This paper will address the question of Islamic fatalism. Survey data will be used to assess Samuel P. Huntington's controversial "Clash of Civilizations" thesis and its emphasis on fatalism as an inherent characteristic of Islamic religion. The concept of fatalism is expanded and theorized as a function of both structural and theological dimensions. Findings here suggest that fatalism in the Islamic world remains a largely misunderstood phenomenon. Christians living in predominantly Muslim countries are no more fatalistic than their Muslim neighbors; and in Indonesia, Christians report higher levels of fatalism than Indonesian Muslims. However, Muslims do indicate a higher level of belief that cosmological forces control life's outcomes than do Christians living in those same Islamic societies. Findings also suggest that the effect of Western influence on fatalism is not as straightforward as that predicted by Huntington's theory. Fatalism in the Muslim world is best understood in light of complex historical, cultural, economic and socio-political processes and not as a direct outcome of Western influence and/or religious denomination alone.
Islamic Fatalism and the "Clash of Civilizations" Thesis
Samuel Huntington places a major emphasis on Islamic beliefs as a source of cultural strain in the modern world. Few essays have generated the volume of scholarly and public debate that Huntington's controversial argument has. The starting point for this study is Samuel Huntington's (1993, 1996) well known "Clash of Civilizations" thesis and the series of often heated debates that Huntington's writings have ignited. As Huntington ominously predicted in his 1993 article, "the fundamental source of conflict in this new world will not be primarily ideological or primarily economic. The great divisions among humankind and the dominating source of conflict will be cultural." (1993:22)
This research aims to consider the validity and consistency of the claims put forth by proponents of the "Clash of Civilizations" thesis. A focus on one specific element of Huntington's thesis will generate specific hypotheses and allow for the C.O.C. theorists' position on Islamic fatalism to be tested. For Huntington and other C.O.C. advocates, Islam is often portrayed as a religion that engenders a fatalistic world view. Before proceeding to a theoretical discussion of fatalism as a sociological concept, I would argue that the notion of fatalism and the fatalistic orientation that Islam allegedly engenders can be seen as a crucial, if underappreciated, component of the C.O.C. perspective.
The fatalistic orientation that C.O.C. theorists often point to is typically described as being at odds with the Westernized, Judeo-Christian vision that fosters the ethic of individual self-empowerment and has influenced progressive social change in the West (Lewis 2003; Manji 2004; Spencer 2002; Warraq 2003). This line of reasoning can be traced back to classic work in the field of religious studies where sacred Qur'anic texts are interpreted by literary scholars as fatalistic. For example, as Helmer Ringgren, the renowned scholar of comparative religion, has bluntly remarked, "[i]t has become a commonplace that Islam is a fatalistic religion which teaches that everything is determined in advance and that man is unable to do anything about it." (1967: 52)
Proponents of Huntington's thesis have also embraced the almost Weberian suggestion that there is a troublesome elective affinity between Islam and a cultural word view that is fatalistic. (1) For C.O.C. theorists, such a fatalistic collective mindset can be described as one that is adverse to the ideals of self-empowerment and individualism that characterize many democratic, Christian nations. Instead, it places the burden of life's outcomes at the hands of omnipotent, metaphysical forces. In fact, according to one critic of Huntington's thesis, such an understanding is consistent with arguments from Huntington and others (e.g., Fukuyama 1992; Lewis 1990) who make the rather uninformed observation that "Muslim countries have not evolved because they have avowed fatalism, authoritarianism, and have not separated the temporal from the celestial." (Kader 1998) Islam thus encourages a rigid collective orientation where submission to the will of Allah results in a superstitious and fatalistic cultural identity.
Several culturalist arguments that either predate or, in more recent examples, illustrate the C.O.C perspective, have argued that fatalism is a general characteristic of traditional value systems that are unreceptive to processes of modernization and economic development (Banfield 1967; Harrison 1985; Harrison 2000; Sosa 1998). As Daniel Etounga-Manguelle, scholar and founder of the Societe Africaine d'Etude, dExploitation et de Gestion, has noted, "Africa, except for the southern tip of the continent, appears to belong entirely to the category of societies with weak control over uncertainty." (Manguelle 2000:68)
For Manguelle and other contributors to the C.O.C. scholarship, control in Islamic societies is often removed from the will of the individual and instead placed at the authority of the sacred. This is a rather crude line of reasoning that foreshadows a more sociologically informed conceptualization of fatalism. But first, can we ask if this shared fatalistic mindset is a characteristic feature of all or most Islamic societies as the C.O.C. theorists have maintained? In their collective scholarship one constant theme is that Islam engenders a spirit of extreme acquiescence and obedience to cosmological forces. Take, for example, comments from the Iranian essayist, novelist and pro-Democracy advocate Amil Imani who has remarked that "one of the greatest subtle, yet important differences between the Muslim's mindset and that of the people in the West is the extent to which Muslims are fatalistic ... The rank and file Muslim has little will of his own. It absolves him of any and all responsibility." (Imani 2006)
Proponents of Huntington's thesis have held firm in their conviction that fatalism--as an underlying tenet of Islamic doctrine--negatively affects the collective ability of national publics to successfully engage the project of modernization and development (e.g., Lewis 1997; Manguelle 2000). Some have expanded on Huntington's thesis to consider such topics as intracultural conflicts between moderate and fundamentalist Muslims (Bilgrami 2003; Timmerman 2003) and the historical origins of the perceived divide between Muslims and non-Muslims (Lewis 2003; Lewis 2004).
Critics have attacked Huntington's work as largely anecdotal and lacking a concrete historical context (Ahmed 2003; Huntington et al. 1996; Qureshi and Sells 2003; Said 2001a; Said 2001b). Others have gone further and labeled the work as politically motivated and maliciously misleading (Bunzl and Said 2004). And in an equally provocative review of the book, Pierre Hassner has gone as far as calling the work "morally objectionable [and] politically dangerous." (Hassner 1996/1997) Other critics have argued that Huntington's thesis actually disguises the failings of Western Democracy that lead to fears of a potential Islamic resurgence (Mahbubani 1993). Whichever stance one takes, there is general agreement that Huntington's writings have fueled heated debates within and outside academic circles. What is less obvious in the C.O.C. literature, however, is a comprehensive analysis that takes the sociological dimensions of the fatalism concept as its starting point. While the fatalistic qualities inherent in Islam are a central feature of C.O.C. theory, what is less apparent is a serious attempt to adequately theorize the fatalism concept in the first place. A major shortcoming of the C.O.C. view of Islamic fatalism, and of much of the literature on fatalism in general, is a failure to fully conceptualize a multidimensional concept.
Disaggretating Fatalism as a Sociological Concept
Durkheim added to his categorization of anomic, egoistic and altruistic suicide by briefly mentioning fatalistic suicides as those occurrences of suicide that result from the power of complete societal coercion over the individual which facilitates the taking of one's own life at a moment of utter hopelessness (see Durkheim 1968:275-76). A host of sociologically informed suicide studies have followed Durkheim's lead and conceptualized fatalistic suicide as resulting from an oppressive, over-regulating social structure (Bearman 1991; Dohrenwend 1959; Douglas 1967; Durkheim 1968; Lester 1994a; Lester 1994b; Pickering and Walford 2000; Taylor 1982).
This research intends to expand previous discussions of Durkheim's fatalism concept (Acevedo 2005; Bearman 1991; Taylor 1982), and will argue that fatalistic worldviews not only develop in response to the type of structural forces that Durkheim identified but also as a result of widely held belief systems that lead adherents to accept life's outcomes. For example, Weber's comparative sociology of religion, and particularly his studies on the religions of India, place a much needed emphasis on the role of cosmological forces such as karma and reincarnation in shaping fatalistic worldviews (Weber 1996). When accepted as compelling explanations of one's place in the world, such concepts may offer forceful "other-worldly" explanations of social order. A concise starting point for this examination can be found in the important theoretical contribution from David Lockwood (1992) in which the author reasons as follows:
"[T]here is after all an important difference between fatalistic beliefs that stem from the individual's realization that he is personally in the grip of circumstances over which he has no control and fatalistic beliefs that are the result of his socialization into an ideology that provides a comprehensive account of why circumstances are beyond his (or anyone else's) control." (Lockwood 1992:44, italics added)
As noted by Lockwood, for Weber, the fatalistic element inherent in the karma doctrine leads to an acceptance of one's social position outside the regulation of material forces and places this acceptance within the realm of religion and culture. Appropriately, Weber interprets the interaction between fatalism and caste as follows: "The caste system and karma doctrine place the individual within a clear circle of duties and offer him a well-rounded, metaphysically-satisfying conception of the world." (Weber 1996:132, italics added) This "metaphysically-satisfying conception of the world" can then be interpreted as an awareness that there exists an internal order and logic to everyday life and that life's outcomes are inevitably dictated by otherworldly powers.
Joseph Elder (1966), in a more recent ethnographic study, has systematized the fatalism concept and proposed three distinct typologies that allow for the specification of well-defined fatalism types. Elder's model offers an ideal theoretical conceptualization that I see as essential for demystifying stereotypical depictions that often mischaracterize Islam as a fatalistic religious system. As will be argued in more detail below, a major flaw in the C.O.C. scholarship is a failure of researchers to account for the subtleties of the fatalism concept as found in Islamic theology and teaching. Instead, C.O.C. theorists often conflate fatalism with the central Islamic principle of "submission." Elder's (1966) ethnographic account of Hindu and Muslim believers lends substantial support for the idea of fatalism as a multidimensional mental construct (see also, Adams 1974).
Of primary interest here are two dimensions of fatalism that Elder introduces. First, Elder acknowledges that fatalism is, in a Durkheimian sense, simply a feeling of powerlessness. Initially, Elder describes the fatalistic orientation as one in which individuals internalize a belief that life's outcomes are "determined by factors over which [they] have little influence, and [the] acceptance of this state of affairs as being correct, natural or just."(Elder 1966:228) This fatalism type echoes Durkheim's cryptic account of the fatalistic condition as one resulting from social conditions where "futures [are] pitilessly blocked and passions violently choked by oppressive discipline." (Durkheim 1968:276) It is this type of fatalism --one characterized by "a belief that empirical phenomenon occur for no comprehensible reason, and [that consequently] they cannot be controlled" --that Elder identifies as "empirical fatalism." (Elder 1966:229)
Elder goes on to propose a second category of fatalism that is less reliant on the disposition of an oppressive social structure and that is more associated with specific cultural belief systems. What Elder calls "theological fatalism" represents a move away from Durkheim and towards Weber's idea of "well-rounded, metaphysically-satisfying conception of the world" noted above (Weber 1996:132). Theological fatalism is, as Elder notes, "the belief that God or some moral order such as karma control's man's destiny and the outcome of his actions."(Elder 1966:229) However, Elder is quick to point out that such theological fatalism may in fact engender specific types of social action that are interpreted by the individual as necessary for the achievement of desired outcomes. So while "man may be powerless in terms of the outcome of any specific action ... over a longer time span man can shape his identity by being virtuous, carrying out God's will, or accumulating merit."(Elder 1966:228)
Why should fatalism, as opposed to other potential differences between Christianity and Islam, be the focus of a test of C.O.C. theory? First, is simply that fatalism plays a significant if underappreciated role in the C.O.C. scholarship. Islam has long been associated with a specific brand of extreme fatalism that is too often depicted as irrational and fanatical--a view that C.O.C. theorists have been quick to capitalize on and around which they develop many of their basic assumptions. Secondly, the discussion offered here is a more refined conceptualization of fatalism as a sociological concept, an approach applicable to fatalism/self-efficacy studies even beyond the current analysis. And finally, the availability of two distinct measures of fatalism (empirical and theological)found in the 2002 Gallup Poll of the Islamic World (GPIW 2003) allows us to (1. empirically assess both the empirical and theological dimensions of fatalism discussed here and (2. assess the validity of C.O.C. claims that it is Islam per se that fosters a sense of personal powerlessness amongst adherents that is not characteristic of Christianity. In keeping with this line of reasoning and realizing that a more comprehensive discussion is outside the scope of the current study, a few words on the nature of Christian fatalism may be useful in helping to frame related C.O.C. hypotheses.
