This is a comparative analysis of the association between populist attitudes and religious affiliation –namely, Protestantism, Catholicism, and Islam. Using the most recent EVS/WVS survey data from 76 countries, we develop a populist attitudes scale consisting of 10 items and covering some major dimensions of populism on which we find general agreement in the relevant literature. A regression analysis of over 94,000 cases and controlling for education and income, reveals significant religious effects on our populist attitudes scale. Further, and confirming expectations, the signs of religion effects on democratic values and attitudes are the opposite of the effects on populist attitudes. This is an indication that, at the attitudinal level, populism and democracy reside in opposing corners.
Since the last decade, populism has become one of the most frequently used – and, if one may say so, overused and abused – terms in political vocabulary. Both journalistic and academic analyses of national, as well as international politics increasingly rely on this concept to describe and sometimes explain current political developments. Undoubtedly, events such as Brexit, the election of Donald Trump as American President, and democratic backsliding in European Union member states such as Hungary and Poland, have contributed greatly to the need to depend on a catchword which could serve as an umbrella that would cover these and similar developments in different corners of the globe. However, despite – or perhaps because of – the immense popularity of the term, a consensus on its precise definition does not seem to be in sight. From time to time, one even gets the impression that any political party, from far left to far right, that one dislikes can get labelled as populist and the same can be said of political leaders as well.
Although the term itself is much older, quite justifiably, the conference organized by the London School of Economics in 1967 with the participation of an impressive rostrum of distinguished international scholars is regarded as the first rigorous attempt at defining populism. The conference proceedings that were put together in an edited volume (Ionescu & Gellner, 1969) started out with a bold and striking statement: “A spectre is haunting the world –populism.” Marx and Engels’ famous opening sentence of The Communist Manifesto seems to have lost its appeal as well as credibility since the collapse of the Soviet Union but Ionescu and Gellner’s prediction cannot be refuted off hand -at least not for the present and probably not for the foreseeable future either.
Commendable as it was, the 1967 LSE conference was concluded in disappointment with regard to reaching a rigorous or even just agreed-upon definition of populism. Indeed, the only consensus that seems to have emerged from this ambitious and high powered meeting was a “no consensus” on the definition of populism. In a rather harsh review of the conference book cited above, Saloutos (1970) concludes that the authors were unable “to come up with an acceptable definition of populism.” Although considerably more complimentary than Saloutos, reviewing the same book, de Kadt (1970) concedes that “populism is a slippery concept.”
In a much more recent commentary, Baker (2019) sums up the feelings concerning the LSE conference in a rather blunt way. “In 1967, when political theorists from around the world gathered at the London School of Economics […] they had a hard time figuring out exactly what they were supposed to be talking about. […] In the end, the conference proceedings failed to clarify the matter at hand. ‘There can be no doubt about the importance of populism,’ read a summary report. ‘But no one is clear what it is.’”
Where do we stand over half a century after the LSE conference? The situation is not nearly as hopeless as it seemed back then, but the definitional issues still have not been completely resolved. The mere subtitle of Baker’s article (2019) should suffice to sum up the present state of affairs: “The battle to define populism.” Kriesi (2018, p. 5) succinctly summarizes both the immense popularity of the term, as well as the ongoing confusion over its definition: “We are living in a time when the term ‘populism’ has become a buzzword that is used by almost everyone in almost every conceivable situation. The concept has never been known for its exceptional clarity and academics have rather characterized it as ‘slippery’, ‘chameleonic’, or worse.”
Nevertheless, there is some agreement on at least few indicators of this multidimensional concept and the analysis I present below takes advantage of this, admittedly shallow, consensus.
