Religion and International Relations Theory

Although there has been a great deal written on religion, there has been relatively little written on the relationship between organized religion and international politics.¹ Perhaps more glaringly, there have been only a few systematic analyses of the relationship between religion and war.² Most students of both international politics and war appear to take for granted the fact that when religion is involved in a conflict, it generally acts as a catalyst: religion makes deterrence more difficult, escalation to violence more likely, and wars more intense and more difficult to halt once under way.³

This chapter explores the connections between religion and war. As defined here, religion can be usefully seen as a system of practices and beliefs sharing most of the following elements: (1) a belief in a supernatural being or beings; (2) prayers or communication with those beings; (3) transcendent realities, including “heaven,” “paradise,” or “enlightenment”; (4) a distinction between the sacred and the profane and between ritual acts and sacred objects; (5) a view that explains both the world as a whole and humanity’s proper relation to it; (6) a code of conduct in line with that worldview; and (7) a temporal community bound by its adherence to these elements.⁴ Like nationalism, religion importantly shares the quality of imagined community and of a community that not only exists across space but also time.⁵ As such, religion can aid in the rationalization of self-sacrifice in support of a larger community and its ends.

Unlike nationalism, in which self-sacrifice can be rationalized in terms of a community only, religious practice holds out the additional possibility of individual salvation; and also unlike nationalism, religion tends to be less attached to specific territory.⁶ A final and crucial difference between nationalism and religion is that nationalists do not proselytize: one is unlikely to encounter a Serb, for example, trying to convince a Frenchman to become a Serb. One either is a member of a nation or one is not (and in some instances cannot become one). By contrast, most religions hold out the prospect of conversion to nonmembers, and some even demand that their members convert others as part of their conception of right practice in faith.

My focus is on the impact of religion on violence, and as such I restrict myself to two core questions. First, are religious adherents, or states whose governments identify closely with a religious tradition, rational actors? Although we can all recognize many differences between established religions, as social scientists we build general theories by abstracting from the details of our chosen unit of analysis. The process of abstraction invariably results in lost information, but if undertaken carefully, the rewards can be great, allowing us to see connections that cannot be appreciated without some distance from the particulars of each individual subject of analysis.

Second, assuming religious actors are rational, how might religion intersect with rationalist explanations for war such as issue indivisibility, commitment problems, private information, and asymmetric time horizons? If religious adherents are rational actors, why shouldn’t war be as inefficient in affirming coveted values as it is for nonreligious actors? I argue that finding answers to both questions will go a long way toward allowing the useful incorporation of religious belief and practice into theories of political interaction, intelligent theory building, and, by extension, into policy making where religion is a central issue.

The global resurgence of religious faith is an empirical trend with important political implications.⁷ This resurgence can be seen in a number of ways. First, by nearly any measure, the distribution of people who identify as “more religious” or as “born again” into a new religion has risen steadily in the past three decades, and shows no sign of slackening.⁸ Muslims and evangelical Christians appear to make up the lion’s share of the phenomenon, with strains of Christianity making headway in formerly Catholic and mainline Protestant areas of the globe (e.g., Latin America and Africa) and Muslims attracting more followers to the faith and having more children on average than their non-Muslim counterparts. Given conversion and population trends, it is not surprising that some predict that the religious shall inherit the earth.

The increase in religious adherents has been accompanied by another trend: democratization. Throughout the globe the pressure for political systems to become more transparent and accountable to their populations has meant that different sectors of society have been able to advance their agendas in the political sphere. What this means in practice is that as religious populations have grown, so too have their influence in politics. A simple metric captures this opening up of politics globally. In just one decade, from 1990 to 2000, the average polity score of states jumped more than two points: whereas the average was .76 in 1990, by 2000 it was 2.83. Admittedly, most states remain on the autocratic side of the spectrum (a score of 6 is typically when scholars code countries as democratic). Nevertheless, increasing levels of democratization at the global level do seem to form a trend, and this trend has allowed for religious actors and issues to enter politics more readily.

A final trend is the decline in the legitimacy of secular ideologies. Although the collapse of colonialism in the late 1940s and 1950s ushered in the departure of the Western powers, the legacies of communism and socialism remained. These ideological systems promised much, but delivered little. By the late 1960s and early 1970s, alternatives to these secular ideologies were being sought. Religion had never disappeared, and thus those elites and organizations that survived were in a good position to deliver what these other systems failed to. In Iran, what started as a middle class struggle against a corrupt state resulted in the transformation of a state and society according to theological principles. Although Poland’s overthrow of communism was not based in theology, the willingness of Pope John Paul II to condemn this political system provided needed legitimacy to challenge these systems and eventually led to their downfall.

What is important to note about each of these indicators of the religious resurgence is that they did not begin in the 1990s. Rather, the global resurgence can be traced back to the late 1960s and early 1970s. This is not a post-Cold War phenomenon. If you consider, for example, the proportion of civil wars with a religious dimension, an upsurge in relation to all types of civil wars began in the 1970s, when religious actors and institutions started to demand changes to their political systems.

Here I address a number of questions about the contemporary relationship between religion and war in the context of substate violence, which, as compared to interstate violence, has become much more common. Although my interest in this section is religion in civil wars, it hardly follows that its implications are limited to war within rather than between states.

For purposes of analysis, a civil war is large-scale violence in which two sets of organized combatants fight it out within the borders of the state, and each side suffers a significant number of casualties, with the war overall resulting in at least 1,000 deaths on average per year of combat. A civil war is a religious civil war if it is the case that (1) the combatants identify with a different faith tradition, as in Sri Lanka, with Buddhist Sinhalese fighting against Hindu Tamils, or (2) the combatants are of the same faith but contest the role of their religion vis-à-vis society and the state, as in the Tajik civil war, which pitted a formerly communist and secular group against a group wanting to impose an Islamic regime. In the first type religion features as a peripheral issue, while in the second, religion is central.

In relation to civil wars more broadly, the proportion of civil wars in which religion is a core issue of contention has risen. From the 1940s to the 1950s, the figure rose from 19 to 30 percent. The 1960s witnessed a modest decline to 22 percent, but in the 1970s this figure grew to 36 percent and continued to climb: in the 1980s to 41 percent, and in the 1990s to 45 percent. Religious civil wars made up a majority of the wars started after 2000.