The word Islam suggests the idea of submission. Too frequently however, this has been associated with an irrational sense of fatalism that Islam allegedly fosters in its adherents. Ideas of fatalism in Christianity, on the other hand, are typically associated with notions of predestination and are juxtaposed with teachings on free will. For example, the most recent entry for fatalism in the digital version of the vast Catholic Encyclopedia contrasts the early Christian notion of fatalism with the fatalism of Islam by arguing that, "several of the early Christian writers were concerned to oppose and refute the theory of fate [and instead placed emphasis on] the principle of man's moral freedom and responsibility ... Consequently, free will is a central fact in the Christian conception of human life."(Catholic Encyclopedia 1907-1912)Islam, on the other hand, reinforces a "tendency to belittle the individuality of man, [and shapes the] development of a theory of predestination approximating towards fatalism." Additionally, the entry foreshadows later C.O.C. arguments on the effect of Islamic fatalism on Middle Eastern publics by stating that "[w]hilst the belief in a predestined lot has tended to make the Moslem nations lethargic and indolent in respect to the ordinary industries of life, it has developed a recklessness in danger which has proved a valuable element in the military character of the people."(Catholic 1907-1912)
Such an assertion mirrors Weber's implicit assertion that Islam fosters a unique form of fatalism that is removed from the more logical and coherent notion of predestination present in Protestant theology. Sociologically, as early as Weber's comparative sociology of religion, the distinction between the principle of Christian predestination and the fatalistic orientation in Islam is mentioned. As Weber noted, "[t]he Islamic belief in predestination easily assumed fatalistic characteristics in the beliefs of the masses."(Weber  1991:205) Here, Weber explicitly associates Islamic theology with irrationality and consequently "argued that the potential for predestination within Islam was transformed into an irrational form of fatalism that worked against modern capitalism."(Lane and Redissi 2004:66)
In sum, Christian fatalism has generally been associated with a subtle and more rational notion of predestination while Islam has been characterized as fostering an extreme form of predestination that sways the theology towards fatalism} Furthermore, the idea of free will, while playing a central role in most interpretations of Christian predestination, is in Islam, generally interpreted as altogether missing. Instead, Islam is said to minimize or even denigrate the significance of the individual will.
Modern Islamic scholarship has offered a more refined interpretation of Islamic fatalism that denotes two important characteristics. First, current Islamic scholarship shows parallels between the notions of fatalism as found in Islamic theology and that of predestination found in Christianity (Belo 2006; Esposito 1997; Esposito 2002). At the same time however, this same scholarship addresses the differences in terms of the Islamic emphasis on personal submission and how this characteristic of the faith may be erroneously interpreted and associated with irrationality, fatalism and powerlessness (Nasr 2001; Peters 1993). In fact, it has been argued that Islam, much like the other major world religions, exhibits the essential quality of all the dominant world religions--that of eliciting quite rational seeking behavior in its adherents (Stark and Finke 2000).
The crux of the matter for the current analysis, however, is the manner in which Huntington and other C.O.C. theorists have depicted Islam as an intrinsically fatalistic cultural worldview that is (1. distinct from that of Christianity and (2. is a causal variable affecting the collective frame of mind shared by modern Muslims. "There are," according to Swiss journalist and author Roger Du Pasquier, "many Westerners for whom Islam can be reduced to three ideas: fanaticism, fatalism and polygamy."(1992:5) Huntington has unabashedly identified the unifying aim of his scholarly work as that of preserving "Western civilization in the face of declining Western power."(quoted in Al-Ahsan 2006:538) For Huntington, this desire is clearly rooted in his identification of Islam as a rival of the West and as the antithesis of the Christian value system. As the contemporary scholar of Islamic thought Abdullah Al-Ahsan has noted, "Huntington clearly advocates a perpetual conflict between Islam and Christianity."(2006:537) As Huntington himself has remarked, "enemies are essential for people seeking identity."(quoted in Al-Ahsan 2006:538)Implicit in his theory is the claim that Islam is the enemy of Christianity. Again, Huntington has remarked that "It]he relations between Islam and Christianity, both Orthodox and Western, have often been stormy. Each has been the other's Other."(quoted in Al-Ahsan 2006:536)
The aim here is to take such a vociferous claim and to put the general C.O.C. expectation to the test. If, as Huntington and his proponents have maintained, there exist such marked differences in the theological conceptualization of fatalism found in Christianity and Islam, we should then expect such theological underpinnings to have an effect on the beliefs shared by Christian and Muslim adherents. And if it is the case that such differences are predicted to exist, we should be able to look to public opinion data to examine the plausibility of the C.O.C. contention that Islam fosters a more pronounced sense of fatalism than Christianity. This line of questioning makes possible the assessment of two interrelated C.O.C. hypotheses that examine the differences between Christians and Muslims living in two predominantly Muslim nations with substantial populations of Middle Eastern Christians.
Hypothesis 1: Christians living in predominantly Islamic societies will be less empirically fatalistic than Muslims living in those same predominantly Islamic countries.
Hypothesis lb: Christians living in predominantly Islamic societies will be less theologically fatalistic than Muslims living on those same predominantly Islamic societies.
The analysis also examines the related proposition that Westernization will have an impact on measures of fatalism. Again, a central concern shared by Huntington and other proponents of his thesis is that Islam and the West are on a cultural collision course that will largely define 21st century global conflict. Some C.O.C. authors have gone as far as arguing that the only resolution to the sluggish pace of development in the Islamic world is a complete embracing of Western values of capitalism, secularization, and technological/scientific evolution. As Daniel Pipes has noted, "To escape anomy, Muslims have but one choice, for modernization requires Westernization ... Islam does not offer an alternative way to modernize ... Secularism cannot be avoided ... Only when Muslims explicitly accept the Western model will they be in a position to technicalize and then develop." (Pipes 1983:197-198)
The key themes of Pipes' argument often resonate and are echoed in many of the writings from Huntington and other C.O.C. theorists (e.g., Harrison and Huntington 2000; Pipes 1981; Pipes 2002). The current analysis allows empirical examination of the validity of such claims as they relate to measures of fatalism. In other words, is it possible to predict that the amount of Western influence on a country would affect levels of fatalism in that nation? If the preceding claims made by C.O.C. theorists are accurate, one would expect that persons living in countries with a longer, more visible history of westernization, modernization and secularization would be less fatalistic than persons living in countries that have purposely evaded Western influence? Likewise, one would expect persons in more westernized and secular Islamic countries to place less emphasis on the role of religion and cosmology in guiding life's affairs.
Turkey presents an interesting case study of the effects of westernization on measures of fatalism. Characterized by a historical legacy of government-sponsored secularism, Turkey has seen more recent trends towards economic privatization, democratic representation and continues negotiations for formal entry into the European Union. One major C.O.C. theorist has gone as far as noting that Turkey represents "the only secular, Islamic country that approaches modern standards of pluralistic governance."(Harrison 2000:301) Consequently, the C.O.C. theoretical perspective should predict that public opinion in a country like Turkey will exhibit features that most closely resemble that of Western publics: a higher sense of individual control over life's affairs and a less pronounced desire to grant control to higher powers.
At the other end of the continuum, Saudi Arabia presents the case for a more puritanical, isolationist Islamic society. Wahhabism--an 18th century reformist movement within Islam that teaches a strict, literalist interpretation of the Qur'an--has led to a path of cultural separatism from the West. (3) Theologically, Wahhabism has been described as focusing "on ritual correctness and punctilious adherence to Islamic law."(Dean 2006:181) The more recent embodiment of Wahhabism has been influenced by Saudi nationalism and anti-Western sentiment in a belief that "the West now possesses subtler, insidious means such as the Internet and satellite television to infiltrate Muslim homes and spread western ideas and values." (Dean 2006:181) Wahhabism in Saudi Arabia has led to a purposeful and often government-enforced minimization of cultural contact with the West. The doctrine remains the official doctrine of the modern Saudi empire (Dean 2006). With these interrelated propositions in mind, we submit:
Hypothesis 2. Muslims living in more Westernized Islamic countries will be less empirically fatalistic than those living in countries that are unreceptive to Westernization and Western influence.
Hypothesis 2b. Muslims living in more Westernized Islamic countries will be less theologically fatalistic than those living in countries that are unreceptive to Westernization and Western influence.
Measuring Empirical Fatalism
The primary data source used for the present study is the 2002 Gallup Poll of the Islamic World (GPIW 2003). The sample design is based on a multistage probability sample with survey collection based on personal-in-home interviews. The weighting variable provided by the principle investigators was used for the current analysis. In order to ensure country-level consistency in the questionnaire items that were available Turkey, Lebanon, Iran, Indonesia and Saudi Arabia are included in the analysis.
One shortcoming of previous self-efficacy literature is the coalescing of fatalism measures that have an intrinsic religious component to them (e.g "Everything that happens is part of God's plan") with those that do not (e.g "We each make our own fate"). And although factor loadings are generally respectable when such measures are correlated, one primary theme of this paper is that there is a conceptual reason to question feelings of fatalism based on social conditions and those based on belief systems (see also, Jacobson 1999). Another benefit to using the GPIW data is the presence of questionnaire items that specifically capture the two dimensions of fatalism (empirical and theological) discussed above.
The GPIW contains one primary measure of empirical fatalism. This item asks "How strongly do you feel that you are in control of what you would like to do in your life?" (GPIW 2003: GPIW, codebook item 16.2) Here respondents replied to a 5-point scale with the number 1 corresponding to the response "No control" and the number 5 to the response "Full control." The numbers 2-4 did not contain corresponding written selections. For ease of interpretation, responses were reversed so that the scale range is 0 through 4 with 4 corresponding to a "no control" response. In short, a higher response corresponds to a higher empirical fatalism score.
