Revisiting the islam-patriarchy nexus: is religious fundamentalism the central cultural barrier to gender equality?

Is Islam a religion that promotes patriarchy? In the academic debate, there are different assessments. On the one hand, there is the thesis of an elective affinity between Islam and patriarchal values. In Muslim-majority countries and among Muslims, support for patriarchal values is most pronounced. On the other hand, there is the antithesis of Islamic feminism, which shows that a significant proportion of devout Muslims support gender equality. It is therefore wrong to describe Islam as a misogynistic religion. What matters is whether the religion is interpreted in an emancipatory manner. This contribution offers a synthesis and argues that religious fundamentalism provides a more valid explanation for patriarchal values than simplistic references to Islam. The 6th and 7th waves of the World Values Survey were analyzed to test this research-guiding hypothesis. Multilevel analyses show that value differences between Muslims and non-Muslims and between Muslim-majority societies and societies with another majority religion turn out to be small or even insignificant when controlling for religious fundamentalism. Fundamentalism is the central driver of patriarchal values and generates uniform effects. At the individual-level, fundamentalism makes both Muslims and non-Muslims more susceptible to patriarchal values. Moreover, Muslims and non-Muslims adapt to the conformity pressures of their societies, resulting in egalitarian as well as patriarchal values, depending on the prevalence of fundamentalism. The high support for patriarchal values in Muslim-majority countries has a simple reason: Religious fundamentalism is by no means a marginal phenomenon in these societies, but rather the norm.

and the Pentagon. And this, in turn, is an important reason why criticism of the deficient situation of women's human rights in the Islamic world defies delegitimization as Islamophobic bigotry (Imhoff and Reicker 2012;Tezcan 2015). Where Islamists succeeded in consolidating their power (e.g., Afghanistan, Iran, Sudan), the consequences for women are indeed disastrous, and the list of injustices is a long one. Islamic fundamentalists demand that women obey their husbands, they tolerate domestic violence, and they discriminate against women in legal proceedings. Divorce, for example, is a privilege of men. In addition, there is the compulsion to wear the hijab and the institutionalization of strict gender segregation in public spaces. Self-appointed vicegerents of Allah and the repression apparatus ensure submission to the rules of the game, and even minor violations of these rules are likely to result in draconian punishments (Schröter 2019, p. 72-78). Obviously, these are the most extreme examples, but they are part of a more generic empirical pattern (Koopmans 2020, p. 103-104). The findings of the latest Global Gender Gap Report (World Economic Forum 2022), which tracks gender inequalities along four key dimensions (economic participation and opportunity, educational attainment, health and survival, and political empowerment), underscore the problematic situation of women in much of the Islamic world. The Middle East and North Africa are the world regions with the most severe gender inequalities, and Muslim-majority societies (including Saudi Arabia, Iran, Pakistan, and Afghanistan) are significantly overrepresented among the ten worst-performing countries in this ranking (World Economic Forum 2022, pp. 7, 10). To cut a long story short: The stereotype of the 'oppressed Muslim woman' is a flawed collectivizing extrapolation-but at the same time it does capture a chunk of reality.
In studies that explore the prevalence and drivers of patriarchal values, there are findings that could be cited as evidence for both of these opposing assessments. Based on the World Values Survey, which is one of the largest and most comprehensive surveys in the social sciences, Inglehart and Norris (2003a, b) substantiate an 'elective affinity between Islam and patriarchal values'. Rejection of equal status for women in life domains such as education, economics, and politics is found to be most prevalent in Muslim-majority societies and among Muslims (see also Alexander and Welzel 2011;Lussier and Fish 2016). Studies on 'Islamic feminism' (e.g., Glas et al. 2018;Glas and Alexander 2020;Glas and Spierings 2019;Masoud et al. 2016) criticize the narratives and framings of the aforementioned inquiries. The rebuke is that the aforementioned studies tend to essentialize Islam as a homogeneous and patriarchal entity. Empirically, this criticism is based on findings that challenge the stereotype that all 'Muslims are misogynists' to highlight another side of the argument. An analysis of survey data reveals that one in four Muslim Arabs are devout supporters of gender equality. Thus, and even if this finding does not show majoritarian support for gender equality, a strong religiosity and egalitarian values are not fully at odds with each other. Islam may therefore turn out to be an 'unlikely ally' in the struggles for women's empowerment once emancipatory interpretations of the Islamic faith gain traction (Glas and Alexander 2020, p. 450;Glas and Spierings 2019, p. 293). 3 The findings of these two lines of research allow for some leeway for elaborating a synthesis that could account for the parallel existence of higher susceptibility to patriarchal values in the Islamic world and among Muslims, as well as the non-negligible number of Muslims that are highly religious and in favor of gender equality. Since there are both supporters and opponents of gender equality among devout Muslims, at least one conclusion appears to be apt: The driving force for susceptibility to patriarchal values does not seem to arise primarily from an individual's religious affiliation or the strength of its religiosity. Hence, it is more useful to engage with religious manifestations that hinder emancipatory interpretations of religion and this contribution argues that it is religious fundamentalism that impedes such progressive readings of religion. An important inspiration for this research-guiding hypothesis is the seminal work of Martin Riesebrodt (1998), in which he portrayed religious fundamentalism as radical patriarchal protest movements. According to him, religious fundamentalism is a reaction to intensified modernization and an attempt to preserve or revitalize patriarchal structures to the greatest possible extent (Riesebrodt 1998, pp. 204-206). This contribution draws on this line of reasoning for a cross-cultural analysis of patriarchal values. To this end, I analyzed the sixth and seventh waves of the World Values Survey to revisit the Islam-Patriarchy nexus. The central research question is: Does religious fundamentalism yield a more plausible explanation for susceptibility to patriarchal values than do references to Islam and Muslims? 2 Theoretical framework 2.1 The islam-patriarchy-nexus: more support for patriarchal values among muslims and in muslim-majority societies?
As already mentioned, Inglehart and Norris (2003a, b) delivered some of the earliest empirical clues to an 'elective affinity between Islam and patriarchal values'. Their seminal study is considered a milestone in the sociology of religion since it turned hitherto unexamined speculations about the effects of religiosity and various religious belief systems on patriarchal values into the subject of sound empirical research (Pickel 2019). The findings show that religion operates as a conservative social force, given that religious individuals are, on average, more likely to deny women equal rights in domains such as education, economics and politics (Inglehart 3 Such empirical findings are in line with the concept of multiple modernities, according to which features of modernity (e.g. gender equality) are achieved not at the expense but through a pragmatic adaptation of tradition (Eisenstadt 2002). In this perspective, the headscarf is thus not a symbol of oppression, but a way of combining conservative gender role requirements of families and women's own desire for a religious lifestyle with autonomous participation in public life (Göle 1995). The sociologist Nilüfer Göle (2000, p. 101) even claims that there might be an 'empowerment through Islamism' within this context. The empirical findings of this contribution contradict such interpretations. Gender equality and the Islamic faith can only be reconciled if religious fundamentalism or Islamism is losing its popularity. But see Tezcan (2019) for a critical acknowledgement of Nilüfer Göle's research. and Norris 2003a, p. 67). However, their empirical results also point out that the specific religious belief system is more important for attitudes toward gender equality than the strength of an individual's religiosity. The sharpest value gap occurs between Christians and nondenominational individuals in affluent postindustrial societies and Muslims in agrarian societies (Inglehart and Norris 2003a, pp. 67-68). These divergences in value orientations are exacerbated by the different trajectories of value change in Western societies and in the Islamic world. While notions of gender roles in Western societies shifted substantially between generations and point into a more egalitarian direction, there are hardly any differences between the youngest cohort, their parents and grandparents in Muslim-majority societies (Inglehart and Norris 2003b, p. 69). Inglehart and Norris (2003b) interpreted these empirical patterns through the lens of the 'clash of civilizations' hypothesis (Huntington 1992), and argued that this conflict is less about (paying lip service to) democracy than about women's equal rights and (her) sexual freedoms. In any case, the authors contend that Islam's religious heritage is a social barrier to women's equality (Inglehart and Norris 2003a, p. 71). Several follow-up studies replicated these findings using more sophisticated multilevel analyses. The results are straightforward and confirm that patriarchal values are more prevalent among Muslims at the individual level and in societies in which Islam is the predominant religion (Alexander and Welzel 2011;Lussier and Fish 2016;Norris 2014). In all cited studies, those empirical patterns are attributed to religious socialization effects and some unique aspects of Islam. Lussier and Fish (2016) argue that exposure to Islamic norms promotes the internalization of patriarchal values because the sacred scriptures of Islam (Quran and Hadith) were written at a time when unequal treatment of women was common practice. As a result, there are some passages in the holy scriptures of Islam that imply unequal treatment of women and that are still instrumentalized today as a source of legitimacy for misogynistic practices. Of course, this is not a unique feature of Islam, and the authors do not ignore the fact that there are many Muslims that interpret their religion in an egalitarian fashion. However, they argue that Islam's distinctiveness resides in its strong tradition of jurisprudence, in which the prestige of an author's exegesis is determined by his temporal proximity to the Prophet Mohammed. As a consequence, patriarchal ideas that emerged centuries ago continue to shape the thinking of Islamic ulama today, a situation that in turn impacts the values of ordinary Muslims as they are usually exposed to conservative spiritual leaders upon whom they rely for the interpretation of their religion (Lussier and Fish 2016, p. 32-33). 4 Alexander and Welzel (2011) point out that mosques play the role of an important socializing institution in this context. The transmission of patriarchal values is favored by par-ticipation in religious ceremonies, since worshippers are repeatedly exposed to the religious norms that are propagated in this setting.