It is more or less commonly agreed that the milestone for the contemporary literature on the subject of populism is the famous article published in 2004 (Mudde 2004). It is now next to impossible to come across to any academic publication on populism that does not include a reference to Mudde’s 2004 article (as of 21 March 2021, Google Scholar reported 4096 cites) and his definition: “an ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’, and which argues that politics should be an expression of the volonté générale (general will) of the people. Populism, so defined, has two opposites: elitism and pluralism” (Mudde, 2004, p. 543; emphases original). Mudde also describes populism as a “thin-centred ideology” (2004, p. 544) in an attempt at justifying the labelling of political parties situated anywhere along the widest possible ideological spectrum as populist.
Mudde’s (2004) widely accepted definition raises a rather fundamental question: is populism a unidimensional concept that can be reduced to the glorification of the “common folk” and the vilification of the “elites?” How about nativism and the related rhetoric of defending the motherland against foreigners, intruders, immigrants, etc.? Or is the yearning for a strong, charismatic leader irrelevant? Are the levels of trust in politicians and/or political institutions to be ignored? Mudde’s response (2004, p. 545) is that “These features facilitate rather than define populism.”
Regardless of whether these features cannot be separated from the definition of the concept, as many including the present author would argue, or be simply – and one might even say casually – regarded as “facilitators,” they still need to be dealt with if the task at hand is the empirical measurement of populism.
Clearly, operationally measuring a concept that is so elusive, vague and even controversial is a task of mammoth proportions. And yet, what cannot be operationally measured cannot be empirically related to anything else. Therefore, the efforts to come up with a list of indicators (given that populism itself has to be treated as a latent variable) must be regarded as “work in progress” and the present chapter should also be read with that caveat.
In a very recent working paper, Norris (2020, p.2) refers to the very same problem and notes that “Unfortunately, systematic, valid and rigorous cross-national measurement of the populist phenomenon has lagged far behind scholarly research.” Then, she details an international project (www.globalpartysurvey. org) aimed at classifying political parties in 163 countries with data obtained through an expert survey. Few attitude questions derived from EVS/WVS questionnaires are intended to relate the positions of the voters of political parties on selected indicators. My aim in this Chapter is to identify populist attitudes at the micro level and to estimate their distribution among adherents of major religions in as many countries as the data will allow.
As noted above, despite the enormous appeal and popularity of the subject, attempts at operational measurement of populist attitudes have been rather limited. We briefly review here some that are relevant for our purposes.
First, we, once more, turn our attention to Mudde and his collaborators whose paper, entitled “Measuring Populist Attitudes” (Hawkins, Riding & Mudde, 2012, p. 7), has direct relevance to our present topic. The authors claim to have developed “four statements that capture the key elements of populism.” These “key elements” to be adopted in survey questionnaires aimed at measuring populist attitudes are, we are told, “a Manichean view of politics, a notion of reified popular will, and, and a belief in conspiring elite.” With these elements in mind, four survey questions are developed. Additionally, the authors draft a module of pluralism (three questions) and use the so-called “stealth democracy” module consisting of four survey questions and drafted by Hibbing and Theiss-Morse (2002). Hawkins et.al. (2012) test the validity of their questions on two large scale surveys carried out in the United States. Two years later, Mudde, this time joined by two different collaborators, published another article proposing scales of populist (eight items), pluralist (3 items) and elitist (three items) attitudes and relating them to political party preferences (Akkerman, Mudde & Zaslove, 2014).
This time their test data came from the Netherlands. As the number of items indicates, the populism scale proposed two years ago was expanded and slightly revised.
