The effect of religiosity on endorsement of democratic values: The mediating influence of authoritarianism

THE EFFECT OF RELIGIOSITY ONENDORSEMENT OF DEMOCRATIC VALUES:THE MEDIATING INFLUENCE OFAUTHORITARIANISM

Daphna Canetti-Nisim

This study addresses the as yet unaddressed question of the nature of the effects ofreligiosity and authoritarianism upon endorsement of abstract democratic values.Findings obtained through the analyses of structural equations show that, despite theunsettled political and security atmosphere, Israelis believe in democratic values to aconsiderable extent. Furthermore, findings do not fully support the theoreticalexpectation that religiosity is likely to decrease support for democratic values; namely,in and of itself, religiosity has a negligible impact on endorsement of democraticvalues. Yet again, analyses lend credence to theory on authoritarianism—the negativeeffect of religiosity on endorsement of democratic values results from the mediation ofauthoritarianism. These findings are understood and discussed with regard to theo-retical implications. The major conclusion, then, in contrast to existing assumptionsregarding religiosity and negation of democracy linkage, is that religiosity cannot beconsidered the sole direct root cause of negation of democracy.

Key words: religiosity; authoritarianism; democratic values.

INTRODUCTION

Democratic government depends on more than formal institutions; namely,on democratic political culture with the commitment of citizens to democraticvalues (Almond and Verba, 1963). Democratic values here refer to placing ahigh value on aspects of the ‘‘democratic creed,’’ such as minority rights, the

Daphna Canetti-Nisim, Department of Political Science, University of Haifa, Mount Carmel31905, Israel ([email protected])

Political Behavior, Vol. 26, No. 4, December 2004 (� 2004)

377

0190-9320/04/1200-0377/0 � 2004 Springer Science+Business Media, Inc. 2004

rule of law and freedom of speech.1 Democratic values may be challenged byreligion, whose documented association with negation of democracy is a causefor concern, especially in light of the rising tide of religiously based politicalviolence. This study, using structural equations, hopes to address the questionof the nature in which negation of democratic values is affected by religiousconvictions. Particularly, its main objective is to answer the question ofwhether religiosity is associated with non-democratic values in and of itself orvia authoritarianism.

The basic religious belief is belief in a supernatural power that created,sustains, and directs the world (James, 1897, 1902). Psychologists of religionpoint to inherent conflicts between strong religious conviction and democracy(Hunsberger, 1995; Schwartz and Huismans, 1995). There is substantial evi-dence of the efforts of strong believers to assert hegemony over other religiousgroups and, at times, to resort to violence to promote their agendas andsuppress those of others (Beit-Hallahmi, 2001; Canetti-Nisim, 2003). Somesociologists contend that religion is an obstacle to the processes of educationand political socialization that shape democratic values (e.g., Bellah, 1976;Lipset, 1981; Stark, 2001).

By and large, empirical studies lend support to such claims. Reports ontrends in the United States in the 1970s and 1980s show that, as religiousaffiliation decreased, political liberalism increased (Nelson, 1988; Wald,Owen, and Hill, 1988). Surveys carried out in Belgium (Duriez, Luyten,Snauwaert, and Hutsebaut, 2002) and in Israel (Arian, Nachmias, Navot, andShani, 2003; Peres, 1995) found that religious orthodoxy predicted non-democratic and intolerant attitudes.

Studies on the association between religiosity and tolerance (e.g., Sullivan,Pierson, and Marcus, 1982; Sullivan, Shamir, Walsh, and Roberts, 1985) raisequestions about this association. Beginning with Stouffer’s (1955) findings inthe 1950s that church attendance predicted political intolerance, a series ofstudies have documented the negative impact of personal piety on politicaltolerance (e.g., Gibson, 1992; Smidt and Penning, 1982; Steiber, 1980).Studies have also found that atheists are more tolerant of political dissent thanbelievers (Herek, 1987; Stouffer, 1955; Sullivan and Transue, 1999). But astudy carried out in the U.S. and Poland found that only theocratic beliefswere negatively related to tolerance (Karpov, 2002). When controlled by otherfactors, religiosity had a negative but non-significant impact on tolerance inIsrael (Sullivan, Walsh, Shamir, Barnum, and Gibson, 1993). However, avariety of studies suggest that these relationships may vary with the natureof religious convictions and with religious denomination (Allport, 1966;Altemeyer and Hunsberger, 1992; Batson and Ventis, 1982; Leak and Randall,1995).

Such variations, as well as the fact that the religions preach love and mercy,have led scholars to ask whether religiosity is related to negation of democracy

378 CANETTI-NISIM

in and of itself, or because of another factor. The one most often suggested hasbeen authoritarianism (e.g., Altemeyer, 1996; Duckitt and Farre, 1994;McFarland, 2000). Hunsberger (1995, p. 125) aptly sums up the literature onthe topic, saying: ‘‘Prejudice is apparently relatively unrelated to the content ofone’s beliefs in terms of belief orthodoxy, but the evidence suggests that it isrelated to religious orientation through right-wing authoritarianism.’’

The construct of authoritarianism has received great attention since thepublication of The Authoritarian Personality (Adorno, Frenkel-Brunswik,Levinson, and Sanford, 1950). Adorno et al. developed an explanation ofgeneralized ethnocentrism in terms of a personality syndrome rooted inFreudian psychodynamics, and a measure of authoritarianism—the F-scale.Authoritarianism literature has been harshly criticized since the original studyappeared, with criticism of both the conceptualization and the empiricalresearch (e.g., Christie and Jahoda, 1954; Martin, 2001; Ray, 1972; Rokeach,1960; Wilson and Patterson, 1970). Recent work by Altemeyer (1981, 1988,1996) gives researchers a sounder empirical footing for authoritarianism (seeForbes, 1985).

The Freudian-based theory has not withstood the test of time (Feldman,2003) and has proved elusive (Altemeyer, 1981, 1988; Duckitt, 1992).Although the literature is replete with data, very little guidance about thestatus of the theory has been provided. Feldman (2003) has offered a noveltheory on authoritarianism, arguing that the tension between a social orderbased on social conformity and the desire to maximize personal autonomy isthe basis of authoritarianism. However, this new conceptualization has not asyet been thoroughly examined.

The proclivity to submission to authority brings authoritarians to seek andblindly follow strong leaders (Adorno et al., 1950). Further, authoritarianattitudes can create a climate of public opinion that promotes undemocraticmovements, and intimidates politicians and opinion leaders (Altemeyer, 1996).Those who strongly value social conformity would require limitation of lib-erties and rights for segments of society (Feldman, 2003). Studies (Altemeyer,1988; Moghaddam and Vuksanovic, 1990) show that authoritarians have dis-played negative attitudes towards human rights, and are prejudiced towardsout-groups (Adorno et al., 1950; Altemeyer and Hunsberger, 1992;Hunsberger, 1995; Wylie and Forest, 1992).

Studies (McFarland, Ageyev, and Abalakina-Paap, 1992; Reisinger, Miller,Hesli, and Hill Maher, 1994) carried out in ex-communist emerging democ-racies have found that authoritarians were hostile towards out-groups andopposed democratization processes. In South Africa, Duckitt (1993), andDuckitt and Farre (1994) found that white authoritarians championed apart-heid and supported censorship, detention without trial, and persecution ofanti-government groups. However, the literature is not unequivocal. Feldman(2003), for example, claims that it is the conservative (conventional) dimension

379THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

of authoritarianism that makes authoritarians more prone to negate demo-cratic values and to wish to limit liberties and rights in democratic societies.