H1 Compared to non-Muslims (nondenominational and members of other religions), Muslims display stronger support for patriarchal values
One of the most important findings, however, is that all people, Muslims and non-Muslims alike (for all the observable differences in their values), tend to align to the predominant norms of their social environment. When it comes to the internalization of patriarchal values, it is therefore less important whether or not an individual is Muslim, but whether Islam is the predominant religion in society (Alexander and Welzel 2011, p. 257). Hence, it follows that methodological individualism is in itself incapable of providing a comprehensive account of the driving forces of patriarchal values. After all, the likelihood of internalizing patriarchal norms and values depends to a large extent on an individual's embeddedness in societies in which unequal gender relations dominate daily life (Lussier and Fish 2016, p. 33). Islam tends to play a problematic role in this context, because it has shaped societal beliefs, attitudes, and norms that seem to legitimize rather than undermine an unequal division of labor between men and women in the household and in public spheres (Norris 2014, p. 258).
H2 Compared to societies with another predominant religion, there is stronger support for patriarchal values in Muslim-majority societies

Structural factors beyond islam: lower levels of human empowerment, rentier economies, and kinship ties
Of course, any analysis must consider a variety of structural factors before proclaiming Islam to be a central pillar of patriarchal values. The gradual liberalization of attitudes about appropriate gender roles in Western societies, for example, is not explainable without changing material conditions and societal modernization processes. In the transition from an agrarian to an industrial society, women became part of the labor market as paid workers. This process was accompanied by increased literacy, more education, and a drastic decline in fertility. The associated changes enabled many women to advance into occupations with higher economic status as societies transitioned to postindustrial settings, which in turn opened the way for more women to enter politics. In most societies with viable and durable democratic institutions, women's representation in parliaments grew substantially in recent decades, and many women head ministries or run government affairs. Although inequalities persist, those largescale changes left their mark on the attitudes of ordinary citizens (Inglehart and Norris 2003b, p. 70). These are the reasons why most studies account for economic development (e.g., women's share of the labor force, wealth) and the presence of democratic institutions to explain support for egalitarian gender roles (Alexander and Welzel 2011, p. 266;Norris 2014, p. 254;Lussier and Fish, pp. 48-49). From a more holistic perspective, a society's rising level of prosperity, its turn toward egalitarian values, and the emergence or consolidation of democratic institutions are components of human empowerment that mutually reinforce each other (Welzel 2013, p. 44). Conversely, low levels of human development (e.g., poverty, low education, and low life expectancy) and socialization under authoritarian regimes are likely to boost support for patriarchal values.

H3
The lower a society's level of human empowerment (lower level of human development and lack of democratic institutions), the higher the support for patriarchal values Moreover, it is important to scrutinize whether it is perhaps other factors that prevail in Islamic societies and that drive support for patriarchal values, but which, by their very nature, are unrelated to Islam itself (Alexander and Welzel 2011, p. 250). One such factor that is given importance within this context relates to the question of whether the societies' prosperity is based on rent-seeking. The key reference point for this debate is Ross's (2008) empirical findings that women's inequality in the Middle East is primarily based on economies that derive their wealth from oil and gas exports. Rent-seeking economies, according to the central argument, tend to invest less into the service and agricultural sectors, which reduces women's participation on the labor market and translates into a lower representation of women in parliaments. This situation then in turn compounds the existing social inequalities between men and women (Ross 2008, p. 111). In other words, patriarchal values prevail in Islamic societies because their economic structures tend to exclude women, and not because they are inhabited by Muslims (Alexander and Welzel 2011, p. 250).

H4 Citizens in rentier economies display stronger susceptibilities to patriarchal values
Since studies either failed to substantiate this assumption (Alexander and Welzel 2011;Norris 2014) or at best yielded mixed results (Lussier and Fish 2016), it could be argued that rentier economies are not the crux when it comes to the drivers of patriarchal values. Charrad (2009, p. 548) objects that patriarchal structures have been in place for centuries and preceded the discovery of oil or gas deposits. Moreover, patriarchal values likewise prevail in Islamic societies that are not classified as rentier states. But that does not make Islam the culprit. In everyday life, patriarchal norms are imposed on women through close-knit tribal and kinship networks. These entail traditional role expectations and notions of chastity, which in extreme cases imply the surveillance of women. As a result, women's activities outside the domestic sphere face daunting limitations (Charrad 2009, p. 549). Quite obviously, organizational and structural principles of family systems offer important insights into society's support of egalitarian principles (Todd 1985, p. 7). A study by Dilli (2015) finds that family systems that emerged in agrarian societies continue to impact values of contemporary societies. It is primarily family systems characterized by patrilineal structures and that tolerate endogamy as well as polygamy that provide a social bedrock for patriarchal values. And since the emergence of these family structures preceded the advent of the Islamic faith, they offer a more valid explanation for the unequal treatment of women than simplistic references to Islam (Dilli 2015, pp. 18, 24). Thus, it cannot be ruled out that the thesis of an 'elective affinity between Islam and patriarchal values' suffers from a bias of omitted variables. K H5 The more intense the kinship ties within a society, the stronger the support for patriarchal values

