Religion and Democratic Attitudes

Michael Hoffman

November 13, 2012

Prepared for Princeton University Comparative Politics Graduate ResearchSeminar

Introduction

Why does religion sometimes promote democratic values and behaviors, but

sometimes do just the opposite? Existing political research on the role of re-

ligion in democracy has produced diverging results: on the one hand, religion

sometimes promotes democratic behaviors and attitudes (as in the resistance

to the Soviet Union, see Ramet 1989); but on the other hand, it is often said

to enhance authoritarianism (for instance, in the Iranian revolution). Even

within the same religious tradition, this duality is present: while the Catholic

Church was cited as a source of stability for authoritarian rule in Franco’s

Spain (Philpott 2007; Johnston and Figa 1988), it was also an important

source of anti-authoritarian mobilization in Latin America during the “third

wave” of democratization (Huntington 1991; Hertzke 2009). These contrast-

ing trends are puzzling from a theoretical point of view: it is evident that the

content of different religious traditions does not explain the variation in their

impact on democracy. I aim to address the question of what does explain

the ambiguous link between religion and democracy.

Motivation

Religion remains an important driver of political behavior, and particularly

so in the developing world (Norris and Inglehart 2004). However, existing

scholarship on the link between religion and comparative political behavior is

rather limited: as Bellin (2008, p. 316) writes, “the subfield has still failed to

reckon with the power of religion as an independent variable, the noninstru-

1

mental aspect of religious behavior, and the malleability of religious ideas,

as well as their differential appeal, persuasiveness, and political salience over

time.” This combination of the continued influence of religion in political life

and the relative lack of attention paid to religion by scholars of comparative

politics is problematic.

Moreover, the events of the Arab Spring highlight the continued impor-

tance of the attitudes of ordinary citizens; views toward regimes are of par-

ticular importance. The revolutions that have spread across the Arab World

suggest that mass mobilization remains a significant potential threat to au-

thoritarian regimes, in contrast to the claims made by much of the transitions

to democracy literature, which focuses primarily on elite behavior (O’Donnell

and Schmitter 1986). The uncertainty about the types of regimes that will

succeed the now-deposed autocrats in several countries in the region makes

understanding citizens’ attitudes about democracy all the more important:

if the people were able to bring down entrenched dictators, what type of

regime will they demand as a replacement?

The Arab Spring brings into focus the major topic addressed by this

project: religion’s unclear impact on democratic attitudes and behaviors.

Even within these protests themselves, observers and analysts have struggled

to grasp the link between religion and pro- or anti-democratic behaviors. On

the one hand, scholars have argued that religion was an indispensable motiva-

tor of anti-regime activity during these revolutions. Benhabib (2011) suggests

that the Islamic notion of shahada, meaning both “witness” and “martyr,”

played an important role in motivating anti-regime activity. Ardıc (2012, p.

38) offers a different account of the role of religion in motivating the Arab

2

Spring, stating that “mosques. . . functioned as a locus of anti-government

agitation and logistical centers of preparation for demonstrations.” In each

of these explanations, religion is said to have provided tools for mobilizing

citizens against the regime, albeit through different mechanisms. Other ob-

servers, however, have downplayed the role of religion in motivating the Arab

Spring. Wright (2011) argues that protesters mobilized not only in opposi-

tion to the existing regimes, but in opposition to extremist ideologies as well,

suggesting that the Arab Spring was a largely secular phenomenon. Likewise,

Lynch (2012, p. 15) warns us to avoid claiming that some “essence of Islam”

was responsible for anti-regime activity in the region.

Taken together, these arguments reveal a fundamental uncertainty about

the pro- or anti-democratic nature of religion in the Arab Spring. I will sug-

gest that this uncertainty extends to many other contexts as well: scholars

of political behavior cannot assume that religion will always either promote

or inhibit democratic attitudes; nor can they find all of the answers to these

questions in the contents of individual religious doctrines. Rather, religious

groups (and individuals) must be considered in the context in which they

operate. The reason why religion has been shown to demonstrate contra-

dictory effects on democracy, I suspect, is that incentive structures dictate

the link between religion and support for democracy. When incentives of

religious groups, leaders, or individuals align with those of pro-democratic

forces, religion will tend to promote democracy; when the opposite incentives

are present, we should expect the opposite effect.

3

Religion and Democracy

After decades of research, the impact of religion on individual-level political

behavior remains cloudy. Campbell (2004) finds that American churches can

either exhibit an effect conducive to civic involvement or exactly the opposite.

Smidt (1999) finds that church attendance fosters involvement in civil society

in Canada and the United States. Driskell, Embry and Lyon (2008) suggest

that identification with a religious group reduces political participation, but

religious practice per se increases political participation. As for political

identities, Putnam and Campbell (2010, p. 389) find that the relationship

between church attendance and partisanship has varied considerably over

time; in other words, the link between religiosity and political identities is

context-dependent.

