Causal beliefs and war termination

Abstract

This article analyzes the length of interstate wars and the process of reaching a mutually acceptable bargaining solution. Rational choice scholarship has mainly sought to explain long wars in terms of commitment problems and private information. This article complements these rational choice perspectives by arguing that causal beliefs – a variable not considered by previous research – can also prolong wars by increasing expectations of battlefield performance and slowing down information updating. It illustrates the role of religiously based causal beliefs with the case of one of the longest interstate wars of modern time, the Iran–Iraq War of 1980–88. Even though commitment problems were present, they do not identify the root cause of Iran’s high expected utility of continuing the war, as religiously based causal beliefs played a more prominent role in prolonging the war. Religious causal beliefs constitute a real word mechanism that not only creates different priors about expected military capacity, but also slows down the process of updating beliefs, as battlefield events are not seen as credible information. Although the prevalence of religious conflicts has increased over time, the formation of beliefs and their effects on wars remains understudied when applying rational choice to real world conflicts.

Keywords

causal beliefs rational choice religion war duration war termination

Introduction

Why are some wars longer than others? This article analyzes the length of interstate wars by creating a model of war termination that explains how causal beliefs – a variable not considered by previous research on coercive bargaining – can prolong wars. The incorporation of this new variable is a supplement to the rationalist research program that has focused, for example, on the role of private information and commitment problems, sometimes even in conjunction with each other (Wolford, Reiter & Carrubba, 2011). The article illustrates the role of religious causal beliefs with the case of one of the longest interstate wars of modern times, the Iran–Iraq War of 1980–88.

Most wars end in a negotiated solution rather than one belligerent’s being completely overrun (Kecskemeti, 1958). Therefore, to solve the puzzle of war duration, rational choice models started to highlight the role of cost–benefit calculations (Wright, 1965; Porsholt, 1966; Wittman, 1979; Bueno de Mesquita, 1981; Pillar, 1983; Iklé, 1991), as wars were seen as coercive bargaining. Most of these models were based on the same theoretical approach in which unitary actors adjust their war aims in accordance with their current or expected battlefield performance, that is, events endogenous to the war. Peace is made when both sides ‘develop similar expectations about the war’ (Stanley & Sawyer, 2009: 652), causing one side to lower its war aims more than the other side raises its war aims, such that a bargaining space is created.

Fearon (1995: 393) famously argued that private information about relative capabilities and resolve, and incentives to misrepresent them, lead to a mismatch between expectations and demands that gives rise to wars. However, Fearon’s insights into the role of private information at the onset of wars do not neatly translate into a model of war termination, that is, of how wars end. According to the theory of endogenous war termination, battlefield events cause states to update their beliefs, providing the information necessary for the combatants to reach a settlement (Slantchev, 2004: 815; see also Filson & Werner, 2002). Events on the battlefield reveal information about the belligerents’ strength and resolve that changes the expectations about the costs and outcome of war, in turn opening up a bargaining space as the minimum demands of both sides start to converge.¹ As soon as a war starts, private information can therefore be expected to disappear. ‘Indeed, war may be the only way to credibly reveal private information’ (Goemans, 2000: 30) and is ‘a source of information less subject to manipulation by adversaries’ (Wagner, 2000: 478). This is puzzling, however, because all wars should then be short, as the mechanisms that prevent agreement before the onset of wars ‘cannot survive prolonged fighting’ (Goemans, 2000: 30).

The theoretical challenge is to identify reasons why a bargaining space does not emerge quickly. Wittman (1979) argues that risk aversion and discount rate affect the calculation of the expected utility of continuing a war. According to Powell (1999), states balance the probability of having their offer rejected at the cost of more fighting against the gains of demanding slightly more. Pillar (1983) suggests that the belligerents may continue hostilities for some time to get the best possible terms. Fearon (1998) also argues that a long shadow of the future gives states an incentive to bargain harder, delaying possible agreements in hopes of getting a better deal. Moreover, a peaceful solution may fail to materialize if the weaker side does not believe that the stronger side will commit to the terms of the peace treaty and refrains from exploiting its advantage in the future (Fox, 1970: 9; Quester, 1970: 31–32; Wittman, 1979: 757; Pillar, 1983: 205; Goemans, 2000: 33; Reiter, 2009; Weisiger, 2013). However, under some circumstances, commitment problems can be bargained away (Powell, 2006; Wolford, Reiter & Carrubba, 2011: 559), which leaves information problems as a more persistent obstacle to finding a bargaining solution.

