and in the Muslim minorities. The denomination of the respondents is measured by a question on self-reported religiosity ('What is your religion?"). Non-believers have been excluded from the table. Figure 1 . 1 Model of democratic values for immigrants Since our survey was framed within the context of immigration it is likely that the respondents (Danes as well as ethnic minorities) read 'neutral' questions through the lens of current immigration debates and attitudes in Denmark. The wording of the question on religiosity may point in the same direction. Part of the question concerned the issue of whether people should have the right to carry religious symbols. Since the wearing of religious symbols is of little significance to most Danes, the perception of the question has probably been coloured by the Muslim headscarf debate, leading respondents to assume that the question only related to Islam. This may explain the low percentage of Danes in favour of religious freedom. An overview of survey evidence (Andersen 2002) has concluded that Danes seem to have particular resentment towards Islam. For many Danes, religion and ethnicity are intimately related. Therefore, an unfamiliar religion such as Islam might be mistrusted because it is perceived as offering an alternative allegiance -a potential negation of what it means to be Danish. lives in a democratic country. Let us apply this general hypothesis to the three groups concerned here: analysis by graphical models is nearly identical withRosenberg's (1968) classic concepts of elaboration and specification in connection with analyses of conditional relationships. Rosenbergs' concept of explanation corresponds to conditional independence of variables. The Markov graph is a statistical model that has no links between variables if the association between the variables can be completely explained by the remaining variables in the model. While Rosenberg and colleagues had to limit the analysis to fairly few variables the use of modern computers and allow for analysis of high-dimensional contingency tables. The statistical analysis uses the DIGRAM computer program(Kreiner 1989) . Partial γ coefficients were used throughout the analysis to measure association among ordinal and/or dichotomous variables. Associations involving nominal variables were investigated using χ 2 tests of conditional independence. The significance of test results was in all cases evaluated by Monte Carlo tests as described inKreiner (1987).This statistical analysis is chosen because of the character of the data. Ethnicity, the core variable, is a nominal variable and the other variables are ordinal: This makes it problematic to use for instance traditional regression analysis. The advantage of the graphical model technique is that there are there are no statistical requirements for the character and distributions of the variables. Another advantage is that all variables can be included in the model from the outset of the analysis including all three dependent variables.. Figure 1 1 shows the result of our statistical analysis. Figure 1 . 1 Model of democratic values for immigrantsThe analysis renders the following result: Only ethnic group influences all threedependent variables. Furthermore, the analysis confirms Camp's findings. The length of the respondent's stay in Denmark correlates positively with language skills, but there is no direct relationship between length of stay and democratic values. The variable language skill only correlates with family democracy. The respondent's sense of belonging to Danish society correlates positively with democratic principles and family democracy, but the attitude towards democratic rights only correlates with ethnicity.This leads us to conclude that ethnicity is the most important determinant of democratic furthermore that the length of stay in Denmark has only an indirect or no impact on respondents' democratic values. In other words, our analysis refutes the socialization hypothesis (a result that supportsPettersson's (2007) findings fromEurope).This result shows that primary socialization seem to be stronger than secondary socialization. Another explanation is that the living conditions of ethnic minorities may limit the impact of secondary socialization from the Danish society. The survey shows that only about half of the immigrants in Denmark have a paid job -in some cases, the immigrant's job will not involve direct contact with Danes at all -and only 30 per cent have regular communication with Danes. Moreover, immigrants are typically married to someone with the same ethnicity as themselves. This indicates that immigrants' Figure 2 . 