Most religious civil wars have occurred in four regions of the world: Asia and the Pacific, the Middle East, Europe, and Africa. From 1940 to 2008, Latin America and North America have experienced no religiously based civil wars. Asia and the Pacific experienced twenty religious civil wars, or 45 percent of all religiously based civil wars, the Middle East eight (18 percent), Europe eight (18 percent), and Africa eight (18 percent).

More than half of Europe’s civil wars (eight of thirteen, or 62 percent) had a religious component. Asia and the Pacific hosted a somewhat smaller proportion: twenty of their forty-one civil wars (49 percent) were religious. The Middle East follows with eight of twenty-one (40 percent), and Africa with eight of forty-six (15 percent). These figures are presented in Table 5.1.

Table 5.1 shows that from 1940 to 2008, Asia and the Pacific experienced a disproportionate number of religious civil wars, and Europe experienced proportionately more religious civil wars than did other regions. These findings, however, do not reveal whether religion was a central or peripheral issue in these wars.

Table 5.2 provides data on the global distribution of wars with religion as a central or peripheral element. Recall that wars in which religion is central are cases in which the parties in the war are fighting for the imposition of a given religious tradition throughout the state. So, for example, the wars in Afghanistan involved an insurgent group, the Taliban, fighting for the imposition of a legal code that strictly adhered to Islam. Wars in which religion is peripheral are those in which the parties follow different religious traditions in the broadest terms. So, for example, Muslims in Bosnia fought against Christian (Catholic) Croats and (Orthodox) Serbs during the Yugoslav wars in the 1990s. In this case religion was not central to the fight, but it was used by elites and masses to identify friend from foe. Most religious civil wars, two-thirds, involve religion as a central issue (n = 27). Only seventeen of forty-four had religion as peripheral.

The third column of Table 5.2 shows that the vast proportion of wars with religion as a central issue took place in Asia and the Pacific (56 percent), the Middle East (22 percent), and Africa (19 percent). Only one of Europe’s religious wars featured religion in a central role. So, although most of Europe’s wars involved religion, religion featured only peripherally. This is not the case for Africa, Asia and the Pacific, and Middle East: in these regions religion featured centrally.

Although there were a total of 44 religious civil wars from 1940–2008, only twenty-four states were responsible for them. Table 5.3 provides the distribution of wars across states. Three states experienced 3 or more wars. India experienced 6 wars, 5 of which involved religion as a central feature, followed by Indonesia and Yugoslavia with 3 each. An additional eleven states experienced 2 of the 44 religious civil wars, and the remaining 10 wars occurred within other states.

If we look at which of the many religions is most apt to be engaged in a religious civil war, a striking finding emerges: one or both parties adhered to Islam in 82 percent of all religious civil wars (thirty-six of forty-four cases);⁹ Christianity was involved in twenty-three cases (52 percent); Hinduism was involved in six cases (14 percent); and other religions were involved in only a handful.

Another way to examine the connection between individual religions and civil war is to look at the religious makeup of the states involved. Table 5.4 shows that 58 percent of the states (14 of 24) that have experienced civil war have Islam as a dominant religion, where dominance is defined in terms of a majority of the population with a given religious identity. Christianity is a dominant religion in 21 percent of these states, while Buddhist, Hindu, Jewish, and Taoist states make up the remainder.

So why is Islam so much more likely than other religions to be involved in religious civil war?¹⁰ Consider religious influence among states and their populations. Of 192 states that are members of the United Nations, 27 (14 percent) have a clear orientation toward Islam,¹¹ while 25 (13 percent) have regimes in which Islam is defined as the state religion.¹² Population data reveal that in 48 countries (about 25 percent of all states), at least 50 percent of the population adheres to Islam, while globally the number of adherents to Islam in mid-2003 was approximately 20 percent of the worldwide population.

In contrast, 32 states (17 percent) have a clear Christian orientation, and 13 (7 percent) have declared Christianity as the official state religion. Population figures show that in 103 countries (54 percent of states worldwide), Christianity is a majority religion, and in mid-2003, Christians represented 33 percent of the worldwide population.¹³

Orientation of states, therefore, does not offer much help in explaining why Islam is disproportionately involved in religious civil wars, because a nearly equal proportion of states have an Islamic or Christian orientation (14 and 17 percent respectively). With regard to regimes that define themselves religiously, nearly twice as many are Islamic—13 percent, compared with 7 percent that are Christian. But this statistic does not offer much help either. Despite the differences, there remain relatively few Islamic- and Christian-oriented countries.

Yet, if the majority of the population of a state adheres to a particular religion, religious minorities are apt to be insecure. This likely explains why we find that as compared to civil wars more generally, religious civil wars were far more likely to involve groups seeking independence from the state. Whereas 66 percent of nonreligious civil wars (sixty of ninety-one cases) involved a centralist fight in the takeover of the government, 73 percent of religious civil wars (thirty- two of forty-four) involved fights over greater territorial autonomy or outright independence.¹⁴ Of these, about half involved Muslim minority groups such as Palestinians, Chechens (twice), Somalis in Ogaden, Moros (twice), Kashmiris (twice), Bosnian Muslims, Abkhaz (the extent of their Muslim background is questionable), Kosovar Albanians, Turkish Cypriots, Achenese, and Hyderabadi Muslims. The remaining cases of self-determination involved a mix of Christians, Hindus, Buddhists, and Sikhs.

Finally, it might be worth considering whether it is “Arab” ethnicity that is driving the finding about Muslims in religious civil wars. For example, it turns out that, overall, twenty civil wars (15 percent) occurred in Arab states. Of the forty-four religious civil wars, only four (9 percent) took place in Arab states. The other sixteen cases involving Arab states did not have a religious dimension. Most religious civil wars involving Islam took place outside of the Arab world. Thus, Arab race/ethnicity is unlikely to be responsible for the finding, since most of these fights took place beyond the Arab world in such places as Bangladesh, Cyprus, Indonesia, and Tajikistan.