Questioning the islam-patriarchy-nexus: islamic feminism and the compatibility of islamic religiosity and support for gender equality
Studies of 'Islamic feminism' (e.g., Glas et al. 2018;Glas and Alexander 2020;Glas and Spierings 2019;Masoud et al. 2016) convey a counter-narrative to the thesis of an 'elective affinity between Islam and patriarchal values'. To avoid misunderstanding at this point: It is not radically questioned that Muslims are on average more susceptible to patriarchal values than non-Muslims (Glas and Alexander 2020, p. 438). However, this result is interpreted differently. Islam is obviously instrumentalized for the unequal treatment of women (Masoud et al. 2016(Masoud et al. , p. 1562, and the all-male club of the ulema has its share for this setting, as its mainstream propagates a patriarchal version of Islam (Glas and Spierings 2019, p. 289). All of these studies, however, share a common objection: It is by no means all Muslims that submit to a patriarchal interpretation of Islam in a passive manner. Hence, it is wrong to essentialize Islam as a patriarchal entity (Glas et al. 2018, p. 687;Glas and Spiering 2019, p. 284). An analysis of the Arab Barometer and the World Values Survey shows that deeply religious supporters of gender equality are by no means a marginal group among Arab Muslims. Based on latent class analysis, it is shown that about one in four respondents can be classified as 'Islamic feminists' (Glas and Alexander 2020, p. 450;Glas and Spierings 2019, p. 293). Masoud et al. (2016) likewise provide evidence that progressive interpretations of the Quran are able to mitigate patriarchal attitudes toward the public role of women. Within the framework of a large survey experiment in Egypt, a subset of respondents was exposed to an argument from the Quran that advocates the inclusion of women in the political arena. This group was significantly more likely to support a political leadership role for women than the group of respondents that was exposed to non-religious arguments for women's suitability for political positions (Masoud et al. 2016(Masoud et al. , pp. 1575(Masoud et al. , 1567. Two important conclusions can be drawn from these findings. First, the role of Islam is ambivalent. Islamic scriptures are all too often instrumentalized in order to discriminate women, and yet they can also be invoked to demand women's equality (Masoud et al. 2016(Masoud et al. , p. 1590. Second, religious socialization is not necessarily accompanied by a passive adoption of patriarchal norms. Many Muslims interpret their own religion in ways that deviate from the prevailing patriarchal mainstream (Glas et al. 2018, p. 687).

H6
In comparison to non-Muslims, being a Muslim and strongly religious does not amplify support for patriarchal values 'Islamic feminism' is certainly more than a scattered anomaly (Glas and Spierings 2019, p. 299), yet it is important to keep in mind that patriarchal values are still commonplace, since unequal treatment of women is endorsed by over 70% of Arab Muslims (Glas and Alexander 2020, p. 450). The persistence of these patriarchal norms is also a product of the androcentrism that prevails in all Islamic schools of law. Men, who hold the reigns within the ulama, are the primary beneficiaries of patriarchal structures and tend to resist progressive reinterpretations of Islam (Engineer 2004, pp. 211-212;Lussier and Fish 2016, p. 35). Given that women consequently vanguard a non-patriarchal exegesis of the Qur'an, the premise of a gendered perspective on the Islam-Patriarchy nexus seems apt. And there is indeed evidence that women are overrepresented within the ranks of 'Islamic feminists' (Glas and Alexander 2020, p. 450), while, conversely, support for patriarchal norms is particularly pronounced among Muslim men (Alexander and Welzel 2011, p. 263).
H7 In comparison to non-Muslims, being Muslim and a man amplifies support for patriarchal values

Bringing religious fundamentalism into the equation: on its defining features and significance for patriarchal values
Since there are both defenders and opponents of unequal gender relations among devout Muslims, any reference to Islam as the driving force of patriarchy seems too simplistic. This brings other questions to the foreground. To what extent do these two groups differ in their interpretation of their religion? And why is the combination of strong religiosity and support for gender equality (still) a minority position in the Arab world? The punchline of this contribution is that religious fundamentalism offers the key role in answering these two questions. There are also several (implicit) hints in the cited literature that lend plausibility to this research-guiding hypothesis. Masoud et al. (2016Masoud et al. ( , p. 1578, for example, pinpoint the existence of political forces in Egyptian society that oppose an equal role for women in politics. Voters of the (outlawed) Freedom and Justice Party, a party that made its name as the parliamentary arm of the fundamentalist Muslim Brotherhood, were more likely to oppose women in political leadership positions than non-voters and the sympathizers of other parties. A finding by Glas et al. (2018) points in a similar direction. At least it matches the problematized role of religious fundamentalism that textualist religiosity severely limits the space for emancipatory reinterpretations of religious scriptures (Glas et al. 2018, p. 701). Lussier and Fish (2016, p. 36) even elaborate explicitly on the hypothesis that a higher presence of fundamentalist groups in a society is associated with a higher level of support for patriarchal values. However, the study's empirical analysis makes no attempt to test whether this is indeed the case. All in all, religious fundamentalism is by no means a blind spot in the studies cited so far, but it has certainly received too little attention compared to its central role vis-à-vis patriarchy.
Arguing for a patriarchy-promoting effect of religious fundamentalism requires, as a first step, a description of its main features-and there are two reasons why this is not a simple undertaking: First, the term fundamentalism arose in the 19th century as a self-description of ultraconservative and militant movements within American Protestantism. Some critical voices therefore claim that the Protestant origin of the terminology disallows its applicability to movements to other religious traditions (Emerson and Hartman 2006, pp. 130-131). Moreover, the term religious fundamentalism is suspected of being a battle cry. In everyday practice, it is often misused to insult people for taking their religion seriously (Riesebrodt 2000, p. 51). The K result, according to critics, is a demonization of religious groups and an obscure term that makes nuanced analysis virtually impossible (Schiffauer 1999). Neither position sounds convincing to me. The first position clings to a provincial outlook and remains indifferent to religious revival movements beyond Protestantism. Given that religious fundamentalism is a reaction to secularization processes, there is indeed evidence that religious revival movements occur in all world religions (e.g., the Abrahamic religions, Buddhism, and Hinduism) and that they also share family resemblances (Brekke 2012). The second position is also ill-conceived. The pejorative and instrumental application of terms is by no means peculiar to religious fundamentalism, and it is an argument for, not against, a scientific specification of the term (Riesebrodt 2000, pp. 51-52).
This contribution focuses on the ideal type of politicized, legalistic-literalist fundamentalism, since Riesebrodt (2000, p. 96) attributes a strong regressive tendency toward women's equality to this variant of fundamentalism. According to his reading, religious fundamentalism is primarily a radical rejection of the value relativism which is one of the most important signatures of modernity (Riesebrodt 2000, p. 93).
Fundamentalists are nevertheless no medieval forces, but both a reaction to modernity and a product of modernity, since their identity develops in opposition to the accompanying trends of modernity (e.g., egalitarianism, individualism, and secularism). This rejection of modernity is not to say, however, that religious fundamentalists forego exploiting the achievements of modernity for their own ends. Thus, they rely on the most cutting-edge means of communication to propagate their messages (Riesebrodt 2000, p. 50). But this hardly changes the radical nature of their positions. Fundamentalists claim exclusive entitlement to the truth and ascribe universal validity to their beliefs in a supremacist manner. There is no inclination to compromise on these issues, and reinterpretation or adaptation of these principles to the circumstances of the time is rejected. Fundamentalists instead demand that these principles must be applied literally and without revision (Riesebrodt 2000, pp. 89-90). In addition, there is the political ambition to make these rules the standard for everyone. Fundamentalists strive for a (revolutionary) transformation of political realities. Politics is to be subordinated to religion in order to achieve a restoration or maintenance of religious rules (Riesebrodt 2000, pp. 89-90). Riesebrodt's (2000) descriptions of the common features of religious fundamentalist movements are in line with the definition put forward for discussion in this special issue by Pollack, Demmrich and Müller (2022). Religious fundamentalism, in this perspective, entails four central components: 1. the claim to exclusive truth, 2. the claim to superiority over all other positions, 3. the claim to universal validity of exclusive truth, and 4. the demand for restoration of the unadulterated, submerged past through radical change of the present.
But why does religious fundamentalism encounter such great demand, and why do patriarchal values play such an important role for fundamentalism? Riesebrodt (2000, p. 92) traced the rise of religious revitalization movements to experiences of crisis that occur in early and intensified phases of modernization. Political cen-tralization, bureaucratization, commercialization, and secularization usually go hand in hand with the marginalization of broad segments of the population and trigger alienation in traditionalist cultural milieus. Whenever the present is experienced as a source of disillusionment, it becomes fashionable to romanticize the past-and it is precisely at this point that fundamentalism generates its demand. Religious fundamentalism enjoys broad popularity because it formulates a critique of society, a diagnosis of its (alleged) causes and remedies for overcoming the crisis (Riesebrodt 2000, p. 53). In this context, fundamentalism's critique of society is directed against modernity, which is equated with moral decay and an attack on the religious identity of its members. The central discourses of fundamentalists therefore usually revolve around the breakdown of families, divorce, adultery, prostitution, homosexuality, pornography, venereal diseases, alcoholism and gambling (Riesebrodt 2000, pp. 86-87). Fundamentalists also name the alleged 'culprits' of these trends. Depending on the context, these might be foreign powers, the political elites, the economic and cultural beneficiaries of the transformation processes, intellectuals, apostates or members of other religions (Riesebrodt 2000, pp. 87-88). Fundamentalists see themselves in an apocalyptic struggle with such groups and refuse to compromise because, in their view, this is tantamount to the destruction of their most cherished values. The formula for overcoming all problems, on the other hand, is quite simple: The establishment of a political order in which the sacred rules will be binding for all (Riesebrodt 2000, p. 89). Patriarchy occupies a central role in this idealized political order, as it is seen as a remedy for crisis-ridden modernity. To make a long story short: Fundamentalism proclaims (a God-given) dualism of men and women. Within the idealized division of labor, women are assigned the role of subordinates. Their task is to bear and raise children, and their natural domain is the domestic sphere. The role of men is conceived in a complementary fashion: Men are not only fathers, but also the breadwinners and patriarchal guardians of the family (Riesebrodt 2000, p. 88). At the end of the day, Riesebrodt (1998Riesebrodt ( , 2000 leaves no doubt that the affirmation of patriarchy constitutes the cross-cultural umbrella of fundamentalist movements. Studies that examine the sources of out-group hostility, as well as sexist and misogynistic attitudes, confirm this hypothesis. Koopmans (2015) shows that religious fundamentalism is the strongest predictor of hostility toward gays, Jews, Muslims (among Christian respondents), and the Western world (among Muslim respondents), while religiosity does not correlate at all with outgroup hostility among Christians and only weakly among Muslim respondents. A study by Kanol (2021) corroborated this finding using a very heterogeneous country sample. Fundamentalist interpretations of religion generate significant effects on hostile attitudes toward religious groups and atheists. This empirical pattern is observed both within and outside the Western world and among members of the Abrahamic religions. There are entirely comparable findings about discriminatory attitudes toward women: Moaddel (2020) demonstrates for several Muslim-majority societies that religious fundamentalism is associated with a rejection of gender equality. Fundamentalists tend to deny women an equal role in important areas of public life, they insist on female obedience, and they are in favor of polygamy (Moaddel 2020, pp. 65-66, 135). This regressive effect of fundamentalism can also be observed in Western societies.
K Hannover et al. (2018) found that Muslims in Germany are more likely to describe themselves to be religious when compared to Christians and nondenominational individuals and more likely to embrace fundamentalist interpretations of religion. In addition, it is observed that Muslim men are more likely to display hostile sexism, meaning that they are more likely to vilify women that deviate from traditional gender roles. But the results are finally similar to the findings of the studies described earlier. Once religious fundamentalism is controlled for, religiosity does not turn out to be an influential predictor of discriminatory attitudes toward women, nor are the observed differences between other religious groups very salient.
H8 The more fundamentalist an individual's interpretation of religion, the stronger the support for patriarchal values These findings match Riesbrodt's (2000) critique of Samuel P. Huntington's (1992) 'clash of civilizations'. In his view, the analytical substance of this thesis (which also inspired some studies on an 'elective affinity between Islam and patriarchal values') suffers from an exaggerated assumption of homogeneity within religious traditions and groups. One argument against this assumption is that there are devout people in all religious communities who are by no means susceptible to fundamentalist worldviews. Moreover, the hostility of fundamentalists is directed less against foreign powers than against elites, religious minorities and apostates in their own country (Riesebrodt 2000, pp. 29, 87). The result is rampant domestic polarization and culture wars, which arise even in societies where fundamentalist forces managed to seize political power. Despite a well-armed repressive apparatus and state propaganda, the theocratic regime in Iran, for example, has never succeeded in convincing the entire population to adopt its ideal of family and its notions of 'appropriate' gender roles (Riesebrodt 2000, p. 137). Conversely, fundamentalist milieus across civilizations and religious groups are much more likely to display affinities and similarities than to share common ground with their nonfundamentalist fellow believers (Riesebrodt 2000, p. 31).
H9 Being Muslim and leaning toward a fundamentalist interpretation of religion amplifies support for patriarchal values to the same extent as among non-Muslims Martin Riesebrodt (2000, pp. 136-137), however, does not downplay the importance of cross-societal divergences. This is because modernization processes and the enormous expansion of women's participation in higher education and their integration into the labor market also left their mark on fundamentalist milieus. Although patriarchal family ideals continue to be ideologically cherished in evangelical circles in the United States, dual-earners and working women are the prevailing norm today. This in turn shapes their socio-moral attitudes. Especially among younger cohorts of evangelicals, attitudes toward women are becoming more egalitarian and more aligned to the mainstream of American society (Riesebrodt 2000, p. 137). It is therefore imperative to consider the social climate that surrounds individuals and that exerts intense conformity pressure (Alexander and Welzel 2011, p. 272). Strong support for patriarchal values is most likely to be found in societies where fundamentalist interpretations of religion display a pronounced societal prevalence (Lussier and Fish 2016, p. 36).