2 Democratic values by religiosity among Pakistani and Turkish descendants. Per cent. hypothesis about identity politics among young Muslims. Fukuyama argues that when Muslims move from traditional Muslims societies to Western societies, they experience an existential disparity between their inner identity as member of a Muslim cultural community and their exterior lives and behaviour in their new European context.According to Fukuyama, this disparity is particularly prominent among descendants. would mean that the democratic values of Muslims are relatively similar irrespective of their ethnicity. This hypothesis is challenged by an ethnicity hypothesis, which claims that differences among ethnic groups are greater than those among denominations due to the impact of common history and destiny. An analysis of these hypotheses can only be conducted among ethnic groups with a varied composition of denomination, i.e. immigrants from Iran, Iraq and West Balkan. Figure 3 3 shows the distribution of democratic values among the mixed ethnicreligious groups. Figure 3 . 3 Democratic values among three groups of Muslim immigrants. Per cent. Figure 1 . 1 Model of democratic values for immigrantsNote: The analysis is based on graphical modelling. A line shows a significant relationship. Figures are partial γ-coefficients. Such cannot be computed with national group because the variable is nominal. In these cases, the measure of association is χ2. Only significant γ's above 0,10 are reported. The analysis includes five immigrant groups (from Iran, Iraq, Pakistan, Turkey and West Balkan). Figure 2 . 2 Democratic values by religiosity among Pakistani and Turkish descendants. Per cent. Figure 3 . 3 Democratic values among three groups of Muslim immigrants. Per cent. Table 2 2 shows the percentage of high the level of democracy values in the four ethnic groups; these groups have been further subdivided into two age groups (18-24 years of age and 25-24 years of age). Table 2 about 2 here. Table 3 3 about here The figures in the table present a picture that we recognize from our general description. The Danes are generally more positive towards democratic rule, but their Table 1 . 1 Democratic values among religious groups in Denmark. Per cent.Notes: The entries show the percentage in the most democratic category. Data is weighted to compensate for bias in the data collection but not according to the size of the national minorities in the Danish population.For operationalization of democratic variables, see text..Table 2. Percentage with high democratic values by age and national group. Democratic Democratic Family Number of principles rights democracy respondents Very high Very high Very high Christian 89 21 64 371-376 Danes Muslim 911-949 descendants 68 53 51 Muslim 1848-1939 immigrants 51 χ 2 = 60.115; χ = 202.664; χ 2 = 46.829; df= 6; df= 6; df= 4; p= 0,000 p= 0,000 p= 0,000 Table 3 . 3 Democratic values among Turkish and Pakistani descendants and Danes aged 18-34. Per cent. Democratic Democratic Family Number of Principles rights democracy respondents Very high Very high Very high Danes 90 23 61 231-235 Pakistan 69 53 49 469-488 descendants Turkey 76 53 56 477-504 descendants Table A.1 Correlations between the variables in the index. Don't know answers are excluded. For wording of questions see text. Correlations between the two Correlations between the two Number of questions on democratic rights questions on family democracy respondents γ p γ p Turkey 0,35 0,002 0,60 0,000 486-504 immigrant Turkey 0,21 0,047 0,53 0,000 498-505 descendant Pakistan 0,36 0,008 0,57 0,000 420-434 immigrant Pakistan 0,12 0,313 0,62 0,000 477-488 descendant West Balkan 0,21 0,001 0,66 0,000 483-503 immigrant Irak immigrant 0,36 0,01 0,67 0,000 491-513 Iran 0,14 0,06 0,70 0,000 492-506 immigrant Danes 0,31 0,00 0,69 0,000 515-521 Note: URL: http://mc.manuscriptcentral.com/rers ethnic@surrey.ac.ukEthnic and Racial Studies Creating a Virtuous Society: Immigration and Australia's Policies of Multiculturalism Gabriel A Almond Sidney Verba JØRGEN GOUL 2002 Danskernes holdning til indvandrere Princeton; ALLAN Princeton University Press ANDERSEN 29 Aalborg: Amid BOROWSKI Learning Democracy