One of the common reasons given for the presumption that an “Arab” factor can explain the violence is the fact that there is much debate about whether there is actually an Arab-democracy deficit and not a Muslim one. Scholars such as Alfred Stepan point to Mali and Indonesia as prime examples of Islam and democracy coexisting quite well. This might be the case; indeed it might also explain why we see Islam disproportionately involved in civil wars. We know civil wars do not happen in highly autocratic states, and that states undergoing transitions are the most likely to suffer from violence. Most Arab states are autocratic; only a few are transitioning to greater openness and accountability (e.g., Algeria). Given this, Arab states should be underrepresented in terms of civil wars. Indeed, consider that there are 23 Arab states (22 in the Arab League plus Palestine), making up 12 percent of the total number of states (assuming 192 states with UN membership). There have been only 9 Arab states involved in the twenty Arab civil wars: Algeria, Egypt, Iraq, Morocco, North and South Yemen, Yemen, Syria, and Jordan. And of these states, three fought wars of independence (Algeria, Morocco, and Tunisia) during times of transition, when we expect violence to occur. Iraq is an anomaly, having suffered five different civil wars. However, an examination of these wars shows that they tended to occur when there was a breakdown in central authority, as in 1991 with the Shia revolt following the Persian Gulf War, and in 2004 with the American invasion.

Given this, there does not seem to be an Arab problem or an Islam problem per se. Rather, it appears to be the interaction of Islam with democracy, secular ideologies, and the right to self-determination, among other things, that implicates religion (or more accurately, allows adherents, elites, and masses to draw on it) in civil wars. In other words, it is not Islam, but Islam in relation to its current trends, trends that began in the 1970s.

One explanation for an increased interest in religious civil wars might be that they are more deadly than nonreligious civil wars. Are religious civil wars costlier than other kinds of civil war, and does it matter whether religion features centrally or peripherally? In terms of the intensity of religious civil wars, three aspects are worth reporting: (1) war duration, (2) whether the war recurred, and (3) noncombatant deaths. Religious civil wars tend to last longer than nonreligious civil wars. Whereas nonreligious civil wars last on average 81 months, religious civil wars last 105 months, or about two years longer.¹⁵ The same holds true if one compares central religious civil wars with all other civil wars: central religious civil wars last on average 92 months; all others last about 87 months. Surprisingly, these data reveal that cases in which religion is peripheral are the longest, lasting on average 120 months (this is not statistically significant, however).

Furthermore, religious civil wars are about twice as likely as nonreligious civil wars to recur: nine of thirty-five (26 percent) recurred, compared with ten of eighty-two (12 percent) nonreligious wars.¹⁶ Wars in which religion features centrally are a bit more likely to recur (six of nineteen, or 32 percent),¹⁷ while religious civil wars in which religion plays only a peripheral role are no less likely to recur than nonreligious civil wars. Thus, when religion is central, we do find evidence for greater intractability, as the parties seem more willing to reengage in war after the first round of fighting is thought to have ended.

In addition, civil wars in which religion is central are two times deadlier to noncombatants than civil wars in which religion is peripheral (31,000 versus 14,000 average deaths per year for central and peripheral religious involvement respectively).¹⁸ To the extent that we care about noncombatant mortality, this finding makes it clear that the movement of religion from a peripheral to a central issue in a civil war can be expected to have dire consequences.¹⁹

Since the Second World War, at least as many human beings have perished in civil wars as did during the six years of world war. Interstate wars—and in particular world wars—are never something to take lightly, but civil wars appear to be the most likely type of war we are apt to see. This, along with the fact that there are simply more cases of civil than interstate war available for comparison and analysis, justifies a close look at the links between religion and civil war.

Three important facts about religion and civil wars emerge from this analysis. First, religious civil wars made up one-third of all civil wars fought from 1940 to 2000, and there is little sign that this proportion will decline any time soon. Second, among the world’s major religions, Islam was involved in just over 80 percent of these civil wars. Third, religious civil wars in some instances are more costly than nonreligious civil wars: they last longer, and when religion is central they are more intractable and deadlier to noncombatants than wars in which religion is peripheral.

Clearly religious engagement in civil war is on the rise, and in many cases this increasing engagement is associated with longer wars, greater harm to noncombatants, and greater difficulties halting ongoing wars (or keeping those halted from reigniting). A key issue is whether these associations are due to an increase in the number of irrational actors in a system or whether something else is going on. From both a theoretical and a policy standpoint, the idea that religious actors are irrational is of little use. Although religious actors are different than secular actors in key ways, they remain rational actors, and this is good news for theorists and policymakers: religious actors’ motivations can be understood, and their behavior manipulated. The bad news is that it will be more difficult to develop theories as broadly general as those based on a more restrictive understanding of rationality.

A key difference between religious actors and secular actors, however, is that the level of generalization and abstraction in anticipating action is not as broad in the case of religious actors. So, for example, in neorealist or neoliberal international relations (IR) theory, one may assume key features of a “state” in the abstract, and still generate stable predictions about how that abstract, idealized state might react to a given stimulus. With religious actors this will not be possible to the same degree: it matters whether the actor in question is a Christian, Hindu, or Jew in a way that has no analogue in secular international relations theory.

A core concern of those dealing with religiously motivated actors is whether such people or groups can act rationally: can, in other words, their behavior be altered by inducements such as threats or bribes? Can religious actors be deterred? For example, after the ouster of the Shah of Iran in 1979, and the subsequent establishment in Tehran of an Islamic theocracy, could the Iranian government be engaged as any other government? Or did the fact that its leaders identified as Muslims leading an Islamic state mean that in terms of coercion and deterrence Iran was a radically different sort of state, with and against which the traditional tools of statecraft did not apply?

For many in the United States and Israel, both often the target of hostile and religiously charged rhetoric from Tehran, this is a crucial question, not least because Iran is on the threshold of developing a nuclear capability—perhaps even the capacity to build a nuclear weapon. If Iran were any other state, one might at least have the confidence that it could be dissuaded from weaponization or, failing that, incorporated into a preexisting structure of nuclear deterrence. In Washington and Tel Aviv, however, the feeling is that Iran may not be a rational state, and that therefore it cannot be allowed to develop a nuclear weapon, even if that means launching a preventive war to halt that development. The stakes in this question of rationality and deterrence are therefore extraordinarily high.