Scholars also claim that authoritarianism is related to religiosity (Huns-berger, 1995; Weller, Levinbok, Maimon, and Shaham, 1975), and Adornoet al. (1950) included the belief in a supernatural power in the F-scale. Thislink, too, is supported by empirical studies, which show strong correlationsbetween authoritarianism and religiosity, whether measured as belief inChristian teachings, in fundamentalism (Altemeyer, 1988; Altemeyer andHunsberger, 1992; Gregory, 1957; Leak and Randall, 1995), or in Judaism(Rubinstein, 1966, 1997).

Notwithstanding studies on religiosity, authoritarianism and tolerance(Hunsberger, 1995), studies on the proposed tripartite relationship betweenreligiosity, authoritarianism, and abstract democratic values, are oftennowhere to be found. Scholars have examined the links between democraticvalues (e.g., human rights, tolerance) and the other two variables separately inthe same study (e.g., Duckitt, 1993; Moghaddam and Vuksanovic, 1990;Rubinstein, 1996, 1997). Altemeyer and Hunsberger (1992) controlled forauthoritarianism in testing the correlation between religiosity and prejudice,and found that authoritarianism reduced the fundamentalist-prejudice cor-relation to non-significant levels. Wylie and Forest (1992) found thatauthoritarianism added significantly to the explanation of racism and homo-phobia, while fundamentalism did not. Laythe, Finkel, and Kirkpatrick (2001)found that while authoritarianism was a potent predictor of both racism andprejudice against homosexuals, fundamentalism became an inverse predictorof racism, but remained a positive predictor of homophobia. In a relatedstudy, Laythe, Finkel, Bringle, and Kirkpatrick (2002) found that fundamen-talism and ethnic prejudice were negatively related when authoritarianism wascontrolled. By and large, these findings support the claim that it is theauthoritarian element which fosters the association between religiosity anddemocratic values, but this assumption has not as yet been thoroughlyexamined.

Distinct from other studies, this study treats authoritarianism as a possiblemediater between religiosity and democratic values.2 As an exercise, it alsoexamines the possibility that religiosity mediates the relationship betweenauthoritarianism and democratic values because the nature of the linkbetween religiosity and authoritarianism is as yet undetermined (Leak andRandall, 1995). Some writers argue that religious training fosters authoritari-anism (Frenkel-Brunswick, 1950); others claim that authoritarian tendenciesfoster stronger religiosity (Weller et al., 1975; Wylie and Forest, 1992), andAltemeyer (1996) proposes that they feed into each other and suggests thatthey may both be outcomes of some third factor, such as rigid parentalupbringing. One attempt to examine the sequential relations argues thatauthoritarian individuals are more attracted to Jewish orthodoxy (Weller et al.,

380 CANETTI-NISIM

1975), but the reverse argument is not really borne out. To test these generallyreverse arguments, I have advanced two main structural equation models. InModel 1, religiosity is related to support for democratic values both directlyand indirectly via authoritarianism, whereas in Model 2 authoritarianism isrelated to support for democratic values directly and indirectly via religiosity.

METHOD

Sample and Procedure

In May–June 2001, a survey was administered to students at 12 colleges(Tel-Chai Academic College; Ariel College, Jezreel Valley College; IsraelCollege, Wingate Institute, Western-Galilee College, Ashkelon NursingSchool) and universities (Hebrew University of Jerusalem, University of Haifa,Tel-Aviv University, the Technion, Bar-Ilan University) in Israel.3 The ques-tionnaires were administered by specially trained assistants, and completed inclass after confidentiality was assured and informed consent obtained. Giventhe cultural nature of the variables, only questionnaires of Israeli born, Jewishstudents were used in this study (Beit-Hallahmi, 1992; Weller et al., 1975). Inthe end, 1,138 questionnaires were analyzed.

The sample consisted of 51.4% (585) women and 48.6% (553) men, with amean age of 27.55 years (SD ¼ 7.19). The majority (54.3%) came from homes ofhigher than average income. About a third (34.3%) of their fathers had academiceducations; 45% of their fathers were native Israelis; 24.1% of the participantswere married. Some 20.3% defined themselves as religious to one degree oranother. The distribution of these variables reflects nationwide distributionamong Israel’s approximately 200,000 students (Statistical Abstracts, 2001).

Despite the fact that a high proportion of the respondents identifiedthemselves as non-religious, there was considerable variance in theirendorsement of the religious beliefs queried (see Measurement Instruments).Defining respondents as believing in a principle when they rated their degreeof belief in it at 4 or over on the 6-point scale that was used, calculationsshowed that 52% of the respondents believe in God, 44.2% in the life of thesoul after death, 37.2% that there is a supernatural power that determines thehistory of the Jewish people, 35.6% that God gave the Torah to Moses inMount Sinai, 35.4% that Jews are the chosen people, and 25.6% that theMessiah will come. This evident variance makes it possible to treat religiosityas an independent variable.

Measurement Instruments4, 5

Religiosity was assessed via items from Ben-Meir and Kedem’s (1979)Religiosity Index constructed for the Jewish population in Israel.

381THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

Confirmatory factor analysis supported the internal consistency of the six-item scale and the coefficient Cronbach’s alpha in the present study was ashigh as .92.6 Also, the scale demonstrated strong correlations with the self-definitions of religiosity noted in the description of the study participants(r ¼ .56, p < .0001).

Authoritarianism was measured by the Hebrew version of Altemeyer’s(1988) scale, validated in Israel (Rubinstein, 2003). Altemeyer’s scale consistsof 30 items and has been found to have high reliability and construct validity(Altemeyer, 1988, 1996; Christie, 1991). To meet the objections that the scaleincludes religious items (e.g., Fiebich and Williamson, 2001; Oyamot andBorgida, 2001), five items deemed to be related to religiosity were notincluded in this scale. In addition, following common practice (e.g.,McFarland, 2001), four items that did not quite fit the current Israeli politicalculture and reduced internal consistency were also excluded.

Following common practice (e.g., Altemeyer, 1996), exploratory factoranalysis was performed and yielded two factors. However, these were distin-guished not by their content but by the direction of the question, with theresponses to positively worded questions forming one factor and those tonegatively worded questions the other factor. Altemeyer (1996, 52–55)obtained similar results from his factor analysis. Following Altemeyer (1996), Ithus view the scale as essentially one-dimensional. In the current sample thescale had a Cronbach’s alpha of .85, with none of the items below this level ofinternal consistency. This is similar to results in Canada, the United States,South Africa, West Germany, and Australia (Altemeyer, 1988, 1996). Tests ofpredictive validity proved that this version of the scale distinguished betweensupporters of the left and right, as well as among supporters of the differentparties; the more right-wing the political orientation of the party, the higherthe mean authoritarianism score of its supporters.

Support for democratic values was measured by a six-item scale. The corescale consists of four items drawn from Yuchtman-Yaar and Peres’ (2000)study of Israeli society and values; one item drawn from Kaase’s (1971) studyof democratic values in Germany, and one item from Miller, Helsi, andReisinger’s (1995) study of democratic values in the United States.7 Confir-matory factor analysis largely supported the internal consistency of the scale.