Existing studies of the relationship between religion (particularly, piety)

and democracy have usually either ignored individual-level characteristics or

cast fairly broad strokes about this relationship without paying attention to

group and state-level characteristics that might influence the impact of piety

on various political attitudes associated with democracy. Stepan’s (2000)

well-known concept of the “twin tolerations” asserts that democracy requires

a certain amount of independence between the political and the religious

and casts doubt upon the claim that certain faiths are incompatible with

democracy, but does not directly address the question of when and how

religious piety makes for good or bad democrats. Using survey data from

the Arab World, Tessler (2002) finds that Muslim piety has little effect on

attitudes toward democracy (and less of an effect on political attitudes in

4

general than is commonly assumed). Ciftci (2010, p. 1442) likewise finds

that “religiosity and Islamic values poorly predict support for democracy

in the Muslim world.” In an impressive cross-national and multi-level study,

Meyer, Tope and Price (2008) find that religiosity–in terms of both belief and

practice–enhances support for democracy, provided that individuals continue

to support separation between religion and politics.

What is clear from the extant literature on this topic is that the rela-

tionship between religion and democracy is complicated. Philpott (2007, p.

510) observes that even within the Third Wave of democratization, religion

has played a tremendously important role in promoting democracy in some

places (e.g. Poland, Lithuania, and Indonesia) but very little role in others

(e.g. Argentina and Senegal). He attributes these differences to divergent

patterns in political theology and religion-state relations: the degree of fu-

sion between religious and political institutions as well as the level of tension

between them. In this sense, the effect of religion on democratization ap-

pears to be conditional not only on the contents and political orientations

of religious groups themselves, but also on these groups’ orientations toward

the state.

If, indeed, religion plays an indeterminate–if not actually “ambivalent”–

role in democratization, what factors push this role in one direction or an-

other? In other words, what are the mechanisms by which religion comes to

play a positive, negative, or neutral role in the process of democratization?

Social capital seems to be a natural starting point. A common explanation of

the transition to democracy in Poland following the fall of the USSR empha-

sizes the role of the Catholic Church in pushing for democracy and building

5

social capital. Communal religious practice would seem to be an especially

potent force for building social capital: through structured, regular interac-

tion with members of the same religious community, it seems plausible that

communal practice could build trust and norms of reciprocity that might,

under the right circumstances, aid in the functioning of democracy.

Clearly, however, this social capital-izing effect of religion is not always

present. The Italian case leads (Putnam 1993, p. 107) to conclude that

“organized religion . . . is an alternative to the civic community, not a part of

it.” In fact, he finds that nearly every form of religious practice is negatively

associated with most forms of civic engagement. He suggests that religion, at

least in Italy, makes believers content with the political environment in which

they live: “they seem more concerned about the city of God than the city

of man” (Putnam 1993, p. 107). Putnam’s claims resemble other common

conceptions about the role of religion in politics. These claims argue that

by focusing believers’ attention on the other-worldly, religion “puts people to

sleep” when it comes to politics. Such arguments often trace their intellectual

heritage to Marx (1970 [1843]), who describes religion as “the sigh of the

oppressed creature, the heart of a heartless world, and the soul of soulless

conditions. It is the opium of the people” (emphasis in original). In this

sense, religion either makes individuals uninterested in politics or actively

supportive of the existing arrangement. In either case, it would appear that

religion often does not provide the social capital necessary to uproot existing

regimes in favor of democracy.

More recently, however, Putnam (2000) has suggested that religion has

played a much different role in the United States. He finds that “religios-

6

ity rivals education as a powerful correlate of most forms of civic engage-

ment,” and that “churches provide an important incubator for civic skills,

civic norms, community interests, and civic recruitment” (Putnam 2000, pp.

66-67). Moreover, he argues that churches have played a crucial role in a

variety of social movements, particularly for minority groups. American re-

ligious organizations it seems, play a starkly different role in building social

capital than do their Italian counterparts. Jones-Correa and Leal (2001) ar-

gue that religious participation plays a similarly crucial civic role as does

participation in other types of organizations. Norris and Inglehart (2004, p.

192) find that religious practice demonstrates rather complicated effects on

social capital: it depresses political interest, political discussion, and social

trust, but increases social tolerance and approval of democracy. Their ev-

idence suggests that communal religious practice builds social capital, but

only in certain ways.

If religion does indeed demonstrate such a social capital-izing effect, this

effect may only build within-group social capital—so-called “bonding” social

capital (Putnam 2000). Uslaner (2002, p. 239) notes that “religion may

mobilize people to take part in their communities, but perhaps only among

their own kind” and that “religious beliefs may also lead people to distinguish

more sharply between their own kind and others.” Moreover, as Jamal (2007)

warns us, social capital can have both pro- and anti-democratic effects: social

capital may, in fact, be regime-sustaining rather than pro-democratic. It is

necessary to consider the contextual factors that impact the effect of the

social capital potentially derived from communal religious practice.