By definition, a rational person balances potential costs and benefits by attempting to arrive at outcomes that result in a net benefit. When religious faith is introduced into the calculation of net benefit, two subsidiary questions attach themselves. First, does the benefit affect primarily the individual or a group of some sort (family, clan, tribe, nation, and so forth)? Leaving aside for the moment the question of variation in how humans actually imagine themselves (we can assume that in some places and times there are no meaningful distinctions between “I” and “we”), this is an important dimension of calculation because, as observed earlier, actions that result in a net cost to individuals might nevertheless confer a net benefit on a family or larger group. Second, there remains the question of whether the benefits and costs taken into consideration have empirical or tangible artifacts. Problems arise when one set of actors imagines that certain types of cost and benefit are necessarily valid and universal. So for example, the idea that a given delimited territory, an amount of rare metal, access to scarce natural resources, or “life” narrowly restricted to a biological component are the only sorts of interests over which a rational person might struggle (and that this is universal or axiomatic) can limit our ability to understand the wide variety of interests and values that have no tangible or existential counterpart—no empirically verifiable referent.

I mention this because when thinking about deterrence as an enduring pillar of political action, the rise of religious faith, which while recognizing the importance of physical existence and values nevertheless systematically discounts these in favor of intangible costs and benefits, can complicate deterrence, especially when one actor—instead of inquiring—attempts to alter another’s behavior by manipulating expected costs and benefits without first understanding how the target actor defines costs and benefits.

In attempting to understand and imagine rationality, however, we are not without a set of useful tools. In the late nineteenth century, the German sociologist Max Weber introduced four ideal-types of rational social action. Two of these are particularly relevant to the discussion here: what Weber called Zweckrationalitat (instrumental rationality) and Wertrationalitat (values rationality). The basic differences between these are that in instrumental rationality, there is no necessary link between the means a given actor chooses to employ and the ends sought. It is therefore possible to judge means separately from ends. In instrumental rationality, the core concern is not limited to a single goal, and is described most succinctly as practical or pragmatic: “maximizing results at a minimum cost.” By contrast, values rationality is concerned with what we today would call identity. As such, means and ends are an irreducible set.²⁰

The key feature then of values rationality, which clearly links it with religious belief and practice, is that the actions themselves constitute the value being sought. We can therefore think of both forms of social action as fundamentally rational. But the most important point is that today we are apt to have forgotten the existence of values rationality. It follows that our ability to understand, predict, and engage peoples for whom values rationality shares a stage with instrumental rationality is compromised.

A chief reason the leaders and publics of advanced industrial states worry about a rise of religion may have to do with their inability to recognize values rationality as a form of rational social and political action. The withering away of Wertrationalitat and its replacement by Zweckrationalitat were not an accident. Wilfred Cantwell Smith, for example, goes to great lengths to explain that understandings of religion that emerged from the eighteenth century skeptical movement are responsible for current, narrow conceptions of religion, faith, and transcendence.

We are now in a position to recognize that it was not fortuitous that the religion concepts, having arisen in Western Europe, are inherently depreciative. For the Christian tradition, particularly in its Protestant form, has historically been unusually disparaging of other religious traditions; and the rationalist academic tradition has been skeptical, if not disparaging, of all . . . This has to do with the fact that Christians have regularly failed or refused to recognize that the faith of non-Christians has that transcendence; that God does in fact encounter men in Buddhist, Muslim, and Hottentot forms, as he does in Christian. Secular academics have regularly failed or refused to recognize that there is a transcendent dimension to human life at all.²¹

Thus Smith not only holds Christian thinkers responsible for denying the faith and logic of other traditions, but he also holds academics with Western—or more accurately, post-Westphalian²²—conceptions of the separation of church and state responsible for steadily eroding our understanding of how religion engages the public, political sphere.

All of this is important because the fading away of values rationality and its replacement by instrumental rationality as the sole conception of rationality led to the assumption that only instrumental rationality counts as true rationality, and all else falls into the error term (is irrational). This has created a blind spot. In other places, values rationality coexists with instrumental rationality (or even predominates). So a better way to reframe our analysis of the relationship between religion and war is to recognize that instrumental rationality is neither natural, inevitable, nor universal, but rather shares the stage with values rationality.

It is therefore unwise and in most cases counterproductive to argue that people of passionate religious faith act irrationally. Instead a more useful set of generalizations, those that are capable of generating effective policy recommendations, focuses on how values rationality works, and how the spread (or perhaps the resurgence) of values rationality in tandem with the spread of religious belief and practice is likely to affect conflict between and within states.

The key thing to remember is that, in terms of forming a stable, predictable system of states (which is what eventually emerged following the signing of the Treaties of Westphalia), the chief advantage of rationality conceived of as interests was that it provided a basis to predict what others would do given a certain stimulus.²³ One need not know the details or history of a given state to understand that all states maintained an interest in survival, for example.²⁴ That stable interest, it so happened, implied other interests. And again, if one accepted a set of simple initial assumptions, a stable set of predictions about how others might react to a given initiative, bribe, threat, attack, conciliation, and so on, could and did emerge.²⁵ One could make predications in the abstract, knowing few if any of the details of a target polity’s culture, language, or history. But values rationality has a similar character: once one takes the trouble to ask and listen and learn, the actions of an ardent nationalist or evangelical Christian each become as predictable and stable as rational action via instrumental rationality. So, for example, learning about the role of Islam in policy making in Egypt will allow us to make stable predictions about Egypt’s foreign and domestic policy. However, it will tell us much less about Syria’s policy unless we ask and answer a similar set of questions. So the stability and predictability are there, but they are not as general as presupposed under instrumental rationality.

From this perspective it becomes more difficult to say whether the implications of resurgent religion are positive or negative vis-à-vis intrastate war. On the one hand, as we will see, religious rationality may act to undermine key pillars of deterrence and accommodation in the contemporary states system, so in that sense the global religious resurgence seems threatening. On the other hand, many religions place limits on what adherents may do, as well as what they may not do. We should therefore, a priori, expect to find cases in which war might have been an ideal response from an instrumental rationality point of view, but was ruled out due to a religious objection, i.e., a values rationality calculation.

