Religion and Support for Democratic Norms: Two Major Debates

The third wave of democratization brought about a lively debate among scholars with regards to the role of religion and religious orientations in democratization. Two major disagreements in the field stand out. The first is with respect to the role of religious traditions in influencing a democratic political culture. Huntington’s famous clash of civilization thesis (aka “the compatibility thesis”) maintained that some religious traditions such as Protestantism have strong prodemocratic elements, whereas Islam, Confucianism, and Eastern Orthodoxy have inherently authoritarian characteristics (Huntington, 1996). Echoing the arguments made by Huntington, Fukuyama (1992) and Kedourie (1992) also contended that Islam fostered antidemocratic attitudes, as democratic norms and institutions such as popular sovereignty, rule of law, or civil society were alien to Muslim political tradition (Kedourie, 1992). These arguments stood in sharp contrast with perspectives emphasizing the role of institutions and structural factors. In one of the most important contributions to the democratization literature, Linz and Stepan (1996) argued that, notwithstanding the role played by the Roman Catholic Church in countries like Poland, Chile, and Brazil in the struggle against authoritarian regimes, religious culture is not a key explanatory factor for democratization.

The implication of the compatibility thesis from the perspective of public opinion studies was whether supporters of some religious traditions had antidemocratic beliefs and values. A second major debate, about the relationship between religion and support for democratic norms, concerns the effects of individual-level religiosity. Whereas seminal works pointed to the fact that religiosity is associated with orientations incompatible with democratic norms, such as political intolerance and extremity (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950; Gibson, 1992; Hunsberger, 1995; Karpov, 2002; Stark, 2001), others viewed it as potentially contributing to the development of civic skills by means of religious social activities (Neiheisel, Djupe, & Sokhey, 2008; Norris & Inglehart, 2004; Putnam, 2000). Based on Putnam’s work on social capital, many authors suggested that religious social activities can have a positive effect on electoral turnout, party membership, and political activism as well as support for democratic norms (Meyer, Tope, & Price, 2008; Norris, 2002; Norris & Inglehart, 2004).

Given the plethora of mixed results and viewpoints regarding the effect of religion on support for democracy, empirical and theoretical scholars alike noted the political ambivalence of religion (Appleby, 2000; Philpott, 2007). The implication of this debate within the context of democratization was clear: the mere process of democratization could foster a return of religion to the public sphere by allowing for religious freedom and encouraging religious political organization and participation. It is thus crucial to know whether allowing for religious freedom and religious participation is a double-edged sword in which democratic regimes encourage antidemocratic values among their citizens.

Viewing religiosity through psychological lenses, current scholarship suggests that a multidimensional conceptualization of both democratic preferences and individual religiosity is necessary to explain the relationship between religion and support for democratic norms (Ben-Nun Bloom, 2013; Ben-Nun Bloom & Arikan, 2012, 2013a, 2013b). Although the literature on the effect of religiosity on different types of orientations related to democracy is extensive and involves religion’s effects on tolerance (Ben-Nun Bloom, Arikan, & Courtemanche, 2015; Carkoglu & Kalaycioglu, 2009; Eisenstein, 2006; Karpov, 2002; Milligan, Andersen, & Brym, 2014; Spierings, 2014, 2018) and political participation and civic engagement (Arikan & Ben-Nun Bloom, 2018; Campbell, 2004; Hoffman & Jamal, 2014; Jamal, 2005; Putnam & Campbell, 2010; Sarkissian, 2012), this article is restricted to the discussion of religion’s relationship to support for democratic norms as citizens’ declaration of general endorsement of a democratic form of government and their understanding of democratic regime principles. It first discusses the multidimensionality of support for democratic norms and individual religiosity, and then reviews the empirical findings concerning the major debates.

