American Political Science Review Vol. 101, No. 3 August 2007

DOI: 10.1017/S0003055407070372

Explaining the Political Ambivalence of ReligionDANIEL PHILPOTT University of Notre Dame

This essay takes on the broad question—–what explains the political pursuits of religious actors?—–byexploring two powerful influences on these pursuits. The first is differentiation, or the degree ofautonomy between religious actors and states in their basic authority. The second is political

theology, the set of ideas that religious actors hold about political authority and justice. Through globalcomparisons across religions, regions, and states, it seeks to establish the effect of both influences on twopolitical pursuits in which religion’s role is hotly debated today: support for democratization and politicalviolence, including communal violence and terrorism. It concludes with lessons learned commonly fromthe analysis of both pursuits.

In 1979, an Islamic revolution in Iran confoundedAmerican foreign policy and inspired an Islamicresurgence in Afghanistan, Kashmir, the Middle

East, and elsewhere. In Turkey, after over sevendecades of rule by a secular nationalist military regime,an Islamic party won elections in 2002, deepeningdemocracy and advocating Turkey’s entry into theEuropean Union. In the 1990s, after four decades ofrule, India’s secular Congress Party yielded power toa Hindu nationalist party that promoted religious lawsand discourse and provoked Hindu-Muslim violence.The teachings of the Catholic Church’s Second VaticanCouncil of 1962 to 1965 encouraged subsequent demo-cratization in the Philippines, Brazil, and Poland, butnot in Rwanda, Argentina, or Hungary. In Sri Lanka, alack of separation between sangha and state has fueledwar between Buddhists and Hindu Tamils, whereasBuddhism in Taiwan and South Korea has promotedhuman rights and religious tolerance. Over the pastgeneration, evangelical Protestants have become apowerful voting bloc in the United States, Brazil,Guatemala, and Kenya.

Defying the erstwhile dominance of the seculariza-tion thesis among western intellectuals, religion haswaxed in its political influence over the past generationin every region of the globe except perhaps WesternEurope (see Berger 1999; Casanova 1994; Stark 1999).

Daniel Philpott is Associate Professor, Department of Politi-cal Science, University of Notre Dame, Notre Dame, IN 46556(philpott.1@nd.edu).

This article was commissioned by the Religion in Global Politicsproject, which is based at the Harvard University Academy for Inter-national and Area Studies, and funded by the Weatherhead Centerfor International Affairs at Harvard University as well as by theSmith Richardson Foundation. I thank these sources of support, aswell as the Edmond J. Safra Center for Ethics at Harvard University.For helpful comments, I thank participants in the Lansing Lee Sem-inar at the University of Virginia, the Religion, Politics, and PublicLife Faculty Seminar Series at Harvard University, the Religion, Po-litical Economy, and Society Seminar at Harvard University; as wellas Jonathan Fox, Bryan Hehir, Samuel Huntington, James Kurth,David Little, Diana Philpott, and Monica Duffy Toft. Extensive con-versations with Timothy Samuel Shah were integral in developingthe paper’s ideas. For research reports I thank Edgar Chen, RobertDowd, Laurie Johnston, Robert Portada, and Eli Sasaran; and forresearch assistance, Stephanie Aberger, Julia Fitzpatrick, ColleenGilg, Stephen Joyce, Benjamin Kaplan, Jessica Lieberman, KevinLoria, Kevin McCormick, Sarah Mehta, Sarah Miller, Paul Nauert,Alexander Pickett, John Skakun, Carolyn Sweeney, Erin Urquhart,and Afiya Whisby.

What form this influence takes has become the subjectof heated debates in the popular media, academia, andpolicy circles. Most voluble are today’s polemicists, forwhom rising religion amounts either to a growth in fun-damentalism, violence, and intolerance or to a welcomebulwark against cultural decadence. More measuredanalyses, along with the previous examples of religiouspolitics, point instead, in Appleby’s phrase (2000), toan “ambivalence of the sacred” that eludes the biva-lence of the culture wars. Religion devastates not onlyNew York skyscrapers but also authoritarian regimes;it constructs not only bellicose communal identities butalso democratic civil society.

Still, is it possible to say anything general? Whatexplains why religion becomes either violent or irenic,a source of terrorism or a contributor to the ruleof law? Scholars have offered a bewildering array ofexplanations for the politics of religions: their theol-ogy, their national and ethnic identities, colonialism,their historical relationships to political authorities,their competition with other religions, their grievances,and a multitude of economic, political, and demo-graphic factors. Yet certain patterns continue to resur-face. Certain features of religious communities proveto be the most powerful bellwethers of their politicalorientation.

Here, two concepts are proposed as particularlypromising. The first comes from modern sociology, buthas earlier antecedents. Known as “differentiation,”it describes how religious and political authority isrelated. Does one dominate the other? Do they col-laborate? Or are they institutionally separate? Thesecond is political theology. Like parties and classes,religious bodies contain shared ideas about legitimatepolitical authority. Situated in core doctrinal teachingsyet adapted to the circumstances of time and place,these ideas prescribe one or another posture towardthe state.

Differentiation and political theology, institutionsand ideas—–these core concepts then yield causalclaims: this kind of differentiation and that politicaltheology lead a religious body to a certain sort of po-litical involvement.1 This essay explores how differen-tiation and political theology shape religious actors’

1 For an earlier version of the thesis, applied to European integrationand the 1989 democratic revolutions in Europe, see Philpott and Shah2006.

505

Explaining the Political Ambivalence of Religion August 2007

pursuit of two ends in particular: democratization andpolitical violence. Both phenomena stand at the centerof today’s heated political debates, both in the UnitedStates and around the world. The proper role of reli-gion in democracy has become a central national issuein Turkey, India, Nigeria, the Palestinian Authority, In-donesia, France, and Poland, and an important issuein the United States and many Latin American states.Now, more than ever before, the United States’ goalof promoting democratization globally depends on anunderstanding of religion’s influence on democracy,not least in Iraq and Afghanistan, where Americanarmed forces are most heavily engaged. More gener-ally, the issue of Islam’s compatibility with democracy,in both its empirical and its normative senses, findspassionate voices on all sides both in the West andthroughout the Islamic world. As for political violence,the role of religion in triggering communal conflictsin places like the former Yugoslavia, Israel, Palestine,Northern Ireland, and Lebanon has been at issue forseveral decades, even before ethnic and religious vio-lence surged globally after the end of the Cold War.Since the attacks of September 11, 2001, religion’sinfluence on terrorism has become a central contro-versy in every country where terrorism has become anissue.

This essay plumbs religion’s relationship to democ-racy and political violence through comparisons acrossglobal regions, religions, and states. What it does notprobe systematically is the influence of differentiationand political theology relative to other shapers of thepolitics of religions—–economic, demographic, cultural,political, or certain traits of religious actors themselveslike their internal hierarchy and unity or the level oftheir members’ commitment. Nor does it test the in-fluence of religious communities on democratizationand political violence relative to other possible causes.Scholars have explained democracy, for instance, asthe product of economic development, ethnolinguisticunity, imperial legacies, and many other factors, includ-ing religion (Geddes 1999; Huntington 1991). Furtherresearch involving such tests would deepen our under-standing of both the causes and the effects of religiousactors’ politics.

But in identifying two major influences on the pos-ture of religious actors toward two vital contemporaryissues in global politics, this essay takes an importantstep toward systematic understanding. In the end, itfinds not only that differentiation and political theo-logy shape religion’s role in each issue area but alsothat their influence in both areas is conceptually linked.By and large, the sorts of differentiation and politicaltheology that lead religious actors to encourage democ-racy are the same ones whose absence tends to result inreligious support for political violence, and vice versa.It is an important result for academic analysts andpolicymakers alike, providing insight into how ideasand institutions make religion a source of promise ora root of peril. The essay’s conclusion draws out thislink and its implications for the normative stability ofdemocracy and the destructive instability of politicalviolence.

HOW RELIGION MEETS THE STATE

Political scientists are most at home when they de-scribe states—–how states make their decisions, howthey interact, and who influences them. They are farless nimble with religions, which are far older than thestate, make claims far larger than the state, entail amembership far wider than most states, and indeedoften accept the legitimacy of states only condition-ally, according to their deepest commitments. Neitherdo religions fit easily into disciplinary boundaries likethat between comparative politics and internationalrelations. Rather, they are transnational. But unlikeKeck and Sikkink’s (1998) transnational networks, reli-gions do not specialize in single issues, but make claimsabout every realm, and even the ultimate condition, oflife. Nor are they limited to networks of activists, butcomprise enormous populations—–1.3 billion Muslims,2.1 billion Christians, 14 million Jews, 900 million Hin-dus, and 376 million Buddhists (Hunter 2007). Express-ing this scale far better is Rudolph’s (1997) concept of“transnational civil society.”

But for all of their depth, width, and breadth, reli-gions do not usually act singly or comprehensively intheir politics. Some tenets of a political doctrine willbe shared across an entire religion. In some religions,on some occasions, a single authority like the Popewill speak for the entire body. Usually, though, reli-gious tenets will be translated and enacted by moreparticular entities, including regional or national levelreligious authorities, masses of believers in a region ornation, ordinary clerics, clerical orders, religious politi-cal parties, or organizations of “lay” believers (Byrnes2001). Sometimes these actors will take cues from theirhigher-ups; sometimes they will act in isolation from,and perhaps at odds with, their religious brethren else-where. Even within the ranks of religious bodies, mem-bers will differ over doctrine and policy. Still, general-ization is possible: at some level of collectivity, leaderswill speak in the name of their followers; a body’smembers may largely tend this way or that. “Religiousactor,” then, denotes any individual or collectivity, localor transnational, who acts coherently and consistentlyto influence politics in the name of a religion.

Differentiation

The ambition and the divergence in the claims of re-ligion and state combine to create potential for fric-tion. One professes truths about the moral order ofthe universe, whereas the other asserts sovereignty, orsupreme authority within a territory, and has succeededin becoming the only form of polity ever to replicateitself across the entire land surface of the globe. Thistension has been resolved from state to state throughsundry balances of authority, with the extremes oftheocracy and atheistic dictatorship flanking a widemiddle ground of diverse schemes and settlements.

The principle that best describes these balances,borrowed from sociologists, is differentiation (Berger1967; Martin 1978). Defined here, differentiation is thedegree of mutual autonomy between religious bodies

506

American Political Science Review Vol. 101, No. 3

and state institutions in their foundational legal author-ity, that is, the extent of each entity’s authority overthe other’s basic prerogatives to hold offices, chooseits officials, set its distinctive policies, carry out its ac-tivities, in short, to govern itself. Religion and staterelationships with a high degree of differentiation cansimply be called “differentiated”; and ones with a lowdegree, “integrationist.”

There are four dimensions by which differentiationis realized to greater and lesser degrees. First, doesthe state grant a singe religion constitutional statusas the official one? Pakistan’s 1956 constitution, forinstance, declares it an Islamic republic, affirming di-vine sovereignty and allowing no law “repugnant toIslam.” Western European countries like England andDenmark have established churches, whereas otherslike Austria, Belgium, and Germany officially recog-nize some religious bodies but not others. Second,does the state exercise the prerogative to promote reli-gious purposes through legislation and judicial powers?This might include providing religious education, col-lecting taxes for religious bodies; sponsoring religiouscourts with jurisdiction over the family or religiouspractice; and passing laws on marriage, burial, dress,speech, and other matters. Third, does the state restrictthe freedom of religion—–either minority religions orall religions—–to operate? Typical strictures have con-strained the ability of religious bodies to appoint of-ficials; worship and practice; educate; build facilities;speak and assemble publicly; convert others to theirfaith; regulate marriage, burial, dress, and artistic ex-pression; and govern civil society institutions such asorphanages, schools, and hospitals. Fourth, does anyreligious body hold express constitutional prerogatives,standing titles, offices, or legal privileges in appointingstate officials or vetoing government decisions? Suchprerogatives are rare today, but they do exist: in Iran,a clerical Council of Guardians can veto legislationand vet candidates for parliament, whereas a clericalCouncil of Experts chooses the Supreme Leader, thecountry’s most powerful authority. It is along these fourdimensions that the degree of differentiation betweenany state and any religious actor can be measured.2

Differentiation should not be equated with con-cepts that are but components of it, like the familiarideas of establishment and religious freedom. BothIran and England establish religion, for instance, butdiffer wildly in the state’s authority over both estab-

2 My measurement of differentiation is drawn from descriptive ac-counts. Precise quantitative assessments of differentiation on a globalscale can be found in at least two major datasets: the InternationalReligion Indexes of Brian J. Grim and Roger Finke (2006), andthe Religion and State Dataset of Jonathan Fox (see Fox 2006 foranalysis; dataset is available on request from him at Departmentof Political Studies, Bar Ilan University). Fox analyzes 152 statesbetween 1990 and 2002, whereas Grim and Finke analyze 196 casesfrom 2003. Unfortunately, the date range of both sets falls later thanthe vast majority of instances analyzed here where differentiation iseither established or undergoes change in a given country. I do makesome use of the Grim and Finke dataset in my analysis of contem-porary terrorist groups below. Generally, both datasets conceive therelationship between religion and state quite similarly to the way Ido.

lished and minority religions and in the authority ofthe established religion over the state. Differentiation,a matter of foundational legal authority, ought not tobe confused, either, with the influence that religion andstate exercise on one another’s policies through per-suasion, ideological influence, or electoral power. Onlythis distinction can explain the paradox of the UnitedStates, where the world’s highest degree of separationbetween religion and state (Fox 2006, 559) coexists withthe strong political activism and influence of religiousgroups.