7

Survey Experiment and Lebanese Case Study

The primary case that I plan to study in order to examine the link be-

tween religion and democracy more closely is Lebanon. One of my main

data sources will be an original survey (including several embedded experi-

ments) conducted on a nationally-representative sample in Lebanon. I hope

to conduct this survey in the summer or early fall of 2013. Lebanon is an

ideal candidate for inclusion in this study for several reasons. First, and most

obviously, Lebanon’s sectarian divisions ensure that religion plays a crucial

role in politics; Harris (2009, p. 9) states that religion is “the primary char-

acteristic” of Lebanon. The importance of religion in Lebanese politics is

beyond doubt, suggesting that a project focusing on religion’s political influ-

ence will not, at least, run into problems finding religious voices in public life.

Religion is not only important because of Lebanon’s institutional structures,

however. Religion, in its various forms, continues to play an important role

in everyday life for a large number of Lebanese citizens. Deeb (2006, pp. 105-

106) quotes a young Shi’ite volunteer in Beirut’s southern suburbs as saying

the following about communal prayer: “Why is prayer in a group? Because

you find everyone, poor, rich, educated, uneducated, all are standing in a

single line praying together.” Religion has a powerful effect on social and

political life in Lebanon independently of its institutional effects. Second,

Lebanon’s semi-democratic history1 provides two important assurances: 1)

citizens have some experience of democratic politics (or something like it);

but 2) democracy is not, in the words of Linz and Stepan (1996, p. 5), “the

1Freedom House (2011) plainly notes that “Lebanon is not an electoral democracy.”

8

only game in town.” Thus, there should be some genuine variation in sup-

port for democracy. Existing surveys in Lebanon suggest that, indeed, while

a majority of Lebanese citizens express support for liberal democracy, strong

democratic attitudes are far from unanimous. Lebanon’s torturous political

history creates a useful setting for analysis of attitudes towards democracy;

its history of relatively liberal policies sets it apart from virtually all of its

neighbors, but Lebanon has also experienced some of the most violent civil

strife in the region. Despite these challenges, Fawaz (2009, p. 33) suggests

that Lebanon “might be the only ray of light coming out of this region.”

Each of these factors makes Lebanon a useful area to study the ambiguous

link between religion and democracy.

Table 1: Communal Prayer and Support for Democracy, by Sect

MuslimsStrongly Supports Democracy Prays at Mosque

No Yes

No 156 7969.03% 50.97%

Yes 70 7630.97% 49.03%

Total 226 155100.00% 100.00%

ChristiansStrongly Supports Democracy Prays at Church

No Yes

No 34 20233.33% 47.42%

Yes 68 22466.67% 52.58%

Total 102 426100.00% 100.00%

9

Using the first wave of the Arab Barometer, I have discovered a puzzling

trend: communal religious practice is associated with higher levels of sup-

port for democracy among Lebanese Muslims, but lower levels of support

among Lebanese Christians. Table 1 displays the levels of (strong) support

for democracy comparing respondents who engage in communal religious

practice with those who do not, separating by sect. Muslims who pray at

mosque are over 18 percentage points more likely to support democracy than

Muslims who do not pray at mosque. Conversely, Christians who pray at

church are 14 percentage points less likely to support democracy than are

Christians who do not pray at church. Thus, communal prayer appears to

have opposite effects on the sects.

Figure 1 displays the changes in predicted probability of support for

democracy due to communal religious practice from a basic logistic regres-

sion. As this figure demonstrates, the model predicts that Christians who

pray at a church are about 15 percentage points less likely to support democ-

racy than are Christians who do not pray at a church. On the other hand,

Muslims who pray at a mosque are almost 20 percentage points more likely to

support democracy than are their non-practicing coreligionists. Thus, when

controlling for these basic characteristics of respondents, the results are at

least as strong as a simple comparison of means across groups.

What explains this puzzling trend? I hypothesize that group interests

explain the divergent effects of religious practice on attitudes towards democ-

racy in Lebanon. Figure 2 displays the average monthly income of individuals

in each of Lebanon’s three major sects using data from the Arab Barometer.

As the figure demonstrates, Christians are substantially wealthier on average

10

Figure 1: Change in Predicted Probability of Support for Democracy fromReligious Practice

Diff

eren

ce in

Pre

dict

ed P

roba

bilit

y of

Sup

port

for

Dem

ocra

cy

●

●

−0.

3−

0.2

−0.

10.

00.

10.

20.

3

Christians Muslims

Note: Figure displays changes in predicted probability of support for democracy comparingan average respondent who engages in communal practice with an average respondent whodoes not, i.e. Pr(Support|Practice) − Pr(Support| ∼ Practice). The predictions comefrom separate logistic regressions for each sect, controlling for age, education, gender,and internet use. Results are similar when controlling for personal piety (importance ofspouse’s prayer), income, and association membership. Bars represent 95% confidenceintervals.

than are Muslims, with Shi’ites being slightly poorer on average than Sun-

nis. In fact, this figure indicates that the average Christian in the sample

earns over $6,000 more per year than the average Muslim, a 40% difference.2

This inequality is magnified further by perceptions of inequality: Christians

have historically been perceived (by both themselves and by others) as the

privileged group in Lebanon, while Muslims have historically had less access

2Following Horowitz (1985, p. 22), we can think of the confessions in Lebanon as“ranked groups”” since there is considerable overlap between class and group identity.