As with authoritarianism, the items divided into two factors in accord withthe direction of their phrasing. An attempt to exclude the contraits from thescale resulted in a miniscule rise in reliability. Thus, all six items were treatedas a unified construct. Reliability was fair with a Cronbach’s alpha of .64 forthe scale as a whole and none of the items below this level of internal con-sistency. The scale was well correlated with authoritarianism (r ¼ ).56,p ¼ .0001), and distinguished well between voters for the right (Ariel Sharon)and voters for the left (Ehud Barak), who were by far more supportive ofdemocratic values.

382 CANETTI-NISIM

Coming on the heels of Feldman’s (2003) argument on the relatedness ofauthoritarianism to concepts it attempts to predict, and given the highcorrelation between authoritarianism and democratic values, it was neces-sary to ensure that the two concepts were empirically distinct. To this end,the path between authoritarianism and democratic values was first exam-ined without a constraint, then with a constraint of 1 (i.e., making the twovariables ‘‘identical’’). These two models were then tested for v2 difference.The constraint produced a significant change (Dv2 ¼ 1033.51, Ddf ¼ 1),indicating that authoritarianism and democratic values, though related, aretwo separate concepts (Byrne, 2001). Overall, this procedure lends supportto the argument that the present constructs measure two differentdimensions.

Data Analysis

The analysis described here is an estimation of structural equations models(SEM) with latent variables, using Version 4.0 of the AMOS program usingfull information maximum likelihood procedure (Arbuckle and Wothke,1999).8 For reasons illuminated elsewhere, I have advanced two full models (1and 2). Each of these was constrained twice to examine direct and indirectrelations nested in it. The resulting models are: 1a (indirect, nested in 1), 2a(indirect, nested in 2), and 1b2b (direct, nested both in 1 and 2).

The quality of the models was evaluated using five criteria. The first, fol-lowing Raykov, Tomer, and Nesselroade (1991), consists of five fit measures:The v2 test is the most basic, but is sensitive to sample size, and is consideredstable and reliable only in samples of 200–400 (Joreskog and Sorbom, 1994) orof 50–500 (Hayduk, 1987). NFI, TLI, CFI, and RMSEA are also reported(Boomsma, 2000). Good fit of the hypothesized model to the observed data,however, does not necessarily mean that it is the correct causality model(Kline, 1998). Hence, the second set of criteria consists of the magnitude anddirection of the path coefficients (Joreskog and Sorbom, 1994). The third isSobel test, performed to test whether a mediator carries the influence of anindependent to a dependent. In examining possible mediation, the Sobel(1982) test offers a significance test in which a z score is calculated such thatwe can determine, whether the indirect effect of the independent variable onthe dependent variable by means of the mediator is significantly different fromzero. The Fourth criterion is a high percentage of explained variance of thedependent variable. Fifth, the best model was selected with the help of v2

difference tests for nested models (Bollen, 1989). This test is conducted on therestricted and unrestricted forms of the full models. It calculates the differ-ence in the usual v2 estimators for restricted and unrestricted models, with thedf equal to the difference in the df’s of the two, and thereby enables selectionof the better model based on comparison of the general goodness of fit. Four

383THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

tests were thus required in this study (Model 1 vs. 1a; Model 1 vs. 1b2b;Model 2 vs. 2a; Model 2 vs. 1b2b).

FINDINGS

Descriptive Overview

This preliminary perusal is three-stepped: means of the computed scales,selected means of prominent items, and correlations. Support for democraticvalues is fairly high in the sample (M ¼ 4.76, SD ¼ .85). The mean level ofauthoritarianism falls in the middle of the scale (M ¼ 2.96 SD ¼ .70), which issimilar to the level previously found among Jews in Israel (e.g., Rubinstein1996, 1997), but higher than that among Jews in the Diaspora (e.g., Altemeyer,1996). The mean level of religiosity was also moderate (M ¼ 2.91, SD ¼ 1.52).Of the democratic values, the right to free speech has won highest endorse-ment (M ¼ 5.15, SD ¼ 1.17). By and large, the authoritarianism items showeda consistent variance, with high support for items tapping such needs as thenecessity for law, order and security (M ¼ 4.48, SD ¼ 1.41), on the one hand,but low support for the suppression of immodesty and especially for physicalpunishment of children. The mean of the item belief in God was aboveaverage (M ¼ 3.66, SD ¼ 1.94).

Correlations among the study variables were significant at the level ofp < .0001. The correlation between religiosity and authoritarianism was positive(.40), whereas the correlations between religiosity and democratic values ().29),as well as between authoritarianism and democratic values ().56), were negative.

Estimating the Models

Models 1 and 2, which were based on a viable measurement model,9 haveadequate fit to the data (see Table 1). As can be seen in Table 2 and Figure 1,in Model 1, both direct and indirect paths between religiosity and authori-tarianism and between authoritarianism and support for democratic valueswere significant. Yet, the coefficient of the direct path ().10) is probably onlystatistically significant because of the large sample size, and by any objectiveindices of effect size magnitude is trivial and may possibly be deleted from themodel. Namely, religiosity hardly reduced support for democratic values on itsown, but through the mediation of authoritarianism. Further comparisonshows that the indirect ().56 · .45 ¼ ).25) path via authoritarianism wasstronger than the direct ().10) one, and Sobel (1982) test (11.66, p < .0001)provides additional evidence to that estimate. This model explained 37% ofthe variance in support for democratic values.

In Model 2, both direct and indirect paths between authoritarianism andreligiosity and between religiosity and support for democratic values were

384 CANETTI-NISIM

TABLE

1.Goodness-of-Fitan

dChi-squaredComparisonofNestedModels

Fit

Mea

sure

sM

odel

Com

par

ison

Mod

elv2

dfX

2/d

fN

FI

TL

IC

FI

RM

SE

ADv

2Dd

fF

ree

Mod

els

131

32.8

6*49

26.

36.9

6.9

6.9

6.0

62

3132

.86*

492

6.36

.96

.96

.96

.06

Con

stra

ined

Mod

els

1a31

39.6

8*49

36.

36.9

6.9

6.9

7.0

6M

odel

1avs

.m

odel

16.

821

1b2b

3335

.11*

493

6.76

.96

.96

.96

.07

Mod

el1b

2bvs

.m

odel

220

2.24

*1

Mod

el1b

2bvs

.m

odel

120

2.24

*1

2a33

02.0

8*49

36.

69.9

6.9

6.9

6.0

7M

odel

2avs

.m

odel

216

9.22

*1

Not

e:p

<.0

01;

Mod

el1:

Fu

llyfr

ee.

Rel

igio

sity

isre

late

dto

sup

por

tfo

rd

emoc

rati

cva

lues

bot

hd

irec

tly

and

ind

irec

tly

via

auth

orit

aria

nis

m.

Mod

el2:

Fu

llyfr

eeA

uth

orit

aria

nis

mis

rela

ted

tosu

pp

ort

for

dem

ocra

tic

valu

esd

irec

tly

and

ind

irec

tly

via

relig

iosi

ty.

Mod

el1a

:P

arti

ally

con

stra

ined

.B

ased

onM

odel

1,d

irec

tp

ath

sco

nst

rain

edto

equ

alze

ro.