Differentiation varies not only in degree but alsoin kind, for it can be achieved through either mutualagreement or heroic resistance to suppression and canbe absent due to either consensual coziness or crueldomination. That is, high and low levels of differentia-tion can each take either a consensual or a conflictualform, yielding a matrix of four types, as Table 1 shows.A consensual arrangement is a stable one, where bothparties are satisfied with the status quo. A conflictualsettlement is one that at least one party wants to re-vamp; any consent it gives is coerced by the other. Ina constitutional democracy where religion and stateare contentedly autonomous, differentiation is highand consensual. The United States is a clear exam-ple. Consensual differentiation is also possible, thoughrarer, under authoritarianism—–many Latin Americanstates in the mid-twentieth century, for example. Inother states, like Communist Poland and Chile un-der General Augusto Pinochet, religious communitieshave achieved autonomy, but only through determinedresistance. Differentiation here is high but conflict-ual. States where the authority of religion and stateare mutually meshed—–contemporary Iran, Saudi Ara-bia, and Sri Lanka, and the now disappeared Catholicestablishment states in Spain, Portugal, and colonialLatin America, for instance—–are ones where differen-tiation is low but consensual. The weaving of institu-tions makes this arrangement integrationist. In a finalvariant, where differentiation is low and conflictual,religious bodies are dominated and suppressed, againsttheir will, sometimes despite their resistance. This re-lationship, too, is integrationist in that the state has“integrated” religion into its authority, as CommunistBulgaria and Romania did to the Orthodox Church,whose choice was to consent or die.

Political Theology

Political theology is the set of ideas that a religiousbody holds about legitimate political authority. Whopossesses it? The state? Or some other entity? To whatdegree ought the state to promote faith? What does jus-tice consist of? What is the right relationship betweenreligious authorities and the state? What are the obliga-tions of religious believers toward the political order?Answers to these questions motivate religious lead-ers and organizations to support, oppose, modify, orthwart the activities of the state. Political theology canbe arcane and scholarly, but also simple and popular.Al Qaeda’s political theology is held by a network ofmilitants, but originated decades earlier with

507

Explaining the Political Ambivalence of Religion August 2007

TABLE 1. Types of Differentiation and Their ExemplarsDegree↓ Kind→ Consensual ConflictualHigh–Differentiated United States

IndiaStates with “Engaged Buddhism”

Communist PolandChile under PinochetKemalist TurkeyPostcolonial Indonesia

Low–Integrationist IranSaudi ArabiaColonial Latin AmericaSri LankaMedieval Christendom

Communist Bulgaria,Romania, SovietUnion, and CzechoslovakiaPostcolonial Arab

nationalist regimes

intellectuals. Sometimes political theology will haveattained the status of “common sense” within a popu-lation, in the way that most Americans regard the doc-trine of separation of church and state. Some planks ofa political theology may be shared widely within a re-ligion, others, by only certain communities or factions.Muslims widely hold that political authority shouldmeet the standards of shari’a, or the pathway to God,but differ radically over its content. Political theologyalso evolves. It is influenced by ancient, formativeteachings, but also by historical development and bythe circumstances of time and place. The rise of “En-gaged Buddhism” all across Asia over the past half-century, for instance, has transformed at least somesectors of Buddhism from otherworldly quiescence toactivism for human rights and democracy. Other Bud-dhist communities, like the one in Sri Lanka, graftedfaith to the modern nation in the late nineteenth andearly twentieth centuries and now aggressively advo-cate a Buddhist nation-state (Queen 2005). To say thatpolitical theology matters is to say that a religious com-munity’s political stance is traceable, at least in part, tothis set of ideas.

The character of these ideas, like that of differen-tiation or of a religious actor’s political activities, isdiscerned through the works of historians, sociologists,anthropologists, religious studies scholars, and politicalscientists. Some actors are described through litera-tures of 10 or more sources, others through only two orthree. How were these sources selected for interpreta-tion? For many religious actors, major sources are infact largely in harmony in characterizing the variablesof concern, even when they differ in their agendas,biases, and interpretations of other matters. In a fewcases, sources disagreed but some interpretations wereclearly better evidenced than others. Generally, then,confidence in characterizing variables is warranted.Space constraints, however, permit only a referenceto one or two sources per actor, usually those that bestdescribe its ideas, institutions, and political activities(see Lustick 1996 on selection issues).

Pathways to the Present

Every religious community’s ideas and its relationshipto the state is like a palimpsest, shaped by the eventsand ideas of its founding, and then centuries of schism,

diaspora, growth, decline, conquest, subjection to con-quest, persecution, migration, growth in institutionalpower, and the spread, rise and ebb of membership.Resonant here are the emphases of the historical in-stitutionalist school of political science, which explainsstate policies as the result of long historical pathwayseked out by evolving institutions and ideas (Piersonand Skocpol 2002). Could not religions experience sim-ilar dynamics?

The pathways through which religious actors’ ideasand relationships with states develop will be diverse.In some religious communities, they wind back manycenturies; in others, their emergence is more recent.Ideas and institutions themselves also interact, causing,incubating, shaping, and enabling each other. Each fac-tor in turn has its own causes. This does not threaten itsexplanatory power, for every causal variable is alsoa dependent variable, ad infinitum. More troublingwould be the discovery that one or both is system-atically reducible to some additional variable, say, theinterests of the owners of the means of production, orthat one of them, say, ideas, flows in lockstep from theother, say, institutions. What the cases show, though,is that political theology and differentiation are rootedin historical circumstances that are identifiable but dif-ficult to generalize about. The unpopularity and theweakness of the Cold War Czechoslovakian CatholicChurch vis-a-vis the state date back to the Habsburgsuppression of nationalist Protestant uprisings duringthe Reformation era. Radical Islamic Revivalism waslaunched by early-twentieth-century intellectuals whoperceived that Islam was decaying due to Western im-perialism and internal corruption. Each religious ac-tor tells its own story. The cases show that ideas donot merely justify, retrospectively explain, or otherwisestrictly conform to institutions, either. Often, the politi-cal theology of single religious actors changes whereastheir differentiation from the state remains constant.Often, too, political theology varies between religiousactors whose differentiation from the state is similar.In cases where political theology leads a religious actorto oppose a political order, its independence is partic-ularly salient.

Generalization is more possible, though, with respectto the sequence in which ideas and institutions emergeand affect each other. At least four pathways are per-ceptible. In the first, institutions—–that is, a religiousactor’s level and kind of differentiation—–take their

508

American Political Science Review Vol. 101, No. 3

FIGURE 1. The Causal Logic of the Argument

Political theology(ideas)

Differentiation(institutional relationship)

Other influences

A religiousactor’s

democratizingactivity orpoliticalviolence(action)

Constitutionalliberal

democracy(achievedinstitution)

orgoal of violence

(e.g., newregime, new

policy)Other variables

shape well before a new political theology arises. Thenew thinking then serves as the proximate cause ofthe change in the religious actor’s political pursuits,though these pursuits will remain empowered or hin-dered by the actor’s prior condition of differentiation.The Catholic Church in Latin America had achieveddifferentiation decades before, and in Poland two cen-turies before, it took up the liberal democratic ideasthat led it to support democratization. Once it didembrace these ideas, its differentiated position em-powered it to pursue them. A Radical IslamicRevivalist political theology led opposition parties inIran and Algeria to take up violence even thoughtheir states had sponsored oppressively integrationistinstitutions for two or more decades. In a second pat-tern, a religious actor takes on a new political the-ology whose very logic then leads it to refashion itsinstitutional relationship with the state. Here, not onlyare ideas independent from institutions, but also theyshape them. Once the leading prelates of the CatholicChurch in Spain accepted the Second Vatican Council’steaching of religious freedom, they withdrew theirdecades-long tight institutional links with the Spanishauthoritarian state, thus enabling them to pressure itto democratize. Successive Islamic political parties inTurkey took on democratic thinking more widely anddeeply over the second half of the twentieth century,out of which they struggled, through fits and starts,to achieve institutional autonomy. In a third pattern,changes in institutions lead the way, eliciting a fairlyquick rise or strengthening of political theology in re-ligious groups. A Communist integrationist regime’sseizure of power in Afghanistan in 1978 spawned orattracted numerous jihad movements, which convergedthere to wage civil war for a decade. Finally, the fourthsequence is one where political theology and differ-entiation both take root decades or even centuries inthe past and remain constant into the present. TheOrthodox Church of the Cold War years espousedand practiced an integrationism that dates back to the“Caesaropapism” of the Byzantine Empire. A slightvariant is the Islamic democracy movement inIndonesia, whose differentiated institutions and ideasfavoring pluralism and separation of religion and stateboth date back to the era of medieval Europe, though

its thinking then became updated through modern lib-eral democratic ideas in the late twentieth century.Although such mutual historical roots and tandemevolution do not reveal how ideas and institutions arecausally linked to each other, still, their integrationistor differentiated character can be compared with thesame factors in other countries and linked with theirrespective political outcomes.

In tracing these pathways, clarity about the nature ofideas and institutions themselves is also essential. Polit-ical theology is a set of propositions about politics thatpeople hold in their minds, share and develop throughlanguage and discourse, and use to persuade and mo-tivate. Differentiation is an institutional relationshipthat embodies legitimate authority. The promotion ofdemocratization and the execution of political violenceare actions through which religious actors further apolitical end. In the case of democratization, that endis a liberal democratic constitution where rights, es-pecially religious freedom, are enshrined in law. In thecase of political violence, ends are more various—–a newregime, perhaps, or the end of imperialism, or revenge.Figure 1 illustrates these causal relationships.

DEMOCRATIZATION

Along with the global resurgence of religion, the pastgeneration has witnessed a historically impressive out-break of democracy. Huntington (1991) documents a“Third Wave” of democratization involving 30 coun-tries between 1974 and 1989. The trend has continued.Freedom House (2004, 2005) reports that over the mostrecent 15 years, the number of electoral democracieshas increased from 69 out of 167 total states (41%) to119 out of 192 (62%), and that between 1994 and 2005,the number of “free” countries in the world increasedfrom 76 to 89; whereas the number of “partly free”countries declined from 61 to 54; and the number of“not free” countries from 54 to 49. Might the two trendsbe related?

Democratization is a natural area where differentia-tion and political theology might explain the politics ofreligious actors. Democracy, after all, itself embodiesdifferentiation, the mutual and consensual distancing

509

Explaining the Political Ambivalence of Religion August 2007

of religious and political authority that Stepan (2001)has described as “the twin tolerations”—–especially ifdemocracy is conceived to embody not only contes-tation and participation but also liberal freedoms likeminority rights and religious freedom. Support for de-mocratization can take the form of several kinds ofcivic, nonviolent modes of resistance, including explicitstatements and actions of protest against authoritarianregimes, conduct of religious ceremonies with an oppo-sitional intent, cooperation with co-religionists acrossborders in defying the regime, and similar collaborationwith parties, unions, and other opposition groups withindomestic civil society.

An initial global glance shows religious actors exer-cising such influence during the Third Wave in placeslike Poland, Lithuania, the Philippines, Indonesia,Turkey, Kenya, South Africa, Chile, and Brazil, butnot in places like Rwanda, Argentina, and Senegal.What explains the varying postures? Those who agitatefor democracy are motivated by a political theologythat endorses liberal democratic institutions. They haveembraced religious freedom and the broad separationof spiritual and temporal authority, abjured standingtemporal authority for their clerics and a legal mono-poly that denies religious freedom to other citizens,and rejected a quietist stance that demands blanketobedience. They have also established differentiationfrom the authoritarian state. Though it may be threat-ened constantly, their autonomy serves as an islandof free space—–a sphere of “moral extraterritoriality,”as Weigel (1992) has put it—–from which they can ac-tively oppose the regime. Conflictual differentiation isthe structural space that empowers liberal democraticpolitical theology. Integrationism, by contrast, deniesreligious bodies the distance to oppose authoritarianregimes. When it is conflictual, they are effectivelysuppressed; when it is consensual, they often enjoyfinancial and legal privileges that make them unwillingto oppose the arrangement. Together, these ideas andinstitutional relationships spur the democratic activityof religious actors, which then helps to bring about theliberal democratic institutions that enshrine consensualdifferentiation stably in the rule of law.

The focus here on democratization—–the transitionfrom an authoritarian regime to a liberal democraticone—–ignores some respects in which religion shapesdemocracy, for instance, through participation and in-fluence on government policy. In addition, the focushere on the Third Wave rules out some cases of democ-ratization. Such circumscriptions exclude some inter-esting cases of religion and democracy—–like Israel, forinstance, which was founded in 1948, before the ThirdWave began, and was a democracy from its conception.Israel’s unique combination of comprehensive Jewishlaw, the lowest degree of separation of religion andstate of any modern democracy (Fox and Sandler 2005,326), and an otherwise liberal democratic constitutionthat affords significant religious freedom to non-Jewsresonates with what Elazar (1995) has called the bib-lical “covenant tradition” in Jewish political thought,a political theology whereby the political communityis founded on divine initiative yet also rests on mutual

consent, equality, joint popular action, and a “federal”separation of powers. Despite omitting states like Is-rael, though, the focus adopted here yields tractability,close comparisons, and thus more confident conclu-sions about religion’s political influence that can thenbe extended to other cases and dimensions of democ-racy through further research.

The Catholic Wave

Huntington (1991, 76–85) observed that the ThirdWave was “overwhelmingly a Catholic wave.” Of the30 countries that made a transition to democracy be-tween 1974 and 1990, roughly three-fourths were pre-dominantly Catholic. It is a striking finding: why wouldcountries the majority of whose population belong toa particular religious community, especially one thathas historically distrusted democracy, compose themotor of a global trend in democratization (Philpott2004)?

Behind the Catholic Wave was a sea change inthe Catholic Church’s political theology: the SecondVatican Council (1962–1965), where the Churchadopted human rights, religious freedom, democracy,and economic development into its teaching and de-clared its withdrawal from temporal prerogatives—–adefinitive, doctrinal embrace of differentiation(Flannery 1975).

True, some separation between temporal and spiri-tual authority has always characterized Christian po-litical theology: “Give to Caesar what is Caesar’s,and to God what is God’s,” taught Jesus. Even afterthe Roman Emperor Constantine officially establishedChristianity on his conversion in 313 A.D., the Churchclung to a principle of separate spheres, expressed fa-mously in Pope Gelasius’s doctrine of the two swordsin 496 A.D. In the Western medieval church, Gelasius’distinction survived, but just barely. In the Respub-lica Christiana, ecclesial and temporal authority weredeeply meshed in both theory and practice, whereasJews, Muslims, and other non-Christians were given nofundamental right to practice their faith (Hehir 2005).