H10
The more religious fundamentalism predominates the societal climate, the stronger the support for patriarchal values Given that (younger generations of) Muslims in Western societies tend to align with the mainstream of their social environment when it comes to the support of gender equality relations (Alexander and Welzel 2011;Norris and Inglehart 2012), there is no 'Muslim distinctiveness' to be anticipated on this front.

H11
Being Muslim and being exposed to a societal climate in which religious fundamentalist interpretations of religion prevail amplifies support for patriarchal values to the same extent as for non-Muslims

Sample description
The central data set of this contribution is the World Values Survey (Haerpfer et al. 2021). The analysis is based upon the sixth and seventh waves of the World Values Survey and thus on population surveys conducted in the last decade (2010-2020). When the populations of the participating countries were surveyed in both waves, I used the most recent data. The combination of the two waves allows for an analysis of 76 highly diverse societies.
The sample includes the most populous nations on all continents (e.g., the United States, Brazil, Germany, Nigeria, India, and China) and the full range of varying levels of human development and political regimes (from closed autocracies to liberal democracies). The sample also encompasses nations whose majority populations cover all the major world religions or religion-like cosmologies (including Christianity in all its versions, Buddhism, Hinduism, and Confucianism). One exception is Israel, but the sample covers Jewish respondents living as minorities in in other countries. A very broad spectrum of Muslim-majority societies is likewise represented within the sample. The sample includes countries from North Africa (Morocco, Tunisia, Libya and Egypt), the Middle East (Turkey, Iraq, Iran), the Arabian Peninsula (Yemen), the Gulf region (Qatar, Kuwait), the Caucasus (Azerbaijan), Central Asia (e.g., Kazakhstan, Tajikistan, Kyrgyzstan) and Southeast Asia (Bangladesh, Malaysia and Indonesia). In addition, there are countries where Muslims account for a substantial share of the population (e.g., India and Nigeria) and countries in which they live as a religious minority (e.g., Germany and Sweden).

Dependent variable: patriarchal values
The dependent variable in this study is patriarchal values. In line with Inglehart and Norris (2003a, b) and Alexander and Welzel (2011), I measure patriarchal values (Cronbach's alpha = 0.665) in terms of affirmative responses to the following three statements: 'University is more important for a boy than for a girl' (D060), 'Men should have more right to a job than women' (C001), and 'Men are better political leaders than women' (D059).

K
The scores on these items, and any other variables of interest, which are mentioned in the following section were recoded into a scale ranging from 0-1.0, whereby in this case a score of 0 indicates absence and a score of 1.0 the strongest support for patriarchal orientations. Intermediate positions beyond the extremes of the scale are represented in decimal values between 0 and 1.0. For all individual-level items that provide rank-ordered response options (e.g., scales of 1-4, 1-7, and 1-10), values above the midpoint of the scale (0.50) are indicative of a tendency to agree with the statements. When the country means of these scales are aggregated, they allow for the same interpretation as percentage averages (Welzel 2013, pp. 63-64). To calculate the patriarchal values index, I added the scores of the three items and divided the sum by three.