Religion can make war more likely in a number of ways, but it may also prevent war when war seems imminent. As most students of conflict and conflict bargaining know, the costs of war generally exceed the benefits for all participants. Even winners often find that the benefits of victory could have been obtained more efficiently without war. This leaves the puzzle of explaining war when key decision makers on both or all sides are rational actors.²⁶ If war always results in a net loss for combatants as compared to other means, why ever resort to it? In 1992, James Fearon’s review of the literature on barriers to conflict resolution short of war highlighted three key explanations, to which I have added a fourth.²⁷

First, rational actors might go to war when they have misrepresented—either deliberately or inadvertently—their own interests in the outcome. For example, the United States did not appear to consider the fate of South Korea to be a vital security interest in 1950, and this made it easier for Kim Il Sung to convince Joseph Stalin to lend him the weapons and diplomatic support North Korea would need in order to forcibly unify the Korean Peninsula. Yet once the attack was under way, the United States revised its calculations and determined that “halting communist aggression in Asia,” whether in South Korea or elsewhere, was in fact a vital U.S. security interest. Had both Stalin and Kim Il Sung understood this clearly, the attack might not have happened, or it might not have happened as or when it did. Fearon calls this a problem of private information.

Second, rational actors might go to war when one or both sides—typically the stronger side—cannot credibly commit to uphold an agreement in the future. The lack of a third party to guarantee compliance with a present agreement may make it rational for even a weaker actor to initiate or continue a war, for fear that it may be even more disadvantaged in a future engagement. Fearon calls this a commitment problem.

Third, rational actors might go to war when they cannot agree that an issue of contention is divisible. Some kinds of values have a continuous quality that allows for their division into subsets that retain their value to those seeking them—money is the classic example. Other kinds of values, however, have a lumpy quality: most can be literally divided, but those seeking such goods no longer consider the thing divided as retaining its value or identity. Children, divine-right titles, and national homeland territories all share this characteristic. Thus, two rational actors who might otherwise agree to share a sought-after value find it impossible to agree on an acceptable division, and choose war in order to gain unfettered control of the person, place, or thing in dispute. Fearon calls this a problem of issue indivisibility.

In addition to Fearon’s summary of three rationalist explanations for war, contemporary work on conflict bargaining has posited a fourth: asymmetric time horizons.²⁸ Rational actors may share information about their true objectives, have the assistance of a third party in guaranteeing their contracts, and consider the issue of contention at least partly divisible; however, if each discounts the present at a different rate, war can appear to be a good gamble. In conflict bargaining, “the shadow of the future” argument posits a situation in which actors choose to forego the capture of short-term gains from defection (and sustain the perceived costs of not obtaining those gains) because in a relationship expected to endure over many exchanges, the benefits of cooperation are expected to far outweigh the short-term gains of defection.²⁹ For example, in the context of his fight with the secular and “corrupt” West, Osama bin Laden speaks of needing two hundred years to overthrow the West’s influence in the Middle East. Given this, we are most likely to find asymmetric time horizons when religious adherents clash with nonreligious opponents.

Religion intersects each of the four obstacles to nonviolent conflict resolution in important ways.³⁰ First—and to take only the Abrahamic religions as an example—the quality of salvation tends to be indivisible. One cannot be partly sinful or partly holy in one’s conduct, and the end state of one’s existence is discrete rather than continuous: one’s conduct is either sanctioned or censured; one goes to heaven or not. In addition to salvation being indivisible, physical locations have come to be seen as sharing this indivisible quality.³¹ Examples include the Q’aba in Mecca; the Temple Mount, or Haram al-Sharif, in Jerusalem; and ancient American Indian sites that contain sacred clays used in religious rituals. Thus, issue indivisibility is closely associated with religion, and may account for the perception that religious belief or motivation makes escalation of a conflict to war more rather than less likely. Finally, interpretations of holy texts often share this indivisible quality. It matters, for example, whether a given Arabic passage of the Quran directs a husband to “beat” his wife or force her to “go away” when he calculates that she has not obeyed the commandments of God.³² Scholars and practitioners who cleave to the former view will often not accept arguments to the effect that the original word in Arabic has another meaning or, more accurately, has another meaning in the specific context in which it was originally written.

Second, the start or resolution of a violent fight may not last when a dominant religion—dominant in terms of numbers of followers and control of government—cannot credibly promise not to renege on governmental agreements or a negotiated settlement with the followers of a minority religion in the same state. If it is the case that a regime has established itself on the basis of a particular faith, yet has a large portion of its population that is not of the same faith, then problems may arise if the state commits itself to upholding and defending only one tradition. This is what happened when the Sudanese government abrogated the Addis Ababa Agreement in 1983 as part of an Islamization campaign through all of Sudan, including the non-Muslim south. The situation does not look better with the recent 2005 Comprehensive Peace Agreement between the still-conservative Islamic government of Khartoum and the Southern resistance (SPLM/A) in Sudan, which continues to fear a Muslim-dominated North.

Third, people of faith often have private information about their true interests or capabilities in a conflict. This may be perhaps religion’s oldest and most frequent association with conflict. Typically, a given correlation of forces makes attack or defense appear hopeless, but the belief that “god is on our side” (private information about a religious sanction) encourages conflict when it otherwise would have been avoided. In Islamic law, for instance, if the leadership is seen to be corrupt and not representing the true faith, the individual has a duty to overthrow the government. This is the rationale behind war in Algeria, Egypt, and Tajikistan. Perhaps the most striking example is the Iranian Revolution and its associated theology, which permitted the Shia to take a more aggressive political role; no longer were they to be victimized martyrs, but hero-martyrs.

Fourth, religious adherents often appear to discount the present in favor of the future. In other words, they may welcome tangible individual or group costs that are frightful in anticipation of future and intangible benefits. For example, a rational individual might seek his own death (a very high cost) under circumstances expected to result in long-term extra-temporal benefits, such as eternal paradise.³³ Note that in neither the “shadow of the future” nor the “martyr” case are the expected benefits guaranteed, yet we may still consider those who calculate in this fashion to be rational actors.