Then, over subsequent centuries, medieval inte-grationism endured a long, slow erosion. A centralagent of differentiation was the continental system ofsovereign states, whose consolidation at the Peace ofWestphalia of 1648 elevated a form of polity whoseauthority was territorial and independent of ecclesi-astical authority and sealed the decline of a unitedChristendom—–eliciting the predictably sharp condem-nation of the Pope (Philpott 2001; for contrasting views,see Osiander 2001; and Krasner 1993). Behind thistransformation, to be sure, were economic, technolog-ical, and demographic forces (Spruyt 1994; Tilly 1992),but also two epochal metamorphoses in ideas and iden-tities: first, the rise of the nation, whose origins, as histo-rians like Hastings (1997) have shown, lie in medievaland early modern Judeo–Christian narratives, and sec-ond, the Protestant Reformation, whose very politicaltheology called for stripping the Catholic Church of itstemporal authority (Philpott 2001, 97–149).

510

American Political Science Review Vol. 101, No. 3

By the nineteenth century, the chief motor of dif-ferentiation was the Church’s struggle with Europeanliberalism. In the name of the rights of man, democ-racy, and the nation, partisans of the French revolution,republicanism, socialism, and Bismarck’s kulturkampfattacked the authority of the Catholic Church, who,in response, clung to its medieval doctrine and con-demned the liberal sovereign state. From the late nine-teenth to the mid twentieth century, though, circles ofCatholic intellectuals and Christian Democratic par-ties in Europe and Latin America came to embracewhat they saw as a friendlier liberalism that envisionedCatholicism to be neither established nor suppressedand that proclaimed religious freedom. In taking thisposition at the Second Vatican Council, the Church infact preserved its censure of absolute state sovereignty,rendering the state’s legitimacy as real but relative to alarger moral order to which the Church would nowdemand conformity from its differentiated position(Hehir 2005, 97–101).

In the wake of the Council, these new teachings re-dounded through the Church via its unusually densetransnational ligatures of authority—–a global networkof bishops united around the Pope, who is understoodtheologically as a visible sign of the Church’s unity.Though some local churches had already sprouted thenew political theology prior to its official promulgation,virtually every Catholic effort to promote democracygained vigor and explicitness once Rome had pro-nounced it officially.

Yet this commonly broadcast seed fell onto soilsof varying fertility. Even in this most centralized ofreligious bodies, national churches came to embraceand promote democracy with divergent levels of vigor(Byrnes 2001). This variance, in turn, corresponds tothe greater and lesser degrees to which churches dif-ferentiated themselves from the state and embracednew political theologies.

Through four broad causal patterns, liberal demo-cratic ideas and differentiation influenced (or failed toinfluence) Catholic churches’ promotion of democra-tization. The first pattern occurred in countries wherethe Church had been institutionally differentiated fromthe state several decades or more prior to the rise of aliberal democratic political theology during the yearssurrounding the Second Vatican Council. Of thesenational churches, those who came to oppose author-itarian regimes most vigorously were those where thenew liberal democratic thinking became most deeplylodged—–among Catholic student movements, labormovements, rural cooperatives, Christian Democraticparties, lay organizations, or grass roots cooperativesknown as Ecclesial Base Communities. But in nochurch could liberal democratic ideas occupy a “centerof gravity,” to borrow Mainwaring’s (1986) phrase, un-less they attained the support of at least a coalition ofbishops.

Best exemplifying this pattern is the Polish CatholicChurch, whose vital role in overthrowing its Com-munist regime in 1989 began with Stefan CardinalWyszinski’s mobilization of opposition during the1950s and 1960s and continued through the pilgrimages

of native son Pope John Paul II, which hundreds ofthousands attended in the late 1970s and early 1980s.This was a church that, through a century and a halfof fending off invaders from Prussia, Russia, and theAustro-Hungarian Empire, had established a strongautonomy from the state, fortified by a deep identifi-cation with the popular nation. Under Communism, itthen defended its differentiation by maintaining con-trol of its own governance, education, and worship.When the teachings of Vatican II arrived, this churchbecame not simply a defender of its own freedom, but amore explicit advocate of human rights and democracy,themes powerfully reinforced by John Paul II, who in-corporated them into his social teachings (Mojzes 1992,294–99; Weigel 1992).

Most like the Polish Church were the Lithuanianand Ukrainian Catholic churches, which also preservedtheir historical autonomy under Communism, alliedwith popular nationalism, took on the teachings ofVatican II, and then strongly challenged their regimesthrough protest and ceremony (Mojzes 1992, 182–256).Likewise, in South Korea, after Vatican II, a CatholicChurch with a long history of conflictual differentia-tion from the ruling state cooperated with Protestantchurches in defying dictator Park Chung Hee throughprotests and political gatherings in the 1970s and the1980s (Yun-Shik 1998).

The pattern is also widely evidenced in LatinAmerica, where Catholic churches share both strongmajorities and a common historical trajectory. Aftermost Latin American states won independence fromSpain and Portugal in the first three decades of the nine-teenth century, they shed the consensual integrationismof the colonial period and came to reproduce Euro-pean rifts between secularizing liberals and Catholicswho wanted to retain as many prerogatives as possible.Between roughly 1850 and 1925, liberals gained theupper hand and won disestablishment and religiousfreedom, thus ensconcing differentiation firmly andearly. Most Latin American churches then adopted a“neo-Christendom” stance by which they sought infor-mal but close ties with the state in order to preservetheir influence in education, marriage law, and otherareas. By midcentury, though, sectors of several ofthese churches began to embrace “progressive” ideasof liberal democracy and economic development forthe poor, concepts that gained great strength in theSecond Vatican Council and in the 1968 Conference ofLatin American Bishops in Medellin, which christenedthem as “liberation theology” (Gill 1998, 17–46).

Variations in ideas then explain variations in pol-itics. It was those Catholic churches where this newpolitical theology took root deepest, widest, and earli-est that came to support democracy most vigorously(Mainwaring and Wilde 1989). Strongest was theBrazilian Church, where urban, rural, and student laymovements, base communities, and a coalition of thenation’s bishops embraced democracy and develop-ment. Though its bishops were divided over militaryrule when it arrived in 1964, by 1970, in good partdue to Vatican II’s influence, they had joined lay dis-sidents and issued a string of pastoral statements and

511

Explaining the Political Ambivalence of Religion August 2007

denunciations that outpaced any church in the region(Mainwaring 1986, 79–141). The Chilean bishops,though also initially divided toward General AugustoPinochet’s coup of 1973, had by 1976 united to createthe Vicariate of Solidarity, whose exposure of humanrights violations rivaled every Latin American Churchsave Brazil’s for the strength of its opposition to dic-tatorship (Fleet and Smith 1997, 111–58). A majorityof Catholic bishops also came to oppose military dic-tatorships by the end of the 1970s in Nicaragua and ElSalvador, where “progressive” ideas had penetratedseveral sectors of the Church at least a decade earlier.A close variant of the same pattern developed in Peru,Ecuador, Panama, and Bolivia, and in Guatemala bythe mid-1980s (Cleary 1997; Klaiber 1998, 121–67, 216–38).

The second pattern is one where national churchesremained consensually integrated with their states wellinto the 1960s, when Vatican II teachings changedtheir thinking and induced them to differentiate them-selves politically. Here, changes in ideas preceded andelicited changes in differentiation. The Spanish Churchfirst took on the Council’s teachings, then abruptlydistanced itself from its longstanding integrationistties with Generalissimo Francisco Franco and becamean impetus for Spain’s democratization following hisdeath in 1975 (Payne 1984, 149–213). Just prior tothis transformation in Spain came democratization inPortugal, also supported by a Catholic church that hadwithdrawn from close collaboration with an authoritar-ian state after taking on Vatican II political theology(Manuel 2002, 81–92). Similarly, a Catholic Church thathad long been closely integrated with the governmentof the Philippines gradually took on the teachings of theCouncil and grew in its opposition to the dictatorship ofFerdinand Marcos, culminating in Cardinal Jaime Sin’sinspiration for the “people’s power” protest movementthat overthrew Marcos in 1986 (Youngblood 1990).

In a third pattern, differentiation and new ideasrose roughly contemporaneously, but neither clearlycaused the other. In Africa during the 1980s and1990s, Catholic churches came to oppose postcolonialregimes in Kenya, Congo, Ghana, Malawi, Zimbabwe,Mozambique, South Africa, and Zambia through bish-ops’ statements, politically suffused worship, educa-tion, papal visits, and cooperation with political par-ties and Protestant churches. In every case, the rise ofdifferentiation and new ideas preceded the Church’sdemocratic activity. Each church came to espouse Vat-ican II’s teachings about human rights and democracy,sometimes with a strong alloy of liberation theology,as in Malawi. In cases like Mozambique, the Churchhad been meshed with the colonial regime, but afterindependence developed a strongly conflictual differ-entiation from the postcolonial government (Gifford1995; Phiri 2001).

A fourth pattern consists of those national churchesthat, by contrast, never or only weakly opposed au-thoritarian regimes. These achieved far less autonomyfrom the state than did those that became oppositionaland espoused liberal democracy far more lukewarmly,if at all. The Czechoslovakian Catholic church came to

oppose its Communist regime less vigorously and muchlater than did the Polish and Lithuanian Churches. Itwas also far less able to govern itself, enjoyed fewerties with other opposition groups, and, at least inCzech lands, was alienated from the nation, whose anti-Catholicism dates back to Habsburg suppression ofseparatism during the Reformation era (Ramet 1998,90–144). The Hungarian Catholic Church, apart fromthe lonely resistance of Cardinal Jozsef Mindszenty,opposed its Communist regime even more weakly,and was also coopted, suppressed, and slow to adoptVatican II (Ramet, 104–22). In Africa, too, severalnational Catholic churches failed to support democ-racy, including in Angola; in Cameroon; in Ugandaunder the Museveni regime after 1986; and in Rwanda,where the Church did little even to stop genocide.These churches were generally neutral in their polit-ical theology, willing to support any sort of regime.In Cameroon, though the Church remained differen-tiated, a quietist theology focusing on personal salva-tion discouraged active political involvement. Othersremained undifferentiated—–either suppressed, as wasthe Angolan Church by its Marxist government, orclosely linked with the regime, as was the Church inRwanda (Gifford 1995; Rittner, Roth, and Whitworth2004).

In Latin America, churches in Argentina andUruguay never became strong democratizing forces,whereas the Church in Paraguay protested only justprior to the fall of dictator Alfred Stroessner in thelate 1980s. Here, neither liberal democratic politicaltheology nor liberation theology took root among thelaity and clergy anywhere nearly as deeply as they did inBrazil, Chile, and elsewhere, whereas hierarchies per-petuated the “neo-Christendom” model of close tiesto military rulers. During the Dirty Wars in Argentinafrom 1976 to 1983, for instance, the Church remainedlargely passive toward the military regime (though withexceptions), failing to advocate democracy until 1981.These churches, because of their ideas and their insti-tutions, stood on the sidelines of the Catholic Wave(Klaiber 1998, 66–120).

Is it possible, though, that despite the relation-ship of differentiation and political theology to theCatholic Wave, another, unexamined factor in factdrove the Church’s stances toward authoritarianism?Although the argument here generally probes theinfluence of two factors rather than testing themagainst alternatives, Gill’s (1998) explanation for theCatholic politics of democratization—–also compara-tive and systematic—–presents a strong enough chal-lenge to warrant consideration.

Covering the decision of twelve national Catholicchurches in Latin America to support or opposeauthoritarianism in the 1960s and 1970s, Gill’s range ofcases converges with a portion of the Catholic Wave.But his explanation focuses on the religious competi-tion that the Church faced from evangelical Protes-tantism. Arguing according to the logic of rationalchoice, he assumes that churches seek to maximizemembership and secondarily financial resources to runtheir organizations. All else being equal, churches have

512

American Political Science Review Vol. 101, No. 3

strong incentives to ally with states, who can providechurches with resources and favorable laws in returnfor their provision of ideological legitimation. Butwhen Protestant churches attract the Catholic Church’smembers, particularly among the poor, and when statesare authoritarian, Gill argues, the Catholic Churchmust then distance itself from such states—–and thussupport democracy—–since the poor regard dictator-ships as oppressors and friends of the wealthy (1998,47–78).

Gill does not ignore the role of ideas. His regressionanalysis uses the age of bishops as a proxy for reformistdoctrines, theorizing that younger bishops will morestrongly favor Vatican II, and finds that it helps to pre-dict the political stance of Catholic churches, thoughwithout statistical significance (1998, 107–108). But heunderestimates the role of ideas. By focusing on bish-ops, he ignores the associations of religious orders andlay people who crucially empowered and embracedliberal democratic ideas. Although a national religion’saggregate favor for ideas is difficult to quantify pre-cisely, it can be assessed feasibly through descriptiveaccounts of organized movements and espicopal opin-ion alike and then compared across countries, espe-cially when only 12 cases (or even up to twenty LatinAmerican states) are involved. As argued previously,such a comparison reveals a relationship between ideasand Catholic politics, stronger than the one Gill allowsfor. Ideas also explain particular cases better. Gill callsGuatemala an outlier, claiming that Protestant com-petition coexisted with a pro-authoritarian stance. Yetin the mid- to late 1980s, the Catholic Church hereproclaimed human rights and democracy and distanceditself from its military regime. What explains this shiftis the death of pro-regime Cardinal Mario Casariego in1984 and his succession by a head prelate who joinedother sectors of the church in favoring Vatican II’s po-litical theology. Neither the fact nor the direction ofthis change can be described through aggregations ofbishops’ ages.