Individual-level independent variables: religious affiliation, religiosity, religious fundamentalism, and gender
The thesis of an 'elective affinity between Islam and patriarchal values' postulates that Muslims are more inclined to ascribe a subordinate status to women. Respondents participating in the World Values Survey were asked about their religious affiliation, and the variable F025 allows a distinction between the adherents of the world's major religions. The studies that address the Islam-Patriarchy Nexus do so by operating with a simple distinction between Muslims and non-Muslims (e.g., Alexander and Welzel 2011;Lussier and Fish 2016;Norris 2014), and I follow this practice. By implication, non-denominational individuals and members of non-Islamic religions serve as the reference category in the empirical analysis. This contribution, however, contends that it is not so much the self-identification as Muslim, but a fundamentalist reading of religion that gives rise to the support of patriarchal values. For this hypothesis to be valid, it must first be ensured that religiosity and religious fundamentalism constitute two separable components. For a detailed account of an individual's religiosity, I use the respondents' self-description as a religious person, the importance they attribute to religion in their own lives, and statements about their religious behavior. More specifically, I use the following items (Cronbach's alpha = 0.803): 'Religious person' (F034), 'Important in life: Religion' (A006), 'How often do you attend religious services' (F028), 'How often do you pray' (F028B) and 'Active membership in a church or religious organization' (A098).
Compared to religiosity, there is a much smaller number of items to tap into fundamentalist interpretations of religion (Cronbach's alpha = 0.706): 'The only acceptable religion is my religion' (F203), 'Whenever science and religion conflict, religion is always right' (F202), and 'Democracy: Religious authorities ultimately interpret the laws' (E225) (see Koopmans 2020Koopmans , p. 37, 2021. While more items are obviously needed for a more detailed measurement of fundamentalism, it is still possible to capture the main components of religious fundamentalism (Pollack, Demmrich, and Müller 2022) with the available instruments. The first item clearly involves an exclusive claim to truth for one's own religion. Moreover, the second item allows respondents to assign universal validity to their religious convictions. They thus place religion above science, even though science holds a de facto monopoly on knowledge since the advent of modernity (Habermas 2003, p. 252). The third point captures fundamentalists' aspiration to subordinate politics to religion. Respondents at least express a sympathy for religious leaders who in their view are supposed to possess 'the ultimate authority' over the interpretation of laws. An analysis of the dimensionality of these items yielded two principal components with Eigenvalues exceeding 1.0 (see Table 1). Keeping in mind that not every religious person adheres to a fundamentalist interpretation of religion, but conversely fundamentalists tend to be religious, there is sufficient reason to assume that the components or factors display a strong correlation.
Based on this line of reasoning, and to simplify the interpretation of the factor structure, I employed the promax-rotation procedure, which belongs to the family of oblique techniques. When a loading criterion above 0.50 is used for interpreting the components, it seems empirically reasonable to distinguish between religiosity and religious fundamentalism. This is not to gloss over the cross-loading of individual items. Fundamentalists and religious people share one (not very surprising) common trait: Both attribute an important role to religion in their lives. And yet, the items on religiosity load particularly strong on the first component, whereas the three fundamentalism items load on the second component. For the empirical analysis, I added the scores of the items capturing religiosity and divided them by five. The same procedure is applied to the items related to religious fundamentalism. The respective scores were added accordingly and then divided by three. 5 5 The following procedure was applied to all indices at the individual-level: If respondents opted to not provide an answer to only a single item, their scores on the indices were constructed using the remaining items. Furthermore, I relied upon the items 'Religious person', 'Important in life: religion' and 'Active membership in a church or religious organization' for both Kuwait's and Qatar's religiosity-scale. The goal of this procedure is to avoid missings on the individual-level.
K Another hypothesis to be tested is that Muslim men exhibit a higher susceptibility to patriarchal values. To test such gendered effects, I draw upon the self-reported sex (X001) of respondents, with women being the reference category in the empirical analysis.

Individual-level control variables
Furthermore, I include several control variables into the analysis. All of them relate to the sociodemographic background of the respondents. One factor of interest is the respondents' marital (X007) and employment status (X028). The reference categories with respect to these two variables are unmarried persons and persons who are neither self-employed nor working in part-time or full-time jobs. Other variables of interest relate to educational resources (X025R) and the age of respondents (X003R2). The analysis differentiates between respondents with low, medium, and high educational resources and membership within three age groups (15-29, 30-49, 50 and older). The reference category is respondents within the youngest age group and respondents with high levels of education.

Societal-level variables: islam, low levels of human empowerment, rentier economies, kinship ties, and a fundamentalist societal climate
Susceptibility to patriarchal values, however, cannot be attributed to individual factors in isolation. Based on the intraclass correlation coefficient, it is even possible to quantify the variance of the dependent variable attributable to the grouping variable or contextual factors. The intraclass correlation coefficient for patriarchal values amounts to 29.2% in the analyzed sample, which is a strong argument for the addition of societal-level factors.
To scrutinize the Islam-Patriarchy nexus, this analysis includes a dummy variable indicating whether Muslims comprise more than 50% of the population within the societies under study. This information is taken from the World Values Survey and cross-checked with the data set of Barro and McCleary (2003). I treat societies with a majority religion other than Islam as the reference group.
Any statement about human empowerment implies information about the material well-being of societies and the existence of democratic institutions. Therefore, the analysis includes both the 2010 human development index (UNDP 2020) and the scores on V-Dem's liberal democracy index (Coppedge et al. 2021) from the year in which the surveys were conducted. The scores of these two indices were added and then divided by two. For the analysis, I use the inverse of the resulting human empowerment index, as low levels of modernization and authoritarian regimes tend to underpin patriarchal values (Pickel 2013).
In addition, a dummy variable is employed to capture the patriarchal effect of rentier economies. The dummy variable provides the information whether states derive more than 40% of their revenues from the export of oil and gas. This information is drawn from Kuru (2014). Non-rentier economies are treated as a reference category in the analysis.
The claim that close kinship ties play an important role in maintaining patriarchal structures has so far hardly entered empirical analyses (Dilli 2015). To investigate their effects, the kinship intensity index of Schulz et al. (2019) provides an extremely valuable instrument. It includes information on cousin marriages, polygamy, coresidence of extended families, lineage organization (patrilineality vs. matrilineality), and endogamy at the community level. To normalize this variable, the lowest score on this index was set to 0 and the highest score to 1.0.
Finally, I utilize the country-specific mean scores on the fundamentalism-scale to shed light on how contextual variations in the prevalence of fundamentalist beliefs affects people's susceptibility to patriarchal mindsets. The country means on the fundamentalism-scale allow statements about the societal climate (Pickel 2009;Welzel 2013).
Combining these different data sets, there are some cases that must be excluded from the analysis due to missing data. V-Dem (Coppedge et al. 2021) does not provide data for Andorra, Macau, and Puerto Rico. Information on the Human Development Index for Taiwan (UNDP 2020) is also missing an the same applies to the kinship intensity index for Singapore (Schulz et al. 2019). In addition, Qatar and Kuwait must be excluded from the analysis because religious affiliation was not queried in the surveys. After removing all respondents with missing data, the dataset includes 69 societies and the response behavior of 96,516 individuals (see Appendix for descriptive statistics).