In addition to issue indivisibility, commitment problems, private information, and asymmetric time horizons, two related factors may also help explain the common association of religious zeal with increased frequency or intensity of war: dehumanization and compulsory proselytization. As a matter of record, most males—and males overwhelmingly make up the world’s fighting forces³⁴—have a difficult time killing other humans.³⁵ Historically, one important way to facilitate killing has been dehumanizing the adversary. The principle is simple and powerful: the more closely the target resembles “me,” the more difficult it will be for me to muster the will to kill. Most people, for example, experience some discomfort killing animals for food, yet very few experience any discomfort killing plants for food. By imagining one’s human adversary as less than human or nonhuman, the moral difficulty in killing is reduced, and as a result killing is facilitated (possibly even difficult to deter, depending on the nature of the characterization invoked by the dehumanizing effort).³⁶

Religious belief has unfortunately facilitated this line of thinking in some cases.³⁷ Some of the earliest “human rights” debates were provoked by the “discovery” of natives by the Spanish Empire in the fifteenth century.³⁸ Spanish Catholic missionaries and soldiers had to wrestle with the status of natives, who were most often referred to as “savages.” From a Christian point of view, each human possesses a soul that is God’s business, not man’s. But the rights that attach to the person as a result generally were held to depend upon acceptance of the Christian faith. Those who heard the word of God and then rejected Christian tenets, or the authority of the church more generally, were dispossessed of all rights and protections under the law. They could be killed without moral or legal consequence to the killer. Moreover, where “unbelievers” were held to be a systematic threat to believers, killing them was not only permissible, but might take on the character of a religious obligation. Such sentiments continued well into the twentieth century. American president William McKinley, for example, justified the conquest, occupation, and conversion of the Philippines as part of a divine plan: “There was nothing left for us to do but to take them all, and to educate the Filipinos, and uplift and civilize and Christianize them.”³⁹ Christianity in its multiple forms is hardly unique in this, and the stripping away of moral consequence and protections from unbelievers and apostates is something many religious faiths share.

Second, and as observed above, in some religions whether one goes to paradise or heaven depends upon a relationship between the individual worshipper and his or her god(s). Sometimes that relationship is mediated by a priest or interpreter, and sometimes not. In other religions, however—in particular Evangelical religions—one’s salvation depends not only on one’s own conduct, but on the successful conversion of others to one’s own faith. Functionally, these types of religion are more likely to be associated with violence, because some proportion of those sought as converts are apt to view the attempt at conversion as an act of coercion even when it has no physical component.⁴⁰

In this chapter I have sketched a number of important ways international relations theorists and foreign policy students can understand the relationship between religion and war. Although, as with nationalism, there is a gap between the degree to which academic and policy audiences are likely to attach the assumption of rationality to actions (policy makers are apt to assume that religiously motivated actors are irrational, academics less so), I argue that religiously inspired actors are rational actors: firstly, they calculate; and secondly, they seek net gains from interactions. A key difference between secular and religious rational actors, however, is the degree to which the latter are apt to count the intangible, as well as the tangible, as significant.

Religiously inspired actors are apt, as individuals, to practice both of the types of rational action Weber introduces as instrumental rationality (Zweckrationalitat), and values rationality (Wertrationalitat). My argument is that once understood (and generalization across religious faiths may or may not be productive in this sense), religiously inspired actors may be seen to act in accordance with a stable set of preferences, implying there will be ways to alter their behavior by manipulating a different set of costs and benefits.

That said, in other ways religious belief and practice as currently interpreted appear to make war more likely than nonreligious practices for two reasons. First, the leaders and states of the advanced industrial world (which I have called “the West” at times, but which includes Japan and South Korea, to name but two states that are industrialized but not Western) have lost their understanding of the systematic bases of religious belief and practice, and are therefore likely to rely on an inappropriate set of cost and benefit inducements in their efforts to secure themselves from perceived threats based in the developing world. By systematically ignoring this noninstrumental rationality, these actors are apt to rely on coercion by threat of physical destruction, which may ultimately make war between the advanced industrial states and their often non-state adversaries more, rather than less, likely.

The second pathway to a higher incidence of war is demographic. Like political affiliation, most young people adopt their parents’ religious views (not unalloyed, but in the main). At the same time, many of the world’s religions—and some of the largest denominations, such as Roman Catholicism—prohibit birth control. Where birth controls are lax, populations increase at a faster rate, and as a result the process of democratization worldwide promises to give members of fecund religious groups increasing political power.⁴¹ Since members of rival religious groups are apt to gain increasing leverage over state foreign and domestic policies, it may be only a matter of time before the nation-state as a core actor in interstate politics is superseded by the “religious state.” Given the similarities between nationalism and religion, and given the intensity of violence that characterized (and continues to characterize) nationalist conflicts, one can only expect the incidence of war—and perhaps its intensity—to rise over time as the world’s religions continue to expand, first demographically and then geopolitically.

With the exception of (parts of) Europe and Japan, religious practice and identity are on the rise globally.⁴² To the extent that intergroup and interstate conflict bargaining may be compromised in the future as a result of religious adherents gaining an increasing share of political power within states, the implications for peace between and within states may be negative, and the likelihood of war may rise.

Currently, the majority of religion’s direct associations with war are represented at the substate level: first in civil wars, such as Sudan’s, between the predominantly Muslim North and predominately Christian and animist South, and Iraq’s, between Shiites and Sunnis; and second in religiously inspired terrorism, such as that directed by Al Qaeda and Islamic Jihad (to name but two groups). But religion affects interstate conflicts of interest as well: most famously, those between predominantly Hindu India and predominantly Muslim Pakistan. Additionally, the Russian Federation’s increasing association with Orthodox Christianity has already complicated its relations with Europe and the United States over the fate of predominantly Muslim Kosovo’s independence from Eastern Orthodox Christian Serbia. On the other hand, as highlighted by the long conflict between the Catholic Irish Republican Army and Protestant Ulster Unionists, and between Israeli Jews and Muslim and Christian Palestinian Arabs, it can often be difficult to disentangle nationalist and religious motives and effects, especially when religion becomes a part of national identity itself. But since nationalism and religion both share key functional aspects—demonization of the “other” and issue indivisibility, for example—once again, the prospects for keeping conflicts from escalating to violence and war may be diminishing.

My argument for why this may be happening is speculative at this point, but focuses on the hidden benefits of an expectation of interstate war. Since there are far more religions and nations than states, and since national self-determination has become an established norm of interstate politics, one puzzle is why there are not more states than there are.