Gill (1998) also overestimates the role of Protestantcompetition. His assumption that the poor demanda church that opposes authoritarianism is empiricallyquestionable. Across Latin America cases where boththe poor and Protestant churches support authoritariangovernments, and even more where both are apoliti-cal, compete with cases where they favor democracy(Freston 2007; Sigmund 1999). His logic also missesa central reason why the poor choose Protestantchurches—–for their ability to provide pastoral services,in contrast to the Catholic Church, where priests havebeen scarce for much of the twentieth century. It isunclear, too, why competition would lead the CatholicChurch to support democracy. Undoubtedly, it hasled some sectors of the Church to reach out to thepoor—–through base communities, for example. Butwhy, in a rational choice logic, would not the Churchalso find it instrumentally rational to urge the author-itarian state to suppress Protestantism, which stateswere perfectly well equipped to do and in fact didwell into the twentieth century? And why not takeother measures like ending priestly celibacy in or-

der to increase pastoral services? (Kalyvas 2000) Theanswer to both questions lies in the Church’s doc-trines. Finally, Gill’s evidence fails to link the tim-ing of Protestant conversions with the Catholic op-position to authoritarianism. Measuring the growthof Protestantism between 1900 and 1970, his statis-tics give little sense of when conversions occurredor were most intense, leaving their link with opposi-tion to authoritarianism ambiguous. In Brazil, the siteof the Catholic Church’s strongest stance for democ-racy, Protestantism grew sharply only following thisstance—–at a rate of 8.8% from 1980 to 2000, as com-pared to 2.6% from 1960 to 1980, according to Brazil’snational census. As for Gill’s qualitative evidence forthe Church’s concern with Protestantism, it is drawnfrom previous decades and does not reveal the motiva-tions behind the Church’s opposition during the 1960s,1970s, and 1980s. Although Gill does establish a cor-relation between Protestant competition and Catholicdefiance, his argument does not supplant the role ofideas, or, for that matter, differentiation, in explainingthe Catholic Church’s stance towards democratizationin Latin America.

Other Christian Churches

Eastern Orthodox and Protestant Christian churchescontributed to the Third Wave far less vigorously thanCatholic churches. Greece’s transition to democracy in1974 received little support from the Greek OrthodoxChurch, which remained close to the preceding militaryjunta, then quickly transferred its loyalty to the victorregime. Nor did the Orthodox Church in Bulgaria andRomania play much of a role in the revolutions of 1989,their small dissident movements surfacing only shortlybefore the transitions. When Ukraine won its indepen-dence from the Soviet Union in 1991, the OrthodoxChurch did little to encourage democratization, in con-trast with the Catholic Church, which had developed anunderground structure and lay pressure groups. Like-wise, the Orthodox Church in Russia has done little toencourage democracy there since the fall of commu-nism (Billington 1999, 56–57; Kuzio and Wilson 1994,89–91; Legg and Roberts 1997, 53–54).

Integrationism penetrates more deeply into the Or-thodox Church than it does in the Latin West, bothin its political theology and its relationship with tem-poral authorities. As western Christianity formed itsGelasian doctrine, the Orthodox Church of easternByzantium developed a “Caesaro-Papism” that fusedspiritual and temporal authority into the same hands.Close integrationism survived under the OttomanEmpire after the fall of Constantinople in 1453 andpersisted into the era of the modern nation-state, whereit has been reinforced by an ecclesiology of “auto-cephaly” that divides the Church’s hierarchy alongnational lines—–again, in sharp contrast to the trans-national centralization of the Catholic Church, whichstrengthens differentiation (Ware 1963, 18–72, 87–101,239–63). Neither have Eastern Orthodox churches

513

Explaining the Political Ambivalence of Religion August 2007

developed a consensus around a political theology ofhuman rights and democracy.

All of these factors characterized the Orthodoxchurches of Bulgaria and Romania, who, in the 1870s,attained national autonomy and came under the con-trol of monarchs in their governance, finances, and rolein civil society, ripening them for the far more thoroughand conflictual integrationism of Communist states inthe 1940s, which killed dissident bishops, priests, andnuns and forced the church into a subservient role ofespousing the regime’s propaganda (Ramet 1998, 181–201, 275–307). The Russian Orthodox Church, too, wasmade susceptible to a harsh Communist takeover in the1920s by a consensual integrationist relationship withRussia’s Tsars dating back to the reign of Peter theGreat in the early eighteenth century. Having attainednational autonomy in 1589, the Russian Church wasalso bereft of transnational links that might fortify itsdifferentiation from the state (Ware 1963, 102–25). TheGreek Orthodox Church also practiced a consensuallyintegrationist relationship with the modern state, whichestablished it as an official national church, managedits fiscal affairs, and even participated in appointing itsclergy (Legg and Roberts 1997, 53–54, 104; Ware, 213).

Protestant churches advanced differentiation far ear-lier than the Catholic Church. Or at least certainProtestant churches did. To claim that the Reforma-tion brought religious freedom or the separation ofchurch and state is both anachronistic and overly gen-eral. The churches of Martin Luther, Ulrich Zwingli,Oliver Cromwell, and those of the Swedish reformationall supported an “Erastian” state that would protecttheir worship, enforce their orthodoxy, and even playa role in their governance. It was rather the churchesof the “radical reformation”—–Mennonites, Quakers,and certain Baptist, Puritan, and Calvinist sects—–thatcalled for a sharp separation between the Christiancommunity and civil authority. Their emphases on in-dividual conscience, the unmediated apprehension offaith, and the autonomy of church authority, and thevery multiplicity of their sects all contributed to theirdemand for religious liberty in seventeenth centuryEngland and the Netherlands, as well as in the Amer-ican colonies, where they laid the foundation forthe United States Constitution’s guarantee of reli-gious freedom. That they often opposed ruling powersand suffered persecution testifies to the autonomy oftheir political theology. During the twentieth century,Protestants have come to advocate religious liberty inAfrica, Latin America, and Asia, where their numbershave grown dramatically—–in Africa, for instance, from2% of the population in 1900 to 27% in 2000 (Wood-berry and Shah 2005, 117–22). Drawing from their po-litical theology, they have likewise become advocatesof the United Nations and international human rightscovenants (Little 2005).

Today, though, Protestant churches are diverseboth in their theology and their differentiationfrom the state, consisting of not only transnational“mainline churches”—–Methodist, Lutheran, Anglican,Presbyterian, and the like—–but also of thousands ofindependent churches, many of them Pentecostal, that

can differ from one another as much as any of themdiffer from Catholicism or Orthodoxy.

A corresponding variation in their support for de-mocratization results. In Europe, Lutheran churchesin East Germany, Latvia, and Estonia fit the Erastianpattern—–a political theology that stresses the authorityof rulers over their justice, a hierarchy willing to submitto the oversight of Communist regimes in exchange fortheir survival, and hence a lack of active support fordemocratization until only a year or so prior to thedownfall of their regimes. In East Germany, a grass-roots “church from below” resisted the regime moreactively than the church’s hierarchy, precisely becauseit espoused a doctrine of human rights that led it todistance itself from the regime in dissent (Kellogg 2001;Monshipouri and Arnold 1996).

In Africa, Asia, and Latin America it was the main-line Protestant churches whose leaders held a strongdoctrine of human rights and democracy and whogoverned themselves independently from the regimethat most actively opposed dictatorships. The relation-ship between ideas and differentiation differs amongthese diverse churches, but especially in Africa it wasoften liberal democratic ideas that led churches todistance themselves from the state. Most formidablewas the South African Council of Churches, whichheld a “contextual theology” that endorsed humanrights, democracy, and racial equality, separated itselffrom the apartheid state, and strongly opposed it fromthe 1960s to the early 1990s (Borer 1998). A simi-lar configuration of political theology, differentiation,and democratizing activity characterized the MalawianPresbyterian Church, the Kenyan Anglican Church,the Mozambican Anglican Church, the GhanianPresbyterian Church, councils of Protestant churchesin Zambia, the National Council of Churches in SouthKorea, the Taiwanese Presbyterian Church, and, toa more moderate degree, evangelical churches inPeru and Nicaragua. Protestant churches who did notfavor democratization either enjoyed a consensual in-tegration with the state, like Protestant churches inUganda, Cameroon, Rwanda, Liberia, and Guatemala,or the Dutch Reformed Church in South Africa, orwere differentiated but held a political theology thatstressed personal salvation over political action, as didPentecostals in Brazil, Chile, South Korea, and Kenya(Freston 2001; Gifford 1995; Yun-Shik 1998).

Islam and Democratization

In the Islamic world, democratization has been rare.Today, out of 47 Muslim-majority countries, only three,Mali, Senegal, and Indonesia, live under regimes thatFreedom House categorizes as “free” (Freedom House2006). In 2001, only Mali was free, whereas 18 Islamiccountries were “partly free” and 28 were “not free.”Eleven of the Muslim countries, or 23%, had electedregimes in 2001, but the failure of 10 of these to fitthe “free” category points to their lack of other demo-cratic features. By contrast, in the non-Islamic worldof 2001, 85 countries were “free,” 39 “partly free,” and

514

American Political Science Review Vol. 101, No. 3

21 “not free,” whereas fully 110 of 145 states, or 76%,had democratically elected regimes. Nor has democ-racy increased in the Muslim world during the pastgeneration. From 1981 to 2001, no Muslim countrymoved into the “free category,” whereas two droppedout of the “partly free” group and ten joined the “notfree” group. The pattern is even starker in the Arabportion of the Muslim world, which altogether lacksan electoral democracy or “free” country (Karatnycky2002, 101–104). Finally, Fish’s (2002, 13) quantitativeglobal analysis reveals a strong statistical relation-ship between Islam and authoritarianism, even whenother demonstrated influences on democracy like eco-nomic development and ethnic uniformity are fac-tored in (see also Donno and Russett 2004; Midlarsky1998).

Are Islam and democracy, then, inherently incom-patible? Since September 11, 2001, analysts have hotlydebated the issue. Whereas few categorically deny thatMuslims can be ruled democratically, skeptics point tothe lack of a basis for constitutionalism, human rights,and democracy in the Islamic tradition of thought andexperience, to the prevalence of fundamentalism, toa privileging of revealed law over legislated law andpopular sovereignty, to gender inequalities, and to eco-nomic and political underdevelopment (e.g., Kedourie1992; Lewis 1996; Pipes 2002). Defenders of Islam’sdemocratic possibilities demur, stressing the multiplic-ity of voices, sources of law, and political traditionsin Islam, historical practices of respect for minoritiessuch as the millet system of the Ottoman Empire, andIslamic concepts that favor democracy such as shurah(consultation), ijma (consensus), and ijtihad (indepen-dent interpretive judgment) (Esposito and Voll 1996;Sachedina 2001; Stepan 2001, 233–46).

The empirical record seems to favor the skeptics. Butit is not the whole story, and the possibility of evolutionmust not be discounted. Prior to the Catholic Wave,Catholicism did not appear compatible with democ-racy, either, and then a sea change occurred. Today,Muslim democratic movements in fact exist in severalparts of the world. That roughly half of the world’sMuslims live under democratic or near democratic con-stitutions also suggests the compatibility of Islam anddemocracy (Stepan 2001, 237).

A paucity of democratic regimes mottled with demo-cratic movements: political theology and differentia-tion explain much about both sides of this profile ofIslam. The prevalence of integrationist political the-ology and institutions matches the rarity of Islamicdemocracy. Fox (2006, 554) confirms that the rate of“government involvement in religion” in Islam is atleast twice that of all other world religions. But thevariations are important, for they represent the possi-bility of Islamic democracy. Behind them, in turn, arevariations in the ideas and institutions of local Islamiccommunities.

Historically, both sorts of possibilities lay in Islamicpolitical thought and institutions, which, like thoseof Christianity, contain a diversity of voices within acommon fealty to the tradition’s founding texts, theQur’an and the Hadith. A fusion of spiritual and tem-

poral authority existed in the first Muslim polity, thecity of Medina, during the latter half of the ProphetMohammed’s life. Well into the period correspondingto the European Middle Ages, Islam’s predominantSunni community favored a caliph, a common spiritualand temporal head of the Islamic people, the umma.The Shi’ite tradition, too, has long conceived of spir-itual and temporal authority as integrated. Yet, con-temporary historians point out that in early Islamic‘Abbasid and Umayyad dynasties, religious authori-ties sought a critical distance from the caliphs, onethat resembled early Christianity’s Gelasian doctrine(Lapidus 1975).

Since the nineteenth century, the issue of separationwithin Islamic political thought has revolved largelyaround the Westphalian state, whose status, accordingto Hashmi (1998), Muslims have viewed in three broadways. The first of these is “statism,” a broadly secu-lar view that embraces the territorial state. A second,more widely held, view is “Islamic internationalism,”which accepts the Muslim state in principle but stressesits Islamic foundation and its pan-Islamic obligations.Finally, “Islamic cosmopolitanism” views the state andthe state system as European impositions and illegiti-mate dividers of the umma. Ayatollah Khomeini of Iranand Abu al-A‘la Maududi of Pakistan, for example,endorsed their states only as instruments for creating aglobalized Islamic order.

Closely resonant with cosmopolitanism is “radicalIslamic revivalism,” the tradition that yields today’smost militant Islamic factions, including Al-Qaeda.Revivalism’s most famous articulators are early andmid-twentieth century theorists and activists likeMaududi, the founder of Jamaat-i-Islami, Hasanal-Banna of Egypt, the founder of the MuslimBrotherhood, and Sayyid Qutb of Egypt, all ofwhom, echoing nineteenth-century Arab Muslim re-vival movements like Salafism and Wahhabism, heldthat Islam had suffered a centuries long, worldwidedecline into jahiliyyah, a state of barbarism. This de-cline, they thought, was not only moral and spiritualbut also social and political, involving the dominationof Islam at the hands of western colonial states, and, inturn, calling for both a moral and spiritual renewal anda reassertion of Islamic governance. It was a secondgeneration of revivalists that made violence a centralplank of their program in the 1950s and 1960s, inspir-ing a proliferation of terrorist groups, many of themeventually converging in Afghanistan’s war against theSoviet Union (Roy 1993). Though revivalists did, andstill do, disagree on the best form of governance—–onwhether it involves a caliph, for instance—–they com-monly demand a strong form of shari’a law upheld bythe state (Abu-Rabi 1996).