Due to limitations of space, I shall forego reviewing additional, but mostly similar, measures drafted by other authors but would like to draw the readers’ attention to an interesting empirical comparison of seven recent scales of populist attitudes (Castanho Silva, B. et. al., 2019) including the Akkerman, Mudde & Zaslove (2014) scale. This article reports the results of the validity tests (along with other major quality checks) of the questions included in these seven scales and propose the items that seem to be the best measures. Their tests are conducted on data from online surveys collected in nine countries.1
Valuable as they are, data limitations of the seven scales mentioned above prevent us from using them in the present Chapter. It will be recalled that our aim is to assess the distribution of populist attitudes among adherents of major world religions. Obviously, to allow meaningful comparisons, this aim necessitates the availability of large-scale international survey data ideally covering as many societies as possible. Among the data sets that these scales are tested with, only the fifth round of the Comparative Study of Electoral Systems (CSES) seems suitable for our purposes. The CSES surveys carried out between 2016 and 2021 cover 20 countries with a total sample size of 35,165. However, a closer examination of the data reveals that Muslims make up only 4.2% of the total number of interviews (n=1,471). And much more importantly, 71% of these 1,471 interviews were conducted in Turkey with an additional 237 coming from Montenegro. Put differently, the overwhelming majority of the interviews are located in only two countries which makes it impossible to disentangle country effects from religion effects.2
Under these circumstances the latest round of EVS/WVS surveys seems to be the only reasonable alternative among international comparative data sets available for analysis. However, EVS/WVS surveys do not include a populism module. To at least partially overcome this limitation, I shall construct a scale with, at best indirect, indicators of populist attitudes from EVS/WVS questionnaires.
Given the deep definitional disagreements, it is doubtful that any scale of populist attitudes can be entirely satisfactory. The scale proposed in this Chapter, on the other hand, is by necessity even more approximate and indirect due to limitations of data.
Sifting through the most recent questionnaires, the following items that are common to both the European Values Study and the World Values Surveys seem to be the best candidates for our scale. These items are:
“Would or would not want a neighbour of a different race?”
“Would or would not want immigrants/foreign workers as neighbours?”
“Agree/disagree with the statement that “when jobs are scarce, employers should give priority to people of this country over immigrants.”
“Trust or do not trust people of another religion?”
“Trust or do not trust people of another nationality?”
“Agree/disagree that it would be a good thing for people to have greater respect for authority?”
“Confidence in political parties?”
“Confidence in parliament?”
“Confidence in the United Nations?”
“Would or would not want a homosexual neighbour?”
The first six items of the proposed scale aim at measuring populist-pluralist dimension in a society. This is important because pluralism, an essential feature of democratic regimes, is regarded as the direct opposite of populism. (Norris, 2019) These six items tap attitudes towards groups that are commonly regarded as the “other.” Thus, unlike pluralists, populists are expected to hold negative views of people of different religions, nationalities and races. Political parties commonly regarded as populist have strong nativist and at times even racist tendencies. (Baker, 2019). Similarly, strong hostile feelings against immigrants are widely regarded as a central characteristic of populism. (Eatwell & Goodwin, 2018; Hawkins, Riding & Mudde, 2012; Joppke, 2020) An emphasis on “dangers” posed by immigrants and support of anti-immigrant policy proposals have been central to the rhetoric and actions of populist leaders around the globe. From Brexiteers over Donald Trump to Viktor Orban, outright hostility towards foreigners in general and immigrants in particular have been a trademark populism and populist leaders.
As Mudde (2004:546) writes “[populists] do not want to change their values or their way of life.” and the objection to having a homosexual neighbor is an indicator of this attitude. Also noted by Mudde (2004:559) is the common observation that populists reject the “political class” and, therefore, lack of confidence in political parties and national parliaments are at least indirect measures of this attitude.
Deep distrust of international organizations and international governance, often highly correlated with nationalistic ideologies, is certainly not a recent or surprising discovery. And although nationalism and populism are not one and the same, both seem to share a profound skepticism of international organizations. As put by Copelovitch & Pevehouse (2019, pp. 169-170) “Around the world, populism and nationalism are on the rise. Everywhere one looks, it seems, the tide is shifting away from globalization and global governance toward economic nationalism and a rising backlash against international organizations.” This observation is shared widely and one is not hard-pressed to cite current examples from different corners of the globe (e.g. Bosco, 2018) The populist challenge to the United Nations was voiced by the organization’s Secretary-General, Antonio Guterres in no uncertain terms: “The U.N. chief painted a grim picture of the state of the world in his opening address to the annual gathering of presidents, prime ministers, monarchs and government officials from the U.N.’s 193 member nations. He pointed to rising polarization and populism, ebbing cooperation, ‘fragile’ trust in international institutions…” (Lederer & Peltz, 2018; emphasis is mine).With these considerations, we can comfortably include our “confidence in the United Nations” variable in our list of indicators of populism and turn to the task of constructing a populism scale using EVS/WVS data.