Measuring Theological Fatalism
The measure of theological fatalism found in the GPIW captures a sense of granting control to cosmological forces that is distinct from the previous measure of empirical fatalism. Respondents were presented the statement, "No matter what I want to do or be, there is much more superior power that fully determines the course of my life."(GPIW 2003: GPIW, codebook item 16.3) Respondents again answered on a 5-point scale with the number 1 corresponding to the response "Totally disagree" and the number 5 to the response "Totally agree." The numbers 2-4 did not contain corresponding verbal selections. Responses were recoded so that the scale range is 0 through 4 with 4 representing a stronger sense that superior powers fully determine the course of one's life, or a higher level of theological fatalism.
Independent and Control Variables
Gender, Age and Place of Residence
A growing body of literature has shown age to be a significant predictor of attitudes in Islamic countries (Haddad and Khashan 2003; Hassan 2002; Hoffman 1993; Ibrahim 1980). More directly related to issues of self-efficacy, Hoffman (1993) has portrayed the psychological profile of many young Muslims as one characterized by feelings of anxiousness, and confusion amidst the insecurity of a changing world. Age for analysis using GPIW data was calculated by taking the ten available survey item midpoints and using the natural log of those midpoints.
Gender has likewise been shown to be an important variable that explains variance in both forms of fatalism discussed thus far. Women have consistently reported having less control over their lives than men and as such show higher levels of empirical fatalism in most major studies addressing this issue (Bandura 1995; Goodwin and Allen 2000; Grenstad 1990). Gender was coded as a dummy variable with "male" serving as the reference category (0 = female, 1 = male).
Finally, there is reason to suspect that place of residence will have an impact on fatalistic views. Peasant life, for example, has often been associated with fatalistic orientations (Redfield 1956; Shanin 1987; Tumin 1952; Wolf 1966). Recent studies reveal a more pronounced sense of fatalism among rural populations in the United States and Europe (Cohen and Nisbett 1998; Marcovitch 2003; Padfield 1980). The GPIW place of residence variable was recoded as a dummy variable with two categories: "urban" and "non-urban" as the reference category (0 = "urban," 1 = "nonurban").
Attitude Toward Religion as Essential
Formation of a "religiosity index" was made difficult because all questions were not asked across the entire set of countries of interest. Thus, I apply Stark's reasoning and suggest that "the best single measure of personal piety is simply to ask people how religious they are." (Stark 2002:496) Recent macro-level analyses have also successfully utilized individual, non-index survey items as dependent measures of underlying religious behaviors and beliefs (e.g., Barro and McCleary 2003; Stark 2002). The GPIW contains a detailed item that states:
"Here are some aspects of life that some people say are important to them. Please look at them and categorize them into three separate categories: those that are essential and you cannot live without, those that are very important, and those that are useful but that you can live without." (GPIW 2003: GPIW, codebook item 12.1)
The aspect of life that is of concern in terms of religiosity is a question that asks how vital "having an enriched religious/spiritual life" is to the respondent. Three ordered categories are kept for the analysis and have been recoded for ease of interpretation so that a higher number denotes a more intense and pronounced sense of religiosity (0 = "Useful, but can live without," 1 = "Very Important," and 2 = "Essential/cannot live without it."(GPIW 2003: GPIW, codebook item 12.1C)
The homogeneity of the Islamic countries represented in the GPIW is reflected in the presence of only two religious categories that are dummy coded as follows: 0 = "Christian," 1 = "Muslim" as the reference category.
Socio-economic status represents an important control variable when assessing measures of perceived fatalism. Higher levels of socio-economic status and education are shown as negatively correlated with fatalism with the more affluent and educated "perceiving themselves as having greater control over their lives."(Goodwin and Allen 2000; see also, Cohen and Nisbett 1998; Kohn 1989; Mirowsky and Ross 1998; Schieman 2001)
For the GPIW data, highest level of education is a categorical variable with seven possible categories. The seven categories have been collapsed and recoded so that 0 = "low level of education," and 1 = "high level of education" as the reference category. Due to the large number of missing cases, a middle range educational category was not feasible using GPIW data. A measure for income asks "Which one of these income brackets comes closest to your household average monthly income?" Thirteen possible monthly income categories were recoded and a new variable to reflect monthly income midpoints (natural log) was used as a continuous variable. Finally, four dummy variables for occupation were coded with 0 = "housewife," 1 = "student," 2 = "unemployed," and 3 = "employed" as the reference category.
For the GPIW data, two variables are used as controls that may have a relevant impact on fatalistic outlooks. The first is a dummy coded variable for marital status where 0 = "single," 1 = "other," and 3 = "married with children." The number of married respondents without children is minimal so these respondents were collapsed with the original "other" category.
The second family structure variable in the GPIW dataset was family size. Respondents were asked to select the actual number of family members. Possible responses were presented as an ordinal scale with a range of "1" to "10+" family members.
Nationalism and Country-Level Effects
Five country dummies from the original GPIW data are used for this analysis. The key consideration in country selection was partly theoretical but also driven by practical concerns. First, these cases reflect countries where the survey items of interest were all asked, so that there is a consistency across all statistical models. Secondly, it did seem crucial to include Turkey and Saudi Arabia in the analysis, as Turkey represents the most Westernized, and democratically established case; Saudi Arabia, the least. Country dummies were constructed with Turkey serving as the reference country (0 = 'Indonesia," 1 = "Lebanon," 2 = "Iran," 3 = "Saudi Arabia," 4 = "Turkey").
Findings & Discussion
Effects of Denominational Variation on Empirical and Theological Fatalism
The first C.O.C hypothesis of interest predicts that Middle Eastern Christians should be less fatalistic than Middle Eastern Muslims. By concentrating on Indonesia and Lebanon--countries with measurable Christian populations--it is possible to gauge the effects of religious denomination in the "geographically controlled" environment of specifically non-Christian countries.
Mean and median fatalism scores are shown in Table 3 and indicate that there is measurable variation in fatalism scores between Muslims and Christians in Lebanon and Indonesia. (4) But do these differences hold when performing more complex statistical analysis that control for relevant individual level characteristics? Table 4 examines the factors that may account for empirical and theological fatalism in Lebanon and Indonesia while holding relevant factors constant. Coefficients from the two-country specific models suggest no significant denominational variation in levels of empirical fatalism between Christians and Muslims in Lebanon. There were actually higher levels of empirical fatalism amongst Indonesian Christians when compared to Indonesian Muslims--a finding that is contrary to C.O.C. expectations. Moreover, while Lebanese Christians are no more or less empirically fatalistic than Lebanese Muslims, the theological fatalism model shown in Table 4 suggests that Lebanese Christians are less theologically fatalistic than Lebanese Muslims. In short, while Lebanese Muslims assign more control to "a higher power" than Lebanese Christians, they simultaneously feel equally empowered on a personal, individual level --a notion that may help shed some light on common C.O.C. depictions of Muslims as somehow intrinsically fatalistic.
In Indonesia, where Christians are more empirically fatalistic than Muslims, there is no statistically significant difference between the two groups in terms of theological fatalism. However, we should note that the coefficient for theological fatalism in Indonesia is in the predicted negative direction and, just exceeds the. 10 cutoff for statistical significance (Pr > [chi square] = .113). (5) So while there is overall support for Hypothesis 1b of greater theological fatalism amongst Muslims when compared to Christians, there is no support for C.O.C. hypothesis la of increased levels of empirical fatalism amongst those same Muslims. So even as empirical fatalism models shown here refute a simplistic C.O.C. expectation of a Muslim fatalistic mindset, higher levels of theological fatalism among Lebanese Muslims may be quite consistent with the theological expectations of Islamic teachings.
Take, for example, the fact that etymologically, Islam refers to ideas of submission and the central tenet of Islam is self-surrender to the will of Allah. And this is where, in my view, it becomes important that we correctly assess the full dimensionality o f fatalism as a sociological concept. The erroneous mischaracterization that Islam lends itself to a fatalistic outlook arises out of two primary shortcomings. First, it is simply the idea that C.O.C. theorists have failed to account for theological fatalism as a multidimensional sociological concept. Secondly, the C.O.C. literature often lacks an objective and systematic interpretation of Islamic theology and is consequently unable to offer a more persuasive explanatory framework from which to consider issues of self-efficacy in the Islamic world.
The findings in Lebanon, in particular, are consistent with the theological underpinnings of Islamic faith whereby the absolute authority of life's affairs rests with the will of Allah and submission to that will is understood as the central tenet of the faith (Esposito 1997; Rahman 1979; Rahman 1989). The Lebanon coefficients suggest that Muslims do not submit in the sense of hopelessness or desperation that might lead to a heightened sense of empirical fatalism but rather in a fully optimistic manner that is divorced from materialist conceptions of self-control. In effect, Islam should not be seen as a correlate of empirical fatalism but, quite the opposite, as a doctrine that teaches absolute control over the self as a reflection of one's submission to a greater force. As the religious historian Vincent J. Cornell (1999:67) has noted,
"Islam is the conscious and rational submission of the contingent and limited human will to the absolute and omnipotent will of God. Islam's advocacy of self surrender should not be thought of as irrational however, or dismissed as the product of a passive or fatalistic mentality. On the contrary, the type of surrender Islam requires is a deliberate, conscious, and rational act."
In Indonesia, contrary to the expected hypothesis, Christians are in fact more, not less empirically fatalistic than Indonesian Muslims. And as Table 4 also shows, unlike Lebanon, there is no statistically significant difference in theological fatalism between Christians and Muslims in Indonesia. So what might account for such an outcome in terms of empirical fatalism in Indonesia? Two possible explanations--one related to historically contingent demographic factors that led to religious pluralism in Indonesia, the other to the more recent tensions resulting from this multireligious landscape--should be mentioned here. And while a fuller treatment is outside the scope of the current analysis, several points are worth mentioning. First, it is important to note that as early as the 16th century, Indonesia's history of exogenous colonization has created a multireligious society that has encountered religious influences ranging from Portuguese Catholicism, Dutch Protestantism and, in more recent times, contact with modern proselytistic denominations such as Jehovah's Witnesses and Seventh Day Adventists (Ricklefs 2001). In more recent periods of Indonesia's history, the influence of various minority faiths, accompanied by growing dissatisfaction with Islamic reform movements, has resulted in waves of conversion to Christianity, and to a lesser extent Hinduism and Buddhism (Ricklefs 2001). Also, the indigenous Javanese traditions often intermingled with existing faiths to create hybrid religious expressions (Geertz 1968; Hassan 2002; Woodward 1989). And finally, a fourth instance of Indonesia's religious pluralism can be seen in the growing presence of well-educated, wealthy, urban Chinese who continue to account for a large portion of Indonesia's Christian population and who have had a considerable impact on Indonesia's economic structure (Ricklefs 2001). Chinese Christians have played a vital role as entrepreneurs in the Indonesian economy and, as an influential historical work on indonesia has suggested, it has been "estimated that around 70 percent of all private economic activity was in Chinese hands in the 1990s." (Ricklefs 2001:393) Such a pattern mirrors consistent historical patterns of religious minority control of economic activity in predominantly Islamic societies (e.g., Kuran 2004a; Kuran 2004b).