How do these approaches, and their institutionalmanifestations, help to explain not only the dearth ofdemocratization but also the exceptional cases of it,within Islam? Ironically, the political theology behindmost of the twentieth-century authoritarian regimesin Islamic states has not been religious at all, but hasbeen a rigidly secular one, framed by nationalism, eco-nomic development, and sometimes socialism. These

515

Explaining the Political Ambivalence of Religion August 2007

are Hashmi’s statist regimes. Though their leadersare Islamic—–sometimes even devout, as was Egypt’sGamal Abdel Nasser—–they sharply separate Islamfrom politics and often favor the import of Westerninstitutions, and even culture, into their countries. Act-ing out of this political theology, they have constructedintegrationist, often highly suppressive, institutions. Inmany cases, they have coopted a moderate Islamic fac-tion, established it as official, and offered it economicand legal support while monitoring its governance andpreventing its political assertion—–consensual integra-tionism. They have marginalized more conservativeor radical Islamic movements by outlawing or pre-venting their political participation—–conflictual inte-grationism. The standard bearer of this kind of regimewas Kemal Ataturk, who established the Republic ofTurkey in 1923 on the basis of nationalism, equality, andsecularism, abolished the last caliphate, and stronglyrestricted the expression of Islam in politics, education,and culture. After World War II, secular authoritarian-ism prevailed as the model for postcolonial Muslimstates in Egypt, Syria, Morocco, Libya, the Iran ofthe Shah, Saddam Hussein’s Ba’athist Iraq, Suharto’sIndonesia, Tunisia, Algeria, Morocco, Turkey, Jordan,Kuwait, Yemen, and in most of the post-Soviet regimesin the Central Asian republics.

These statist, integrationist institutions have con-tributed to a correspondingly undemocratic posture inIslamic movements that have arisen to oppose them. Itis too strong to argue that statist Islamic regimes havespawned radical Islamic revivalists ex nihilo, but cer-tainly these regimes have radicalized already conser-vative Muslim movements by suppressing their legal,nonviolent participation, convincing them that theirregime is hostile to Islamic justice, and sequesteringthem from the moderating influences of democraticcompetition, compromise, and public argument. Inmany cases, regimes and opposition movements haveplayed what Almond, Appleby, and Sivan refer to asa continual cat and mouse game of suppression, loos-ening of control, resurgence, and further suppression(Almond, Appleby, and Sivan 2003, 217–18). This dy-namic of suppression and radicalization, where in-stitutions foster ideas, is observable all across statistregimes, but most recently in Central Asia, where post-Soviet secular authoritarian governments’ curtailmentof the freedoms of Muslims, including nonmilitant andeven nonpolitical ones, has spawned the rise of mili-tant Islamic revivalists throughout the region (Hunter2001).

The other sort of Islamic integrationist regime isone where integrationist movements have triumphedover statist governments. In the last quarter century,several movements with a Radical Islamic Revivalistpolitical theology have succeeded either in revoltingand establishing their own highly integrationist regimeor in pressuring their ruling regime to promote theirpolitical theology through integrationist institutions.Here, ideas clearly shape institutions. The standardbearer is revolutionary Iran, where a Shi’ite move-ment overthrew the statist regime of the Shah in 1979,established a regime as integrationist as any in the

world, and called for similar revolution throughout theumma. Such regimes have also reigned in Sudan, inAfghanistan under the Taliban, in 12 out of 36 statesin Nigeria, and in Saudi Arabia, where Islamists exertstrong influence over the monarchy.

But Islam is not authoritarian everywhere. Muslimdemocratic movements exist; where they do, they againevidence the importance of ideas and institutions. InMali, one of the three “free” majority Muslim countriesin the world today, a syncretic Islamic community witha half-millennium old tradition of differentiation fromthe state in both its thinking and its practice supportedthe country’s first multiparty elections in 1992. In neigh-boring Senegal, the second of these “free” states, bycontrast, a far more integralist Islam supported demo-cratization hardly at all. Democratic Islamic move-ments have also surfaced in statist regimes like Egyptand Jordan, which relaxed some of their controls in thelate 1980s (Esposito and Piscatori 1991, 428–30).

Two other Islamic democratic movements, though,are more impressive in their size and influence.Indonesia’s Nahdlatul Ulama (NU) represents a tra-dition of what Hefner (2000) calls “civil Islam”—–aninstitutional separation between religious and politi-cal authorities complemented by a doctrine that sanc-tions such a separation and a culture of religiouspluralism—–that is at least six centuries old. In the lastgeneration, the NU’s political theology has evolvedto embrace modern democracy, which then led it topromote more differentiated institutions. Empoweredby an Islamic resurgence in the 1970s and 1980s, theNU came to join a coalition for democracy that wasinstrumental in overthrowing the dictatorship of HajiMohamed Suharto, encouraging multiparty electionsin 1999, and transforming Indonesia into the third“free” Islamic state.

In Turkey, too—–the very prototype of a statistregime—–an Islamic movement exercised force fordemocracy. Its political theology springs from the mod-ernist Islam of the Nurcu and Naksibendi movements,rooted in the urban middle class and business elite, in-fused with a Sufi spirituality, and committed to influenc-ing society and politics as a civil society actor. Thoughlong accompanied by a more integralist Islamist strain,this new, democratic, thinking has risen to dominateTurkish Islam, embodied in a succession of politicalparties that have then pursued a differentiated politicsin which they might participate democratically. But theKemalist military has long made this differentiationconflictual though its regular interventions to suppressIslam, most recently in the soft coup of 1997 in whichit overthrew a government in which the Welfare Partyhad become the first Islamic coalition partner in thehistory of the republic. In the more recent electionsof 2002, though, a new Islamic party—–the Justice andDevelopment Party (AKP)—–came to power, espous-ing loyalty to European Union standards of humanrights and democracy and indeed advocating Turkishmembership in the European Union. At least for themoment, a democratically thinking Islamic movementhas strengthened Turkey’s differentiation in the form ofdemocratic institutions (Yavuz 2003). Here, ideas have

516

American Political Science Review Vol. 101, No. 3

evolved within, and come to challenge, integrationistinstitutions.

Finally, the sources of Islamic democracy can be per-ceived in the politics of three movements—–Jamaat-e-Islami in Pakistan, the Islamic Party of Malaysia, andHamas in the Palestinian Authority—–that favor theirrespective states’ complex combination of differenti-ation in their practice of electoral politics and inte-gration in their constitutional provisions that stronglypromote Islam’s place in society and allow little reli-gious freedom (Esposito and Voll 1996, 102–23; Nasr1995). All of these movements also proclaim a politi-cal theology that sanctions such a combination and doso in the differentiated setting of electoral competi-tion. These cases of “illiberal democratic” movements(Zakaria 2004) further reveal the relationship amongideas, institutional position, and the political stances ofreligious actors.

Democratization in Hinduism and Buddhism

Hinduism and Buddhism offer far less grist for judgingthe sources of the politics of religions. They are a ma-jority in far fewer states; only a portion of these havedemocratized. Historically, they are far more eclectic inthe sources of their political theology—–and indeed oftheir general doctrines—–than Islam, Judaism, or Chris-tianity. Arguably, these traditions came to be system-atized as “world religions” at all only in the eighteenthand nineteenth centuries, and largely under Westerncolonial influence. As Kitagawa observes of Buddhism,though it contains universal principles and a moralcode, it provides few “middle principles” for politics,economics, and society (Kitagawa 1980, 100; Queen2005).

Still, both traditions contain lessons for religion andpolitics. India, the world’s largest Hindu state and,since 1947, the world’s largest democracy, is foundedon religious freedom and a “secularism” by which thestate is to remain at an “equal distance” from—–butnot uninvolved in—–the country’s religious affairs. Howdoes Hinduism appraise this arrangement? In theHindu tradition, political rule is grounded in the di-vine cosmos but differentiated from spiritual func-tions, as set forth in the Law Code of Manu and inKautilya’s Arthashastra, both dating back to the firstfour centuries, B.C. (Madan and Juergensmeyer 2005).The modern Hinduism of Mahatma Gandhi and theCongress Party have supported differentiation, too. Afar more integrationist Hinduism, centering on the ideaof a Hindu rashtra, or nation, and the idea of Hindunationalism, or hindutva, emerged in the context ofnineteenth century Indian nation-building. Developingfirst as a cultural movement, sponsored most promi-nently by the Rashtriya Swayamsevak Sangh (RSS), amilitant organization founded in 1925, Hindu nation-alism had by the 1960s become a political movement,eventually led by the Bharatiya Janata Party (BJP),founded in 1980. Though they do not contest the consti-tution’s basic secularist provisions, Hindu nationalistspromote Hinduism as a national religion and urge sym-

bolic activities like constructing a Hindu temple on thesite of the Babri Mosque in Ayodhya, the alleged birth-place of the god Ram, which Hindu rioters destroyed in1992. Once it became the dominant partner in a coali-tion government in 1998, the BJP passed laws in somestates prohibiting Christian and Muslim conversions,limited quotas for economically disadvantaged Mus-lim minorities, and even sanctioned pogroms againstMuslims in Gujarat. Here, then, is a case where a reli-gious actor’s integrationist ideas actually curtailed thedifferentiation of institutions. But it is also a case wheredemocratic institutions—–the imperatives of competingin elections and forming governing coalitions—–forcedthe same religious actor to temper its message. In 2004,the Indian electorate then voted the BJP-led coali-tion out of power at the center (Hansen 1999, 90–199;Jaffrelot 1993).

Buddhist political traditions originated roughlyaround the same time that Hindu political traditionsdid with Emperor Asoka Maurya, who conquered vasttracts of India in the third century B.C., then convertedto Buddhism and established a rule of religious toler-ation and political forbearance. The tradition of righ-teous kingship that developed in South and SoutheastAsia during the time of the European Middle Ages,though, was one that closely integrated the authorityof the monarch and the sangha, or the body of clergy.Today, Buddhism contains both differentiated and in-tegrationist orientations, often within the same state.In at least five countries, sangha and state are closelyintegrated. Integrationism may be consensual, as in SriLanka and to some degree in Thailand, where thesangha legitimizes and offers its ongoing policy ad-vice to the state while the state supports Buddhismfinancially and legally. Or it may be conflictual, as inVietnam, Burma, and Laos, where the state tightly andcoercively controls the leadership, doctrines, and activ-ities of the sangha. In their political theology, integra-tionist sanghas typically espouse either quietism, call-ing monks away from political pursuits, or a Buddhistnationalism that resembles Hindu nationalism in India.The nineteenth- and early-twentieth-century fusion ofBuddhism with nationalism in Sri Lanka formed thebasis of a close integration of sangha and state anda suppression of the minority rights and autonomy ofHindu Tamils after independence in 1948, a clear caseof ideas shaping institutions. In the past half-century,though, a competing doctrine, “Engaged Buddhism,”has arisen, which combines ancient Buddhist notionsof peace and toleration with Western political con-ceptions like human rights, democracy, nonviolence,and environmentalism. Found in Cambodia, Vietnam,Thailand, Korea, Japan, and Taiwan, communities withthis view have sought to shape the policies of theirgovernments, sometimes in the face of intense oppo-sition (Queen 2005; Swearer 1995, 66–75). Like otherfaiths, Buddhism manifests both likenesses and varia-tions.

Arrayed around political theology and differenti-ation, these likenesses and variations explain simi-larities and differences in democratization, not onlyamong religious traditions but also among actors within

517

Explaining the Political Ambivalence of Religion August 2007

religious traditions. The most comprehensively con-trasting cases are the Catholic Wave and contempo-rary Islam. The Catholic Wave illustrates the power ofnew ideas to transform a religion globally, though withdifferent effects on national churches depending onhow extensively they embraced the ideas and on howdifferentiated they were. Islam, by contrast, has experi-enced no such global onset of liberal democratic ideas.Integrationism, either secular nationalist or radical re-vivalist, is far more common. Democratizing Islamicactors are correspondingly far rarer. Where they doexist, though, they demonstrate the influence of lib-eral democratic political theology and differentiation.Orthodox churches, generally nondemocratizing, andthe Hindu BJP party, which has sought to dilute liber-alism in India, offer negative evidence for the influenceof ideas and institutions, whereas Protestant churchesand Buddhist sanghas illustrate this influence compar-atively through the varying stances of their variouscommunities.

POLITICAL VIOLENCE

As the headlines continually cry, religion has a darkervalence, one most vividly unveiled when believerstake up the gun, both on the large scale of com-munal conflict—–in Sudan, Afghanistan, Kashmir, andChechnya—–and through the more targeted violenceof terrorism. Might their political theology and degreeof differentiation from the state help explain why re-ligious communities engage in these deadly pursuits?The logic mirrors what brings religious bodies to en-courage democracy, a form of regime, after all, based onthe nonviolent resolution of disputes. Religious com-munities are prone to violence when they hold a polit-ical theology that interprets their scriptures, traditions,and divine commands so as to favor an integrationiststate, one that both makes its religion official and sup-presses other faiths. They also tend toward belligerencewhen they are faced with laws and institutions—–eithersecular or sponsored by another faith—–that suppresstheir own practice and expression. Either cause mayoperate alone, but the two may also interact, reinforc-ing each other. The obverse manifestation of the samefactors that explain religious actors’ pursuit of demo-cratic politics—–ideas and institutions—–also account fortheir pursuit of lethal politics.

Communal Conflict

Sometimes, religious violence occurs on a large scale,involving popular insurrections, rebel militia offen-sives, and their suppression by the armed forces ofthe state, all of these justified, urged, and sanctified bythe honor, ideals, and claims of religious communities.Every religion on the planet has carried out such vio-lence. Toft conceives this type of conflict as “religiouscivil war,” involving at least two groups, one of whom isthe state, a contest over the governance of the politicalunit, and at least 1,000 annual battle deaths on average(Toft 2007, 112–113). Another political scientist, Fox,drawing on the Minorities at Risk data set, identifies

them as “ethnoreligious conflicts,” a subset of ethnicconflicts where the warring groups are of different re-ligions (Fox 2002, 70–71; Gurr 2000).