Assuming I have justified our list of 10 populism indicators based on the most recent EVS/WVS questionnaires, I now proceed to construct a scale of populism. However, I must warn the reader once again that the proposed scale is an incomplete (it does not cover all dimensions that are generally agreed upon in the relevant literature) and indirect (items were not originally formulated as indicators of populism) measure. With these caveats in mind, the task now is to construct the scale and test our hypothesis.
With each item coded as 0 and 1, our simple additive scale ranges between 0 and 10 with higher values indicating higher degrees of populism.3 The frequency distribution of our populism scale4 for 94,006 cases spread over 76 countries (mean=4.90; median=5.00 and standard deviation=2.15) is given in Table 17.1.
Table 17.1 Frequency Distribution of 10-item Scale of Populism
REL. FREQ. (%)
Among the 76 countries for which data are available, on the average, Scandinavian countries have by far the lowest levels of populist attitudes. Table 17.2 gives five countries in our sample of 76 which are at the lowest and five countries at the highest end of our scale. The five countries with the highest scores on the scale give us an indication about the relationship between religion and the prevalence of populist attitudes. In fact, macro level correlations point out to this correlation. However, this chapter is about mass values and attitudes and, therefore, calls for an analysis of micro level data.
Table 17.2 Populism Scale Means in Selected Countries (Lowest and Highest Populism Scores)
In a previous paper (Esmer, 2013), using EVS data from 47 member states of the Council of Europe, I had shown that there was a significant association between religious affiliation and democratic values. Furthermore, this correlation was still strong after controlling for income and education. More specifically, Protestants in Europe, on the average, had the highest and Muslims the lowest scores on a scale of democratic values.
I hypothesize that the same relationship will hold with our populism scale as well. Put differently, Protestants will have the lowest and Muslims will attain the highest scores with Catholics in between the two. I further hypothesize that, similar to democratic values, the relationship will hold when income and education are controlled for. As stated above, to test this hypothesis I use the last round of EVS/WVS data which covers 76 countries.5
The estimates of the regression equation with populism scale as the dependent; Protestant, Catholic and Muslim affiliations (all coded as dummy variables) as independent and income (10 levels) and education (ISCED) as control variables are given in Table 17.3.
Table 17.3 Regression Populist Attitudes Scale on Religious Affiliation and Control Variables
St. err. of B
Although the populism scale is, as I have repeatedly warned, rather imperfect, we observe that all the coefficients in Table 17.3 are in the expected direction and all are highly significant. Furthermore, the effects are quite similar to the ones on the democratic values scale mentioned above (Esmer 2013). More specifically, even after controlling for income and education, being a Muslim has positive, while being a Catholic or Protestant has negative effect on populist attitudes scale. Viewing the standard coefficients, Protestantism has the largest effect. Put differently, being Protestant significantly decreases an individual’s likelihood of espousing populist attitudes.
To our knowledge, this is the first analysis linking attitudes – albeit mostly indirectly – associated with populism with major world religions in the largest possible number of countries located in all corners of the world. In fact, our analysis covers 76 countries although, as emphasized a number of times, the list of available indicators leaves much to be desired and fails to cover some important dimensions of populism. Nevertheless, results indicate a clear religion effect over and above the effects of income and education. Further, as expected, our populist attitudes scale is negatively correlated with democratic values.
We find that adherents of Islam – compared to Catholics and Protestants – have closer affinity to populist values while, Protestants, on the average, score much lower in our scale. To give but one example, Egypt’s score (7.08) is more than three times that of Sweden’s (1.98).