in Mexico and the United States Roderic A Camp Mexican Studies/Estudios Mexicanos 19 1 How People Think, Reason, and Feel about Rights and Liberties José Casanova American Journal of Political Science 37 3 University of Chicago Press Public Religions in the Modern World The Sociology of Religion, London: Sage FLEDELIUS, CARSTEN 2006 Personal communication FONER, NANCY 1997 'The Immigrant Family: Cultural Legacies and Cultural Changes Grace Davie Interethnic Migration Review 31 4 Citizens and the State Dieter Fuchs Giovanna Guidorossi Svensson Palle Support for the Democratic System HANS-DIETER KLINGEMANN and DIETER FUCHS Oxford Oxford University Press Identity, Immigration, and Liberal Democracy Francis Fukuyama Journal of Democracy 17 2 Etniske gruppers vaerdier, København: Ministeriet for flygtninge, indvandrere og integration HUNTINGTON, SAMUEL P. 1993 'The Clash of Civilizations? Anthony Giddens Nørregård-Nielsen Esther The Third Way: The Renewal of Social Democracy Cambridge Policy Press GUNDELACH 73 The Clash of Civilizations? The Debate _1997 The Clash of Civilizations and the Remaking of the World Order Simon & Schuster __ London; New York Universal Values and Muslim Democracy Anwar Ibrahim Journal of Democracy 17 3 Human Values and Beliefs -a Cross-Cultural Source Book Based on the 1999-2002 Values Surveys Ronald Inglehart Tørre Tal Om Troen MARGIT WARBURG and BRIAN JACOBSEN Mexico; København Forlaget Univers Muslimer i Danmark -En kritisk vurdering af antal opgørelser Critical Citizens. Global Support for Democratic Governance Hans-Dieter Klingemann Mapping Political Support in the 1990s: A Global Analysis PIPPA NORRIS Oxford Oxford University Press Computerized exploratory Screening of Large-Dimensional Contingency Tables Svend Kreiner Compstat 1986. Proceedings in Computational Statistics. 7 th Symposium Antoni, F de, Lauro, H. and Rizzi, A. Heidelberg, Wien Physica Verlag User Guide to DIGRAM. A Program for Discrete Graphical Modeling Svend Kreiner Copenhagen University of Copenhagen An Informal Introduction to Graphical Modelling Svend Kreiner H.C Mental Health Service Evaluation Knudsen and G. Thornicroft Cambridge University Press Udlaendingeservice NASR, VALI 2005 'The Rise of Muslim Democracy Steffen L Lauritzen Udlaendingestyrelsen ___ 2006b Tal og fakta -Udlaendinges tilknytning til arbejdsmarkedet og uddannelsesystemet Oxford; København; København Oxford University Press 16 MINISTRY FOR REFUGEE, IMMIGRATION AND INTEGRATION AFFAIRS 2006a Tal og fakta på udlaendingeområdet Constants and Contexts in Pupil Experience of Learning and Schooling: Comparing Learners in England, France and Denmark Mehmet Necef Ümit Comparative Education 37 3 Centre for Contemporary Middle East Studies, Odense University OSBORN Ethnic Identity and Language Shift among Young Turks in Denmark Muslim Immigrants in Western Europe: Persisting Value Differentiation or Value Adaptation Thorleif Pettersson MANSOR MOADDEL Palgrave Macmillan New York Values and Perceptions of the Islamic and Middle Eastern Publics Worrisome Signs, Modest Shifts Puddington Aili Piano Journal of Democracy 16 1 Basic Books SAPIRO, VIRGINIA 2004 'Not Your Parent's Political Socialization: Introduction for a New Generation Morris Rosenberg Annual Review for Political Science 7 The Logic of Survey Analysis Statistics Denmark Demokrati Palle Svensson Danskernes saerpraeg PETER GUNDELACH København Hans Reitzels Forlag Do Islamic Orientations Influence Attitudes towards Democracy in the Arab World? Anthony D ; Smith , Nationalism Cambridge RONALD INGLEHART de Sitter Willowdale Islam, Gender, Culture, and Democracy Citizens and the State Jacques Thomassen Support for Democratic Values HANS-DIETER KLINGEMANN and DIETER FUCHS Oxford Oxford University Press ADDRESS: Department of Sociology, 5 Østerfarimagsgade, DK-1014 Copenhagen K, Denmark. Email: pg@soc ku.dk Note: Ellipses show significant difference between descendents and immigrants in the same ethnic group (χ 2 , p<0,05) when controlled for age (xsd:string)
?:linkScore	0.0 (xsd:float)
?:linkSource	GEFUREX pipeline (xsd:string)
?:linkingMethod	search (xsd:string)
?:mainEntity	<https://data.gesis.org/gesiskg/resource/study_program_90>
?:name	WVS (xsd:string)
is ?:referenceMetadata of	<https://data.gesis.org/gesiskg/resource/ssoar_reference_gesis-ssoar-22042_study_program_90>
?:toSource	gesis (xsd:string)
rdf:type	<https://data.gesis.org/gesiskg/schema/LinkMetadata>