The successes experienced by Indonesia's religious minorities have not, however, come without a price. Starting in the mid 1960s, Islamic reformers throughout Indonesia pressed for a more rigorous and fundamentalist application of Islamic doctrine, resulting in an increased hostility to Indonesia's existing religious minorities. According to M.C. Ricklefs, a leading scholar of Indonesian history, "In April 1967 a series of violent incidents began when Muslims attacked Christian churches in Aceh. In October very serious anti-Christian riots broke out in Makasar (Ujung Pandang) and religious violence was reported from Java and Sumatra." (2001:355)
Considering this disquieting reaction to Indonesia's modern religious pluralism, the statistical results for increased empirical fatalism among Indonesian Christians may be at least partly explained by the greater sense of structural oppression that Indonesian Christians experience compared to Lebanese Christians. As a substantial minority population exhibiting high levels of socio-economic attainment and political empowerment, Lebanese Christians should exhibit lower levels of empirical fatalism. Indonesian Christians, while often members of Indonesia's wealthier elite, are a substantially outnumbered minority that continues to experience often harsh persecution and discrimination in Indonesian society. 6 While higher levels of theological fatalism amongst Lebanese Muslims may be consistent with Islamic teaching, empirical fatalism seems to be affected by a wide range of social forces; these forces exist apart from any ideological motivation that C.O.C. theorists consistently point to as its cause. While religiosity is clearly a powerful and driving force in Indonesia, nationalism and political context also play a vital role in Indonesian social life and public opinion (e.g., Bertrand 2004).
Again, the C.O.C. expectation falters and findings suggest the need for more refined interpretations of Islam and its effect on fatalistic orientations amongst adherents of the faith. Instead of assuming a monolithic "Islam effect," C.O.C. theorists should consider the type of structural characteristics of specific Islamic countries and the historical trajectories experienced by distinct Islamic societies. Indonesian Christians and Muslims differ on a host of relevant characteristics beyond religious affiliation. It is this essential point that C.O.C. theorists often overlook.
Western Influence and the Fatalistic Imagination
Much of the C.O.C. literature places the burden of sluggish societal "progress" on the inability of Islamic nations to follow a Western model of societal development. One reason for this lack of development is the inherent fatalistic mentality, according to C.O.C. theorists. As a logical consequence, one should expect a country such as Turkey--a Middle Eastern country that Huntington cites as the model Westernized, secular, democratic Islamic state (Huntington 1996)--to exhibit lower levels of both empirical and theological fatalism? If it is the case that fatalism is a function of Islamic influence, shouldn't the influence of Westernization affect levels of fatalism in a country like Turkey?
In order to test this claim, we can look to both descriptive data and to multinational regressions that use country dummies as a means of examining country-of-origin effects on empirical and theological fatalism. Table 5 shows descriptive statistics for the two measures of fatalism included in the GPIW data. In terms of empirical fatalism, the data indicate that it is in Turkey and Lebanon where respondents feel the least amount of control over life; a finding that runs counter to the C.O.C. expectation. Theological fatalism, on the other hand, is moderate in Turkey, which lends partial support to the C.O.C. hypothesis. However, it must be noted that theological fatalism is lower in Saudi Arabia than in Turkey which refutes the hypothesis of lower theological fatalism in highly Westernized Islamic countries. Beyond these basic descriptive statistics though, a clearer picture emerges when we consider full regression models.
As shown in Table 6, coefficients from ordered logistic regression models suggest that the size and direction of the "Westernization effect" on fatalism is not as straightforward as that predicted by C.O.C. theory. In order for C.O.C. hypothesis 2a to be supported, empirical fatalism should be lowest in Turkey and highest in Saudi Arabia. On the whole, Turkey seems to be characterized by relatively moderate levels of empirical fatalism and high levels of theological fatalism. (7) It is also evident by looking at full statistical models that Saudis represent the most empirically fatalistic group, a finding that does support the C.O.C. expectation of higher fatalism in Islamic societies that are less receptive of Western influence. However, this finding is buttressed somewhat by the fact that levels of empirical fatalism are higher in Turkey than in Indonesia.
Iran is a specific case requiring a cursory discussion. While this modern Islamic theocracy seems to symbolize the archetypal anti-Western Islamic regime, Iran may, in fact, be more like Turkey than a theocratic Wahhabist state such as Saudi Arabia. Iran's historical trajectory includes 20th century attempts at Westernization mirroring that of Turkey. Under the leadership of Reza Shah, Iran experienced an expansive project of modernization that spanned several decades starting in the early 1920s. Iran's transformation under Reza Shah was sweeping in scope and included infrastructural improvements, economic reforms and enhancements in public health and education. Furthermore, much of Iran's development under Reza Shah took shape with the cooperation of the United States and other European powers (Ansari 2003). It is this complex history that may help explain lower levels of empirical fatalism in Iran when compared to Turkey, and it is exactly this type of historical nuance that C.O.C. theorists often overlook.
Country dummies for theological fatalism are also shown in Table 6. Findings for theological fatalism are more clear-cut than is the case for empirical fatalism and suggest little support for the C.O.C. theorists' proposition that Westernization and an associated increase in religious secularism will be related to lower levels of theological fatalism. With Indonesia as the sole exception, Turkish respondents place more control in the hands of cosmological forces than either Iranian, Saudi or Lebanese respondents, while perceiving the most cosmological control over life's outcome. Here we find little evidence for C.O.C. Hypothesis 2b that predicts an inverse relationship between Westernization and theological fatalism.
A reasonable criticism that could be raised at this point has to do with the limited selection of cases used for the analysis. Instead of the five nations analyzed here, why not explore the dimensions of fatalism from a broader cross-cultural perspective? I would respond by arguing that the primary analytical focus of this paper requires the operationalization of fatalism around both its empirical and theological dimensions--a strategy that the GPIW survey data allows for. However, it might be useful at this point to isolate the question of empirical fatalism by looking at data that provides the necessary number of cases to carry out a more ambitious cross-cultural analysis. The most recent publicly available wave of the World Values Survey (WVS 2004) includes increased participation from Islamic countries and could have been considered as a source for a more extensive cross national analysis. Unfortunately, the WVS data set contains only one dependable measure of empirical fatalism but does not offer a separate item that captures the idea of theological fatalism.
Empirical Fatalism from a Cross-Cultural Perspective
The World Values Survey contains one clearly worded empirical fatalism item that asks the following:
"Some people feel they have complete free choice and control over their lives, while other people feel what they do has no real effect on what happens to them. Please use this scale where 1 means "none at all" and 10 means "a great deal" to indicate how much freedom of choice and control you feel you have over the way your life turns out."
(WVS 2004: codebook mnemonic A173)
This questionnaire item does not contain an intrinsic religious component to it (e.g., "Everything that happens is part of God's plan") and is restricted to feelings of freedom and control over life's outcomes, thus it represents a well-defined measure of empirical fatalism. Respondents answered on a 10-point scale. The number 0 corresponds to the response "a great deal" and the number 9 corresponds to the response "None at all." The numbers 1-8 did not contain written selections. The scale range is 0 through 9 with a higher response representing a greater level of perceived fatalism. For the wave of WVS data used here, the fatalism measure was asked in 80 nations.
The overall descriptive statistics indicate generally low levels of fatalism around the world (mean = 3.32, median = 3, S.D. = 2.49, N = 110,832 cases for all 80 countries). In terms of individual country variance, several main points stand out. First, it is evident that empirical fatalism scores in both the WVS and GPIW surveys are consistent. As with GPIW data, WVS results show high levels of empirical fatalism in Turkey (mean = 4.52, median = 5, S.D. = 3.32, N = 3,397), low levels of fatalism in Indonesia (mean = 2.75, median = 3, S.D. = 2.02, N = 945), and middle range fatalism scores for Iran (mean = 3.38, median = 3, S.D. = 2.30, N = 2,395) and Saudi Arabia (mean = 3.39, median = 3, S.D. = 2.22, N = 1,473) - a fact that indicates a level of reliability to the two distinct fatalism measures found in two separate surveys. Unfortunately, Lebanon was not included in the cases that included the fatalism measure.
Secondly, when aggregated based on national religious tradition, higher mean fatalism scores exhibited by Muslim respondents may, at first glance, lend support to the notion that Islamic publics exhibit higher levels of fatalism and powerlessness than respondents from other world regions. The strategy here is to attempt further empirical verification of the key arguments, with particular emphasis on empirical fatalism. The first step in the process will be to aggregate the WVS cases into two distinct categories. The first category will compare Muslim vs. non-Muslim nations (0 = "Muslim," 1 = "non-Muslim" as the omitted category), while a second category will highlight the dominant religious tradition of each country and create a series of dummy variables based on dominant national religion (0 = "Muslim," 1 = "Eastern European Muslim," 2 = "Orthodox," 3 = "Catholic," 4 = "Protestant" as the omitted category). See Appendix A for a list of the specific cases used.
As shown in Table 7, higher mean fatalism scores in Muslim nations, at first glance, lend support to the notion that Islamic publics exhibit higher levels of fatalism and powerlessness than respondents from other world regions. However, it is also important to note the higher standard deviation amongst those same Islamic publics, implying a greater level of variance in the Muslim world. For example, where Pakistan (mean = 5.32), Turkey (mean = 4.52) and Egypt (mean = 4.52) score quite high on fatalism, fatalism scores in Indonesia (mean = 2.75) and Jordan (mean = 2.83) are comparable with those of many industrialized, Western democracies such as Norway (mean = 2.82), Denmark (mean = 2.66) and Great Britain (mean = 2.85). Such aggregate level descriptive data implies a wide range of public opinion in Islamic countries and again, contradicts an overly simplistic reading of Islamic public opinion as monolithic and static, as the C.O.C perspective would suggest. And while a higher representation of Islamic countries is found above the mean score for all countries, it is important to note that Hungary (mean = 3.8), Poland (mean = 3.86), and France (mean = 3.53) are also represented amongst countries scoring high on the WVS fatalism measure while, as noted above, Jordan and Indonesia score low on fatalistic attitudes.
Another avenue to explore the question of empirical fatalism from a cross-national perspective is to move away from descriptive data and towards individual level statistical models that control for relevant attributes. Table 8 shows results of the ordered logistic regression of fatalism scores around the world. Models 1 and 2 examine the variation between Muslim and non-Muslim countries, while models 3 and 4 examine national religion effects. Models 5 and 6 include interaction effects for religiosity and denomination. Models 1 and 3 do not include control variables and establish a baseline of variation between countries. The details of coding for all variables and additional information relevant to this supplementary analysis are provided in Appendix B.