Honouring Loek Halman’s academic legacy (see e.g. Halman & Van Ingen, 2015), it is well documented that religious faith plays an important role in shaping values of adherents. This Chapter shows that we can safely conclude that populist attitudes are not an exception to that rule.
Akkerman, Agnes, Mudde, Cas and Zaslove, Ankrej. 2014. “How populist are the people: Measuring populist attitudes in voters.” Comparative Political Studies. 47(9), pp. 1324-1354.
Baker, Peter. C. 2019. “’We the people’: the battle to define populism.” https://www.theguardian.com/ news/2019/jan/10/we-the-people-the-battle-to-define-populism accessed Feb 15, 2021.
Bosco, David. 2018. “For the U.N. a rise in populism reveals an old challenge.” The Wilson Quarterly, Fall.
Castanho Silva, B., Jungkunz, S., Helbling, M., & Littvay, L. (2019). An Empirical Comparison of Seven Populist Attitudes Scales. Political Research Quarterly, 73(2), pp. 409-424.
Copelovitch, Mark & Pevehouse, Jon C. W. 2019. “International organizations in a new era of populist nationalism.” Review of International Organizations, 14(1), pp. 169-186.
De Kadt, Emanuel. 1970. Populism: Its Meanings and National Characteristics. International Affairs, 46(1), pp. 138-139.
Eatwell, Roger and Goodwin, Mathew. 2018. National Populism: The Revolt Against Liberal Democracy. London: Pelican Books.
Esmer, Yilmaz. 2013. “Democracy, civil society and Islam” in de Hart, Joep, Dekker, Paul and Halman, Loek, eds. Religion and Civil Society in Europe. Springer. pp. 267-84.
Halman, L., & Van Ingen, E. (2015). Secularization and Changing Moral Views: European Trends in Church Attendance and Views on Homosexuality, Divorce, Abortion, and Euthanasia. European Sociological Review, 31(5), pp. 616-627.
Hawkins, Kirk, Riding, Scott and Mudde, Cas. 2012. “Measuring populist attitudes.”http://citeseerx.ist. psu.edu/viewdoc/download?doi=10.1.1.965.5278&rep=rep1&type=pdf. Accessed Mar 24, 2021.
Hibbing, John R. and Theiss-Morse, Elizabeth. 2002. Stealth Democracy: Americans’ Beliefs about How Government Should Work. New York: Cambridge University Press.
Ionescu, Ghita and Gellner, Ernest, eds. 1969. Populism: Its Meaning and National Characteristics. London: The Macmillan Company.
Joppke, Christian. 2020. “Immigration in the populist crucible: comparing Brexit and Trump”. Comparative Migration Studies. Vol. 49. https://link.springer.com/article/10.1186/s40878-020-00208-y. Accessed Mar. 30, 2021.
Kriesi, Hanspeter. 2018. “Revisiting the populist challenge.” Politologick ý č asopis / Czech Journal of Political Science. Vol.1. pp.1-24.
Lederer, Edith M. and Jennifer Peltz. 2018. “In most global of settings, U.N. ponders populism’s problems.” https://apnews.com/article/699abba4c15e4bdea7c7963dec19a3a4, accessed Mar. 14, 2021.
Mudde, Cas. 2004. “The populist zeitgeist.” Government and Opposition, 39(4), pp. 541-563..
Norris, Pippa. 2019. “Varieties of populist parties.” Philosophy and Social Criticism, 45(9), pp. 981-1012..
Norris, Pippa. 2020. “Measuring populism worldwide.” Faculty Research Working Paper Series. Harvard Kennedy School, Feb. 2020. Accessed Mar. 31, 2021.
Saloutos, Theodore, 1970. Reviewed Work: Populism: Its Meaning and National Characteristics by Ghita Ionescu, Ernest Gellner. Technology and Culture, 11(2), pp. 327-329.