Toft reports that from 1940 to 2000, 42 out of 133 civilwars, or 32%, have involved religion (Toft 2007, 97).According to Fox (2002, 71), 105 out of 268 disputesinvolving ethnic minorities, or 39%, are ethnoreligious.Religious conflicts have only increased in proportion.In the 1940s, by Toft’s measure, they made up 19%of civil wars, then 29% in the 1950s and 21% in the1960s. But in the 1970s, they rose to 36%, then rose to39% in the 1980s, then rose up to 43% in the 1990s.Since 2000, 50% of civil wars have been religious(Toft 2006, 9) Both Fox and Toft report that Islamappears disproportionately often in religious conflicts.Of the 42 civil wars that Toft describes as religious,30, or 71%, involved Islamic practice as an issue (Toft,personal communication). In 34 of the 42 religious civilwars, or 81%, one or both parties were Muslim. Of allstates that have fought civil wars, 58% have majorityMuslim populations (Toft 2007, 113–114). And inthose 10 religious civil wars fought between groups ofthe same faith, nine have involved Muslims (Toft 2006,15). Fox also finds strong evidence for a rise in conflictamong Muslims during the 1990s (Fox 2004, 68).

What exactly does it mean that a conflict is reli-gious? Analysts often debate whether a war betweenreligious communities in Northern Ireland, Bosnia, SriLanka, or Kashmir is “really” religious or rather aboutsomething else—–land, oil, ethnicity, or historical mem-ories. In fact, religion fuels conflict in two broad ways.First, it shapes the identities and loyalties of warringcommunities—–Serbs, Northern Irish Loyalists, or Bud-dhist Sinhalese. This influence is pervasive: Juergens-meyer (1993) argues that all over the world both beforeand after the end of the Cold War, religion becamefused with nationalism, replacing an earlier secular na-tionalism with new loyalties and motives for conflict.And it can be powerful: Sells (1996) has shown how, inYugoslavia’s wars of the 1990s, nationalist demagoguesused religious language and symbols to inflame the bel-licosity of communities whose faith had become “folkreligion”—–theologically desiccated, but rich in ritual,lore, and ethnicity. Once religion has shaped communalidentity, something other than religious aims then mo-tivates conflict—–nationalist campaigns for autonomyor independence, memories of historical injustice, andeconomic ends. Political theology, here, plays little role.Religion is a source of communal loyalty and not aset of propositions about right authority. Of Toft’s 42post-1940 religious civil wars, 17, or 40%, were of thisvariety: religion shapes identity (Toft 2007, 103).

In a second logic, religion fuels conflict more directlyby defining not only the identities and loyalties of com-munities, but also their very political goals, which thenbecome casus belli. In some conflicts, religion evendefines ends but not communities—–intra-Muslim dis-putes in Iran and Algeria, for example. It is in thoseconflicts where religion also defines ends that politicaltheology matters. Toft reports that 25 out of 42, or60%, of religious civil wars are of this sort: religionshapes ends as well as identities (Toft 2007, 103). It is

518

American Political Science Review Vol. 101, No. 3

the integrationist character of these ends that forcesan armed clash. One or more belligerents—–the stateor the opposition—–promotes a political theology thatdemands the denial of differentiation, calling for aregime that discriminates against a separate, usuallyminority, religious group. My own analysis of these 25conflicts shows that 18, or 72%, involved at least onecombatant with integrationist goals. The importance ofintegrationist political theology suggests, in turn, whyIslam—–whose history has yielded globally widespreadintegrationist thinking—–is represented so largely in re-ligious conflicts. Toft notes that in cases where Islamwas involved in a religious civil war, 18 out of 34, or53%, of conflicts entailed religion as a central issue,whereas when other religions were involved, religionwas a central issue in only three out of 11, or 27%, ofcases (Toft, personal communication). Similarly, Foxdiscovers a sharp increase in the role of religious issuesin conflicts between Muslims over the years from 1965to 2001 (Fox 2004, 69).

In some conflicts, religious ends and identities min-gle. Kashmir’s many separatist groups, for instance,vary among secular nationalist and radically integra-tionist, those that want independence and those thatwant to join Pakistan, all of them differing in theirmethods, goals, and longevity (Sikand 2004). In manyconflicts, religious ends and identities mix with non-religious goals, too—–economic greed and grievance,nonreligious discrimination, and self-determination.

Where conflict involves a religious end, it is commonfor integrationist groups, if they succeed in capturingthe state, to impose a set of integrationist laws and insti-tutions that suppress or discriminate against a rival re-ligious group, causing it to fear for its faith: ideas shapeinstitutions. The logic is similar to that by which au-thoritarian regimes cause opposition groups to be lessdemocratic; here, these groups become violent as well.Ruling integrationist groups often will be what BruceLincoln calls “religions of the status quo,” ones desiringa symbiosis of a state that promotes faith and religiousauthorities who hold political powers (Lincoln 2003,79–83). Twelve out of 25 conflicts over religious ends,according to my assessment of Toft’s cases, involvedintegrationist states. Civil war raged for 21 years inSudan, where an authoritarian Islamist state sought toimpose a harsh version of shari’a law on the Christiansouth. In Sri Lanka, civil war between Buddhists andminority Hindu Tamils results partly from the fusionof religion and nationalism in Sinhalese Buddhism, butalso a political theology by which the sangha wieldsformidable political authority while the state promotesa Buddhist homeland through its constitution, its edu-cation policies, and its funding of religious activities.Although Tamil separatism is itself largely secular, it iscrucially propelled by the Tamils’ marginalization—–politically, economically, religiously—–in a Buddhiststate (Tambiah 1993).

Integrationist regimes that beget violence throughmarginalizing religious groups are not always them-selves religious. Communist regimes, for instance, holda political theology of their own, a set of presuppo-sitions about religion that beget a low and conflict-

ual differentiation. When the Soviet Union invadedAfghanistan in 1979 to sustain a Communist coupagainst an Islamic regime, it provoked an 11-year up-rising that united traditional Afghan Islamic leaderswith radical revivalist jihadis from all over the Muslimworld (Roy 1993, 492). Similarly, China’s Communistregime suppressed Tibetan Buddhists, who respondedwith civil war during the early 1950s.

Integrationist regimes might also be secular or mildlyreligious non-Communist dictatorships, whose ideo-logy also shapes the nature of their institutions. Inpostcolonial Algeria, the authoritarian government ofthe National Liberation Front sought to control Is-lam by allying with a reformist sect, promoting it andregulating it while marginalizing more conservativeforces. In reaction to this as well as to economic under-performance and political corruption, Islamist move-ments rose up and gained political strength during the1970s and 1980s. When the government finally allowednational elections in 1991, the Islamic Salvation Frontgained victory in the first round, only to have the gov-ernment cancel the second round and revive militaryrule, bringing a civil war that is estimated to havetaken over 100,000 lives in the 1990s (Malley 1996).The authoritarian government of Iran under the Shahsimilarly sought to suppress forms of Islam that chal-lenged its modernizing, socially liberalizing goals, giv-ing rise to the Islamism that triumphed in the 1979revolution.

In many conflicts over religious ends, oppositiongroups are themselves integrationist, also driven byideas. In Lincoln’s terms, they arise as “religions ofresistance” and then become “religions of revolution”(Lincoln 2003, 82–91). They are found in the abovementioned struggles in Algeria, Iran, and Afghanistan,as well as in Muslim opposition forces in Central Asianrepublics and in Chechnya. In not every case is the statewhich they oppose integrationist—–witness the MoroIslamic Liberation Front, which fights the Philippinesgovernment, or Islamic revivalist groups in Kashmirwho war against India. Sometimes it is other grievancesthat propel them. Of the 21 post-World War II conflictswhere religion has shaped ends, my research into Toft’scases reveals that nine have involved opposition groupswith an integrationist political theology, all of themMuslim.

Terrorism

Terrorism is distinguished by its killing of civilians forpolitical purposes (Hoffman 1998, 13–44). It may occurwithin the course of communal violence or entirelyseparately. It is usually more episodic and targeted thancommunal violence, though it may well be repeated andits targets sizable and widespread. Though terrorismmay be practiced by governments, the focus here is onopposition groups.

Religious terrorists are those who declare for them-selves religious aims and identities. In fact, their pur-suits are rarely limited to religion, as Jessica Sterncautions, but range from spiritual to temporal, from

519

Explaining the Political Ambivalence of Religion August 2007

ideological to profit-driven, and from instrumental,where terrorism is a means to a concrete end, to expres-sive, where terror is itself a mode of communicatinga message (Stern 2003, 6–8). But even in this rangeof activities, religious terrorists always proclaim theirreligious purposes.

Until the nineteenth century, religious terroristswere the only sort of terrorists there were. The En-glish words assassin, thug, and zealot indeed arisefrom ancient and medieval Islamic, Hindu, and Jew-ish terrorism (Rapaport 1984). By 1968, though, secu-lar ends had supplanted religious ones as the motiveof all of the world’s 11 terrorist groups, according toHoffman (1998). It was not until 1980 that religiousterrorists inched their way back into global politics,then amounting to two out of 64 terrorist groups in theworld. Then, their numbers began to swell. By 1992, 11terrorist groups were religious; in 1994, 16 of 49, or 33%of groups fit the description; by 1995, the numbers hadclimbed to 26 out of 56, or 46% of all terrorist groups(Hoffman 1998, 90–94). Today, according to my exam-ination of the Terrorism Knowledge Base, 95 out of262, or 36% of known terrorist groups, are identifiablyreligious.3

Why do religious terrorists take up the gun?Scholars including Stern (2003); Ranstorp (1996);Juergensmeyer (2003, 183–85); Pape (2003); andAlmond, Appleby, and Sivan (2003) emphasize a rangeof motivations: adventure; profit; heavenly reward orretribution; the alienation and humiliation of popu-lations on the down side of globalization, urbaniza-tion, and economic progress; legacies of colonialismand Western imperialism; oppressive political regimesthat choke off religious and political expression; for-eign military occupations; the breakdown of secularnationalism as a source of legitimacy and loyalty; andthe loss of masculinization.

But each of these analysts likewise recognizes thatsuch motivations fall short of accounting for reli-gious terrorism. Poverty, oppression, and the like applyto whole populations that dwarf the tiny pockets ofpeople who form and join religious terrorist groups.Nor do these factors explain diversity among militantmovements. The variegated separatist movements ofKashmir, for instance, exist in a common political, eco-nomic, and demographic environment that does littleto explain their differences.

Most analysts, then, also stress the centrality of reli-gious terrorists’ beliefs, that is, how they interpret theirtexts, their traditions, their truths, and their historicalmoment so as to take up violence urgently, vehemently,and without regard for traditional laws of war. Severalmotifs recur: religious terrorists commit violence as asacramental or divine duty; they claim divine sanctionto commit indiscriminate killing on a grand scale; theirconstituency, or audience, is often their own follow-

3 The knowledge base is presented by the National Memorial Insti-tute for the Prevention of Terrorism (MIPT), and can be found athttp://www.tkb.org/Home.jsp. I performed my analysis of its data inJuly 2005. These data do not necessarily correspond to the data fromwhich Hoffman derives his numbers.

ers; they view themselves as opposed not merely to agiven regime’s policies, but to an entire, irrevocablycorrupted order; often they act out of an apocalyp-tic vision (Hoffman 1998, 94–95). Near the center ofvirtually every religious terrorist group’s beliefs alsolies a political theology. They believe that one or moreregimes is illegitimate for having defiled and failed topromote authentic faith, and should be replaced byone where political authority is tightly meshed withreligious authority, which actively promotes right re-ligion, and that thereby subordinates other religiouscommunities.

Empirical patterns bear out political theology’s im-portance. My analysis of the Terrorism KnowledgeBase shows that 93% of all religious terrorist groupshold an integrationist political theology. They havetaken up the gun to replace corrupted, secularizedorders with ones where political authority is rightlyoriented. Whatever their differences, they share theseideas about politics, which then lead them to changeinstitutions. Hoffman quotes a description of Iranian-backed Shi’ite terrorists: “[They] do not believe in thelegitimate authority of secular governments . . . SinceIran is the only state to have begun to implement ‘true’Islam, however, it is thought to be the world’s only legit-imate state with a unique obligation of facilitating theworldwide implementation of Islamic law. Force andviolence are not only acceptable but necessary meansof doing so” (Hoffman 1998, quoting Zonis and Brum-berg 1984). For all Radical Islamic Revivalists, politicalauthority consists either of a caliphate or of an Islamicgovernment that enforces a strongly integrationist ver-sion of shari’a law. Currently, fully 91% of all religiousterrorist groups are Radical Islamic Revivalist, theirideas drawn from this movement’s critique of Islam andits relationship to the modern world. Christian whitesupremacist groups in the U.S. and Jewish extremistmovements like the Kach movement of Rabbi MeirKahane hold roughly equivalent notions.

If political theology is behind religious terrorism,then what of differentiation? Perhaps religious terror-ism, like communal violence, arises through the samelogic by which integrationist regimes make oppositiongroups less democratic. Almond, Appleby, and Sivan(2003) make just such a case: when states prevent re-ligious groups from expressing their doctrines, raisingfunds, and recruiting, such groups are likelier to turn vi-olent. Regimes may sometimes succeed in suppressingthem, but they rarely do away with these phoenixes. Tobe sure, oppressive regimes do not alone give rise to re-ligious terrorists, many of whose similarly situated fel-low believers do not choose violence. But such regimesdo encourage religious terrorists. The obverse side ofAlmond, Appleby, and Sivan’s logic is that religiousterrorists will be far rarer under differentiated institu-tions, that is, democracies. Here, the arguments aboutdemocracy and political violence converge. Within ademocracy, religious groups can operate but will alsoface exposure, competition, and the imperative of ally-ing with moderates, all of which temper, attenuate, andeven factionalize these groups (Almond , Appleby, andSivan 218).