11

to both wealth and social resources (Mackey 1989, pp. 13-14). Historical

perceptions, of course, remain sticky – even if Muslim wealth were to in-

crease relative to that of Christians, it would likely take some time before

perceptions of inequality were updated completely, particularly since Chris-

tian “advantages” in Lebanon stretch back at least several centuries (Fawaz

2009, pp. 27-28). The prevalence of kin-based organizations as providers

of social insurance likewise suggests that the economic fates of members of

particular religious groups in Lebanon are linked together in significant ways

(Baylouny 2010). Since these kin ties overlap considerably with sect, mem-

bers of the same sect (usually in smaller communities) frequently rely on each

other (through such associations) for social services. These structures pro-

vide a further incentive for Lebanese citizens to consider group rather than

only individual interests.

Figure 2: Average Monthly Income, by Sect

Christian Sunni Shi'ite

Ave

rage

Mon

thly

Inco

me

(In

LBP

)

050

0,00

01,

500,

000

2,50

0,00

0

Recent literature in the field of comparative democratization suggests

12

that democracies tend to redistribute more than non-democracies, and that

wealthy elites therefore have incentives to prevent democratic transition while

the poor have incentives to favor it (Boix 2003; Acemoglu and Robinson

2006). I hope to synthesize this literature with the literature on religion

described above. Given the available evidence, it is understandable that

religion might have a pro-democracy effect among Lebanese Muslims but

an anti-democracy effect among Muslims. If religion promotes pro-in-group

behavior, then it should also push individuals’ attitudes about democracy

in a direction dictated by the status of their religious group. Since Muslims

tend to be much poorer than Christians, religion should therefore induce

pro-democratic sentiments among Muslims, while having the opposite effect

among Christians.

My proposed survey will confront this claim in a more rigorous manner

than is possible using existing data. Respondents will be asked about their

perceptions of linked fate as well as the importance of religion in public

life, in order to examine exactly how Lebanese citizens believe religion and

politics to be related, and if religious piety (in its various forms) affects these

perceptions. Moreover, the experimental portion of the survey will help to

“exogenize” piety to some extent. While we cannot assign respondents to

“religious” or “not religious” backgrounds, we can hope to trigger a limited,

but considerable religious response in order to ascertain how religion causes

certain political attitudes and behaviors. Thus, these new data will prove

a much more thorough test of whether or not the proposed relationship is

causal.

My questionnaire will include a more thorough battery of questions about

13

religious beliefs and behaviors than do existing surveys of Lebanon, including

questions not only about individual religiosity, but also about characteristics

of the respondent’s local religious community, their beliefs about the state

of their religious group in the country, and notions of “linked fate” (Harris

1994). Furthermore, I will include several experiments in this questionnaire.

The primary experiment for the purposes of this study will be a religious

priming experiment. At its most basic level, this type of experiment at-

tempts to “prime” respondents with religious cues, thereby inducing a more

religious state of mind, in order to create exogenous variation in the religios-

ity of respondents. Existing studies employing religious priming frequently

use simple word games to prime respondents: for example, respondents are

told to unscramble sentences in which the order of words had been changed,

or are told to identify which word among a set of words is not actually a word.

In both cases, the control group is given neutral words (not clearly related to

religion or any other concept), while the treatment group is given religiously-

themed words. The aim of these treatments is to trigger a more “religious”

state of mind in respondents. While religious piety is clearly endogenous to

any number of political variables, and cannot be created or eliminated in a

laboratory setting, a number of experiments in social psychology and behav-

ioral economics have found religious priming to be effective. Wenger (2004)

finds that religious priming “automatically” activates religious concepts, sug-

gesting that the primes demonstrate the desired effect on respondents’ states

of mind.

This category of experiments has identified a number of interesting links

between religion and various outcomes, both positive and negative. Shariff

14

and Norenzayan (2007) find that religious priming induces pro-social behav-

ior in an economic game, a result that has been replicated elsewhere (Pichon,

Boccato and Saroglou 2007). Randolph-Seng and Nielsen (2007) observe that

religious priming reduces cheating, and Carpenter and Marshall (2009) find

that religious priming reduces moral hypocrisy. However, Johnson, Rowatt

and LaBouff (2010) find that priming Christian religious concepts increases

racial prejudice, and Toburen and Meier (2010) note that priming God con-

cepts increases anxiety (as well as task persistence). Likewise, Saroglou,

Corneille and Van Cappellen (2009) observe that religious priming promotes

submissive behaviors and attitudes. The wide range of findings regarding

the impact of religious priming is surprising considering the relative weak-

ness of the treatments involved. These findings suggest that religious priming

techniques are powerful enough to demonstrate impacts on respondents for

a variety of outcomes.