Mod

el1b

2b:

Par

tial

lyco

nst

rain

ed.

Bas

edon

Mod

el(1

and

2),

ind

irec

tp

ath

sco

nst

rain

edto

equ

alze

ro.

Mod

el2a

:P

arti

ally

con

stra

ined

.B

ased

onM

odel

2,d

irec

tp

ath

sco

nst

rain

edto

equ

alze

ro.

385THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

significant. Yet, here also, the path coefficient of ).10 is probably only sta-tistically significant. Put differently, religiosity only elusively mediated thenegative impact of authoritarianism on support for democratic values, yet thedirect impact of authoritarianism was significantly stronger. Further com-parison shows that the direct ().56) path was much stronger than the indirect(.45 · ).10=).04) one via religiosity. Accordingly, Sobel (1982) test (2.62,p < .01) produced a significant yet fragile result, weaker than that of Model 1.This model explained 37% of the variance in support for democratic values.

Thus far, the models were examined separately. The next step was tocompare them so as to select the best model. The first comparison was aimedat analyzing the uniqueness of the models to determine whether the two full

TABLE 2. Structural Coefficients and Squared Multiple Correlations of theResearch Variables

Model

Parameters 1. 2. 1a. 1b2b. 2a.

AuthoritarianismfiReligiosity

.45*** .45***

ReligiosityfiAuthoritarianism

.45*** .45***

AuthoritarianismfiDemocratic values

).56*** ).56*** ).61*** ).56***

ReligiosityfiDemocratic values

).10*** ).10*** ).14*** ).35***

R2

Religiosity .21 .21Authoritarianism .21 .21Democratic values .37 .37 .38 .33 .12

Note:**p < .01,***p < .001.

.45***

Model 2

-.56***

-.10**Religiosity

Democraticvalues

Authoritarianism

.45***

Model 1

-.56***

-.10**Religiosity

Democraticvalues

Authoritarianism

(Sobel test=11.66, p<.0001) (Sobel test=2.62, p<.01)

FIG. 1. Full models and their path coefficients.

386 CANETTI-NISIM

models could be eliminated in favor of the more parsimonious constrainedmodels that are nested in them—the assumption being that more parsimo-nious explanations are preferable (Byrne, 2001; Kline, 1998). To this end, fourcomparisons of the nested models were carried out. Model 1 was compared toModels 1b2b and 1a; Model 2 was compared to Models 1b2b and 2a. Table 1shows significant differences between Models 1 and 1b2b and between Model2 and both Model 1b2b and Model 2a. No significant difference, however, wasfound between Models 1 and 1a, meaning that these were not unique. From astatistical standpoint, Model 2 cannot be replaced by either Model 1b2b or 2a,but Model 1 can be replaced by Model 1a. In other words, the only model thatcan be eliminated on the basis of these comparisons is Model 1.

The second comparison was among all the models.10 Fit measures andexplained variance of Models 1b2b, 1a, 2, and 1 are similar, but Model 2a canbe eliminated on the grounds of low explained variance. Yet, further scrutinysuggests that Model 1a, in which the impact of religiosity, on democratic valuesis mediated by authoritarianism, is preferable to the others on four grounds:First, all its path coefficients are significant and strong, which is not the casefor Models 1b2b, 2, and 1. Second, it is more parsimonious and refined thanModel 1. Third, it verifies the assumption that the trivial coefficient of the pathbetween religiosity and support for democratic values which appears in Model1, can be eliminated. Fourth, it is an embodiment of the results of the Sobeltest of the relations described in Model 1, meaning that there is evidence thatthe relationship between religiosity and support for democratic values is sta-tistically reduced when controlling for authoritarianism and the magnitude ofthis reduction is statistically significant. Such a result calls for a plausible,refined set of relations, such as depicted in Model 1a. Thus, its statisticallytested superiority compared to Model 1, along with its general excellencecompared to the rest of the models, merit its selection as the best model (seeFigure 2). Notwithstanding, it cannot be ruled out that Model 2 highlightsrelevant information; namely, that authoritarianism not only mediates theimpact of religiosity on support for democratic values, but also has a strongdirect impact of its own. Thus, it may be viewed as an addendum to Model1a.11

SUMMARY AND CONCLUSIONS

This study addresses fundamental concerns about the sources of negation ofdemocratic values and, more generally, the role of religiosity and authoritar-ianism in democratic theory. The objective of the study was to untangle thenature of the relations between religiosity, authoritarianism, and democraticvalues. The results are completely consistent with expectations. The mainfinding is that the association between religiosity and democratic values isalmost entirely mediated by authoritarianism. The findings also highlight the

387THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

very potent impact of authoritarianism on democratic attitudes, not only as amediating variable but also on its own. These findings bear likeness to claimsand findings that there are strong associations between authoritarianism andintolerance (e.g., Adorno et al., 1950; Duckitt and Farrre, 1994), as well aswith findings showings that authoritarianism is a stronger predictor of intol-erance than religiosity (e.g., Altemeyer and Hunsberger, 1992; Laythe et al.,2002; Wylie and Forest, 1992). Yet, in the way it is designed, this study baressubstantial methodological and theoretical highlights that are innovative andare clearly different than others.

Persons with strong authoritarian tendencies are likely to negate abstractpolitical basics of democratic political cultures, whether they are religious ornot. Yet, persons with strong religious convictions are likely to negate thesesame democratic basics because of the strong association between religiosityand authoritarianism. As in previous studies (e.g., Adorno et al., 1950; Alte-meyer, 1981, 1988, 1996; Laythe et al., 2002; Weller et al., 1975), a closenessfound between religiosity and authoritarianism was affirmed. Yet idiosyncraticfrom other studies, the present findings shed light on the nature and nurtureof this close relationship and its implication on negation of abstract democraticvalues. They suggest that what makes religiosity an engine of negation ofdemocratic values is the tendency of persons with strong religious convictionsto become authoritarian in inclination. In a sense, the findings also lendcredence to William James’ (1902) observation over a century ago that tol-erance is best predicted not by religiosity itself but by the way that people holdtheir religion.

Hence, the results lead to an appealing conclusion that authoritarianismseems to mediate the relationship between religiosity and support for demo-cratic values, rather than religiosity mediating the relationship betweenauthoritarianism and support for democratic values. There is, however, afurther conclusion of no less importance. In supporting the model inwhich religiosity leads to authoritarianism and not the model showing the

-.61***

.45***

Religiosity

Democraticvalues

Authoritarianism

FIG. 2. Authoritarianism mediates the relationship between religiosity and supportfor democratic values.

388 CANETTI-NISIM

opposite, the data could very tentatively be seen as not consistent with thewidely held view in classical authoritarianism literature of authoritarianism asbeing a personality variable. Instead, these findings seem more consistent withinterpretations of authoritarianism scales as measuring social or ideologicalattitude dimension (Altemeyer, 1996; Duckitt, 2001).