Even if accurate information about battlefield performance were available, fear of domestic punishment may prevent decisionmakers from lowering unreasonably high war aims, which is necessary for the creation of a bargaining space (Goemans, 2000). Although receiving little empirical support (Levy, 1989, 1998; Gelpi, 1997; Chiozza & Goemans, 2003), the related diversionary war theory, which suggests that problematic domestic circumstances motivate a country’s leader to divert attention from sources of popular discontent by launching a militarized international crisis, can be used to explain why belligerents do not lower their war aims when the battlefield events are negative. Moreover, in Jackson & Morelli’s (2007) model, hostilities are possible because leaders keep a disproportionate share of the gains of war. Powell (2012) argues that even weaker states can bear the cost of fighting and run the risk of defeat to prevent an adverse shift in the distribution of power, and Fearon’s (2013) model suggests that continued fighting can be used to screen weak state types. However, the reasons why the creation of a bargaining space can sometimes take a long time remain undertheorized (Stanley & Sawyer, 2009: 653).

This article contributes to the theoretical and empirical analysis of war termination in three ways. First, many empirical investigations have treated ‘war’ or ‘war-year’ as the unit of observation (Bennett & Stam, 1996; Reiter & Stam, 1998; Werner, 1998; Slantchev, 2004; Stanley & Sawyer, 2009; Nilsson, 2012). However, as Ramsay (2008: 851) points out, the dominant rational choice model instead assigns causality to battlefield events. While some game theoretic models treat war as a series of interactions (Bennett & Stam, 2004), a case study focusing on battlefield events brings theory and empirical analysis closer together.

Second, in a study of battle-level data to test rational choice hypotheses, Ramsay (2008: 872) argues that ‘the informative battle hypothesis’ finds little support. Moreover, in the literature on conflict, it is often suggested that the very long duration of some wars ‘argues against private information as an important part of the explanation for such conflicts’ (Fearon, 2013: 2; see also Powell, 2006 and Blattman & Miguel, 2010). Reiter’s (2009) empirical examination of key decision times in six interstate wars shows that often state leaders do not change their aims or even raise them after major battlefield setbacks because of, for example, overconfidence. This empirical investigation argues that religious causal beliefs can be a source of overconfidence and can make it difficult for the belligerents to create a bargaining space during a war.

Third, an analysis of religious causal beliefs entails an original departure from the common-priors assumption, which means that if two actors both see the same information, for example, on the battlefield, their beliefs about the state of the world should converge. This assumption is ‘a poor representation of the real world’ (Smith & Stam, 2004: 786), as priors are not always the same for different actors. The complexity of making military estimates can sometimes allow two boundedly rational decisionmakers to be mutually optimistic about their chances of winning.² Smith & Stam (2004) argue that different theoretical perspectives on the nature of war and military technology make belligerents interpret battlefield events differently and create ‘heterogenous beliefs’ about the expected outcome of the war. The argument of this study is somewhat similar to Smith & Stam’s (2004) theory, as it gives another reason why information is handled differently by decisionmakers. However, this study focuses on religious causal beliefs, which can make decisionmakers discredit battlefield events as a reliable source of information. Thus, they have a more fundamental disagreement about causality.

Religious causal beliefs and expected utility

Expected utility calculations are central to a rational choice analysis, not only of whether to start a war but also of decisions to end or continue warfare. The expected value of an action equals the sum of the products of the probabilities (p) and values for each possible outcome (u) minus the cost of action (c): p(win) × u(win – c) + p(loss) × u(loss – c). Bargaining space refers to any agreement that the belligerents prefer to fighting.