520

American Political Science Review Vol. 101, No. 3

The evidence that democracy diffuses terrorism,however, is mixed. Some scholars even conclude theopposite. Pape (2003), for instance, finds that suicideterrorist attacks are more likely to be directed againstdemocracies than against authoritarian regimes, andthat since 1980 all such campaigns have been directedagainst democracies. His argument, however, is notthat democratic regime structures breed terrorists, butthat terrorists attack states whom they perceive asoccupying their lands, which happen to be democra-cies. His sole focus on suicide terrorists also limitsgeneralizations about terrorism and democracy fromhis argument. More straightforwardly, Weinberg andEubank (1998) found that during 1994 and 1995, ter-rorism occurred more often in democracies than else-where. Gause (2005) corroborates the claim with aState Department Report showing that between 2000and 2003, 269 major terrorist incidents took place incountries that Freedom House ranks as “free,” whereas199 occurred in “partly free” countries and 138 in “notfree” countries.

Against these findings, though, and supportingAlmond, Appleby, and Sivan’s (2003) argument, isAlberto Abadie’s (2004) report showing a strong cor-relation between terrorist risk and authoritarianism,an inverse correlation between terrorism and politicalfreedom, and a lack of correlation between terrorismand poverty, once regime type is factored into the anal-ysis. He cautions, though, that terrorist risks are higherin countries that are between authoritarian and free,suggesting the vulnerability of the transition to democ-racy. The link between terrorism and democracy, then,is ambiguous. Gause (2005) concludes that on balance,the evidence is inconclusive in either a positive or anegative direction.

Perhaps a connection between authoritarianism andterrorism, though, emerges less in where terrorists op-erate than in where they incubate. Even if terrorists tar-get democracies, they may well emerge in authoritariansettings. The claim indeed finds support in a FreedomHouse study that connects the lack of political rightsand civil liberties to the origins of terrorist movements.Between 1999 and 2003, 70% of deaths attributable toterrorism were caused by terrorists whose origins lie in“not free” countries. In comparison, only 8% of deathsfrom terrorism were caused by terrorists with originsin “free” countries (Freedom House 2005).

None of these findings, though, specifically focuseson religious terrorists, who, again, make up 36% of allterrorists. Given their religious ends, one might expectauthoritarian regimes, especially integrationist ones, torouse them especially strongly. An analysis of the Ter-rorism Knowledge Base shows a positive relationshipbetween authoritarianism and religious terrorism withrespect to the site of operation: of 95 current religiousterrorist groups, only 31, or 32%, operate in “Free”countries, whereas 42, or 43%, operate in “Not Free”countries, and 20, or 21%, in “Partly Free” countries.Evidence that the countries where these groups oper-ate are integrationist is also found in the InternationalReligion Indexes of Grim and Finke (2006). The“Government Regulation Index” of these countries,

which measures government interference in religionon a scale from 0 to 10, averages out to 5.75, com-pared to an average of 3.07 among the total of 196countries in the data set; their “Government FavoritismIndex,” which measures direct government support forreligions, averages 6.92 for countries where terroristsoperate, compared to 4.34 for the entire dataset. Datashowing where religious terrorists originate are unfor-tunately scarce.

A form of evidence for the integrationist sourcesof religious terrorism is less direct and moresubjective—–the perceptions of religious terroriststhemselves. Recall that 91% of them are RadicalIslamic Revivalists. All revivalists believe that out-siders are attacking and eroding their faith (thoughthey identify internal sources of decline, too). Eventhough not all of today’s revivalist groups originated oroperate in an oppressive setting, they commonly tracetheir parentage to intellectuals who perceived all ofIslam in a state of defensive embattlement—–Maududi,al-Banna, Qutb. In part, what revivalists want to de-fend are Muslim homelands. The Terrorist KnowledgeBase shows that 32% of today’s religious terroristgroups mix their religious ends with self-determination;11% of them fight for the Palestinian cause. But re-vivalists also hold that oppressive governmental in-stitutions prevent the realization of Islamic law andmorality. They identify such institutions historicallyas colonial regimes imposed by the West and con-temporarily as secular, nationalist, authoritarian Arabregimes and communist regimes—–the strongest em-bodiments of integrationism. It is in response towhat they perceive as attempts to marginalize themand their vision that revivalist terrorists wage theirjihad.

Political theology and institutions, both of an in-tegrationist sort, then, plausibly fuel the pursuit ofboth communal violence and terrorism among reli-gious actors. Communal violence is advanced by groupswith integrationist political theologies, sometimes sec-ular in character, who capture the state and imposeintegrationist institutions upon minority faiths whothen rebel, and by integrationist religious groups whotake up opposition to states. Religious terrorism re-sults from integrationist political theology, the sup-pression of religious communities by integrationistauthoritarian regimes, and the perceptions of reli-gious terrorists themselves that they operate in hostilesurroundings.

The analysis of political violence is shorter than thatof democratization. In part, this is because it draws onthe description of the historical development of ideasand institutions in the world religions that appears inthe democratization section. But it is also due to arelatively thinner literature on religion and violence,especially when it involves religious ends and identi-ties rather than identities alone. Additional researchis needed, then, on the role of political theology andintegrationism in particular cases of religious violence,on global trends such as the nature of regimes wherereligious terrorists originate, and on the sources ofpolitical violence in Islam.

521

Explaining the Political Ambivalence of Religion August 2007

CONCLUSIONS

Religion matters. More specifically, religions matter.Variations within and between local religious commu-nities matter, too. Two features of these communitiesexplain much about their politics.

Political theology and the relationship between re-ligion and state will not explain everything. Amongviolent conflicts, they explain only those where re-ligious ends, not just identities, matter. In explain-ing both violence and democratization, other aspectsof religious communities matter, too—–their size andthe centralization of their hierarchy, for instance.Some evangelical and Pentecostal communities inLatin America and Africa have held liberal demo-cratic ideas but were too small or decentralized toshape their state. Ethnic and religious pluralism makesa difference as well (Dowd 2006). Finally, religios-ity itself is important—–the degree to which membersadhere to their community’s beliefs and participatein its activities. The Catholic Church in CommunistPoland, a famous democratizer, stood out for the de-votion of its members, whereas the Catholic Churchin Czechoslovakia, a comparatively passive democra-tizer, registers the lowest rates of religiosity on theplanet, at least in what is now the Czech Republic (PewGlobal Attitudes Study 2002). But these variables alsohave their limitations. The importance of religiosity,for instance, is confounded by the Greek OrthodoxChurch, whose members are among the most religiousin the world, but which failed to support democra-tization. It is this church’s statist political theologyand lack of differentiation from the state that explainwhy.

At a time when religion’s impact on politics has be-come contentious both in the United States and aroundthe world, what lessons arise from the importance ofpolitical theology and differentiation? The normativestatus of a political theology and set of institutions, ofcourse, depends on a given evaluator’s commitments.Where the analysis can offer firmer conclusions is inthe related matter of normative stability. That is, whatconfiguration of political theology and institutions ismost likely to attain lasting legitimacy?

The most troubled configuration is conflictual inte-grationism. Here, a regime with an integrationist po-litical theology suppresses religion, denying its auton-omy and even more so its political participation. Re-ligions accept this denial only because they are forcedto. Such regimes are typically secular. Sometimes theyruthlessly suppress religion, as Communist regimes inBulgaria, Romania, Russia, and Czechoslovakia did forover a generation. In other cases, as with most Arabnationalist regimes and Communist China, they per-mit religious groups only insofar as they conform tothe regime’s ends. But if religion can come to resist,conflictual integrationism will totter. Some oppositiongroups will radicalize their own integrationist ends andbecome violent, as Radical Islamic Revivalism has overthe past generation. But other groups, usually ones witha liberal democratic political theology, will succeedthrough nonviolent struggle in securing a measure of

autonomy despite the regime’s opposition. They trans-form a country’s institutional configuration to one ofconflictual differentiation.

Consensual integrationism, an undemocratic andilliberal arrangement in which religion and state aremutually and contentedly meshed in their institutionalauthority, might even last longer, as it did in medievalChristendom or colonial Latin America, or at least im-pressively long, as it has in Saudi Arabia, Iran, andSri Lanka. It is consensual, after all. But two sorts ofthreats can undermine it. One is a minority or dissent-ing religious group suppressed by the integrationistregime—–Sri Lankan Tamils, Sudanese Christians,Tibetan Buddhists, or Islamic revivalists under Arabregimes. The result is often a violent conflict that movesthe country to a state of conflictual integrationism. Theother is a change in political theology so that at leastone party no longer favors the arrangement. Within acentury after Latin American countries attained theirindependence, liberal parties managed to win dises-tablishment and religious freedom. Gradually, nationalchurches themselves came to favor the arrangement,solidifying a consensual differentiation in which theypursued their favored laws and polices through infor-mal links.

Conflictual differentiation is contested by definition.Here, religion has carved out significant autonomy inits own governance and practice and challenges itsregime in the name of its liberal democratic politi-cal theology, usually non-violently. The regime resists,either because it desires integrationism, as KemalistTurkey did, or is differentiated but still authoritarian,as were Latin American military dictatorships in the1960s and 1970s. The contest may last for decades,but if and when the religion’s struggle succeeds, asit did in many countries in Eastern Europe, LatinAmerica, Africa, and Asia during the Third Wave,the result is the consensual differentiation of liberaldemocracy.

Consensual differentiation is likely to be the mostnormatively stable. By definition, it is uncontested.But unlike consensual integrationism, it guaranteesreligious freedom, and so is less likely to provoke therevisionism of minorities. It is most likely to result whenreligious communities hold a liberal democratic politi-cal theology. They renounce institutional prerogatives,which they might otherwise regard as guarantees oftheir preferred politics, and permit minority religionsto compete for followers and influence. But they do notabjure political influence, which they may now exertfrom a differentiated position. The state, for its part,agrees to respect the autonomy of religious groupsand to allow their participation in politics, even whengroups promote ends that some elites disfavor. Suchmutual indulgence and forbearance are Stepan’s (2001)twin tolerations. Consensual differentiation, though,can be fragile. If a religious group becomes inte-grationist in its political theology and politically em-powered, it can elicit a more integrationist state, asHindu nationalists in India have attempted. Statesthemselves become more integrationist when theymore and more restrict religion, as France has done

522

American Political Science Review Vol. 101, No. 3

through its policies of “laıcite.” Consensual differen-tiation requires a religion that seeks influence, butnot standing constitutional authority, and a state thatallows its religious communities—–all of them, includingminorities—–to practice and participate. Herein lies thepossibility of liberal democracy.

REFERENCES

Abadie, Alberto. 2004. Poverty, Political Freedom, and the Roots ofTerrorism. Cambridge, MA: Harvard University Press and NBER.

Abu-Rabi, Ibrahim. 1996. Intellectual Origins of Islamic Resurgencein the Modern Arab World. Albany, NY: State University of NewYork Press.

Albright, Madeleine. 2006. The Mighty and the Almighty: Reflec-tions on America, God and World Affairs. New York, NY: HarperCollins Publishers.

Almond, Gabriel, R. Scott Appleby, and Emmanuel Sivan. 2003.Strong Religion: The Rise of Fundamentalisms Around the World.Chicago, IL: University of Chicago Press.

Appleby, R. Scott. 2000. The Ambivalence of the Sacred: Reli-gion, Violence, and Reconciliation. Lanham, MD: Rowman andLittlefield.

Berger, Peter. 1967. The Sacred Canopy: Elements of a SociologicalTheory of Religion. New York: Random House.

Berger, Peter L. 1999. The Desecularization of the World: ResurgentReligion and World Politics. Washington, DC: Eerdmans/Ethicsand Public Policy Center.

Billington, James. 1999. “Orthodox Christianity and the RussianTransformation.” In Proselytism and Orthodoxy in Russia: TheNew War of the Souls, ed. John Witte and Michael Bourdeaux.Maryknoll, NY: Orbis.

Borer, Tristan Anne. 1998. Challenging the State: Churches As Politi-cal Actors in South Africa, 1980–1994. Notre Dame, IN: Universityof Notre Dame Press.

Byrnes, Timothy A. 2001. Transnational Catholicism in Postcommu-nist Europe. Lanham, MD: Rowman and Littlefield.

Casanova, Jose. 1994. Public Religions in the Modern World. Chicago,IL: University of Chicago Press.

Cleary, Edward. 1997. The Struggle for Human Rights in LatinAmerica. Westport, CT: Praeger.

Donno, Daniela, and Bruce Russett. 2004. “Islam, Authoritarianism,and Female Empowerment: What Are the Linkages?” WorldPolitics 56 (July): 582–607.

Dowd, Robert. 2006. “Religious Pluralism and Peace: Lessons FromSub-Saharan Africa,” ms.

Elazar, Daniel. 1995. Covenant and Polity in Biblical Israel: BiblicalFoundations & Jewish Expressions. New Brunswick, NJ: Transac-tion Publishers.

Esposito, John L., and James Piscatori. 1991. “Democratization andIslam.” The Middle East Journal 45 (Summer): 427–40.

Esposito, John L., and John O. Voll. 1996. Islam and Democracy.New York: Oxford University Press.

Fish, M. Steven. 2002. “Islam and Authoritarianism.” World Politics55 (October): 4–37.

Flannery, Austin, O. P., ed. 1975. Vatican Council II: The Conciliarand Post Conciliar Documents. Northport, NY: Costello PublishingCompany.

Fleet, Michael, and Brian H. Smith. 1997. The Catholic Church andDemocracy in Chile and Peru. Notre Dame, IN: University ofNotre Dame Press.

Fox, Jonathan. 2002. Ethnoreligious Conflict in the Late 20th Century:A General Theory. Lanham, MD: Lexington Books.

Fox, Jonathan. 2004. Religion, Civilization, and Civil War: 1945Through the Millennium. Lanham, MD: Lexington Books.

Fox, Jonathan. 2006. “World Separation of Religion and State Intothe 21st Century.” Comparative Political Studies 39 (June): 537–69.

Fox, Jonathan, and Shmuel Sandler. 2005. “Separation of Religionand State in the Twenty-First Century: Comparing the MiddleEast and Western Democracies.” Comparative Politics 37 (April):317–35.

Freedom House. 2005. Freedom in the World 2005: The AnnualSurvey of Political Rights and Civil Liberties.