Results
Before testing the hypotheses in detail, it makes sense to throw a descriptive glance on the data. The question is whether Muslim-majority societies are indeed strongholds of patriarchal values and whether Muslims are more supportive of these values when compared to other religious denominations.
The heat map in Fig. 1 visualizes the intensity of support patriarchal values across the analyzed sample. For the sake of a complexity reduction, the heatmap differentiates between three groups of societies: (1) countries in which only a minority of the population supports patriarchal values (light gray), (2) countries in which less than half of the population is receptive to these values (medium gray), and (3) countries in which patriarchal values are supported by most of the population (dark gray). The empirical patterns replicate the 'values clash' between the Western and Islamic world that was identified by Inglehart and Norris (2003a). Whereas only a minority of the population in Western societies (Western Europe, Scandinavia, the USA, Canada, Australia and New Zealand) still supports patriarchal values, the very opposite is observable in Muslim-majority societies. In North Africa, the Middle East, the Gulf States, the Arabian Peninsula, the Caucasus, Central Asia, and Southeast Asia (e.g., Malaysia and Indonesia), support for patriarchal values is the prevailing norm. This is not to insinuate, however, that patriarchal values are an exclusive characteristic of Muslim societies. In predominantly Christian (e.g., Ghana), Hindu (e.g., India) and Buddhist (e.g., Myanmar) societies, a majority of the population speaks out against gender equality by the same token. And yet, it is hard to deny that Muslim-majority The findings of Inglehart and Norris (2003b) also hold at the individual-level. The violin plots in Fig. 2 visualize the support for patriarchal values among the non-denominational and members of different religions. They clearly show that patriarchal values meet the highest approval ratings among Muslims. In view of the violin plots, one could of course argue that allegiance to a religious denomination is not a matter of fate when it comes to patriarchal values. Across all groups, there are both individuals in favor of gender equality and individuals advocating patriarchal hierarchies between men and women.
But such an interpretation misses the crucial point: On average, Muslims express the strongest support for patriarchal values, and there are significant median differences compared to members of other Abrahamic creeds, members of other religions, and to individuals who do not feel affiliated with any religious denomination. Thus, if empirical research focuses on the Islam-Patriarchy nexus, it is not out of a prejudiced bias, but rather because a reality-based problem is being scrutinized-and this is intense support for patriarchy in Muslim-majority societies and among Muslims.
Such descriptive visualizations, however, do not address the crucial question: Are we dealing with a spurious correlation? To test the hypotheses, I rely on multi-level modeling. This procedure is appropriate for my research interests as it allows to isolate the effects of individual and societal-level parameters. In addition, it enables me to test the hypothesized (cross-level) interaction effects (Hox 2002). The results of the first model (see Table 2) are in line with the descriptive findings. The first and second hypotheses turn out to be both plausible. Muslims (β = 0.082, p = 0.0001) display a higher tendency to support patriarchal values when compared to non-Muslims. It should be emphasized, however, that these value differences between Muslims and non-Muslims do not amount to a sharp chasm. The stronger support for patriarchal values among Muslims turns out to be a strict relative finding and one that is tremendously sensitive to the specific contexts (see Alexander and Welzel 2011). Arguing in favor of this is the fact that a predominant Muslim population within a society (β = 0.169, p = 0.0001) represents a more powerful parameter than whether (or not) an individual self-identifies as being Muslim. The societal-level effect of a predominant Muslim population is remarkable: 41.7% of the observed variance in patriarchal values between societies can be explained by this factor on its own. Models 2 and 3 corroborate that the nexus between Islam and patriarchy is quite robust. Model 2 reveals that the effect of self-identification as Muslim on support for patriarchal values (β = 0.064, p = 0.0001) persists after adding almost all individuallevel control variables. It is worth mentioning, however, that other parameters trigger more pronounced effects. Rather unsurprisingly, patriarchal values find greater appeal among men (β = 0.097, p = 0.0001) than among women. Moreover, religiosity (β = 0.094, p = 0.0001) and lower levels of education (β = 0.099, p = 0.0001) are shown to underpin patriarchal values. Model 3 adds the structural variables discussed in the theory section. This does not alter the robustness of the effect of a majority Muslim population on patriarchal values, however. Although the effect (β = 0.075, p = 0.026) attenuates compared to model 1 and 2, it is still significant. Hypothesis 3 can be confirmed in this and all subsequent models. In line with the theory, lower levels of human empowerment (i.e., poverty and living in authoritarian regimes) sustain patriarchal values (β = 0.398, p = 0.0001). The fact that democracies are a rarity in the Islamic world (e.g., Huntington 1992; Fish 2002; Koopmans 2021) is one of the key reasons why the effect size of a Muslim-majority population is dwindling in this model. Hypotheses 4 and 5, by contrast, need to be rejected. All other things being equal, there is no evidence of higher susceptibility to patriarchal values in K Table 2 Multi-level explanations for patriarchal values.  The table shows the results of several multilevel models. Parameters were calculated using maximum likelihood estimation N (number of observations) are 96,516 respondents at the individual-level and 69 countries at the societal level Entries are the unstandardized regression coefficients with robust standard errors in parentheses Except for the dummy variables, all individual-level variables were centered on the country mean The same applies to the society-level variables, which, however, were centered at the global mean Calculated models allow for random slopes for Muslim respondents Explained variances are calculated from the change in the random variance component relative to the baseline model Estimates were computed using the xtmixed command of STATA (version 16.1) ** < 0.05, *** < 0.01 K rentier economies nor in societies with intense kinship ties. In bivariate analyses at the society level, these two variables are indeed strongly correlated with patriarchal values. However, neither variable is a viable candidate to explain away the patriarchy-promoting effect of a Muslim-majority population (see Model 3). Islam's cultural imprint on societies seems to be more conducive for the preservation of patriarchal values than the existence of lootable mineral resources-and this finding is also logically sound for at least two reasons. First, the emergence of patriarchy preceded the discovery of oil and gas resources. And second, high support for patriarchal values can also be observed in Muslim-majority societies that do not qualify as rentier economies (Charrad 2009). On the surface, Norris (2014) is not wrong in arguing that patriarchal values are more likely to be inspired by 'Mecca than by petroleum'. Conversely, it is rather surprising that the kinship intensity index fails to render the Islam-patriarchy nexus insignificant. Polygamy and endogamy were common practices throughout the Arabian Peninsula well before the advent of Islam, and Islam even introduced certain limits to these practices. However, these customs were not completely abolished either and hence these practices still occur in many contemporary societies. During conquests and the accompanying conversions to Islam, various ruling dynasties succeeded in establishing patriarchal structures and values within societies that lacked strong kinship ties in pre-Islamic times (Engineer 2004). The cases of Indonesia and Malaysia are illustrative examples for this trend. Within the investigated sample, these two countries are the ones with the lowest scores on the kinship intensity index (Schulz et al. 2019) of all societies with Muslim-majority populations. There is a simple reason for this: Prior to the conversion to Islam, matrilineal household structures were not an uncommon phenomenon in Indonesia and Malaysia (Schröter 2021, p. 119-123). One indication of Islam's strong influence on societies is that support for patriarchal values in Malaysia and Indonesia reaches similar levels to that in the Arab world. On balance, model 3 substantiates the thesis of an 'elective affinity between Islam and patriarchal values' (Alexander and Welzel 2011;Inglehart and Norris 2003a;Lussier and Fish 2016). Even when controlling for important structural factors, Islam's ability to shape societies accounts for an explanatory surplus when it comes to patriarchal values. 6 This story, however, gets an entirely new twist once religious fundamentalism enters the equation. As evidenced by Model 4, religious fundamentalism is by far the strongest predictor of patriarchal values. I find evidence of this effect at both the individual and societal-level. Consequently, hypotheses 8 and 10 are not rejected. Respondents that interpret their religion in a fundamentalist fashion (β = 0.248, p = 0.0001) are the strongest supporters of patriarchal values. Accounting for religious fundamentalism, the effect sizes of individual religiosity (β = 0.024, p = 0.011) and the self-identification as Muslim (β = 0.028, p = 0.012) drop consider-6 All in all, most Islamic societies find themselves in a vicious circle. All factors that reinforce patriarchal values are particularly prevalent in these societies. Muslim-majority societies display low levels of human empowerment (r = 0.470); they are overrepresented among rent-seeking economies (r = 0.461); they score high on the kinship intensity index (r = 0.672); and fundamentalist interpretations of religion are widespread (r = 0.729). A regression of all these factors on patriarchal values shows that one can rule out a problematic degree of multicollinearity despite these high correlations. The VIF score is 2.21.