Private information about expected military capacity affects estimates of p(win) and p(loss) and the prospects of creating a bargaining space. Thus, the belligerent states’ estimated probabilities of winning or losing the war do not need to sum to 1 ‘because they have different information available’ (Wittman, 1979: 745), which complicates the finding of common ground in peace negotiations. Causal beliefs, though not information as understood in the rationalist research program, can have the same effect of increasing the length of wars by inflating estimates of p(win) and deflating estimates of p(loss). In addition to creating different priors about estimated probabilities of winning or losing, they also slow down the process of information updating, which together yield different posteriors.

Causal beliefs are based on the ideational view of international politics held by constructivism, which has a starting premise that ‘the material world is indeterminate and is interpreted within a larger context of meaning. Ideas thus define the meaning of material power’ (Tannenwald, 2005: 19). According to Tannenwald (2005: 16), ‘Causal beliefs are beliefs about cause–effect […] and provide guidelines or strategies for individuals on how to achieve their objectives’, such as winning a war. Individuals make decisions among available alternatives and choose the option that best serves their ranking of outcomes, given their beliefs about the world. However, constructivists, as a rule, do not subscribe to mechanical positivist conceptions of causality and causal beliefs do not need to be based on empirical observations. While causal beliefs may be formulated using religious, political, or nationalistic terminology, they all rely on the same principle: some force beyond the grasp of those outside the chosen religion, political party, or nation will ultimately grant victory. In religion, this force may be an act of God, in nationalism the nation’s natural superiority, and in political ideology immutable forces of history, such as the inevitable defeat of capitalism and capitalist states in Marxist historical materialism.

Such causal beliefs promise the realization of territorial or other aims when the balance of power on the battlefield says the opposite, or when the expected future changes in the balance of power in terms of available weapons and contingents are not promising. They can therefore prolong the duration of wars even in the absence of information asymmetries about current battlefield events or about the number of troops and the military technology available in the future. They can also diminish the role of commitment problems, which have been used to explain why weaker states start or continue fighting (Fearon, 1995; Powell, 2006; Leventoglu & Slantchev, 2007), because they make even weaker states believe that they are strong.

If decisionmakers observe during the war that their previous priors, that is, beliefs about the probability of winning, were overestimated, they can rationally be updated according to Bayesian reasoning, given the new evidence. Jervis (1976: 143) argues that cognitive biases can make actors assimilate information to their pre-existing beliefs, which can slow down the updating process. However, religious causal beliefs can more efficiently slow down the process of updating beliefs relating to the expected p(win). They promise the realization of territorial or other aims even when the information derived from the battlefield is not assimilated to pre-existing priors about p(win). Battlefield events are simply discredited as a reliable source of information about relative strength, as compared to interpretations of God’s will. The war aims are therefore not lowered to a level justified by battlefield events, and the expected utility of continuing the war remains high.

Like private information, religious causal beliefs must be found in the decisionmaking elite if they are to affect war duration. These causal beliefs need not be private information, but they must be asymmetric, that is, the enemy is simply unlikely to agree that they are reasonable, which is also the case with Smith & Stam’s (2004: 787) ‘heterogenous beliefs’. If one state does not base its assessment of its p(win) on current battlefield events or on shared information about increasing future military capacity, the gap between the belligerents’ minimum demands is likely to be so wide that no bargaining space will be created. As for private information, asymmetries will disappear with the introduction of new, awaited military technology or tactics that reveal previously private information about relative strength so that the belligerents can agree on a mutually acceptable bargaining solution. However, asymmetries caused by religious causal beliefs are not easily changed by battlefield events as there is no mutual understanding of what counts as evidence of a future ability to emerge victorious, p(win).

Such immaterial variables not only affect the expected probability of winning, p(win), but also the utility of winning, or stakes, u(win), and the perceived costs of warfare, c, as the case of the Iran–Iraq War will illustrate. The idea that religion can prolong conflict resolution is not new. Svensson & Harding (2011: 133) write that once conflicting parties raise religious demands, then the possibility of solution wanes, as religious demands entail a certain degree of non-negotiability. Hassner (2003) argues, based on an analysis of the status of Jerusalem in the Palestine–Israel conflict, that the indivisibility of sacred spaces makes conflict resolution difficult. Problems relating to non-negotiability and indivisibility often derive from high estimates of u(win). However, as religiously based causal beliefs only affect p(win), they function in a different manner to prolong warfare.