The Multidimensionality of Support for Democratic Norms

Classic works view support for democratic regime principles as a key requirement for democracy and democratic consolidation. These works distinguish between different dimensions of support for democratic norms: diffuse or overt support for democracy as a regime on the one hand, and more specific endorsement of particular elements of the democratic system on the other (Easton, 1965; Muller & Jukam, 1977; Norris, 2011). At the very basic level, support for democracy involves “abstract” or “overt” support, that is, citizens’ declaration of general endorsement of the democratic regime and of deeming it desirable for their country (see Inglehart, 2003; Klingemann, 1999; Norris, 2011; Welzel, 2007). This type of support is usually captured with survey questions tapping the extent to which citizens prefer a democratic political system and reject authoritarian alternatives such as having a strong leader, or having military rule. Other common measures of overt support capture the extent to which individuals believe democracy is a well-functioning system by asking them to indicate their level of support with statements such as “Democracies are not good at maintaining order;” “In democracy, the economic system runs badly;” “Democracies are indecisive and have too much squabbling.”

While overt support for democracy is often viewed as critical for the thriving of democratic culture (Almond & Verba, 1963; Dalton, 2004; Diamond, 1999; Easton, 1975; Inglehart, 2003), it does not necessarily manifest a genuine support for the democratic creed (Schedler & Sarsfield, 2007). Rather, citizens may equate democracy to an efficient regime, and may therefore utter support for the democratic regime in the hope that it will bring about beneficial outcomes, such as lower crime, state aid, or economic prosperity. In other words, support for democracy may be “instrumental” in nature, rather than being “intrinsic” (Bratton & Mattes, 2001). In addition, some researchers suspect that “paying lip service to democracy” may have become common among citizens as democracy has become a universal aspiration (Dalton, 2004; Inglehart, 2003). Respondents may provide favorable assessments of democracy due to social desirability and without even knowing its real content in terms of procedures and values (Inglehart, 2003; Klingemann, 1999; Norris, 2011; Schedler & Sarsfield, 2007). Thus, overt support for democracy does not fully capture to what extent citizens endorse liberal democratic ideals or whether they think of democracy as being essentially instrumental in terms of delivering material benefits and/or law and order.

Therefore, scholars identified another dimension support: Intrinsic support for democracy is substantive support for principles and normative values upon which democracy is founded. It consists of procedural understanding of democracy—a conception of democracy as being primarily associated with liberal-democratic norms and institutions, rejection of authoritarian interpretations such as religious leaders interpreting laws and military taking over, and non-instrumental understanding of democracy whereby support is not be contingent on beneficial outcomes such as economic growth or social order (Schedler & Sarsfield, 2007). Accordingly, measures of intrinsic support for democracy often include both a positive dimension, the endorsement of democratic procedures and norms such as free elections, civil rights, gender equality, and freedom of speech as well as a negative dimension, by which citizens reject authoritarian interpretations as well as instrumental outcomes that are not integral to liberal democracy, such as economic prosperity, relative security, lower crime levels, and redistribution of wealth, and by that accept that their endorsement of the regime is not necessarily contingent on its performance (Norris, 2011; Schedler & Sarsfield, 2007).

International surveys such as the World Values Survey (WVS) or Arab Barometer have been increasingly including measures of intrinsic support for democracy. Still, to date, most research on religion and support for democratic norms has largely considered overt support for democracy, with some recent works paying attention to procedural and non-instrumental understandings of democracy as well. Thus, most works discussed in this article focus on the relationship between religiosity and overt support for democracy.

Individual religiosity is also often conceptualized as being multidimensional. Therefore, “The Multidimensionality of Individual Religiosity” considers the different dimensions of religiosity and briefly discusses their relevance to democratic norms.

The Multidimensionality of Individual Religiosity

The literature commonly views religiosity as a multidimensional phenomenon that consists of belief, behavior, and a sense of belonging, dubbed the 3Bs approach (Guth et al., 1988; Layman, 2000; Smidt, Kellstedt, & Guth 2009; Wald & Smidt, 1993; Wald & Wilcox, 2006). First, the belonging dimension refers to identification with a particular organized denomination and/or a religious movement or trends within a denomination. Belonging could be conceptualized and measured as congregational (Djupe & Gilbert, 2006, 2009) or sectarian (Carkoglu, 2005) affiliation, as well as affiliation with a major religious tradition, such as Catholicism, Eastern Orthodoxy, or Islam (Ben-Nun Bloom et al., 2015). Much of the research on the relationship between religion and democratic norms focused on this latter dimension and tested whether belonging to certain religious traditions is associated with higher or lower levels of support for democracy. Some works examined the link between religious affiliation and support for democratic political system at the aggregate level (Al-Braizat, 2002; Norris, 2011; Norris & Inglehart, 2004), whereas the majority of works looked at the extent to which belonging to certain religious traditions was associated with the likelihood of supporting democratic norms.