It is possible, however, that the effect of religious priming will vary across

subgroups. Van Cappellen, Corneille, Cols and Saroglou (2011) find that reli-

gion’s tendency to promote “submissiveness” is only present for respondents

who were already predisposed to submissiveness. Inzlicht and Tullett (2010)

note an interesting result: religious priming decreases neurological response

to errors among believers, but actually increases this response among non-

believers. Likewise, McKay, Efferson, Whitehouse and Fehr (2011) find that

religious priming increases punishment of unfair behavior, but only among

subjects who had previously donated to a religious charity. These results

indicate an important possibility: among respondents with low levels of reli-

giosity, religious primes may demonstrate no effect or even the opposite effect

15

compared to what would be present for high-religiosity respondents. It is rea-

sonable to expect, therefore, that religious priming may not have a universal

effect across subgroups of respondents.

Studies employing religious priming that have paid attention to the poten-

tially non-uniform effect of such priming have presented a number of findings

that will be important in guiding this project. Johnson, Rowatt and LaBouff

(2012) find that religious priming promotes both in-group favoritism and

out-group derogation, suggesting that these primes do not simply lead to

greater pro-social behavior across the board. Using a field experiment in

Ghana in which respondents played an economic game, Parra (2011) finds

that religious priming makes the religious affiliation of the other player a

major factor in determining subjects’ behavior, leading to inter-group bias.

Similarly, Blogowska and Saroglou (2011) note that religion promotes pro-

social behavior only towards in-group members. These findings suggest an

important consideration: religion appears to induce in-group favoritism, and

should therefore lead to political attitudes more favorable towards in-groups

than out-groups.

Historical Cases

While the survey experiment will be highly useful for identifying the causal

effects of religion (as captured by religious priming), it is also important to

consider the role played by religion in actual democratization processes. In

other words, it is crucial to engage in some process tracing (Eckstein 1975;

Gerring 2007) in order to understand how religion has led to different political

16

outcomes in real-world cases. As Coppedge (2012, p. 115) observes, most

of what scholars of comparative politics know about democratization comes

from in-depth analysis of particular cases. However, such techniques also pose

inferential difficulties, so case selection must be done carefully (Coppedge

2012).

It is worth asking, therefore, how we might best approach the question

of the Janus-faced nature of religion and democracy in the context of histor-

ical case studies. In order to minimize the methodological difficulties from

small-N comparative analysis and maximize their analytical leverage, I pro-

pose a comparative study of a handful of Latin American countries during

the “third wave” of democratization (Huntington 1991). Such a design is

ideal because of the considerable number of major factors that are held con-

stant. First, and perhaps most importantly, the dominant religious tradition

in these countries is Roman Catholicism; consequently, we can dismiss the

possibility that different religious traditions led to different results in these

cases3. Second, “period effects” should be mostly equal: since several Latin

American countries made transitions to democracy in a relatively short pe-

riod of time, it is unlikely that the differential impacts of religion in each

of these cases can be attributed to gradual change over time. Relatedly,

this design eliminates the possibility that Vatican II was responsible for a

complete overhaul of the Church’s approach to democracy: since all of these

transitions occurred well after the completion of Vatican II, any remaining

variation in the Church’s stance towards democracy cannot be explained by

3As Hagopian (2009, p. 6) notes, “when the Latin American Catholic Church laysclaim to a regional Catholic culture, it is essentially correct.”

17

the theological changes resulting from Vatican II itself.

It is important to note that in this section, as in all other sections of the

dissertation, the outcome of interest is not democratization or democracy

per se. Rather, I am interested in the determinants of the link between reli-

gion and democracy. Thus, for the Latin American cases, I wish to explain

why the Catholic Church enthusiastically supported democratization in some

countries, but was much more conflicted in others. This variation has been

identified by a number of scholars, including Johnston and Figa (1988) and

Mantilla (2010), but no consensus has emerged on why the Church actively

supported democracy in some cases but not in others. Johnston and Figa

(1988) cite Church resources/organization and the level of development of

the country in question as the key determinants of Church behavior, while

Mantilla (2010) argues that cultural frames, mobilizing resources, and politi-

cal opportunity structures explain the Church’s response to democratization

movements.

My working hypothesis in this portion of the project draws from the

hypothesis I have proposed for Lebanon. In this case, I suggest that the vari-

ation in the Catholic Church’s support for democracy during the Third Wave

in Latin America is attributable to interests: when the Church’s interests in

a given country were better-suited to democracy, it would be more inclined

to push for democratization. Thus, the logic outlined above can be applied

at more than one level: leaders of religious communities may be influenced

by similar considerations as their followers when it comes to support for

democracy. In both cases, it is conceivable that individuals (and groups) will

develop instrumental regime preferences; if a regime will benefit their group,

18

they will support it. Religion, therefore, has an ambiguous effect on support

for democracy. When the interests of the religious community in question

are aligned with those of democracy, then religion will promote democracy;

and when the group’s interests clash with democracy, religion will impede

democracy.