Caution is called for. The student sample is a potential limitation, but alsoan opportunity. On the one hand, this is not a representative sample of adults,so the specific results presented here may not be generalized to a broaderpopulation. On the other hand, it is an adequate sample to explore the centraldynamics of religiosity, authoritarianism, and support for abstract democraticvalues. Moreover, Hollinger and Smith (2002) have shown that student pop-ulations are much more similar to their own society than to students in othercountries; namely, as far as relationships between variables in this area areconcerned, findings from students do typically generalize well to their popu-lations. Laythe et al. (2002) suggest that the use of student samples may confersome advantages. It facilitates the use of large scales at reasonable levels ofreliability and validity, which would be difficult and costly to administer to alarge representative sample. It enhances internal consistency by a priorilimiting the number of confounding variables. Many studies have foundsubstantial variation in authoritarianism and intolerance in student samples.Most of Altemeyer’s research has involved such samples, as has a great deal ofother research in this area (e.g., Feldman, 2003). Hence, such sample facili-tates comparison with other studies.

Questions of generalizability are also raised by the fact that the study wascarried out in Israel on an all-Jewish sample. Israel differs from most otherWestern societies in two key respects. It is a self-declared Jewish state, inwhich religious authorities wield considerable power, so that the separationbetween religion and state is relatively blurred (Beit-Hallahmi, 1992;Liebman, 1995; Liebman and Don-Yehiya, 1983) and religion and politicsmix to a high degree (Altemeyer, 1996; Wald and Martinez, 2001). It is alsoa state whose security has been constantly threatened by the unremittingconflict with its Arab neighbors (Bar-Tal, 2000). In addition, it containswithin its borders a sizeable Palestinian minority with political and culturalattachment to Palestinians, and other Arabs, beyond its borders (Horowitzand Lissak, 1989). Moreover, Judaism, in its orthodox form, is a religionwith a particularly large number of required precepts, which makes it amore encompassing way of life than orthodox Christianity (Weller et al.,1975).

On the one hand, research consistently indicates that threat perceptiongenerally leads to intolerance and denigration of civil rights (Doty, Petersonand Winter, 1991; Feldman and Stenner, 1997; Huddy, Feldman, Capelos,and Provost, 2002; Marcus, Sullivan, Thesis-Morse, and Wood, 1995); andclaims have been made that religious Judaism is more inclined to authori-

389THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

tarianism than Christianity (Weller et al., 1975) and does not encouragedemocratic values (Liebman, 1995). On the other hand, the sample evi-denced the same high support for democratic values as samples in otherdemocratic counties (Farnen and Meloen, 2000; Laythe et al., 2002), eventhough, it may be noted, the questionnaire was administered during the fullforce of the Al-Aqsa Intifada. Moreover, both its scores on authoritarianismand the correlations between authoritarianism and religiosity were wellwithin the ranges found elsewhere, as depicted in the Introduction. Inaddition, the extension of the terror threat to Europe and especially theUnited States since September 11, 2001 has reduced this difference betweenIsrael and other democratic countries. Finally, to my knowledge, there islittle literature on authoritarianism in Israel that has been published inEnglish. There is probably no other study with a large Israeli sampleexamining the effect of religiosity and authoritarianism on democratic values.Further research is recommended to cross-validate the study findings inother populations.

Yet another caveat is in order. Religiosity has been measured by tapping anumber of core Jewish principles, which in many ways bear likeness to fun-damentalism. Religiosity may be defined by different orientations, such asintrinsic-extrinsic, attendance at religious services, depth of conviction, orfundamentalism (Hill and Hood, 1999). While there may be a certain corre-spondence among different orientations of religiosity (Laythe et al., 2002), thecorrespondence may not be total and findings may vary with the dimensionwhich is measured (Leak and Randall, 1995).

In summary, the study contributes to the development of theoretical andsubstantive knowledge in political psychology. Initially, it adds to the still smallnumber of empirical studies which simultaneously examine the relationshipsbetween religiosity and authoritarianism, on the one hand, and abstractdemocratic values, on the other. To my knowledge, it is the only publishedstudy to date, that methodically examines both direct and mediated rela-tionships, and the only study of all three variables that measures democraticvalues in terms of abstract political beliefs. Previous studies, of which there aremany, have been carried out either from a psychological or a political per-spective. Studies carried out by psychologists examined the impact of religi-osity and/or authoritarianism on intolerance and prejudice. Studies by politicalscientists sought to explain support for non-democratic political values bymeans of religiosity, but did not consider authoritarianism. This study bringstogether the separate approaches of the two fields to try to explain support forpolitical dimensions of democratic values using psychological variables.Greater attention to the interface between psychology and political attitudes isparticularly pertinent in our times in light of the rise of anti-democratic bel-ligerency by religious groups of all faiths. I hope that, whatever further study

390 CANETTI-NISIM

may demonstrate, the end result will be a better understanding of the originsof negation of democracy.

APPENDIX

Religiosity MeasureTo what extent do you believe in the life of the soul after death?To what extent do you believe that the Messiah will come?To what extent do you believe that there is a supernatural power that

determines the history of the Jewish people?To what extent do you believe God gave the Torah to Moses on Mount

Sinai?To what extent do you believe that the Jews are the chosen people?To what extent do you believe in God?

Authoritarianism measure

The way things are going in this country, it’s going to take a lot of ‘‘strongmedicine’’ to straighten out the troublemakers, criminals, and perverts.

It’s wonderful that young people today have greater freedom to protestagainst things they don’t, like and to ‘‘do their own thing.’’*

It’s always better to trust the judgment of the proper governmental andreligious authorities than to listen to the noisy rabble-rousers in our societywho try to create doubt in people’s minds.

It would be best for everyone if the proper authorities censored magazinesand movies to keep trashy material away from youth.

It may be considered old-fashioned by some, but having a decent,respectable appearance is still the mark of a gentleman and, especially, a lady.

The facts on crime, sexual immorality, and the recent public disorders allshow we have to crack down harder on deviant groups and troublemakers ifwe are going to save moral standards and preserve law and order.

It is important to protect fully the rights of radicals and deviants.*Obedience and respect for authority are the most important virtues children

should learn.Once our government leaders and the authorities condemn the dangerous

elements in our society, it will be the duty of every patriotic citizen to helpstamp out the rot that is poisoning our country from within.

‘‘Free speech’’ means that people should even be allowed to make speechesand write books urging the overthrow of the government.*

Some of the worst people in our country nowadays are those who do notrespect our flag, our leaders, and the normal way things are supposed to bedone.

391THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

In these troubled times laws have to be enforced without mercy, especiallywhen dealing with the agitators and revolutionaries who are stirring things up.

Young people sometimes get rebellious ideas, but as they grow up theyought to get over them and settle down.

If a child starts becoming unconventional and disrespectful of authority, it ishis parents’ duty to get him back to the normal way.

In the final analysis the established authorities, like parents and our nationalleaders, generally turn out to be right about things, and all the protesters don’tknow what they are talking about.

A lot of our rules on modesty and sexual behavior are just customs which arenot necessarily any better or holier than those which other people follow.*

There is absolutely nothing wrong with nudist camps.*The real keys to the ‘‘good life’’ are obedience, discipline, and sticking to the

straight and narrow.It is best to treat dissenters with leniency and an open mind, since new ideas

are the lifeblood of progressive change.*Our country will be great if we honor the ways our forefathers, do what the

authorities tell us to do, and get rid of the ‘‘rotten apples’’ who are ruiningeverything.

One reason we have so many troublemakers in our society nowadays is thatparents and other authorities have forgotten that good old-fashioned physicalpunishment is still one of the best ways to make people behave properly.