ably. Although the effects remain statistically significant, they are not substantial in content given their minuscule effect sizes. In any case, the results do not permit the impression of an irreconcilable antagonism between Muslims and Non-Muslims or between secular and religious citizens. An even more important factor than an individual's personal interpretation of religion is the societal climate of its surrounding environment. The societal prevalence of fundamentalist interpretations of religion (β = 0.313, p = 0.0001) is the most powerful driver of patriarchal values. One result that deserves emphasis is that the patriarchal effect of a predominantly Muslim population (β = 0.041, p = 0.258) turns out to be insignificant under control for the prevalence of religious fundamentalism. It is by no means wrong that patriarchal values find their strongholds in Muslim-majority societies. But it appears that previous research somewhat oversold the link between Islam and patriarchy. The central reason for patriarchal values is not the inalterable nature of Islam, but rather societal susceptibility to a fundamentalist version of Islam.
The Models 5-8 complement the previous findings by testing the hypotheses that involve assumptions about (cross-level) interactions. To simplify the interpretation of these results, Fig. 3 provides a visualization of the corresponding marginal effect plots (Helmdag 2017). As shown in model 5 and Panel A in Fig. 3, the effect of self-identification as Muslim on patriarchal values is amplified by stronger religiosity (β = 0.061, p = 0.022). Religiosity tends to unleash conservative effects among Muslims and hence increases their susceptibility to patriarchal values. Among non-Muslims, this patriarchy-promoting effect of religiosity is less accentuated. There are no substantial differences between secular and more devout individuals in the non-Muslim reference group. This finding is more easily conceived if one considers the balance of power between religious supporters and opponents of patriarchal values across the religious denominations. Among Muslims in the investigated sample, there is indeed a fraction of 'Islamic feminists' or religious supporters of egalitarian gender relations. In line with the findings of Glas and Alexander (2020), it is almost one out of four Muslims (23.1%) to be classified in this camp. However, most devout Muslims are still in favor of patriarchy (54.1%).
The size ratios of these two groups are significantly different for members of the other religions. Within members of non-Abrahamic religions (e.g., Hindus and Buddhists), there is a stalemate between religious supporters (35.3%) and opponents of patriarchy (35.1%). And in the group of non-Islamic Abrahamic religions (i.e., Christians and Jews), religious proponents of gender equality (50.5%) are even in a majority position vis-à-vis devout defenders of patriarchy (24%). 7 Consequently, hypothesis 6 must be rejected. Among Muslims, religiosity is more likely to fuel an internalization of patriarchal values than to cause an agentic questioning of traditional gender roles. At the same time, this effect should not be overinterpreted. The strength of the interaction effect is not particularly impressive and does not add to the explanatory power at the individual-level. Model 6 furthermore fails to corroborate that Muslim men (β = 0.012, p = 0.298) display any significant stronger inclination toward patriarchal values when compared to their non-Muslim reference group. Thus, hypothesis 7 must be rejected as well. Panel B in Fig. 3 is simply indicating that men in general are more inclined to subscribe to patriarchal values than women. In other words: While women demand equality, men insist on their privileges, regardless of their religious affiliation. The ongoing struggle for gender equality in the Islamic world must therefore reckon with resistance from men, just as it does in the rest of the world. It is not very likely that men will surrender their privileges on a voluntary basis, and it is rather men than women that instrumentalize Islam in order to lend their privileges a sacred patina.
Religious fundamentalism occupies the pivotal role for these ideological ambitions. Since religious fundamentalism generates uniform effects (β = -0.035, p = 0.146), there are no striking particularities unique to Muslims. The more individuals lean toward a fundamentalist interpretation of their religion, the more they support patriarchal values (see Model 7 and Panel C). But even more important than an individual's interpretation of religion is the prevalence of fundamentalism on the societal-level. The societal climate creates tremendous conformity pressure on individuals, and both Muslims and non-Muslims (β = 0.083, p = 0.106) adjust to this group pressure by adopting egalitarian or patriarchal attitudes towards women (see Model 8 and Panel D). Consequently, hypotheses 9 and 11 are not rejected.
The density plot of Panel D in Fig. 3 shows why many studies were able to substantiate the thesis of an 'electoral affinity between Islam and patriarchal values' (Alexander and Welzel 2011;Inglehart and Norris 2003a;Lussier and Fish 2016). The underlying reason for this result is the fact that Muslims account for the bulk of the population in most of the countries displaying high levels of support for fundamentalism. This empirical pattern looms up particularly clear in the scatterplot in Fig. 4. On the one hand, it is evident that religious fundamentalism is a formidable predictor of patriarchal values at the societal-level (r = 0.841, p = 0.0001). The second finding is that Muslim-majority countries are strongholds of accentuated approval ratings towards religious fundamentalism, which in turn translates into a higher susceptibility to patriarchal values. Religious fundamentalism is not a marginal fringe phenomenon in the Islamic world, but the general norm. Azerbaijan is the only country that breaks ranks in this regard. It is the only Muslim-majority society in the sample in which less than half of the population is susceptible to religious fundamentalism. I see at least three compelling reasons why religious fundamentalism provides a better explanation for patriarchal values than simplistic reference to Islam. To begin with, fundamentalism and its regressive patriarchal ideologies only gained momentum in recent decades in societies such as Indonesia and Malaysia. In analytical terms, a static reference to a Muslim majority population fails to deliver a convincing explanation of this cultural drift. The observations of country experts on this subject are more concise. They suggest that fundamentalist movements started to gain popularity in more recent times and that they owe their burgeoning popularity to generous funding from the Gulf states (e.g., Saudi Arabia) (Schröter 2019, pp. 52-62). The misogynistic effects of religious fundamentalism are secondly by no means unique to Islamic societies. After all, it is possible to observe very similar empirical patterns in Ghana, the Philippines, India and Myanmar, to give just a few examples. A third argument is that the highly simplistic reference to Islam is incapable of explaining the large differences between Muslim-majority countries. For example, support for patriarchal values differs significantly between Lebanon and Yemen and covaries with the respective society's susceptibility to fundamentalism. Of course, all these arguments do not change the fact that patriarchal values are rather the norm in most Muslim-majority countries. But in contrast to the static reference to Islam, religious fundamentalism is not an inevitable destiny, but open to value change. Evidence for this possibility is provided by sizeable proportions of devout Muslims expressing support for gender equality (Glas and Alexander 2020;Glas and Spierings 2019).
The studies on 'Islamic feminism', however, do not employ a cross-cultural comparative research design. Inevitably, this leaves one important question untouched: Why are devout supporters of gender equality so severely underrepresented among Muslims compared to other religious denominations? The scatter plot in Fig. 5 illustrates that religious fundamentalism is one piece of this puzzle. At the societallevel, the prevalence of fundamentalist beliefs goes hand in hand with the proportion of respondents being both religious and in favor of patriarchal values (r = 0.916, p = 0.0001). Religious fundamentalism thus offers an explanation why the combination of strong religiosity and support for gender equality is still a minor phenomenon in most Islamic societies. The logic underlying this pattern is rather simple: Religious fundamentalism is a regressive ideology that severely curtails the space for emancipatory interpretations of religion, which in turn reinforces the discrimination of women in societies. The fate of women in Islamic societies is the subject of heated public debate, and academic research likewise offers contrasting assessments of the situation. Within this context, the thesis of an 'elective affinity between Islam and patriarchal values' (Inglehart and Norris 2003b;Alexander and Welzel 2011;Lussier and Fish 2016) encounters criticism from the antithesis of 'Islamic feminism' (e.g., Glas et al. 2018;Glas and Alexander 2020;Glas and Spierings 2019). The first line of research reveals strong support for patriarchal values in the Islamic world and among Muslims. A robust