Some discussion of these findings is in order and sheds some light on the question of empirical fatalism from a cross-national perspective. At first glance, models 1 and 2 clearly indicate that respondents in Muslim countries are more empirically fatalistic than respondents in non-Muslim countries and, as shown by model 2, these findings are consistent when controlling for important individual-level attributes. However, models 3 and 4 introduce additional religious denomination dummy variables and may tell a more nuanced story. In terms of country-level denomination, the size and direction of the coefficients suggest that while Muslims are more fatalistic than any other group, it is evident that there is substantial variation in fatalism scores within Christian denominations as well. Greek Orthodox and Protestant countries, for example, seem to exhibit higher levels of empirical fatalism than Catholic countries. In fact, the data indicate that, contrary to what one might expect, Catholic countries exhibit the lowest levels of overall fatalism. However, it should be noted, that this finding may be the consequence of a measurable "Latin America" effect on overall fatalism scores. Of the 10 least fatalistic countries in the data, five are Latin American, Catholic countries (El Salvador, Colombia, Mexico, Venezuela and Puerto Rico). The data also include large Catholic countries such as Brazil and Argentina and smaller countries such as the Dominican Republic, which are below average on fatalism.
In keeping with this line of reasoning, it is my sense that the high standard deviation among Muslim countries implies that there is reason to suspect that outlying cases may also affect high fatalism scores in the Muslim cases represented here. As noted above, fatalism is high in several Islamic nations. In fact, within the top five most fatalistic countries in the data--Pakistan, Egypt and Turkey--are from predominantly Islamic societies. As a result, an additional statistical model was examined with these three cases removed, leaving only Iraq, Morocco, Saudi Arabia, Iran, Algeria, Jordan and Indonesia in the model. (8) While fatalism in Muslim countries is still higher than in Protestant countries, with the outlier Muslim cases removed, fatalism is actually higher in both Eastern European Muslim countries and in Greek Orthodox countries than in Middle Eastern Islamic nations. In the previous model with the three outlying observations included, the size of the coefficient indicated that the probability of a higher fatalism score was highest in Middle Eastern Muslim countries where with the outliers removed it is highest in Eastern European Muslim countries followed by Orthodox countries. This finding reinforces what other studies have suggested which is a consistent high level of fatalism in Eastern Europe regardless of national religion (e.g., Goodwin and Allen 2000).
While the main focus of the current analysis is on denominational differences, a cursory discussion of religiosity effects is in order and in keeping with literature that examines the association between levels of religiosity and fatalistic attitudes. Results for interaction effects are shown in Table 8, models 5 and 6. The evidence here suggests that more religious Muslims are also more empirically fatalistic. And while religiosity is consistently used as a control variable throughout the analysis, what we can gauge from the interaction effect is that religious intensity does seem to affect fatalistic orientations among Muslims to a greater extent than other faiths; a finding that speaks to competing theories that seek to explain the effects of religiosity on measures of fatalism. (9) In fairness to C.O.C. theorists, such a finding does seem to resonate with their claims that Islam fosters a fatalistic orientation. On the other hand, the consistent significance of variables including income and education in all models suggests that religious orientation alone does not fully account for levels of fatalism, but that other factors also affect feelings of control over one's life regardless of denomination and religiosity.
The central concern of this article is to demystify notions of a "pure" Islamic fatalism that is distinct from other notions of fatalism and higher and more persistent in the Muslim world. I propose a more nuanced interpretation that takes into account the theological dimensions of fatalism.
With WVS data it is possible to look at country-specific denominational variation and assess the impact of this religious heterogeneity on empirical fatalism. Analysis of individual country variation not shown here suggests further problems for C.O.C. theory. When analyzing within country variation among 10 religiously heterogeneous nations with large Muslim populations, support for a simplistic C.O.C. interpretation once again falters. (10) India represents the only case in which Muslims are more fatalistic than the other religious group(s) in that country. In India, Muslims seem to be more fatalistic than Hindus (coefficient = .167, Pr > [chi square] = .043) which is understandable given India's history. All other countries show no statistically significant effects of being Muslim when compared to other religious groups in that particular country. It should also be noted that effects do exist, and in some cases, fatalism is significantly affected by religious denomination, but other than in India, not consistently by an Islamic effect. In South Africa for example, membership in the Independent African Church seems to be associated which a much higher probability of being more fatalistic than all other South African religious groups (coefficient = .431 ; Pr > [chi square] = .0001).
The results of WVS data confirm findings from GPIW data that Islam per se should not be regarded as a cause of fatalistic orientations. While empirical fatalism measures do seem to be higher in the Muslim world, specific country level variation between Muslim countries exists. Also, fatalism is not monolithic in the Christian world either with some Christian faiths exhibiting higher levels of fatalism than others; a fact the C.O.C. view of Islamic fatalism does not address. Finally, within country differences in nations with substantial Muslim populations do not indicate a direct causal link between Islamic beliefs and empirical fatalism. Future research could clearly assess the impact of Islam on theological fatalism from a cross-cultural perspective; a strategy that was not possible with the currently available WVS data.
Taken in their totality then, these findings once again suggest common misunderstandings and the reliance on impressionistic claims made by C.O.C. theorists. Western-style modernization and secularization do not seem to have the explicit effect of mitigating the fatalistic mindset that C.O.C. theorists report to exist in the Islamic world. In fact, it could be argued by looking at the data presented here, that Western influence has mixed effects on fatalism depending on the specific historical experiences shared by Muslims in different Islamic countries. Furthermore, without adequately conceptualizing fatalism in the first place, the C.O.C. agenda falls short of providing a theoretically informed explanatory framework. While in Saudi Arabia high levels of empirical fatalism seem to confirm the C.O.C. expectation, theological fatalism in Saudi Arabia is significantly lower than in Turkey--a more secularized Islamic society. And in Turkey, where levels of empirical fatalism are moderate, theological fatalism is relatively high.
This is not to say, however, that Westernization, and religious denomination do not have any independent effect on levels of fatalism in predominantly Islamic societies. To think otherwise would be analogous to the type of shallow theorizing that I am arguing C.O.C. theorists are guilty of. Instead, fatalism in the Muslim world is best understood in light of complex historical, cultural, economic and socio-political processes and not as a direct outcome of Western influence, and/or religious denomination alone. Muslim public opinion is filtered through the lens of not only individual level differences such as socio-economic status, age and gender but also through the prism of national culture and the diverse experiences of Christians and Muslims that is so influential in both Muslim and non-Muslim societies. (11)
Islamic Theology, Qismah and Theological Fatalism
This study proposes an alternative possibility to a misleading interpretation of Islam as an inherently fatalistic religious system. What is mistaken for "Islamic fatalism" may be best interpreted as a greater acceptance for central authority and a relinquishment of life's outcomes to an omnipotent deity. The primary error in interpreting Islam as a fatalistic religion is in not properly addressing the cosmologically oriented dimensions of personal efficacy and the reliance that individuals may place on metaphysical powers to determine worldly outcomes.
Where earlier scholars interpreted Islamic religious texts as having "contributed much to the development of the fatalistic attitude in Islam,"(Ringgren 1967: 60)contemporary scholarship interprets the Qur'anic understanding of fate, or Oismah, as "a matter of ongoing and continuous interaction between human will and God's will." (Esposito 2003:254) This stands in stark contrast to an erroneous understanding of Islamic theology as a cultural force that eliminates notions of free will, proposing instead a highly rationalized interaction between human action and cosmological determinism. As Cornell writes, "[f]or the Muslim, belief in God's determination of affairs is not fatalism but common sense. A believer feels liberated in knowing his or her limits, because the acceptance of what can never be changed removes the worry and frustration of striving in vain and opens the door to constructive engagement with the possible." (1999:71)
What we should expect however is a more pronounced expectation on the part of Muslims that God's will interacts with and inevitably determines human action. In comparison to Islam--a faith that retains a strong sense of the collective (ummah)--one of Protestantism's influences on modern Christianity, particularly in the West, has been the emphasis on the authority of personal initiative and individualism (MacCulloch 2004). If this is correct, we should expect Christians to relegate less control over life's affairs to cosmological forces than Muslims, an expectation that is at least partly supported by the findings presented here. However, this in no way implies that Muslims are the empirical fatalists caricatured by proponents of the clash of civilizations thesis.
In closing, four over-arching miscues can be identified in the C.O.C. analysis of Islam and its "fatalistic tendencies." The first is an erroneous interpretation of both Islam and fatalism that stems from a shallow understanding of Islamic theology and its principle tenets. Secondly, while the C.O.C. literature places an emphasis on the beliefs of Muslims, little empirical attention is paid to the Islamic public opinion which is at the heart of the C.O.C. critique (see, Inglehart and Norris 2003). Third, what C.O.C. theory offers is an often a historical analysis that does not account for the global diversity of Muslim and Christian experiences in the modern word. And finally, I have argued here that there is a need to re-conceptualize fatalism as a multidimensional cognitive orientation that includes both feelings of perceived personal control as well as culturally influenced orientations that look to cosmological forces as sources of mastery over life's outcomes. Maybe then can we begin to fully address the important dynamics of Islamic beliefs and the role that this complex belief system plays in shaping the mindset of members of one of the world's major religions.
Appendix A. Description of World Values Countries and "National Religion" Categories Muslim Greek Muslim (Eastern Europe) Orthodox Catholic Algeria Albania Armenia Argentina Bangladesh Azerbaijan Belarus Austria Egypt Kyrgyzstan Bulgaria Belgium Indonesia Georgia Brazil Iran Greece Chile Iraq Moldova Colombia Jordan Romania Croatia Dominican Morocco Republic Pakistan El Salvador Saudi Arabia France Turkey Ireland Italy Lithuania Luxembourg Malta Mexico Peru Philippines Poland Portugal Puerto Rico Slovakia Slovenia Spain Uruguay Venezuela "Non-Muslim" Muslim Protestant Only Algeria Denmark Australia Bangladesh Finland Bosnia-Herz Egypt Great Britain Canada Indonesia Iceland Czech Rep Iran Norway Estonia Iraq Sweden Germany Jordan USA Hungary Morocco India Pakistan Japan Saudi Arabia Latvia Turkey Macedonia Netherlands New Zealand N. Ireland Korea Russian Fed. Serbia Mont. Singapore South Africa Switzerland Taiwan Tanzania Uganda Ukraine Vietnam Zimbabwe Note: The table above shows the 80 nations used to create aggregate categories and corresponding "national religion" and "Muslim/non-Muslim" classifications. For the "Muslim/non-Muslim" categories, please note the following: Eastern European Muslim countries are excluded from the "Muslim only" category. Religiously pluralistic nations in the "non-Muslim" only column are not included in any of the national religion categories but along with Greek Orthodox, Catholic and Protestant, comprise the "non-Muslim" category.
Appendix B. WVS Variable Codings and Elaboration of Sample Size Across Models
The discussion throughout the main body of the paper offered an analytically framed rationale for the use of specific control variables. For analysis of WVS data, the same rationale is in order with account taken of the consistent availability of control variables for such a large number of cases. For the sake of brevity and in keeping with editorial constraints, the coding for control variables for WVS fatalism models will be presented here.
Country dummies were created and coded as follows: For Muslim vs. non-Muslim: 0 = "Muslim," 1 = "non-Muslim" as the reference category. For National Religion: 0 = "Muslim," 1 = "Eastern European Muslim," 2 = "Greek Orthodox," 3 = "Catholic," 4 = "Protestant," as the reference category.