My answer, in keeping with realist tradition, is that only a few homogeneous nation-states are viable in security terms (to say nothing of economies of scale). Thus, so long as invasion and conquest are a real fear, small nations and religious groups have a strong incentive to bury their differences and coexist productively within states. As noted above, however, different nationalities and religious groups have different capacities for coexistence.⁴³ More to the point, given that since the end of the Cold War the apparent threat of interstate war (invasion and conquest from another state) has diminished (or appeared to), it may be less surprising that in the absence of a common threat, groups within states—and here I focus on religious groups—may feel more confident in seeking a greater voice as groups. For the state’s part, the persecution or restriction of religious group access to political power may be weakening as well, and for the same reasons.

Finally, and most broadly, the implications of the rise of religious identity and practice worldwide demand more focused scholarly attention—especially, I would argue, from students of conflict and violence. While religion has often been associated with war, there is an equally long association between religious practice and peace and nonviolence. A key issue will be whether religion hinders or catalyzes escalation of conflicts to violence. The answer is far from clear. Consider that while many religions call for peace and condemn violence, these same religions are often involved in violence of the most extreme sort, including torture and mass murder. In terms of philosophy, religious prohibitions against violence (and war) count as tacit acknowledgment that “men” often seek violence and delight in conquering and subjugating other men. Thus, the rise in religious identity and its marriage to politics worldwide may be a measure of a rise in the number of persons with a desire to conquer others, who are at the same time seeking a way to govern or discipline that fundamental human desire. Moreover, when religion disciplines violence, its success depends on how closely the person so restricted may be identified with the faith in question. Unbelievers, in other words, do not benefit from prohibitions on violence to the same degree that co-religionists do.

All in all then, we may greet the rise of religion with regard to the likelihood of intergroup war with the same trepidation we would greet a similar rise in nationalism. While, like nationalism, religious faith per se is no indicator of an increased likelihood of a will to harm others, some percentage of religious adherents will cleave to a faith that functionally duplicates what we think of as hypernationalism, and with the same violent outcome: an increased likelihood of war. Understanding which religions are most likely to cultivate such adherents, and under what circumstances, should be a major preoccupation of students of interstate and intrastate violence.

1. An exception is Scott M. Thomas, The Global Resurgence of Religion and the Transformation of International Relations: The Struggle for the Soul of the Twenty-first Century (New York: Palgrave MacMillan, 2005). See also Douglas Johnston, Religion: The Missing Dimension of Statecraft (Oxford: Oxford University Press, 1994); the special issue of Millennium: Journal of International Studies 29, no. 3 (December 2000); and Daniel Philpott, Revolutions in Sovereignty: How Ideas Shaped Modern International Relations (Princeton: Princeton University Press, 2001).

2. Exceptions include Kalevi J. Holsti, Peace and War: Armed Conflicts and International Order, 1648—1989 (Cambridge: Cambridge University Press, 1991); and Samuel P. Huntington, The Clash of Civilizations and the Remaking of World Order (New York: Simon and Schuster, 1996). On civil war, see Jonathan Fox, Religion, Civilization, and Civil War: 1945 Through the New Millennium (Lanham, MD: Lexington, 2004); and Monica Duffy Toft, “Getting Religion? The Puzzling Case of Islam and Civil War,” International Security 31, no. 4 (Spring 2007): 97–131.

3. One of the best treatments of the “ambivalence” of religion can be found in R. Scott Appleby, The Ambivalence of the Sacred: Religion, Violence, and Reconciliation (New York: Rowan and Littlefield, 2000).

4. William P. Alston, “Religion,” Encyclopedia of Philosophy, vol. 7 (New York: Macmillan, 1972), 140–145.

6. Christianity and Hinduism, for example, each acknowledge holy places, yet the control or possession of a specific territory is not necessary to the practice of either faith. By contrast, many in Islam hold that territory does matter: control of the lands surrounding the Q’aba in Mecca and the Haram al-Sharif in Jerusalem, for example, are material to the practice of Islamic faith and often analogized to the body. This is why for many Muslims, even the nonviolent occupation of territories linked to the holy sites can be akin to an act of violence or sacrilege: a blind-spot for members of other religious communities who seek to control or share the same territories.

7. Timothy Samuel Shah and Monica Duffy Toft, “Why God is Winning,” Foreign Policy, July/August 2006, 39–43.

8. The most comprehensive assessment of religious values can be found in Pippa Norris and Ronald Inglehart, The Sacred and the Secular (Cambridge: Cambridge University Press 2003). An exception to this rule is some of the Western European states; they are outliers relative to the rest of the world.

9. For a similar finding about ethnic conflicts in the early 1990s, see Huntington, The Clash of Civilizations, 269.

11. Assaf Moghadam, “A Global Resurgence of Religion?” (Working Paper 03–03, Weatherhead Center for International Affairs, Harvard University, Cambridge, MA, August 2003), 46–54. Religious orientation refers to states in which a given religion is accorded preferential treatment in terms of a disproportionate allocation of state funds to religious institutions or groups when compared with state allocation to other religions and groups.

18. To calculate whether religious civil wars are more brutal toward civilians, I subtracted the average number of battlefield deaths per year for each war from the total deaths per year, thereby obtaining an average nonbattlefield annual death count. Note that nonreligious civil wars seem to brutalize civilians even more than central religious wars, accounting for 51,000 nonbattle deaths per year on average.

19. I confirm this finding about the deadliness of religious civil wars in a paper in which I examine religiously based violence in the Caucasus region of Russia from 2000–2008: religiously based acts of violence produced far more deaths than those that did not have a religious basis. See Monica Duffy Toft and Yuri Zhukov, “Religious Violence in the Caucasus: Global Jihad or Local Grievances?” (paper presented at the annual conference of the International Studies Association, New Orleans, February 17–20, 2010).

20. David Little, “Max Weber and the Comparative Study of Religious Ethics,” Journal of Religious Ethics 2, no. 2 (Fall 1974): 5–40; Bradley Starr, “The Structure of Max Weber’s Ethic of Responsibility,” Journal of Religious Ethics 27, no. 3 (Fall 1999):407–434.