Support for Democratic Values Measure

The State of Israel must provide equal social and political rights to allcitizens, regardless of religion, race or gender.

Open and public criticism of the government is permissible even in states ifemergency.

The slightest threat to the security of the state is enough to justify seriousrestrictions on democracy.*

I prefer a democratic government whose actions and views are opposed tomine over a non-democratic government with whose views I agree.

Every citizen has the right to take his convictions to the street if necessary.Public participation of the people is not necessary if decision making is left

in the hands of a few trusted leaders.*

NOTES

1. Before establishing the theoretical grounds for the study on democratic values, distinctionbetween democratic values and political tolerance is called for. Political tolerance is at theheart of a modem conception of democratic values and citizenship. It embodies specificprinciples vis-a-vis abstract democratic values (Shamir, 1991), and the former are at timespredicted by the latter (Sullivan, Walsh, Shamir, Barnum, and Gibson, 1993). Unlike

392 CANETTI-NISIM

democratic values, political tolerance may be defined as the willingness to extend the rights ofcitizenship to all members of polity—that is, to allow, political freedoms to those who arepolitically different (Gibson and Bingham, 1982). In contrast to extensive research on politicaltolerance, few studies on democratic theory explore democratic values and their religious andauthoritarian correlates.

2. In principle authoritarianism may be either a mediator or a moderator of the associationbetween religiosity and democratic values. This study examines its role as a mediator.According to Baron and Kenny (1986), it is appropriate to test mediation when the predictorand dependent variables are strongly correlated, and moderation when their correlation isweak or inconsistent. Since the literature cited above consistently shows correlations betweenboth authoritarianism and religiosity, on the one hand, and democratic values, on the other, Iargue that a mediating relationship is more likely than a moderating one.

3. As in most studies on authoritarianism to date, students were chosen as the research popu-lation because of their relative accessibility. Studies dealing with ingroup–outgroup(e.g., racism, prejudice, authoritarianism) concepts have found that there are almost no dif-ferences between results based on student samples and general population samples (Pen-tony, Peterson, Phillips, Leong, Harper, Bakowski, Steward, and Gonzales, 2000; Sears,1993).

4. All measures were tested in a pilot study carried out among Israeli students. Where necessary,modifications were made in accord with the findings. Following McFarland, Ageyev, andAbalakina (1992), who studied similar variables, all items were coded on a 6-point scale: 1denoting the least agreement with the item and 6 the most. For the purpose of analysis, thecontraits were reversed.

5. Either the original scales were unbalanced or necessary processes have made them so (seeAppendix for specific item). Yet, where possible, the set of relationships tested here wasexamined in a pilot study using a balanced version of the scale, but this did not change theresults materially (results available upon request). At the same time, it should be kept in mindthat the scholarly debate on the ultimate way to design scales is far from exhausted (Krosnickand Fabrigar, 2001).

6. Confirmatory factor analyses reported in this section represent only the final stage of acomprehensive preliminary work to grant validity and reliability of each measure. In mostcases, such as with the authoritarianism, a long exploratory process took place, before the finalmeasure was confirmed. For space considerations, only a brief treatment of the measurementdesign is reported here.

7. As illustrated elsewhere in this paper the conceptualization of democratic values is differentfrom that of political tolerance. In a similar vein, whereas this measure of democratic valuestaps abstract aspects of the ‘‘democratic creed,’’ measures of tolerance usually tap support forcivil liberties in concrete cases (Sullivan and Transue, 1999).

8. The SEM consists of confirmatory latent-variable structural models (Bentler, 1990), which arepowerful tools for dealing with the systematic and unsystematic measurement errors (Block,Croon, and Hagenaars, 2004; Bollen, 1989; Boomsma, 2000). It provides simultaneous esti-mation of the hypothesized regressions using the estimated covariance matrix generated onthe basis of the observed covariance matrix of the measured variables (Kline, 1998). Theestimated matrix is also used for evaluating the goodness of fit between the data and themodel. A covariance matrix among the research variables formed the input for the pathanalysis, and the results were calculated on the basis of all indicators comprising the latentvariables.

9. v2 ¼ 3132.86; RMSEA ¼ .06; NFI ¼ .96; TLI ¼ .96; CFI ¼ .97.10. As implied elsewhere, because of the theoretical rationale and thus the structure of the models,

not all dyads of models can be compared by means of Chi-square difference. Therefore, ageneral comparison based on the remaining criteria, is called for.

393THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

11. To rule out the likelihood of moderation, democratic values were regressed on religiosity,authoritarianism, and the interaction between them. The findings show that the model as awhole was significant (F(3,1128) ¼ 176.175, p < .0001), explaining 32% of the variance.However, the only variable that contributed significantly to the variance was authoritarianism(b(t) ¼ ).48 (8.94), p < .0001). Neither religiosity nor the interaction had a significant impact,thereby ruling out the possibility of moderation (Baron and Kenny, 1986).

REFERENCES

Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., and Sanford, R. N. (1950). TheAuthoritarian Personality. New York: Harper and Row, Publishers.

Aiken, L. S., and West, S. G. (1991). Multiple Regression: Testing and InterpretingInteractions. Newbury Park: Sage Publications.

Allport, G. W. (1966). The religious context of prejudice. Journal for the ScientificStudy of Religion 5: 447–457.

Almond, G., and Verba, S. (1963). The Civic Culture. Princeton: Princeton UniversityPress.

Altemeyer, B. (1981). Right-wing Authoritarianism. Manitoba: The University ofManitoba Press.

Altemeyer, B. (1988). Enemies of Freedom. San Francisco: Jossey-Bass Publishers.Altemeyer, B. (1996). The Authoritarian Specter. Cambridge: Harvard University

Press.Altemeyer, B., and Hunsberger, B. (1992). Authoritarianism, religious fundamentalism,

quest, and prejudice. The International Journal for the Psychology of Religion 2:113–133.

Arbuckle, J. L., and Wothke, W. (1999). AMOS User’s Guide 4.0. Chicago: Smallwa-ters.

Arian, A., Nachmias, D., Navot, D., and Shani, D. (2003). Auditing Israeli Democracy2003. Israel Democracy Institute (Hebrew).

Bar-Tal, D. (2000). Shared Beliefs in a Society: Social Psychological Analysis. ThousandOaks: Sage Publications.

Baron, R. M., and Kenny, D. A. (1986). The moderator-mediator variable distinction insocial psychological research: conceptual, strategic, and statistical considerations.Journal of Personality and Social Psychology 6: 1173–1182.

Batson, C. D., and Ventis, W. L. (1982). The Religious Experience, A Social Psycho-logical Perspective. New York: Oxford University Press.

Beit-Hallahmi, B. (1992). Despair and Deliverance: Private Salvation in ContemporaryIsrael. New York: SUNY Press.

Beit-Hallahmi, B. (2001). Fundamentalism. In J. Krieger (ed.), The Oxford Companionto Politics of the World. New York: Oxford University Press.

Bellah, R. (1976). The new religious consciousness and the crisis of modernity. In C.Glock and R. N. Bellah (eds.), The New Religious Consciousness, pp. 133–152.Berkeley-Los Angeles: University of California Press.

Ben-Meir, Y., and Kedem P. (1979). Religiosity index for the Jewish population inIsrael. Megamot 24: 353–362 (Hebrew).