Church attendance is a variable with eight ordered categories, 7 = "More than once a week," 6 = "Once a week," 5 = "Once a month," 4 = "Only on special holy days/Christmas/Easter days," 3 = "Other specific holy days," 2 = "Once a year," 1 = "Less often," 0 = "Never practically never."
Subjective religiosity is measured by an item that asks, "Independently of whether you go to church or not, would you say you are ..." Four categories are provided in the original survey, 1 = "A religious person" 2 = "Not a religious person," 3 = "A convinced atheist," 4 = "Other answer." Responses were collapsed into a binary variable with "not a religious person" and "a convinced atheist" serving as the low religiosity category (0 = "low religiosity") and "a religious person (1 = "religious person") as the high religiosity, reference category.
Education level is a three level index recoded (Highest educational level attained, mnemonic X025) on an individual country basis by original researchers. (0 = "low," 1 = "mid," 2 = "high" as the reference category.
Income (mnemonic X047R) asked: "Here is a scale of incomes. We would like to know in what group your household is, counting all wages, salaries, pensions and other incomes that come in. Just give the letter of the group your household falls into, before taxes and other deductions. Recoded by original researchers on an individual country basis 0 = "Lower," 1 = "Middle," 2 = "Upper" as the reference category.
Gender: 0 = female, 1 = male as the reference category. Marital status: 0 = "married," 1 = "divorced/separated," 2 = "other," 3 = "single" as the reference category. Number of children: nine ordered categories, 0 = "No child" to 8 = "8 or more children."
It may be useful to offer a brief note on the large reduction in N across models: Models 2 and 4 introduce control variables that may not have been answered by some respondents and, as is typical in survey research, the N is lower. The larger issue, however, has to do with the coding of collapsed variables for "national religion" and "Muslim/non-Muslim" categories. Models 3 and 4 include the "national religion" variable while 1 and 2 use the "Muslim/non-Muslim" variable. Countries coded as "non-Muslim" included all countries that were not Muslim regardless of religious composition. For the "national religion" variable, on the other hand, only countries that were predominantly made up of one dominant religious category were included. For instance, Colombia is both a non-Muslim AND a Catholic country where a country like the Czech Republic is included as non-Muslim but NOT in the national religion category because the Czech Republic is partly Catholic but largely "unaffiliated." In short, the "Muslim/ non-Muslim" variable includes more cases where the national religion variable only included countries where a definite majority belongs to one particular faith.
I would like to thank Jeffrey Alexander, Ivan Szelenyi, Philip Smith and Hannah Brueckner for helpful comments on early drafts of this paper. I am indebted to three anonymous reviewers at Social Forces for invaluable comments and suggestions. Thanks also to Raquel Marquez, Derral Cheatwood and John P. Bartkowski for the institutional support and practical guidance throughout the revision process. Finally, I owe a special word of gratitude to Christian Smith for the observant reading and comprehensive suggestions that strengthened the overall argument. Use of the privately held Gallup data set was made possible by the cooperation of the Yale University Social Science Libraries & Information Services. An earlier version of this article was presented at the American Sociological Association meeting in Philadelphia, 2005. Direct correspondence to Gabriel A. Acevedo, Assistant Professor of Sociology, University of Texas at San Antonio, Department of Sociology, One UTSA Circle, San Antonio, Texas 78249-0655. E-mail: firstname.lastname@example.org.
Acevedo, Gabriel. 2005. "Turning Anomie on Its Head: Fatalism as Durkheim's Concealed and Multidimensional Alienation Theory." Sociological Theory 23(1):75-85.
Adams, Bert N. 1974. "Urban Skills and Religion: Mechanisms for Coping And Defense Among the Ugandan Asians." Social Problems 22(1 ):28-42.
Ahmed, Akbar S. 2003. Islam Under Siege. Living Dangerously in a Post-Honour World. Polity Press.
Al-Ahsan, Abdullah. 2006. "Political Discourse of the Organization of the Islamic Conference." Pp. 527-43. The Blackwell Companion to Contemporary Islamic Thought. Ibrahim Abu Rabi, editor. Blackwell Publishing.
Ansari, Ali M. 2003. A History of Modern Iran Since 1921: The Pahlavis and After. Longman.
Bandura, Albert. 1995. "Excercise of personal and collective efficacy in changing societies." Pp. 1-45. Self-efficacy in Changing Societies. Albert Bandura. Cambridge University Press.
Banfield, Edward C. 1967. The Moral Basis of a Backward Society. Free Press.
Barro, Robert J., and Rachel M. McCleary. 2003. "Religion and Economic Growth Across Countries." American Sociological Review 68(5):760-81.
Bearman, Peter S. 1991. "The Social Structure of Suicide." Sociological Forum 6(3):501-24.
Belo, Catarina. 2006. "Ibn Rushd on God's Decree and Determination (Al-qada Wa-l-qadar)." Al-Qantara 27(2):245-64.
Bertrand, Jacques. 2004. Nationalism and Ethnic Conflict in Indonesia. Cambridge University Press.
Bilgrami, Akeel. 2003. "The Clash Within Civilizations." Daedalus 132(3):88-93.
Bunzl, John, and Edward Said, editors. 2004. Islam, Judaism, and the Political Role of Religions in the Middle East. University Press of Florida.
Catholic, Encyclopedia. 1907-1912. Entry for Fatalism. Catholic Online Catholic Encyclopedia Digital version, Robert Appleton Company. Accessed Dec. 25, 2007 at: http://www.catholic.org/encyclopedia/.
Cohen, Dov, and Richard E. Nisbett. 1998. "Are There Differences in Fatalism Between Rural Southerners and Midwesterners?" Journal of Applied Social Psychology 28(23): 2181-95.
Cornell, Vincent J. 1999. "Fruit of the Tree of Knowledge: The Relationship Between Faith and Practice in Islam." Pp. 63-105. The Oxford History of Islam. John L. Esposito, editor. Oxford University Press.
Dean, David Commins. 2006. The Wahhabi Mission and Saudi Arabia. I.B. Tauris.
Delong-Bas, Natana J. 2004. Wahhabi Islam: From Revival and Reform to Global Jihad. Oxford University Press.
Dohrenwend, Bruce P 1959. "Egoism, Altruism, Anomie, and Fatalism: A Conceptual Analysis of Durkheim's Types." American Sociological Review 24(4):466-73.
Douglas, Jack D. 1967. The Social Meanings of Suicide. Princeton University Press.
Du Pasquier, Roger. 1992. Unveiling Islam. Islamic Texts Society.
Durkheim, Emile. 1968. Suicide, a Study in Sociology. Free Press.
--. 1968. Suicide, a Study in Sociology. Free Press.
Elder, Joseph W. 1966. "Fatalism in India: a Comparison Between Hindus and Muslims." Anthropological Quarterly 39(3):227-43.
Esposito, John L. 1997. Islam. The Straight Path. 3rd edition. Oxford University Press.
--. 2002. What Everyone Needs to Know About Islam. Oxford University Press.
--. Editor. 2003. The Oxford Dictionary of Islam. Oxford University Press.
Francis, Leslie J., Stephen H. Louden and Christopher J.F. Rutledge. 2004. "Burnout Among Roman Catholic Parochial Clergy in England and Wales: Myth or Reality?" Review of Religious Research 46(1):5-19.
Fukuyama, Francis. 1992. The End of History and the Last Man. Free Press
Geertz, Clifford. 1968. Islam Observed, Religious Development in Morocco and Indonesia. Yale University Press.
Goodwin, Robin, and Peter Allen. 2000. "Democracy and Fatalism in the Former Soviet U n ion." Journal of Applied Social Psychology 30(12):2558-74.
GPIW. 2003. "2002 Gallup Poll of the Islamic World." A CD-Rom produced by The Gallup Organization. New Haven, CT: Yale University Library and Electronic Services, distributor.
Grenstad, Gunnar. 1990. "Europe by Culture: An Exploration of Grid/Group Analysis." Unpublished doctoral thesis, Department of Comparative Politics, University of Bergen, Norway.
Haddad, Simon, and Hilal Khashan. 2003. "Lebanese Muslim Views of September 11." Journal for the Scientific Study of Religion 42(2):291-98.
Harrison, Lawrence E. 1985. Underdevelopment is a State of Mind: the Latin American Case. Lanham, MD: Center for International Affairs Harvard University and University Press of America.
--. 2000. "Promoting Progressive Social Change." Pp. 296-307. Culture Matters. How Values Shape Human Progress. Lawrence E. Harrison and Samuel P. Huntington. Basic Books.
Hassan, Riaz. 2002. Faithlines. Muslim Conceptions of Islam and Society. Oxford University Press.
Hassner, Pierre. 1996/1997. "Morally Objectionable, Politically Dangerous." The National Interest 46(Winter):63-69.
Hoffman, Valerie J. 1993. "Muslim Fundamentalists: Psychological Profiles." Pp. 199-230. Fundamentalisms and Society: Reclaiming the Sciences, the Family, and Education. Martin E. Marty and R. Scott Appleby, editors. University of Chicago Press.
HRW. 2001. Human Rights Watch World Report 2001. New York, NY: Human Rights Watch Homepage. Accessed Dec. 30, 2007 at: http://www.hrw.org/ wr2k1/asia/indonesia.html.
Huntington, Samuel P. 1993. "The Clash of Civilizations." Foreign Affairs 72(3):22-50.
--. 1996. The Clash of Civilizations and the Remaking of World Order. Touchstone.
Huntington, Samuel P., Fouad Ajami, Kishore Mahbubani, Robert L. Bartley, Pinyen Liu, Jean J. Kirkpatrick, Albert Loren Weeks, Gerard Piel and Council on Foreign Relations. 1996. The Clash of Civilizations?: The Debate. New York, NY: Foreign Affairs.
Ibrahim, S.E. 1980. "Anatomy of Egypt's Militant Islamic Groups: Methodological Note and Preliminary Findings." International Journal of Middle East Studies 12(4):423-53.
Imani, Amil. 2006. "A Look at the Muslims' Mindset." Washington, DC: Danielpipes.org. Accessed July 29, 2007 at: http://www.danielpipes.org/ comments/53446.
Inglehart, Ronald, and Pippa Norris. 2003. "The True Clash of Civilizations." Foreign Policy 135(March/April): 63-70.
Jacobson, Cardell K. 1999. "Denominational and Racial and Ethnic Differences in Fatalism." Review of Religious Research 41(1):9-19.
Kader, Zerougui Abdel. 1998. "The Clash of Civilizations and the Remaking of the World Order: Book Review." Arab Studies Quarterly 20(1):89-92.
Kohn, Melvin L. 1989. "Social Structure and Personality: A Quintessentially Sociological Approach to Social Psychology." Social Forces 68(1):26-33.
Kuran, Timur. 2004a. "The Economic Ascent of the Middle East's Religious Minorities: The Role of Islamic Legal Pluralism." Journal of Legal Studies 33(2):475-515.
--. 2004b. "Why the Middle East is Economically Underdeveloped: Historical Mechanisms of Institutional Stagnation." The Journal of Economic Perspectives 18(3):71-90.
Lane, Jan-Erik, and Hamadi Redissi. 2004. Islam. Faith and Rationality. Ashgate Publishing.