21. Wilfred Cantwell Smith, The Meaning and End of Religion (1962; repr., Minneapolis, MN: Fortress Press, 1991).

22. The Romans, to cite but one important example, maintained neither a strict separation of church and state nor a bright line between mortality and divinity. See Christopher Kelly, The Roman Empire: A Very Short Introduction (New York: Oxford University Press, 2006).

23. Albert O. Hirschman, The Passions and the Interests:Political Arguments for Capitalism Before Its Triumph (Princeton: Princeton University Press, 1997).

24. Kenneth Waltz, Theory of International Politics (New York: McGraw Hill,1979).

25. John Mearsheimer, The Tragedy of Great Power Politics (New York: Norton, 2001).

26. This is one of the main puzzles that motivates most international relations theorists. For a classic treatment, see Geoffrey Blainey, Causes of War (New York: Free Press, 1988)

27. James D. Fearon, “Rationalist Explanations for War,” International Organization 49, no. 3 (Summer 1995): 379–414; Monica Duffy Toft, “Issue Indivisibility and Time Horizons as Rationalist Explanations for War,” Security Studies 15, no. 1 (January-March 2006): 34–69.

28. Toft, “Issue Indivisibility and Time Horizons”; Jack S. Levy and Philip Streich, “Time Horizons, Discounting, and Intertemporal Choice,” Journal of Conflict Resolution 51, no. 2 (April 2007): 199–226.

29. Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984).

30. I suspect that the range of circumstances under which commitment and private information obstacles apply to religious disputes is narrower than for secular disputes. Most religions, for example, hold that an all-knowing, all-powerful supreme being guarantees their agreements, so that a temporal third-party guarantee to enforce contracts would not be necessary. Private information, while clearly affecting some religious conflicts, is also less of a problem, since the stakes in a religious conflict are generally zero sum, such as salvation or paradise.

31. See, for example, Ron E. Hassner, “To Halve and to Hold: Conflicts Over Sacred Space and the Problem of Indivisibility,” Security Studies 12, no. 4 (Summer 2003): 1–33; and Toft, “Issue Indivisibility and Time Horizon.”

32. Lelah Bakhtiar, The Sublime Quran (New York: Kazi, 2007). Bakhtiar, an American woman who has translated the Quran into English, has stirred controversy over her arguments concerning the meaning of the Arabic word most often translated as “to beat,” which she argues should be translated as “to go away.”

34. Joshua S. Goldstein, War and Gender (Cambridge: Cambridge University Press, 2001); Joanna Bourke, An Intimate History of Killing: Face to Face Killing in 20th Century Warfare (New York: Basic Books, 1999).

35. On this startling line of argument, see David Grossman, On Killing (New York: Little and Brown, 1996). Grossman shows that up until the Korean Conflict (1950–1953), only one in two U.S. infantry soldiers fired their weapons at the enemy while in combat. S. L. A. Marshall’s research in World War II confirmed this finding, and the U.S. military subsequently overhauled its training program so that by the time of the Vietnam War (1965–1973), well over 98 percent of U.S. combat infantrymen were able to fire their weapons at the enemy. Grossman estimates that in the general population, only two percent suffer no psychological or emotional discomfort in killing other humans. The other 98 percent do suffer some degree of psychological or emotional injury as a result of killing.

36. The Nazis, for example, worked hard to indoctrinate average German citizens and convince them of the existential threat posed by Jews (among others). Even where such indoctrination was successful, however, the murder of defenseless human beings took a fearful psychological and emotional toll on most perpetrators. Two factors facilitated what we now think of as the Holocaust: (1) the organization in charge, the SS, worked tirelessly to identify and retain that tiny minority of Germans who were not traumatized by the act of killing; and (2) the industrialization of murder turned what began as a labor-intensive effort by Hitler’s Einsatzgruppen into the capital-intensive effort (e.g., gas chambers) most people commonly associate with the Holocaust. See Richard Rhodes, Masters of Death: The SS-Einsatzgruppen and the Invention of the Holocaust (New York: Vintage, 2003).

37. Religion is hardly alone in this aspect. Nationalist and other secular training systems seek similar results, some more successfully than others. On the power of communist and leftist ideologies to dehumanize adversaries, see Elisabeth Jean Wood, Insurgent Collective Action and Civil War in El Salvador (Cambridge: Cambridge University Press, 2003). On the war-causing effects of hypernationalism, see John J. Mearsheimer, “Back to the Future: Instability in Europe After the Cold War,” International Security 15, no. 1 (Summer 1990): 25, 35, 55–56; and Stephen van Evera, “Hypotheses on Nationalism and War,” International Security 18, no. 4 (Spring 1994): 5–39.

38. Anthony Pagden, The Fall of Natural Man (Cambridge: Cambridge University Press, 1987).

39. As recounted and quoted in Madeleine Albright, The Mighty and the Almighty: Reflections on America, God, and World Affairs (New York: Harper Collins, 2006), 23.

40. See, for example, the most recent efforts of the Russian Orthodox Church and the Russian state in partnership to limit the activities of non-Russian Orthodox traditions in the country. Clifford J. Levy, “At Expense of All Others, Putin Picks a Church,” New York Times, April 24, 2008, A1.

41. Monica Duffy Toft, “Differential Demographic Growth in Multinational States: The Case of Israel’s Two-Front War,” Review of International Affairs 56, no. 1 (Fall): 71–94, and “Population Shifts and War: A Test of Power Transition Theory,” International Interactions 33 (2007): 243–269; Eric Kaufmann, “Breeding for God,” Prospect Magazine, November 2006; Monica Duffy Toft, “Wombfare: The Politics of Fertility,” in Political Demography: Identity, Institutions, and Conflict, ed. Jack Goldstone, Eric Kaufmann, and Monica Duffy Toft (Boulder, CO: Paradigm, 2011).

42. Pippa Norris and Ronald Inglehart, The Sacred and the Secular: Religion and Politics Worldwide (New York: Cambridge University Press, 2004); Shah and Toft, “Why God is Winning.”

43. For a general theory of the conditions under which violence will emerge when identity groups challenge states, see Monica Duffy Toft, The Geography of Ethnic Violence: Identity, Interests, and the Indivisibility of Territory (Princeton: Princeton University Press, 2003).