Bentler, P. M. (1990). Comparative fit indexes in structural models. PsychologicalBulletin 107: 238–246.

Bolck, A., Croon, M., and Hagenaars, J. (2004). Estimating latent structure models withcategorical variables: one-step versus three-step estimators. Political Analysis 12: 3–27.

394 CANETTI-NISIM

Bollen, K. A. (1989). Structural Equation with Latent Variables. New York: Wiley.Boomsma, A. (2000). Reporting analyses of covariance structures. Structural Equation

Modeling 7: 461–482.Byrne, B. M. (2001). Structural Equation Modeling with AMOS: Basic Concepts,

Applications, and Programming. Mahwah: Lawrence Erlbaum Associates.Canetti-Nisim, D. (2003). Two religious meaning systems, one political belief system:

Religiosity, quasi-religiosity, and political extremism. Totalitarian Movements andPolitical Religions 43(3): 35–54.

Christie, R. (1991). Authoritarianism and related constructs. In Robinson, P. R. Shaverand L. S. Wrightsman (eds.), Measures of Personality and Social PsychologicalAttitudes Vol. 1, 2nd Ed. New York: Academic Press.

Christie, R., and Jahoda, M. (eds.) (1954). Studies in the Scope and Method of ‘‘TheAuthoritarian Personality.’’ Glencoe, IL: Free Press.

Doty, R. M., Peterson, B. E., and Winter, D. G. (1991). Threat and authoritarianism inthe United States, 1978–1987. Journal of Personality and Social Psychology 61: 629–640.

Duckitt, J. (1992). The Social Psychology of Prejudice. New York: Praeger.Duckitt, J. (1993). Right-wing authoritarianism among white South African students: its

measurement and correlates. The Journal of Social Psychology 133: 553–564.Duckitt, J. (2001). A dual-process cognitive-motivational theory of ideology and pre-

judice. Advances in Experimental Social Psychology 3: 41–113.Duckitt J., and Farre B. (1994). Right-wing authoritarianism and political intolerance

among whites in the future majority-rule South Africa. Journal of Social Psychology134: 735–741.

Duriez, B., Luyten, P., Snauwaert, B., and Hutsebaut D. (2002). The importance ofreligiosity and values in predicting political attitudes: evidence for the continuingimportance of religiosity in Flanders (Belgium). Mental Health, Religion and Culture5: 35–54.

Farnen, R., and Meloen, J. (2000). Democracy, Authoritarianism, and Education: ACross-National Empirical Survey. New York: St. Martin’s Press.

Feldman, S. (2003). Enforcing social conformity. Political Psychology 24: 41–74.Feldman, S., and Stenner, K. (1997). Perceived threats and authoritarianism. Political

Psychology 18: 741–770.Fiebich, C., and Williamson, I. (2001). Cutting out the Fat: Modifying the

Authoritarian Personality Scale to Avoid Tautologies. Paper presented at the July2001 International Society for Political Psychology conference in Cuernavaca,Mexico.

Forbes, H. D. (1985). Nationalism, Ethnocentrism, and Personality. Chicago: Uni-versity of Chicago Press.

Frenkel-Bruswick, E. (1950). Comparative scores and summary of interview results. InAdorno, T. W., E. Frenkel-Brunswik, D. J. Levinson, and R. N. Sanford (eds.), TheAuthoritarian Personality, pp. 468–486 New York: Harper and Row, Publishers.

Gibson, J. L. (1992). The political consequences of intolerance: cultural conformity andpolitical freedom. American Political Science Review 86: 338–356.

Gibson, J. L., and Bingham, R. D. (1982). On the conceptualization and measurementof political tolerance. American Political Science Review 76(3): 603–620.

Gregory, W. (1957). The orthodoxy of the authoritarian personality. Journal of SocialPsychology 45: 217–232.

Hayduk, L. A. (1987). Structural Equation Modeling, with LISREL: Essentials andAdvances. Baltimore: Johns Hopkins University Press.

395THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

Herek, G. M. (1987). Religious orientation and prejudice: a comparison of racial andsexual attitudes. Personality and Social Psychology Bulletin 13: 34–44.

Hill, P. C., and Hood, R. W. Jr. (1999). Measures of Religiosity. Birmingham: ReligiousEducation Press.

Hollinger, V. F., and Smith, T. (2002). Religion and esotericism among students:a cross-cultural comparative study. Journal of Contemporary Religion 17: 229–249.

Horowitz, D., and Lissak, M. (1989). Trouble in Utopia: The Overburdened Polity ofIsrael. Albany: New York Press.

Huddy, L., Feldman, S., Capelos, T., and Provost, C. (2002). The consequences ofterrorism: disentangling the effects of personal and national threat. Political Psy-chology 23: 485–509.

Hunsberger, B. (1995). Religion and prejudice: the role of religious fundamentalism,quest and right wing authoritarianism. Journal of Social Issues 51: 113–129.

James, W. (1956) [1897]. The Will to Believe. New York: Dover Publications.James, W. (1958) [1902]. The Varieties of Religious Experience. New York: New

American Library.Joreskog, K., and Sorbom, D. (1994). Structural Equation Modeling with the SIMPLIS

Command Language. Chicago: Scientific Software International.Kaase, M. (1971). Demokratische einstellung in der bundesrepublik Deutschland. In

R. Wildenman (ed.), Sozialwissenschaftliches Jahrbuch fur Politik. Vol. 2, pp. 119–326. Munich: Gunter Olzog Verlag.

Karpov, V. (2002). Tolerance in the United States and Poland. Journal for the ScientificStudy of Religion 41: 267–288.

Kline, R. B. (1998). Principles and Practice of Structural Equation Modeling. NewYork: Guilford Press.

Krosnick, J. A., and Fabrigar, L. R. (2001). Designing Good Questionnaires: Insightsfrom Psychology. New York: Oxford University Press.

Laythe, B., Finkel, D. G., and Kirkpatrick, L. A. (2001). Predicting prejudice fromreligious fundamentalism and right-wing authoritarianism: a multiple regressionapproach. Journal for the Scientific Study of Religion 29: 442–462.

Laythe, B., Finkel, D. G., Bringle, R. G., and Kirkpatrick, L. A. (2002). Religiousfundamentalism as a predictor of prejudice: a two-component model. Journal for theScientific study of Religion 41: 623–635.

Leak G. K., and Randall, B. A. (1995). Clarification of the link between right wingauthoritarianism and religiousness—the role of religious maturity. Journal for theScientific Study of Religion 34: 245–252.

Liebman, C. S. (1995). Religion and democracy in Israel. In S. Deshen, C. S. Liebman,and M. Shokeid (eds.), Israeli Judaism: The Sociology of Religion in Israel, pp. 347–366. New Brunswick, NJ: Transaction.

Liebman, C. S., and Don-Yehiya, E. (1983). Civil Religion in Israel. Berkeley: Uni-versity of California Press.

Lipset, S. M. (1981). Political Man: The Social Basis of Politics. Baltimore: The JohnsHopkins University Press.

Marcus, G. E., Sullivan, J. L., Thesis-Morse, E., and Wood, S. L. (1995). With MaliceToward Some: How People Make Civil Liberties Judgments. New York: CambridgeUniversity Press.