Lester, David. 1994a. Emile Durkheim: le Suicide One Hundred Years Later. Charles Press.
--. 1994b. Suicide From a Sociological Perspective. Charles C. Thomas.
Lewis, Bernard. 1990. "The Roots of Muslim Rage." The Atlantic Monthly 266(3):47-60.
--. 1997. "The West and the Middle East." Foreign Affairs 76(1):114-30.
--. 2003. The Crisis of Islam: Holy War and Unholy Terror. Modern Library.
--. 2004. From Babel to Dragomans: Interpreting the Middle East. Oxford University Press.
Lockwood, David. 1992. Solidarity and Schism. the Problem of Disorder in Durkheimian and Marxist Sociology. Oxford University Press.
MacCulloch, Diarmaid. 2004. The Reformation, A History. Viking.
Mahbubani, Kishore. 1993. "The Dangers of Decadence: What the Rest can Teach the West." Foreign Affairs 72(4):10-15.
Manguelle, Daniel-Etounga. 2000. "Does Africa Need a Cultural Adjustment Program?" Pp. 65-77. Culture Matters. How Values Shape Human Progress. Lawrence P. Harrison and Samuel R Huntington, editors. Basic Books.
Manji, Irshad. 2004. The Trouble With Islam: a Muslim's Call for Reform in her Faith. St. Martin's Press.
Marcovitch, Harvey. 2003. "Farming and 'Fatalism' May be Related." British Medical Journal 326(7397):1005-11.
Mirowsky, John, and Catherine E. Ross. 1998. "Education, Personal Control, Lifestyle and Health." Research on Aging 20(4):415-49.
Nasr, Seyyed Hossein. 2001. Ideals and Realities of Islam (Revised U.S. Edition). Chicago, II: ABC International Group.
Padfield, Harland. 1980. "The Expendable Rural Community and Denial of Powerlessness." Pp. 159-85. The Dying Community. Art Gallaher and Harland Padfield, editors. University of New Mexico Press.
Peters, F.E. 1993. A Reader on Classical Islam. Princeton University Press.
Pickering, W.S.F., and Geoffrey Walford. 2000. Durkheim's Suicide.. a Century of Research and Debate. Routledge.
Pipes, Daniel. 1983. In the Path of God: Islam and Political Power. Basic Books.
Qureshi, Emran, and Michael Anthony Sells. 2003. The New Crusades. Constructing the Muslim Enemy. Columbia University Press.
Rahman, Fazlur. 1979. Islam. University of Chicago Press.
--. 1989. Major Themes of the Quran. Bibliotheca Islamica.
Ramos, Blanca M. 2004. "Culture, Ethnicity, and Caregiver Stress Among Puerto Ricans." Journal of Applied Gerontology 23(4):469-86.
Redfield, Robert. 1956. Peasant Society and Culture, an Anthropological Approach to Civilization. University of Chicago Press.
Ricklefs, M.C. 2001. A History of Modern Indonesia Since c. 1200. 3rd edition. Stanford University Press.
Ringgren, Helmer. 1967. "Islamic Fatalism." Pp. 52-62. Fatalistic Beliefs in Religion, Folklore, and Literature. Papers Presented at the Symposium on Fatalistic Beliefs, Held at Abo Sept. 7-9, 1964. Helmer Ringgren, editor. Almqvist & Wiksell.
Said, Edward. 2001a. "A Clash of Ignorance." The Nation 273(12):11-13.
--. 2001b. "We All Swim Together." New Statesman 14(678):20-23.
Schieman, Scott. 2001. "Age, Education, and the Sense of Control--A Test of the Cumulative Advantage Hypothesis." Research on Aging 2(2):153-78.
Shanin, Teodor. 1987. Peasants and Peasant Societies: Selected Readings. Blackwell.
Sheffield, John. 1996. "Warning: Your Theology May be Fatal to Your Kids." U.S. Catholic 61(7):33-37.
Smith, Christian. 1994. "The Spirit and Democracy: Base Communities, Protestantism, and Democratization in Latin America." Sociology of Religion 55(2):119-43.
Sosa, Lionel. 1998. The Americano Dream: How Latinos Can Achieve Success in Business and Life. Penguin.
Spencer, Robert. 2002. Islam Unveiled. Disturbing Questions About the World's Fastest Growing Faith. Encounter Books.
Stark, Rodney. 2002. "Physiology and Faith: Addressing the 'Universal' Gender Difference in Religious Commitment." Journal for the Scientific Study of Religion 41(3):495-507.
Stark, Rodney, and Roger Finke. 2000. Acts of Faith: Explaining The Human Side of Religion. University of California Press.
Taylor, Steve. 1982. Durkheim and the Study of Suicide. St. Martin's Press.
Timmerman, Kenneth R. 2003. Preachers of Hate. Islam and the War on America. Crown Forum.
Tumin, Melvin Marvin. 1952. Caste in a Peasant Society: A Case Study in the Dynamics of Caste. Princeton University Press.
Warraq, Ibn. 2003. Leaving Islam. Apostates Speak Out. Prometheus Books.
Weber, Max. 1998. "The Protestant Ethic and the Spirit of Capitalism." Roxbury Pub. Co.
--.  1991. The Sociology of Religion. Beacon Press.
--. 1996. The Religion of India, the Sociology of Hinduism and Buddhism. Munshiram Manoharlal.
Wolf, Eric R. 1966. Peasants. Prentice-Hall.
Woodward, Mark R. 1989. Islam in Java: Normative Piety and Mysticism in the Sultanate of Yogyakarta. University of Arizona Press.
WVS. 2004. "World Values Survey and European Values Survey, 1999-2001, ICPSR 3975." University of Michigan Institute for Social Research [producer]. Accessed Aug. 5, 2005 at: http://www.icpsr.umich.edu/ICPSR/.
(1.) Although Weber never completed his intended writings on Islamic religion, he does make several explicit references to Islamic fatalism and contrasts it to the highly rationalized conception of predestination stemming from Calvinist theology. In a footnote to the Protestant Ethic, Weber notes that "fatalism is, of course, the only logical consequence of predestination."(Weber 1998:232) The empirical reference point is clearly Islam when he states that, "It]he Islamic belief in predestination easily assumed fatalistic characteristics in the beliefs of the masses." (Weber  1991:205) However, for Weber, Calvinism takes a rationalistic and empiricist turn away from fatalism and looks to worldly manifestations for verification of God's omnipotence. It is the Calvinist belief that it requires "evidence" of salvation --not works leading to salvation--that removes the fatalistic tendencies from the Calvinist worldview. These tendencies are, in Weber's view, retained in Islam and in the Hindu karma concept. As Weber goes on to say, "on account of the idea of proof the psychological result [of Calvinism] was precisely the opposite [of fatalism.]" (Weber 1998:232)What Weber missed is the possibility of a similar process of rationalization occurring in Islam. In fact, a central unifying theme of Islamic theology states that logical reasoning and the quest for what the Qur'an calls "'clear evidences' (bayyinat) of God's presence in the world" (Cornell 1999:64) are required of all Muslims.
(2.) It is important to note that certain stands of Catholicism (particularly in Latin America) are often associated with a more pervasive and intense sense of fatalism than is found in Protestantism. As Christian Smith has noted, "The spiritual introspection, methodical self-discipline, application of faith to everyday experience, means-end mentality, and personal responsibility involved in the conservative Protestant sanctification experience all engender an ethos of rational individualism. In these and other ways, Latin American Protestantism carries and inculcates a complex of beliefs and practices that are much more compatible with and supportive of democracy than the collectivism, fatalism, mysticism, and traditionalism found in much of Catholic, monistic-corporatist Latin America."(1994:133)See also, Francis, Louden and Rutledge 2004; Ramos 2004; Sheffield 1996.
(3.) Wahhabism is a complex cultural, religious phenomenon. Full discussion is beyond the scope of the present analysis. As with many strands of religious doctrine, Wahhabism has experienced a series of complex historical developments and in more recent times, Saudi Wahhabism has been influenced by geopolitical conflict, regional military hostilities and economic strain. For a sympathetic account of Wahhabism and its theological origins please see, Delong-Bas 2004.
(4.) Simple cross tabulations and chi-square tests confirm statistically significant differences in fatalism scores as follows: Indonesian Christians show higher levels of empirical fatalism (Pr > [chi square] .006) and lower levels of theological fatalism (Pr > [chi square] -.066) than Indonesian Muslims. In Lebanon, Christians are no more or less empirically fatalistic than Muslims (Pr > [chi square] .780), but are less theologically fatalistic than Lebanese Muslims (Pr > [chi square] = .029). Results are available upon request.
(5.) Models were also run using dummy variables for income categories in place of the logged midpoints (Lebanon and Indonesia). In each case, dummy variables for income did not significantly alter the coefficients shown here. Statistical significance and the direction of coefficients for religion/ denomination in Lebanon and Indonesia remained unchanged indicating that the effects shown are robust regardless of the income controls used. Results are available upon request.
(6.) Human Rights Watch has catalogued the ongoing ethno-religious tensions in Indonesia noting the continuing escalation of armed Christian-Muslim conflicts in many regions of the country. While the exact number of wounded and dead is difficult to ascertain, HRW places their estimates in the thousands (HRW 2001).
(7.) As Table 6 shows, country dummies for empirical fatalism Model 1, are not statistically significant. However, it should be noted that all coefficients just exceed the .05 level of confidence and are in the same direction as the country dummy coefficients in empirical fatalism shown in Model 2, which includes relevant statistical controls. Furthermore, chi-square tests are significant and indicate a noteworthy country-level effect on empirical fatalism.
(8.) Raw output was provided to the editorial staff at Social Forces and is available upon request.
(9.) In an unpublished paper (available upon request), I consider the question of religiosity and empirical fatalism in more detail. Suffice it to say that discussions on this topic are typically centered on two competing hypotheses. The resource compensation hypothesis predicts that religiosity will serve as a cognitive buffer against structural constraints whereby individuals will exhibit lower levels of fatalism than their less religious counterparts. The resource amplification perspective, on the other hand, predicts that religiosity will foster a sense that otherworldly forces control life's outcomes, leading to higher levels of fatalism. The main concern here is to examine what effect, if any, religiosity has on measures of fatalism and to contrast that effect in distinct regions of the world.
(10.) Country-specific ordered regression models with relevant controls were carried out for 10 nations that included substantial Muslim and non-Muslim populations. The countries analyzed were Albania, Bangladesh, Bosnia-Herzegovina, Egypt, India, Macedonia, Nigeria, South Africa, Tanzania and Uganda. Raw output was provided to the editorial staff at Social Forces and is available upon request.
(11.) Future research could also consider aggregate level analysis to include macro-level data. In particular, a control for regime type may shed an important light on the effect of political structure on fatalistic attitudes. Inclusion of such a variable in a multi-level model may speak to the significant variation in fatalism present in the Islamic nations studied here.