Freedom House. 2006. Freedom in the World 2006: The AnnualSurvey of Political Rights and Civil Liberties.

Freedom House Press Release. 2004. Russia Downgraded to ‘NotFree.’

Freston, Paul. 2001. Evangelicals and Politics in Asia, Africa, andLatin America. Cambridge, UK: Cambridge University Press.

Freston, Paul, Editor. 2007. Evangelicals and Democracy in LatinAmerica. New York, NY: Oxford University Press.

Gause III, F. Gregory. 2005. “Can Democracy Stop Terrorism?”Foreign Affairs 84 (September/October): 62–76.

Geddes, Barbara. 1999. “What Do We Know About Democratiza-tion After Twenty Years?” Annual Review of Political Science 2(June): 115–44.

George, Alexander. 1979. “Case Studies and Theory Development:The Method of Structured, Focused Comparisons.” In Diplomacy:New Approaches in History, Theory and Policy, ed. Paul GordonLauren. New York: Free Press.

Gifford, Paul. 1995. The Christian Churches and the Democratisationof Africa. Leiden: E.J. Brill.

Gill, Anthony. 1998. Rendering Unto Caesar. Chicago, IL: Universityof Chicago Press.

Grim, Brian J., and Roger Finke. 2006. “International ReligionIndexes: Government Regulation, Government Favoritism, andSocial Regulation of Religion.” Interdisciplinary Journal of Re-search on Religion 2 (1): 3–40.

Gurr, Ted Robert. 2000. People Versus States: Minorities At Risk inthe New Century. U.S. Institute of Peace: Washington, DC.

Hansen, Thomas Blom. 1999. The Saffron Wave: Democracy andHindu Nationalism in Modern India. Princeton, NJ: PrincetonUniversity Press.

Hashmi, Sohail. 1998. “Islamic Ethics in International Society.” InInternational Society: Diverse Ethical Perspectives, ed. David R.Mapel and Terry Nardin. Princeton, NJ: Princeton UniversityPress.

Hastings, Adrian. 1997. The Construction of Nationhood: Ethnicity,Religion and Nationalism. Cambridge, UK: Cambridge UniversityPress.

Hefner, Robert. 2000. Civil Islam: Muslims and Democratization inIndonesia. Princeton, NJ: Princeton University Press.

Hehir, J. Bryan. 2005. “The Roman Catholic Church and WorldOrder Issues: Ideas, Structures, and Relationships.” Report pre-pared for the Religion in Global Politics project, Harvard Uni-versity.

Hoffman, Bruce. 1998. Inside Terrorism. New York: ColumbiaUniversity Press.

Hunter, Preston. 2007. “Major Religions of the World Rankedby Number of Adherents.” http://www.adherents.com/ReligionsBy Adherents.html (June 13, 2007).

Hunter, Shireen. 2001. “Religion, Politics, and Security in CentralAsia.” SAIS Review 21 (Summer/Fall): 65–89.

Huntington, Samuel P. 1991. The Third Wave: Democratization in theLate Twentieth Century. Norman, OK: University of OklahomaPress.

Jaffrelot, Christophe. 1993. The Hindu Nationalist Movement inIndia. New York: Columbia University Press.

Juergensmeyer, Mark. 1993. The New Cold War? Religious Nation-alism Confronts the Secular State. Berkeley, CA: University ofCalifornia Press.

Juergensmeyer, Mark. 2003. Terror in the Mind of God: The GlobalRise of Religious Violence. 3rd ed. Berkeley, CA: University ofCalifornia Press.

Kalyvas, Stathis. 2000. “Rendering unto Caesar: The Catholic Churchand the State in Latin America.” American Political Science Review94 (March): 213–14.

Kalyvas, Stathis. 1996. The Rise of Christian Democracy in Europe.Ithaca, NY: Cornell University Press.

Karatnycky, Adrian. 2002. “Muslim Countries and the DemocracyGap.” Journal of Democracy 13 (Janaury): 99–112.

Karatnycky, Adrian. 2005. Civic Power and Electoral Politics,Freedom House.

Keck, Margaret, and Kathryn Sikkink. 1998. Activists BeyondBorders: Advocacy Networks in International Politics. Ithaca, NY:Cornell University Press.

Kedourie, Elie. 1992. Politics in the Middle East. London, UK: FrankCass.

523

Explaining the Political Ambivalence of Religion August 2007

Kellogg, Michael. 2001. “Putting Old Wine Into New Bottles:The East German Protestant Church’s Desire to ReformState Socialism.” Journal of Church and State 43 (Autumn):747–72.

Kitagawa, Joseph. 1980. “Buddhism and Social Change: An Histori-cal Perspective.” In Buddhist Studies in Honor of Walpola Rahula.London, UK: Gordon Frazer.

Klaiber, Jeffrey. 1998. The Church, Dictatorship, and Democracy inLatin America. Maryknoll, NY: Orbis Books.

Krasner, Stephen D. 1993. “Westphalia and All That.” In Ideas andForeign Policy: Beliefs, Institutions, and Political Change, ed. JudithGoldstein and Robert O. Keohane. Ithaca, NY: Cornell UniversityPress.

Kuzio, Taras, and Andrew Wilson. 1994. Ukraine: Perestroika to In-dependence. New York, NY: St. Martin’s Press.

Lapidus, Ira. 1975. “The Separation of State and Religion in theDevelopment of Early Islamic Society.” International Journal ofMiddle East Studies 6 (October): 363–85.

Legg, Keith, and John Roberts. 1997. Modern Greece: A Civilizationon the Periphery. Boulder, CO: Westview Press.

Lewis, Bernard. 1996. “Islam and Liberal Democracy: A HistoricalOverview.” Journal of Democracy 7 (April): 52–63.

Lincoln, Bruce. 2003. Holy Terrors: Thinking About Religion AfterSeptember 11th. Chicago, IL: University of Chicago Press.

Little, David. 2005. “Protestant Christianity and the Problems ofWorld Order.” Report prepared for the Religion in Global Politicsproject, Harvard University.

Lustick, Ian S. 1996. “History, Historiography, and Political Sci-ence: Multiple Historical Records and the Problem of SelectionBias.” The American Political Science Review 90 (September):605–18.

Madan, T. N., and Mark Juergensmeyer. 2005. “Notes on the Con-cept of ‘World Order’ in the Hindu Tradition and is Relevance toForeign Relations.” Report prepared for the Religion in GlobalPolitics project, Harvard University.

Mainwaring, Scott. 1986. The Catholic Church and Politics in Brazil.Stanford, CA: Stanford University Press.

Mainwaring, Scott, and Alexander Wilde. 1989. “The ProgressiveChurch in Latin America: An Interpretation.” In The ProgressiveChurch in Latin America, ed. Scott Mainwaring and AlexanderWilde. Notre Dame, IN: University of Notre Dame Press.

Malley, Robert. 1996. The Call From Algeria. Third Worldism,Revolution, and the Turn To Islam. Berkeley, CA: University ofCalifornia Press.

Manuel, Paul Christopher. 2002. “Religion and Politics in Iberia:Clericalism, Anti-Clericalism, and Democratization in Portugaland Spain.” In Religion and Politics in Comparative Perspective:The One, the Few, and the Many, ed. Ted Gerard Jelen. Cambridge,UK: Cambridge University Press.

Martin, David. 1978. A General Theory of Secularization. New York:Harper & Row Publishers.

Midlarsky, Magnus. 1998. “Democracy and Islam: Implications forCivilizational Conflict and the Democratic Peace.” InternationalStudies Quarterly 42 (September): 485–511.

Mojzes, Paul. 1992. Religious Liberty in Eastern Europe and theUSSR: Before and After the Great Transformation. Boulder, CO:East European Monographs.

Monshipouri, Mahmood, and John W. Arnold. 1996. “The Christiansin Socialism—–and After: The Church in East Germany.” Journalof Church and State 38 (Autumn): 751–73.

Nasr, S. V. R. 1995. “Democracy and Islamic Revivalism.” PoliticalScience Quarterly 110 (Summer): 261–85.

Osiander, Andreas. 2001. “Sovereignty, International Relations, andthe Westphalian Myth.” International Organizations 55 (June):251–87.

Pape, Robert. 2003. “The Strategic Logic of Suicide Terrorism.”American Political Science Review 97 (August): 343–61.

Payne, Stanley G. 1984. Spanish Catholicism: A Historical Overview.Madison, WI: University of Wisconsin Press.

Pew Global Attitudes Study. 2002.Philpott, Daniel. 2001. Revolutions in Sovereignty: How Ideas Shaped

Modern International Relations. Princeton, NJ: Princeton Univer-sity Press.

Philpott, Daniel. 2004. “The Catholic Wave.” Journal of Democracy15 (April ): 32–46.

Philpott, Daniel, and Timothy Samuel Shah. 2006. “Faith, Freedom,and Federation: The Role of Religious Ideas and Institutions inEuropean Political Convergence.” In Religion in an ExpandingEurope, ed. Timothy A. Byrnes and Peter J. Katzenstein.Cambridge, UK: Cambridge University Press.

Phiri, Isaac. 2001. Proclaiming Political Pluralism: Churches andPolitical Transitions in Africa. London, UK: Praeger.

Pierson, Paul, and Theda Skocpol. 2002. “Historical Institutional-ism in Contemporary Political Science.” In Political Science: TheState of the Discipline, ed. Ira Katznelson and Helen V. Milner.New York: Norton.

Pipes, Daniel. 2002. In the Path of God: Islam and Political Power.New Brunswick, NJ: Transaction Publishers.

Queen, Christopher. 2005. “Buddhism and World Order.” Reportprepared for the Religion in Global Politics project, HarvardUniversity.

Ramet, Sabrina P. 1998. Nihil Obstat: Religion, Politics, and SocialChange in East-Central Europe and Russia. Durham, NC: DukeUniversity Press.

Ranstorp, Magnus. 1996. “Terrorism in the Name of Religion."Journal of International Affairs 50 (Summer): 41–60.

Rapaport, David C. 1984. “Fear and Trembling: Terrorism inThree Religious Traditions.” American Political Science Review78 (September): 658–77.

Rittner, Carol, John K. Roth, and Wendy Whitworth, eds. 2004.Genocide in Rwanda: Complicity of the Churches? St. Paul, MN:Paragon House.

Roy, Olivier. 1993. “Afghanistan: An Islamic War of Resis-tance.” In Fundamentalisms and the State, ed. Martin E. Martyand R. Scott Appleby. Chicago, IL: University of ChicagoPress.

Rudolph, Susanne Hoeber. 1997. “Introduction: Religion, States, andTransnational Civil Society.” In Transnational Religion and FadingStates, ed. Susanne Hoeber Rudolph and James Piscatori. Boulder,CO: Westview Press.

Sachedina, Abdulaziz Abdulhussein. 2001. The Islamic Rootsof Democratic Pluralism. New York: Oxford UniversityPress.

Sells, Michael A. 1996. The Bridge Betrayed: Religions andGenocide in Bosnia. Berkeley, CA: University of CaliforniaPress.

Sigmund, Paul. ed. 1999. Religious Freedom and Evangelization inLatin America: The Challenge of Religious Pluralism. Maryknoll,NY: Orbis Books.

Sikand, Yoginder. 2004. “Kashmir: From National Liberation toIslamist Jihad?” In Muslims in India Since 1947. London, UK:Routledge.

Spruyt, Hendrik. 1994. The Sovereign State and Its Competitors.Princeton, NJ: Princeton University Press.

Stark, Rodney. 1999. “Secularization, R.I.P.” Sociology of Religion60 (Fall): 249–74.

Stepan, Alfred. 2001. “The World’s Religious Systems and Democ-racy: Crafting the “Twin Tolerations.” In Arguing ComparativePolitics. Oxford, UK: Oxford University Press.

Stern, Jessica. 2003. Terror in the Name of God: Why Religious Mili-tants Kill. New York, NY: Harper Collins.

Swearer, Donald K. 1995. The Buddhist World of Southeast Asia.Albany, NY: State University of New York Press.

Tambiah, Stanley J. 1993. “Buddhism, Politics, and Violence in SriLanka.” In Fundamentalisms and the State: Remaking Polities,Economies, and Militance, ed. Martin E. Marty and R. Scott Ap-pleby. Chicago: University of Chicago Press, 589–618.

Tilly, Charles. 1992. Coercion, Capital, and European States, AD 990-1992. Oxford, UK: Basil Blackwell.

Toft, Monica Duffy. 2006. “Religion, Civil War, and International Or-der,” BCSIA Discussion Paper, Discussion Paper 2006-03, BelferCenter for Science and International Affairs, July 2006.

Toft, Monica Duffy. 2007. “Getting Religion? The Puzzling Caseof Islam and Civil War,” International Security 30 (Spring): 97–131.

Ware, Timothy. 1963. The Orthodox Church. London, UK: PenguinBooks.

Weigel, George. 1992. The Final Revolution: The Resistance Churchand the Collapse of Communism. Oxford, UK: Oxford UniversityPress.

524

American Political Science Review Vol. 101, No. 3

Weinberg, Leonard, and William L. Eubank. 1998. “Terrorism andDemocracy: What Recent Events Reveal.” Terrorism and PoliticalViolence 10 (Spring): 108–18.

Woodberry, Robert D., and Timothy S. Shah. 2005. “The PioneeringProtestants.” In World Religions and Democracy, ed. Larry Dia-mond, Marc F. Plattner, and Philip J. Costopoulos. Baltimore, MD:Johns Hopkins University Press.

Yavuz, Hakan. 2003. Islamic Political Identity in Turkey. Oxford, UK:Oxford University Press.

Youngblood, Robert L. 1990. Marcos Against the Church. CornellUniversity Press: Cornell University Press.

Yun-Shik, Chang. 1998. “The Progressive Christian Church andDemocracy in South Korea.” Journal of Church and State 40(Spring): 437–65.

Zakaria, Fareed. 2004. The Future of Freedom: Illiberal DemocracyAt Home and Abroad. New York: Norton.

Zonis, Marvin, and Daniel Brumberg. 1984. “Behind BeirutTerrorism.” The New York Times. October 8.

525