Even if they identify with the same religious tradition, individuals may differ in their intensity of religious beliefs and their level of engagement with the religious community. At the same time, individuals belonging to different traditions can otherwise resemble each other in their strength of belief and frequency of practice. Although the lion’s share of the debate on religion and democratic norms focused on the effect of religious belonging on support for democracy, the strength of religious belief and variation in religious social behavior are two other dimensions of individual religiosity linked to democratic support. In fact, some authors argued that understanding the effects of these two dimensions is crucial to resolve the major debate concerning the “political ambivalence of religion” in support for democratic norms.

The belief dimension of religiosity encompasses the set of fundamental beliefs, values, and symbols associated with an understanding of the divine and humanity’s relationship to it, and may entail belief in God, heaven and hell, life after death, or the tendency among people to characterize themselves as religious (Layman, 2000). Religious belief is associated with traditional values and authoritarian orientations (Schwartz & Huismans, 1995; Wald & Smidt, 1993) that are seen to be incompatible with democratic expectations such as open-mindedness and tolerance for diverse worldviews (Ben-Nun Bloom & Arikan, 2012, 2013a, 2013b; Canetti-Nisim, 2004; Canetti-Nisim & Beit-Hallahmi, 2007). The behavior component consists of private practice such as prayer or reading of the holy texts, and social practice, which involves participation in religious networks and communities and attendance at places of worship. Most works in the literature focus on the social aspect of religious behavior, which influences political attitudes and behavior via fostering in-group attachment that increases the salience of group interests and the development of civic skills (Djupe & Gilbert, 2009; Jelen & Wilcox, 2002; Norris, 2002; Putnam, 2000) that have consequences for various political attitudes including attitudes toward democracy.

The Compatibility Hypothesis

From the perspective of public opinion research, the compatibility hypothesis implied that belonging to a specific religious tradition was crucial in predicting support for democracy at the individual level (Anderson, 2004). The centrality of the debate coupled with the increasing availability of cross-national survey data, and most notably WVSs, led various researchers to test whether members of some religious traditions hold attitudes and orientations that are hostile to democracy. Although a number of works focused on the compatibility of Eastern Orthodoxy (Anderson, 2004; Hoffman, 2004; Radu, 1998) and evangelical identity (Shah, 2004), a large number of works focused on Islam’s compatibility with democracy (Al-Braizat, 2002; Bratton, 2003; Ciftci, 2011, 2013; Gu & Bomhoff, 2012; Hoffman, 2004; Jamal & Tessler, 2008, 2012; Rose, 2002; Spierings, 2014; Tessler, 2002a, 2002b). Most of the studies testing the compatibility hypothesis use similar measures of overt support for democracy, as well as comparable religiosity indicators, but rely on different analytical strategies. In many of these studies, the focus is on Islam perhaps “in response to the observation that the Muslim-dominated regions of the world have proved particularly resistant to democratization” (Anderson, 2004, p. 201).

To examine Huntington’s seminal clash of civilizations hypothesis, a body of research relied on the comparison of Muslim identifiers to non-Muslims (usually Catholics and Eastern Orthodox) in select regions or countries. These studies showed no significant differences in mean levels of overt support for democracy between Muslims and non-Muslims (Bratton, 2003; Gu & Bomhoff, 2012; Rose, 2002), although with regards to some indicators, Muslims were found to be more supportive of democracy as an ideal, compared to Eastern Orthodox identifiers (Hoffman, 2004).