I expect that the mechanisms of this relationship will work somewhat

differently in the Latin American cases, particularly because they are taking

place primarily at the level of Church hierarchy rather than at the level of

the individual believer. In these cases, the calculus made by Church leaders

is as follows: will Church interests be so damaged by democratization that

supporting the incumbent dictator will be worth the cost of alienating adher-

ents? Gill (1998) has identified a similar choice facing Church leaders in these

settings. On the one hand, Church leaders presumably wish to preserve the

benefits in privileges they often receive through arrangements with author-

itarian governments (such as state subsidies, special access to charities and

media sources, and control over education; see Hagopian 2009, pp. 28-35).

On the other hand, supporting repressive governments comes at a cost: such

behavior may push members towards other faiths (Gill 1998). Under this

framework, the decision to support or oppose the incumbent regime comes

down to the relative costs of losing privileges or losing members.

This portion of the dissertation will consist largely of qualitative data

gathered from secondary historical sources as well as relevant statements

from Latin American bishops in and around the periods when democratic

transitions were on the horizon. Since the goal of this section is different

from that of the survey experiment, the techniques must also be different.

19

Tentatively, I plan to pair two cases in a “most-similar” design (Przeworski

and Teune 1970). The motivation behind this design is to “control” for other

variables as much as possible while maintaining variation on the dependent

variable (in this case, the response of religious groups to democratization

movements). Two likely candidates for case selection are Argentina and

Brazil. Although these countries are not by any means identical in terms

of other potentially confounding variables (no two countries could be), they

share many key social, economic, and political characteristics, and have been

paired together in other studies employing a most-similar systems design

(see Manzetti 1999; Weyland 2002). Despite these similarities, the Catholic

Church’s approach to democracy in these countries was quite different: while

Brazilian bishops were outspoken in their opposition to the authoritarian

regime, in Argentina the Catholic leadership was much more reserved and,

in some cases, even vocally supportive of military rule (Klaiber 1998; Mantilla

2010). These divergent responses, I suspect, are attributable to the different

incentives facing the Brazilian and Argentine churches.

Conclusion

My dissertation aims to shed light on the “big question” of religion’s support

for (or opposition to) democracy in several contexts. While the issues facing

individual countries are numerous and diverse, I believe that the logic out-

lined above can be applied usefully to a variety of cases. As a supplement

to the survey experiment and case studies described in previous sections, I

will conduct a cross-national statistical analysis examining how religion fuels

20

support for or opposition to democracy in various contexts, and how these

contextual factors influence this relationship. Straightforward tests of the

plausibility of the claims I have made are possible using data from the Afro-

barometer as well as the World Values Survey. These datasets provide a

considerable amount of public opinion data from countries where: a) democ-

racy is not consolidated; and b) religion is politically salient. It is possible,

then, to use these data in order to determine the external validity of my

hypotheses, if not a perfectly rigorous test of their mechanisms.

Taken together, I believe that these different research strategies can com-

bine in such a way as to contribute to several literatures. First, my findings

will be relevant to scholars of democratization, who have so far only been

able to speculate about the role of religion in mobilizing public opinion for

or against democracy (see Linz and Stepan 1996). This project will also

address some of the interests of political behavior scholars, who have long ar-

gued about the types of values that religion underwrites. Finally, I anticipate

that this project will speak to the interests of scholars of ethnicity, since it

aims to explain how social divisions can promote or undermine democracy. I

do not aim to make any claims about the contents of any particular religions

(in fact, I have suggested that such differences between faiths are perhaps

less important than commonly claimed); rather, I wish to suggest that the

incentives faced by religious groups in the political settings under which they

operate influences their approaches to democracy in significant ways.

21

References

Acemoglu, Daron and James Robinson. 2006. Economic Origins of Dicta-torship and Democracy. New York: Cambridge University Press. [13]

Ardıc, Nurullah. 2012. “Understanding the Arab Spring: Justice, Dignity,Religion and International Politics.” Afro Eurasian Studies 1(1):8–52. [2]

Baylouny, Anne Marie. 2010. Privatizing Welfare in the Middle East: KinMutual Aid Associations in Jordan and Lebanon. Bloomington: IndianaUniversity Press. [12]

Bellin, Eva. 2008. “Faith in Politics: New Trends in the Study of Religionand Politics.” World Politics 60:315–347. [1]

Benhabib, Seyla. 2011. “The Arab Spring: Religion, Revolution, and thePublic Sphere.” Public Sphere Forum . [2]

Blogowska, Joanna and Vassilis Saroglou. 2011. “Religious Fundamentalismand Limited Prosociality as a Function of the Target.” Journal for theScientific Study of Religion 50(1):44–60. [16]