Gabriel A. Acevedo, University of Texas at San Antonio
Table 1: Distribution of Survey Respondents by Country Cumulative Cumulative Country Frequency % Frequency % Indonesia 1,050 19.5 1,050 19.5 Lebanon 1,050 19.5 2,100 39 Iran 1,501 27.9 3,601 67 S. Arabia 754 14 4,355 81 Turkey 1,019 18.9 5,374 100 GPIW: 2002 Table 2: Distribution of Lebanese and Indonesian Respondents who are Christians and Muslims Lebanon Indonesia % Christian 43 (N = 447) 5 (N = 49) % Muslim 57 (N = 603) 95 (N = 982) Total N 1,050 1,031 Table 3: Empirical and Theological Fatalism Descriptive Statistics by Country and Religion EF EF Country and Religion Mean Median N Indonesian Christian .839 1 49 Indonesian Muslim .566 0 933 Lebanese Christian 1.36 1 442 Lebanese Muslim 1.34 1 595 TF TF Country and Religion Mean Median N Indonesian Christian 3.24 3 49 Indonesian Muslim 3.30 3 930 Lebanese Christian 2.19 2 434 Lebanese Muslim 2.45 3 572 Table 4: Empirical and Theological Fatalism in Two Islamic Countries: Ordered Logistic Regression Estimates Empirical Indonesia Indonesia Model 1 Model 2 Age (survey item midpoints logged) -.966 *** (.286) Female -.323 *** (.082) Beliefs Views on religion as essential -.386 *** (.108) Religion: Christian .341 * .465 ** (.141) (.147) Location Sampling point classification = urban .091 (.066) Socio-economic Status Level of education low .061 (.079) Avg. monthly income (item midpoints logged) -.862 *** (.197) Occupation Housewife -.053 (.163) Student .124 (.169) Unemployed .357 * (.157) Family Structure Other 1.05 (.574) Single -.790 * (.316) Family Size .109 ** (.036) Pseudo [R.sup.2] .005 .072 Proportional odds assumption ([chi square]) 4.30 60.74 Model df 3 39 p .230 .014 Weighted N 1,003 1,001 Empirical Lebanon Lebanon Model 1 Model 2 Age (survey item midpoints logged) -.047 (.209) Female -.068 (.074) Beliefs Views on religion as essential -.001 (.084) Religion: Christian 0.008 .045 (.056) (.058) Location Sampling point classification = urban -.222 *** (.062) Socio-economic Status Level of education low .031 (.063) Avg. monthly income (item midpoints logged) -.112 (.110) Occupation Housewife .051 (.149) Student .199 (.158) Unemployed .017 (.159) Family Structure Other .290 (.288) Single -.201 (.175) Family Size -.001 (.032) Pseudo [R.sup.2] .000 .020 Proportional odds assumption ([chi square]) 1.74 83.56 Model df 3 39 p .626 <.0001 Weighted N 1,037 1,021 Theological Indonesia Indonesia Model 1 Model 2 Age (survey item midpoints logged) .220 (.273) Female .255 ** (.078) Beliefs Views on religion as essential .405 *** (.105) Religion: Christian -.140 -.225 (.137) (.142) Location Sampling point classification = urban -.044 (.063) Socio-economic Status Level of education low -.134 (.075) Avg. monthly income (item midpoints logged) .678 *** (.189) Occupation Housewife .187 (.158) Student -.332 * (.164) Unemployed -128 (.153) Family Structure Other -.831 (.551) Single .570 (.303) Family Size -.077 * (.035) Pseudo [R.sup.2] .001 .054 Proportional odds assumption ([chi square]) 5.98 107.85 Model df 3 39 p .112 <.0001 Weighted N 1,000 998 Theological Lebanon Lebanon Model 1 Model 2 Age (survey item midpoints logged) -.106 (.211) Female .105 (.075) Beliefs Views on religion as essential .221 ** (.084) Religion: Christian -.148 ** -.165 ** (.056) (.059) Location Sampling point classification = urban .061 (.063) Socio-economic Status Level of education low -.058 (.063) Avg. monthly income (item midpoints logged) -.102 (.112) Occupation Housewife .010 (.151) Student -.191 (.160) Unemployed .180 (.162) Family Structure Other -.149 (.288) Single -.021 (.175) Family Size .037 (.033) Pseudo [R.sup.2] .006 .029 Proportional odds assumption ([chi square]) 3.96 62.18 Model df 3 39 p .265 .010 Weighted N 1,006 989 Notes: Standard errors are in parentheses * p < .05 ** p < .01 *** p < .001 Table 5: Descriptive Statistics for Empirical and Theological Fatalism by Country Country EF Mean EF Median N Turkey 1.94 2 986 Lebanon 1.35 1 1,037 S. Arabia 1.31 1 720 Iran 1.25 1 1,451 Indonesia .555 0 1,046 Country TF Mean TF Median N Indonesia 3.34 4 1,043 S. Arabia 2.46 3 702 Turkey 2.38 3 982 Iran 2.36 3 1,396 Lebanon 2.34 3 1,006 GPIW: 2002 Table 6: Fatalism in Five Islamic Countries: Ordered Logistic Regression Estimates EF Model 1 EF Model 2 Age -.278 ** (.102) Female .011 (.036) Beliefs Views on religion as essential -.082 * (.040) Sampling point classification = urban -.058 * (.028) Socio-economic Status Level of education low .068 * (.030) Avg. monthly income -.078 (.043) Occupation Housewife .165 * (.067) Student .047 (.076) Unemployed .153 * (.066) Country Indonesia -1.13 *** -1.13 *** (.054) (.056) Iran -.061 -.167 ** (.044) (.051) Lebanon .082 .142 * (.050) (.055) Saudi Arabia .097 .231 *** (.058) (.065) Family Structure Other .129 (.139) Single -.084 (.086) Family Size .017 (.013) Pseudo [R.sup.2] .124 .148 Proportional odds assumption ([chi square]) 281.334 356.83 Model df 12 48 p <.0001 <.0001 Weighted N 5,240 4,195 TF Model 1 TF Model 2 Age .202 * (.102) Female .122 *** (.036) Beliefs Views on religion as essential .258 *** (.039) Sampling point classification = urban .009 (.028) Socio-economic Status Level of education low .069 * (.030) Avg. monthly income -.049 (.043) Occupation Housewife .116 (.067) Student -.235** (.075) Unemployed .073 (.066) Country Indonesia .852 *** .900 *** (.052) (.055) Iran -.235 *** -.248 *** (.045) (.051) Lebanon -.226 *** -.248 *** (.050) (.055) Saudi Arabia -.178 ** -.188 ** (.058) (.065) Family Structure Other -.220 (.139) Single .163 (.086) Family Size .031 * (.013) Pseudo [R.sup.2] .057 .086 Proportional odds assumption ([chi square]) 317.32 381.48 Model df 12 48 p <.0001 <.0001 Weighted N 5,129 4,778 Notes: Standard errors are in parentheses * p < .05 ** p <.01 ** p <.001 Table 7: Descriptive Statistics for Empirical Fatalism by National Religion Standard National Religion Mean Median Deviation N Muslim 4.02 4 2.807 19,726 Muslim (E. Europe) 3.90 4 2.771 3,912 Greek Orthodox 3.84 4 2.547 8,857 Catholic 2.86 3 2.288 32,640 Protestant 2.53 2 1.845 7,303 WVS 2004 Table 8: Empirical Fatalism in 80 Nations: Ordered Logistic Regression Estimates Model 1 Model 2 Muslim or non-Muslim Country Muslim country .296 *** .328 *** (.006) (.008) National Religion Muslim Eastern European Muslim Greek Orthodox Catholic Religiosity Church Attendance -.012 *** (.002) High religiosity -.022 ** (.007) Socio-economic Status Low level of education .143 *** (.009) Mid-level education .035 *** (.008) Standardized Household Income Categories Low income .183 *** (.009) Middle income -.013 (.008) Demographic Controls Female .084 *** (.006) Age .004 *** (.000) Family Structure Married .106 *** (.010) Divorced/separated -.005 (.019) Other -.018 (.016) Number of children -.041 *** (.004) Interaction Effects Muslim country X high religiosity National Religion Muslim X high religiosity Eastern European Muslim X high religiosity Greek Orthodox X high religiosity Catholic X high religiosity Pseudo [R.sup.2] .016 .038 Proportional odds assumption 2636.85 4446.78 ([chi square]) Model df 8 104 P <.0001 <.0001 Weighted N 101,435 82,517 Model 3 Model 4 Muslim or non-Muslim Country Muslim country National Religion Muslim .372 *** .410 *** (.012) (.015) Eastern European Muslim .298 *** .359 *** (.023) (.025) Greek Orthodox .274 *** .269 *** (.016) (.018) Catholic -.397 *** -.419 *** (.011) (.012) Religiosity Church Attendance -.019 *** (.003) High religiosity -.030 ** (.009) Socio-economic Status Low level of education .193 *** (.011) Mid-level education -.010 (.010) Standardized Household Income Categories Low income .153 *** (.010) Middle income .007 (.009) Demographic Controls Female .098 *** (.007) Age .004 *** (.000) Family Structure Married .100 *** (.013) Divorced/separated -.009 (.024) Other -.060 ** (.019) Number of children -.036 *** (.005) Interaction Effects Muslim country X high religiosity National Religion Muslim X high religiosity Eastern European Muslim X high religiosity Greek Orthodox X high religiosity Catholic X high religiosity Pseudo [R.sup.2] .045 .071 Proportional odds assumption 3463.12 4503.82 ([chi square]) Model df 32 128 P <.0001 <.0001 Weighted N 72,298 59,849 Model 5 Model 6 Muslim or non-Muslim Country Muslim country .305 *** .372 *** (.011) (.019) National Religion Muslim .372 *** (.019) Eastern European Muslim .341 *** (.030) Greek Orthodox .285 *** (.020) Catholic -.393 *** (.015) Religiosity Church Attendance -.011 *** -.018 *** (.002) (.003) High religiosity 0.003 -0.015 (.011) (.011) Socio-economic Status Low level of education .142 *** .192 *** (.009) (.011) Mid-level education .036 *** -0.009 (.008) (.010) Standardized Household Income Categories Low income .183 *** .153 *** (.009) (.010) Middle income -.013 .007 (.008) (.009) Demographic Controls Female .085 *** .099 *** (.006) (.007) Age .004 *** .004 *** (.000) (.000) Family Structure Married .105 *** .100 *** (.010) (.013) Divorced/separated -.005 -.010 (.019) (.024) Other -.018 -.060 ** (.016) (.019) Number of children -.041 *** -.037 *** (.004) (.005) Interaction Effects Muslim country X high religiosity .033 ** (.011) National Religion Muslim X high religiosity .051 ** -.019 Eastern European Muslim X high .037 religiosity (.030) Greek Orthodox X high religiosity -.024 (.020) Catholic X high religiosity -.042 ** (.014) Pseudo [R.sup.2] .038 .071 Proportional odds assumption 4589.19 4675.61 ([chi square]) Model df 112 160 P <.0001 <.0001 Weighted N 82,454 59,766 WVS 2004 Notes: Standard errors are in parentheses * p < .05 ** p < .01 ** p < .001