A second line of works relied exclusively on Muslim contexts and tested whether more religious Muslims are less supportive of democracy than less religious ones. These studies used different dimensions of religiosity as independent variables, along with additional control variables including demographics, social capital variables (interpersonal and political trust, associational involvement), attitudes toward gender equality, and education, to predict overt support for democracy in selected Muslim-majority countries. In line with some of the previous findings, religious belief (Al-Braizat, 2002; Ciftci, 2011; Spierings, 2014; Tessler, 2002b) and religious social behavior (Jamal, 2006; Tessler, 2002a) were not found to have consistent effects on support for democracy among Muslim identifiers, although some differences across contexts emerged (see “Religious Context”).

Outside of the 3Bs framework, researchers also considered another indicator, political Islamism, as potentially influencing support for democracy among Muslim identifiers. Political Islamism is measured by the extent to which survey respondents want public officials to be pious and/or approve the influence of clergy over the decisions of government. Whereas some scholars found no consistent effects of this variable on support for democracy (Spierings, 2014; Tessler, 2002a, 2002b), Ciftci (2011, 2013) found it to be associated with lower levels of overt support for democracy and higher levels of support for sharia. Still, Ciftci argues that the substantive effect of political Islamism is much smaller than variables such as political trust, perceptions of gender equality, and education as well as satisfaction with one’s personal financial situation. In addition, he suggests that supporters of democracy “are able to find the middle ground between Islamic values and democracy” (Ciftci, 2013, p. 790; also see Norris, 2011). These assertions are supported by Jamal (2006), who suggests that those who embrace political Islamism do not necessarily reject democratic institutions. In fact, both supporters and opponents of political Islam seem to have similar levels of openness to diverse political ideas and racial tolerance, and support for equal job opportunities for women, suggesting that elements of democratic political culture are also shared by political Islamists as well (Jamal & Tessler, 2008; Tessler, Jamal, & Robbins, 2012).

Although empirical tests of the compatibility hypothesis with regards to other religious traditions has been relatively limited, in his analysis of Catholic and Protestant identifiers in Chile and Argentina, Patterson (2004) found that support for democracy was slightly higher among devout Catholics and that Catholics generally supported democracy more than Protestants. Still, as was the case with Muslim respondents, it was factors other than religious affiliation, such as education and socioeconomic status, that consistently predicted support for democracy (Rose, 2002; Tessler, 2002a, 2002b; Tezcur & Azadarmaki, 2008). In even more comprehensive analyses that use a larger set of countries, no difference between identifiers of different religious traditions in terms of overt support for democracy emerged (Al-Braizat, 2002; Fish, 2011; Norris & Inglehart, 2004). Thus empirical analysis generally found insufficient support for the compatibility thesis, at least when examining overt support for democracy by different religious groups.

Whereas these results join Linz and Stepan’s (1996) and Halliday’s (1996) in suggesting that religious traditions, and particularly Islam, are not consequential for democracy, there is some evidence that intrinsic support may differ across identifiers of religious traditions. For example, in his study of democratic orientations in sub-Saharan African countries, Bratton (2003) found Muslims to be more likely to associate democracy with instrumental goals such as equality and justice, compared to non-Muslims. Such instrumental support for democracy among Muslim populations was also confirmed by Tessler et al. (2012), with half or more of survey respondents in Arab/Muslim countries stating that economic characteristics such as redistribution are most essential in defining democracy. Again, in Arab nations, personal piety measured with the frequency of Quran reading was associated with greater support for instrumental democracy (Doherty & Mecellem, 2012).