Boix, Carles. 2003. Democracy and Redistribution. New York: CambridgeUniversity Press. [13]

Campbell, David. 2004. “Acts Of Faith: Churches And Political Engage-ment.” Political Behavior 26(2):155–180. [4]

Carpenter, Thomas and Margaret Marshall. 2009. “An Examination of Re-ligious Priming and Intrinsic Religious Motivation in the Moral HypocrisyParadigm.” Journal for the Scientific Study of Religion 48(2):386–393. [15]

Ciftci, Sabri. 2010. “Modernization, Islam, or Social Capital: What Ex-plains Attitudes Toward Democracy in the Muslim World?” ComparativePolitical Studies 43(11):1442–1470. [5]

Coppedge, Michael. 2012. Democratization and Research Methods. New York:Cambridge University Press. [17]

Deeb, Lara. 2006. An Enchanted Modern: Gender and Public Piety in Shi’iLebanon. Princeton: Princeton University Press. [8]

Driskell, Robyn, Elizabeth Embry and Larry Lyon. 2008. “Faith and Poli-tics: The Influence of Religious Beliefs on Political Participation.” SocialScience Quarterly 89(2):294–314. [4]

22

Eckstein, Harry. 1975. Case Studies and Theory in Political Science. InHandbook of Political Science: Strategies of Inquiry, ed. F.I. Greensteinand N.W. Polsby. Boston: Addison-Wesley pp. 79–138. [16]

Fawaz, Eli. 2009. What Makes Lebanon a Distinctive Country? In Lebanon:Liberation, Conflict, and Crisis, ed. Barry Rubin. New York: PalgraveMacmillan pp. 25–34. [9, 12]

Freedom House. 2011. Freedom in the World: Lebanon. Technical report. [8]

Gerring, John. 2007. Case Study Research: Principles and Practices. NewYork: Cambridge University Press. [16]

Gill, Anthony. 1998. Rendering Unto Caesar: The Catholic Church and theState in Latin America. Chicago: University of Chicago Press. [19]

Hagopian, Frances. 2009. Religious Pluralism, Democracy, and the CatholicChurch in Latin America. Notre Dame, IN: University of Notre DamePress. [17, 19]

Harris, Frederick. 1994. “Something Within: Religion as a Mobilizer ofAfrican-American Political Activism.” Journal of Politics 56(1):42–68. [14]

Harris, William. 2009. Reflections on Lebanon. In Lebanon: Liberation, Con-flict, and Crisis, ed. Barry Rubin. New York: Palgrave Macmillan pp. 9–24.[8]

Hertzke, Allen. 2009. The Catholic Church and Catholicism in Global Pol-itics. In Routledge Handbook of Religion and Politics, ed. Jeffrey Haynes.London and New York: Routledge pp. 48–63. [1]

Horowitz, Donald. 1985. Ethnic Groups in Conflict. Berkeley: University ofCalifornia Press. [11]

Huntington, Samuel. 1991. The Third Wave: Democratization in the LateTwentieth Century. Norman, OK: University of Oklahoma Press. [1, 17]

Inzlicht, Michael and Alexa Tullett. 2010. “Reflecting on God: ReligiousPrimes Can Reduce Neurophysiological Response to Errors.” PsychologicalScience 21(8):1184–1190. [15]

Jamal, Amaney. 2007. Barriers to Democracy: The Other Side of SocialCapital in Palestine and the Arab World. Princeton: Princeton UniversityPress. [7]

23

Johnson, Megan, Wade Rowatt and Jordan LaBouff. 2010. “Priming Chris-tian Religious Concepts Increases Racial Prejudice.” Social Psychologicaland Personality Science 1(2):119–126. [15]

Johnson, Megan, Wade Rowatt and Jordan LaBouff. 2012. “Religiosityand Prejudice Revisited: In-Group Favoritism, Out-Group Derogation, orBoth?” Psychology of Religion and Spirituality 4(2):154–168. [16]

Johnston, Hank and Jozef Figa. 1988. “The Church and Political Oppo-sition: Comparative Perspectives on Mobilization Against AuthoritarianRegimes.” Journal for the Scientific Study of Religion 27(1):32–47. [1, 18]

Jones-Correa, Michael A. and David L. Leal. 2001. “Political Participation:Does Religion Matter?” Political Research Quarterly 54(4):751–770. [7]

Klaiber, Jeffrey. 1998. The Church, Dictatorships, and Democracy in LatinAmerica. Maryknoll, NY: Orbis Books. [20]

Linz, Juan and Alfred Stepan. 1996. Problems of Democratic Transition andConsolidation: Southern Europe, South America, and Post-CommunistEurope. Baltimore: Johns Hopkins University Press. [8, 21]

Lynch, Marc. 2012. The Arab Uprising: The Unfinished Revolutions of theNew Middle East. New York: Public Affairs. [3]

Mackey, Sandra. 1989. Lebanon: A House Divided. New York: W.W. Norton.[12]