Martin, J. L. (2001). The authoritarian personality, 50 years later: what lessons arethere for political psychology? Political Psychology 22 (1): 1–26.

McFarland, S. G. (2000). Autoritarismus und die entwicklung eines demokratischenbewuatseins in Russland (Authoritarianism and the Development of Democratic

396 CANETTI-NISIM

Consciousness in Russia). In S. C. Rippl, C. Seipl, and A. Kindervater (eds.), Kon-troversen und Ansaetze der Aktuellen Autoritarismusforschung (Controversies andApproaches in Authoritarianism Research Today), pp. 173–197. Opladen, Germany:Leske Bundrich.

McFarland, S. (2001). Explicit and Implicit Generalized Prejudice. Paper presented atthe July 2001 International Society for Political Psychology Conference in Cuerna-vaca, Mexico.

McFarland, S., Ageyev, V., and Abalakina-Paap, M. (1992). Authoritarianism inthe former Soviet Union. Journal of Personality and Social Psychology 63: 1004–1010.

Moghaddam, F. M., and Vuksanovic, V. (1990). Attitudes and behavior towardhuman-rights across different contexts—the role of right-wing authoritarianism,political-ideology, and religiosity. International Journal of Psychology 25: 455–474.

Nelson, L. D. (1988). Disaffiliation, decentralization, and political values. In D. G.Bromley (ed.), Falling From the Faith: Causes and Consequences of ReligiousApostasy. Newbury Park, CA: Sage.

Oyamot, C. M., and Borgida, E. (2001). Understanding the Ambivalent Authoritarian:The Role of Competing Value Orientations and Attitudes Toward Immigrants. Paperpresented at the July 2001 International Society for Political Psychology conferencein Cuernavaca, Mexico.

Pedahzur, A. (2002). The Israeli Response to Jewish Extremism and Violence:Defending Democracy. Manchester: Manchester University Press.

Pentony, J. F., Peterson, K. S. E., Phillips, O., Leong, C., Harper, P., Bakowski, A.,Steward, S., and Gonzales, R. (2000). A comparison of authoritarianism in theUnited States, England, and Hungary with selected nonrandom samples. EuropeanPsychologist 5: 259–268.

Peres, Y. (1995). Religious adherence and political attitudes. In S. Deshen, C. S.Liebman, and M. Shokeid (eds.), Israeli Judaism: The Sociology of Religion in Israel,pp. 87–106. New Brunswick, NJ: Transaction.

Piereson, J., Sullivan, J. L., and Marcus, G. (1980). Political tolerance: an overview ofsome new findings. In J. C. Pierce and J. L. Sullivan (eds.), The Electorate Recon-sidered, pp. 157–178. Beverly Hills: Sage Publications.

Raijman, R., Semyonov, M., and Schmidt, P. (2003). Do foreigners deserve rights?Determinants of public, views towards foreigners in Germany and Israel. EuropeanSociological Review 19: 379–392.

Ray, J. J. (1972). A new balanced F scale—and its relation to social class. AustralianPsychology 7: 155–166.

Raykov, T., Tomer A., and Nesselroade, J. R. (1991). Reporting structural equationmodeling results in psychology and aging: some proposed guidelines. Psychology andAging 6: 499–503.

Reisinger, W., Miller A. H., Hesli, V. L., and Hill Maher, K. (1994). Political values inRussia, Ukraine and Lithuania—sources and implications for democracy. BritishJournal of Political Science 24: 183–223.

Rokeach, M. (1960). The Open and Closed Mind. Basic Books.Rubinstein, G. (1996). Two people in one land: a validation study of Altemeyer’s right-

wing authoritarianism scale in the Palestinian and Jewish societies in Israel. Journalof Cross-Cultural Psychology 27: 216–230.

Rubinstein, G. (1997). Authoritarianism, political ideology, and religiosity among stu-dents of different faculties. The Journal of Social Psychology 137: 559–567.

397THE EFFECT OF RELIGIOSITY ON ENDORSEMENT OF DEMOCRATIC VALUES

Rubinstein, G. (2003). Authoritarianism and its relation to creativity: a comparativestudy among students of design, behavioral sciences and law. Personality andIndividual Differences 34: 695–705.

Schwartz, S. H., and Huismans, S. (1995). Value priorities and religiosity in fourWestern religions. Social Psychology Quarterly 58: 88–107.

Sears, D. O. (1993). Symbolic politics: a socio-psychological theory. In S. Iyengar andW. J. McGuire (eds.), Explorations in Political Psychology pp. 113–149. Durham,NC: Duke University Press.

Shamir, M. (1991). Political intolerance among masses and elites in Israel: a reevalu-ation of the elitist theory of democracy. The Journal of Politics 53: 1018–1043.

Smidt, C. E., and Penning, J. M. (1982). Religious commitment, political conservatism,and political and social tolerance in the United States: a longitudinal analysis.Sociological Analysis 43: 231–246.

Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equationsmodels. In S. Leinhart (ed.), Sociological Methodology 1982, pp. 290–312. SanFrancisco: Jossey-Bass.

Stark, R. (2001). Reconceptualizing religion, magic, and science. Review of ReligiousResearch 43: 101–120.

Statistical Abstracts of Israel (2001). Jerusalem: The Central Bureau of Statistics.Steiber, S. R. (1980). The influence of religious factor on civil and sacred tolerance,

1958–1971. Social Forces 58: 811–832.Stouffer, S. A. (1955). Communism, Conformity, and Civil Liberties. Garden City, NY:

Doubleday & Co. Inc.Sullivan, J., and Transue, J. (1999).The psychological underpinnings of democracy: a

selective review of research on political tolerance, interpersonal trust, and socialcapital. Annual Review of Psychology 50: 625–650.

Sullivan, J. L., Shamir, M., Walsh, P., and Roberts, N. S. (1985). Political Tolerance inContext: Support for Unpopular Minorities in Israel, New Zealand, and the UnitedStates. Boulder, CO: Westview Press.

Sullivan, J. L., Piereson, J. J., and Marcus, G. (1982). Political Tolerance and AmericanDemocracy. Chicago: The University of Chicago Press.

Sullivan, J. L., Walsh, P., Shamir, M., Barnum, D. G., and Gibson, J. (1999). Whypoliticians are more tolerant: selective recruitment and socialization among politicalelites in Britain, Israel, New Zealand and the United States. British Journal ofPolitical Science 23: 51–76.

Wald, K. D., Owen, D. E., and Hill, S. S. (1988). Churches as political communities.American Political Science Review 82: 531–548.

Wald, K. D., and Martinez, M. D. (2001). Jewish religiosity and political attitudes in theUnited States and Israel. Political Behavior 23: 377–397.

Weller, L., Levinbok, S., Maimon, R., and Shaham, A. (1975). Religiosity andauthoritarianism. The Journal of Social Psychology 95: 11–18.

Wilson, G. D., and Patterson, J. R. (1970). Manual for the Conservatism Scale. NFERPublishing Co.

Wylie, W., and Forest, J. (1992). Religious fundamentalism, right-wing authoritarian-ism and prejudice. Psychological Reports 7l: 1291–1298.

Yuchtman-Yaar, E., and Peres, Y. (2000). Between Consent and Dissent: Democracyand Peace in the Israeli Mind. Lanham: Rowman & Littlefield.

398 CANETTI-NISIM