Figure 1 shows mean levels of overt and intrinsic support for democracy among members of different religious traditions. Data come from Wave 6 of the WVSs collected between 2010 and 2014 from Algeria, Azerbaijan, Argentina, Australia, Armenia, Brazil, Belarus, Chile, China, Taiwan, Colombia, Cyprus, Ecuador, Estonia, Georgia, Germany, Ghana, Hong Kong, India, Iraq, Japan, Kazakhstan, Jordan, South Korea, Kyrgyzstan, Lebanon, Libya, Malaysia, Mexico, Morocco, Netherlands, New Zealand, Nigeria, Pakistan, Palestine, Peru, Philippines, Poland, Romania, Russia, Rwanda, Singapore, Slovenia, South Africa, Spain, Sweden, Thailand, Trinidad and Tobago, Tunisia, Turkey, Ukraine, the United States, Uruguay, Uzbekistan, Yemen, and Zimbabwe. Major religious traditions are coded following the guidelines in Arikan and Ben-Nun Bloom (2018). Overt support for democracy is an additive index combining the extent to which respondents embrace democracy as an ideal and reject authoritarian alternatives (namely, having a strong leader, having the army rule, and having experts make decisions). Intrinsic understanding of democracy is a summative index that captures the extent to which respondents endorse a procedural understanding of democracy (i.e., associate democracy with democratic procedures such as free elections, political equality, and civil rights) while also rejecting authoritarian decision-making (religious rule and military leadership) and dissociating democracy from instrumental benefits such as redistribution and state aid. All measures were recoded to vary between 0 and 1. Note that 95% confidence intervals are very small and thus almost invisible due to the large number of observations for groups with the exception of Jewish identifiers.

As can be seen from Figure 1, mean levels of overt support for democracy are highest among Protestant identifiers. The difference in mean levels of support among Protestants and other religious groups is statistically significant. Protestants are followed by Buddhist and Jewish identifiers in terms of their support for democracy. Although members of both traditions in this data set come from a large number of countries, the number of Jewish respondents is very small (n = 178), so these results should be interpreted with care. Catholic and evangelical identifiers have almost identical levels of overt support and are followed by Orthodox, Muslim, and Hindu respondents, respectively. The mean differences between these groups are all statistically different from zero.

Open in new tab

Figure 1. Religious belonging and support for democratic norms.

Data source: World Values Surveys, Wave 6. All indices were coded to vary between 0 and 1.

Findings presented in Figure 1 also show that there are substantive differences in levels of overt and intrinsic support for democracy across members of all religious traditions. Members of all religious traditions included in the data set have significantly less intrinsic support for democracy compared to overt support. Mirroring the findings for overt support, Protestants as well as Jewish respondents have the highest level of intrinsic support. These groups are followed by Catholic, evangelical, and Buddhist respondents. The lowest levels of intrinsic support are observed among Orthodox, Hindu, and Muslim identifiers. Thus, at the aggregate level, there is some evidence that support for democracy is lowest among Muslim and Orthodox respondents as well as the Hindu. It should be noted that 88% of Hindu respondents in the data set come from India, and therefore these findings are largely representative of the Hindus living in India.

It is not obvious from such aggregate analysis whether the differences across traditions can be explained by theology itself or whether they are attributable to differing levels of religious belief and religious social behavior among the identifiers of different traditions. For example, at the aggregate, Muslims have the highest levels of religious belief, and Protestants have the highest levels of religious social behavior. A more detailed analysis of the relationships between religiosity dimensions is necessary to understand how religious identification influences support for democratic norms across members of different religious traditions. Belonging to a certain religious tradition, such as Islam or Christian Orthodoxy, also entangles a variety of relevant contextual elements. For instance, for the Middle East and North Africa (MENA) region, which the Muslim respondents in the data set mainly come from, being Muslim equals being a religious majority member as well. Thus, there are many factors that may explain the variance in levels of support among members of different religious traditions that could be dependent on contextual factors.

To test the compatibility thesis, researchers generally considered whether religious affiliation or related aspects such as political Islamism were associated with antidemocratic regime preferences. More recent studies also consider other dimensions of individual religiosity, such as the strength of religious beliefs or the frequency of religious involvement (Ben-Nun Bloom & Arikan, 2012, 2013a, 2013b; Canetti-Nisim, 2004; Meyer, Tope, & Price, 2008). In fact, these studies suggest that such an approach could provide useful insights into the compatibility debate as well as helping to settle the debate concerning the ambivalence of religion.

The Ambivalence of Religion

Conclusions and Future Directions