Mantilla, Luis Felipe. 2010. “Mobilizing Religion for Democracy: ExplainingCatholic Church Support for Democratization in South America.” Politicsand Religion 3:553–579. [18, 20]

Manzetti, Luigi. 1999. Privatization South American Style. New York: Ox-ford University Press. [20]

Marx, Karl. 1970 [1843]. Marx’s Critique of Hegel’s Philosophy of Right. NewYork: Cambridge University Press. [6]

McKay, Ryan, Charles Efferson, Harvey Whitehouse and Ernst Fehr. 2011.“Wrath of God: Religious Primes and Punishment.” Proceedings of theRoyal Society of Biological Sciences 278:1858–1863. [15]

Meyer, Katherine, Daniel Tope and Anne Price. 2008. “Religion And Sup-port For Democracy: A Crossnational Examination.” Sociological Spec-trum 28:625–653. [5]

24

Norris, Pippa and Ronald Inglehart. 2004. Sacred and Secular: Religion andPolitics Worldwide. New York: Cambridge University Press. [1, 7]

O’Donnell, Guillermo and Philippe Schmitter. 1986. Transitions from Au-thoritarian Rule: Tentative Conclusions about Uncertain Democracies.Baltimore: Johns Hopkins University Press. [2]

Parra, Juan Carlos. 2011. “The effects of religion on social cooperation:Results from a field experiment in Ghana.” Working Paper . [16]

Philpott, Daniel. 2007. “Explaining the Political Ambivalence of Religion.”American Political Science Review 101(3):505–525. [1, 5]

Pichon, Isabelle, Giulio Boccato and Vassilis Saroglou. 2007. “NonconsciousInfluences of Religion on Prosociality: A Priming Study.” European Jour-nal of Social Psychology 37:1032–1045. [15]

Przeworski, Adam and Henry Teune. 1970. The Logic of Comparative SocialInquiry. New York: Wiley-Interscience. [20]

Putnam, Robert. 1993. Making Democracy Work: Civic Traditions in Mod-ern Italy. Princeton: Princeton University Press. [6]

Putnam, Robert. 2000. Bowling Alone: The Collapse and Revival of Amer-ican Community. New York: Simon and Schuster. [6, 7]

Putnam, Robert and David Campbell. 2010. American Grace: How ReligionDivides and Unites Us. New York: Simon and Schuster. [4]

Ramet, Pedro. 1989. Religion and Nationalism in Soviet and East EuropeanPolitics. Durham: Duke University Press. [1]

Randolph-Seng, Brandon and Michael Nielsen. 2007. “Honesty: One Ef-fect of Primed Religious Representations.” International Journal for thePsychology of Religion 17(4):303–315. [15]

Saroglou, Vassilis, Olivier Corneille and Patty Van Cappellen. 2009. “Speak,Lord, Your Servant is Listening: Religious Priming Activated SubmissiveThoughts and Behaviors.” International Journal for the Psychology of Re-ligion 19:143–154. [15]

Shariff, Azim and Ara Norenzayan. 2007. “God is Watching You: PrimingGod Concepts Increases Prosocial Behavior in an Anonymous EconomicGame.” Psychological Science 18(9):803–809. [14]

25

Smidt, Corwin. 1999. “Religion and Civic Engagement: A ComparativeAnalysis.” Annals of the American Academy of Political and Social Science565:176–192. [4]

Stepan, Alfred. 2000. “Religion, Democracy, and the “Twin Tolerations”.”Journal of Democracy 11(4):37–57. [4]

Tessler, Mark. 2002. “Islam and Democracy in the Middle East: The Impactof Religious Orientations on Attitudes toward Democracy in Four ArabCountries.” Comparative Politics 34(3):337–354. [4]

Toburen, Tina and Brian Meier. 2010. “Priming God-Related Concepts In-creases Anxiety and Task Persistence.” Journal of Social and Clinical Psy-chology 29(2):127–143. [15]

Uslaner, Eric. 2002. “Religion and Civic Engagement in Canada and theUnited States.” Journal for the Scientific Study of Religion 41(2):239–254.[7]

Van Cappellen, Patty, Olivier Corneille, Stephanie Cols and VassilisSaroglou. 2011. “Beyond Mere Compliance to Authoritative Figures: Re-ligious Priming Increases Conformity to Informational Influence AmongSubmissive People.” The International Journal for the Psychology of Re-ligion 21:97–105. [15]

Wenger, Jay. 2004. “The Automatic Activation of Religious Concepts: Im-plications for Religious Orientations.” International Journal for the Psy-chology of Religion 14(2):109–123. [14]

Weyland, Kurt. 2002. The Politics of Market Reform in Fragile Democracies:Argentina, Brazil, Peru, and Venezuela. Princeton: Princeton UniversityPress. [20]

Wright, Robin. 2011. Rock the Casbah: Rage and Rebellion Across the IslamicWorld. New York: Simon and Schuster. [3]

26