finding that cannot be explained away even when controlling for structural and individual confounding factors (e.g., Alexander and Welzel 2011;Lussier and Fish 2016). The second line of research criticizes the accompanying framings of these findings, contending that it is wrong to describe the essence of Islam as hostile to women. The existence of 'Islamic feminists' or devout supporters of gender equality contradicts this narrative. Hence, women's rights and Islam are not mutually exclusive. Improvements of the situation of women in the Islamic world depends instead on an emancipatory interpretation of the Islamic faith (e.g., Glas et al. 2018;Glas and Alexander 2020;Glas and Spierings 2019). This contribution connects to these ideas and offers a synthesis. Its central argument suggests that religious fundamentalism provides a better explanation for regressive gender norms than simple references to Islam. Moreover, emancipatory interpretations of the Islamic faith are not ruled out either, though it is argued that the prevalence of fundamentalism severely shrinks its playing field. Studies of the Islam-Patriarchy nexus need to be reconsidered once religious fundamentalism enters the equation. The key finding suggests that the 'elective affinity between Islam and patriarchal values' is somewhat overestimated. This is not to deny that Islamic societies are indeed strongholds of support for patriarchal values, nor that Muslims display a particular high susceptibility to patriarchal values in comparison to different religious denominations. It is just that the story takes a new twist after the effects of religious fundamentalism are accounted for. Controlling for religious fundamentalism, there is no significant nexus between a Muslim population majority and support for patriarchal values. At the individual-level, the differences between Muslims and non-Muslims remain significant, but the effect size of this parameter is far too small to invoke scenarios of a vicious clash of values. Once again, it is important to emphasize that the general finding of an 'elective affinity between Islam and patriarchal values' is not wrong on the surface. Islamic societies do display a remarkably high level of support for patriarchal values. The reason for this, however, is not the unchangeable nature of Islam, but religious fundamentalism. In this context, there is some good news and some bad news. The bad news is that religious fundamentalism is by no means a fringe phenomenon in the Islamic world. The good news, on the other hand, is that religious fundamentalism is not a constant, and that it can lose its popularity if a shift in values sets in. The existence of devout Muslims that reject patriarchal discrimination against women is a telling indicator of such developments (e.g., Glas et al. 2018;Glas and Alexander 2020;Glas and Spierings 2019). The thesis of an 'elective affinity between Islam and patriarchal values' is also hardly convincing from a normative perspective. Religious fundamentalists are strengthened in their position if it is claimed that patriarchal interpretations are a logical consequence of Islamic faith. Another argument against such an assessment is offered by the fact that it is primarily (Muslim) men, and not (Muslim) women, that support patriarchal values. People dedicated to a critique of religion are thus well advised to address their rebukes more precisely. Such criticism would be more credible if it were not directed against Islam (or any other religions) in the abstract, but against specific actors that exploit religion for their own agendas. A good audience for this criticism is those imams and parts of the ulema that promote patriarchal interpretations of Islam to ensure privileges for men. One reason why this would be appropriate is that religiosity among Muslims tends to contribute to the preservation of patriarchal norms. Compared to the reference category of non-Muslims, devout Muslims display higher support for patriarchal values than their fellow believers, with a more secular lifestyle. This is not to deny the existence of devout supporters of gender equality among Muslims. Yet it is also important not to overlook the current balance of power. Proponents of patriarchal values do still constitute a clear majority among devout Muslims. This is another observation for which religious fundamentalism provides an explanation. Its strong prevalence curtails the playing field for emancipatory interpretations of the Islamic faith throughout most societies in the Islamic world. At the end of the day, religious fundamentalism turns out to be a deeply misogynistic ideology. Hence, religious fundamentalism offers the strongest account for the support of patriarchal values at both the individual and the societal-level. Controlling for religious fundamentalism, there is a clear leveling of the value gaps between Muslims and non-Muslims. If they subscribe to a fundamentalist interpretation of their religion, they are equally likely to support patriarchal values. Muslims and non-Muslims also adapt alike to the conformity pressures of their environment. In societies with a low prevalence of fundamentalist interpretations of religion, they tend to hold more egalitarian attitudes. Conversely, they are equally inclined to patriarchal values if they live in societies where fundamentalist ideologies predominate. These findings are entirely congruent with the theoretical assumptions and empirical research of Martin Riesebrodt (1998Riesebrodt ( , 2000, in which patriarchal values are described as the cross-cultural common ground of various fundamentalist movements. This contribution suggests that his theoretical insights, informed by case studies of fundamentalist movements, are also applicable to a cross-cultural comparative analysis of patriarchal sentiments. The key source of inspiration for the present empirical analysis is the ideal type of politicized, legalistic-literalistic fundamentalism (Riesebrodt 2000). Its central components also found their way into the definition of fundamentalism proposed by Pollack, Demmrich, and Müller (2022) in this Special Issue. In my opinion, the merit of this definition lies in the fact that the central characteristics of this ideal type are highlighted and accentuated. It is emphasized that fundamentalists claim access to an 'exclusive truth'; that they declare a 'superiority over all other positions'; and that they attest a 'universal validity' to their conception of truth. The politicization of this truth claim is also accentuated by Pollack, Demmrich, and Müller (2022), emphasizing that fundamentalists aspire a 'restoration' and a 'radical change of the present'. These accentuations of the components of fundamentalism are helpful in distinguishing fundamentalists from traditional and orthodox religious groups. Another asset over the much-cited definition of fundamentalism by Altemeyer and Hunsberger (1992, p. 118) is that the fundamentalist claim to truth is not confined to concrete deity conceptions. This specification of the definition is preferable for cross-cultural analyses, given that not all world religions share notions of deity. This is also a good opportunity to pinpoint a potential weakness of the present analysis. Namely, the fact that only three items were available to tap into fundamentalism. It could also be debated whether respondents aspire for a radical change of political conditions if they are seeking for 'religious authorities to interpret the laws'. Having said that, it deserves to be emphasized that the fundamentalism scale used in this study meets a nomological validity criterion in full clarity. After all, the multi-items scale entails an immense effect on its expected correlate of patriarchal values (Welzel et al. 2021). However, it is also clear that more items for each component would be the ideal case. It would thus be possible to assess the fit between the factual dimensionality of these items against the four components of fundamentalism (Pollack et al. 2022). In addition, it would be feasible to examine whether the dimensionality of these items is equivalent across different religious groups (Rippl and Seipel 2015). The empirical validity and generalizability of the fundamentalism definition, as well as the presented results of this contribution, might thus be subjected to more rigorous tests.
This contribution, however, provides important hints that there is a compelling need to capture fundamentalist beliefs and to make a stronger distinction between religiosity and religious fundamentalism. The reason for this is quite easy. There is a tendency to rashly blame regressive tendencies such as patriarchal values and other forms of discriminatory attitudes on Islam, Muslims and religious individuals, even though religious fundamentalism is the crux of the issue. It may sound oversimplified, but fundamentalists are usually very religious, but obviously not all religious people are also fundamentalists. By not including religious fundamentalism into the equation, there is a risk of falling prey to a spurious correlation. This contribution exemplified this possibility using the Islam-patriarchy nexus as an illustrative example. Riesebrodt (2000) has anticipated another stimulating exercise for empirical analysis. Following his portrayals, fundamentalist evangelicals in the United States of America were grudgingly accepting the integration of women into the labor market. Their last bastion since then has been sexual morality-or rather their concept of it. There are thus sound reasons to suspect that fundamentalism is also the main culprit when it comes to the demonization of homosexuality and the obsession with virginity. Table 3 Descriptive statistics. (Source: Own calculations based on the World Values Survey (Haerpfer et al. 2021); Barro and McCleary (2003); UNDP (2020); V-Dem (Coppedge et al. 2021 These are descriptive statistics of the dataset that underlies the multilevel models in Table 2 K