Religion can arguably inhibit learning and updating beliefs about p(win). Horowitz (2009) suggests that during the five centuries of crusading, religious beliefs contributed to the initiation of new crusades despite the failures of previous ones. Yet, the literature lacks a comprehensive model that explains how non-material

Figure 1.

A model of war terminations

variables can affect the expected utility calculations of the warring parties during specific wars that can be seen as costly bargaining processes. As this study analyzes religious causal beliefs within the framework of a rational choice model and treats their impact as the same as that of private information, the incorporation of expectations such as the will of God is a complement to the rationalist research program and its expectations of how beliefs are updated.

Figure 1 presents a model of war termination illustrating how a bargaining space is created, as belligerents calculate their expected utility of continuing the war and adjust their war aims. The model is general in that religion is only one of many possible immaterial variables that can affect causal beliefs about how to win the war, p(win), the level of the stakes fought for, u(win), and the perceived costs of warfare, c, specified in the model. Commitment problems, that is, fear that a peace treaty will not lead to a lasting peace, are included in the stakes, as the utility of winning increases if defeating or at least weakening the enemy is expected to diminish the risk of future aggression.

After calculating its expected utility of continuing the war based on p(win), p(loss), u(win), u(loss), and c, a state will accordingly formulate its minimum demands (i.e. what it prefers to continued fighting), expressed as war aims. A high expected utility of continuing the war gives rise to high war aims. The state will then decide whether to continue the war by comparing its war aims with the enemy’s war aims to see whether they form a bargaining space (i.e. a mutually acceptable bargaining solution). If they are incompatible, the enemy’s war aims will not be accepted. Continued warfare then simply promises a higher payoff than opting for peace and the war will be long. For the war to end and a bargaining space to open up, at least one state must lower its expected utility of fighting and moderate its war aims.

I will illustrate the model, and especially the war-prolonging effect of religiously based causal beliefs, with the case of the Iran–Iraq War of 1980–88. This long war is especially suitable for the purpose because it involved only two states, which simplifies the analytical task. Moreover, as one of the longest wars, it has previously been used to analyze the role of commitment problems (Weisiger, 2013; Reiter, 2009; Shirkey, 2016), suggesting that Iran was not interested in a peace treaty because it feared that Saddam Hussein would not stop threatening Iran. However, the present study argues that even though such commitment problems were present, they do not identify the root cause of Iran’s high expected utility of continuing the war, as religiously based causal beliefs played a more prominent role in prolonging the war.

It can also be argued that the revolutionary regime in Iran used the war as a means of consolidating its domestic political power. Moreover, if leaders keep a disproportionate share of the gains of the war, warfare can be rational even when they have accurate intelligence about military capabilities (Jackson & Morelli, 2007). These arguments are to some extent plausible. Chubin & Tripp (1988) argue that Iran sought to use the war both to spread its revolutionary message and to reshape its internal polity. ‘Initially, the war helped the Iranian government suppress its opposition and rally the people around the flag’ (Moshiri, 1991: 132). During the first year of the war, it was easy to label as traitors the leftist Mujahedin activists fighting the new central government and thus to diminish their popular support. Yet, by 1983 Tehran had effectively succeeded in suppressing the rebels’ armed struggle (Hiro, 1990: 69; Moshiri, 1991: 130) and Iran’s economic development then fell back by 20 years (Moshiri, 1991: 133). As the new leadership had consolidated its power and prestige, the remaining five years of costly and risky warfare call for a different explanation of the expected utility in continuing the war, an explanation going beyond the explanatory power of domestic political reasoning or the leaders potentially keeping a disproportionate share of the gains. Moreover, while Jackson & Morelli’s (2007) model assumes that decisionmakers see battlefield events as correct information and still neglect them because of personal benefits from warfare, this study suggests that they discredit them as a source of information.

Shirkey (2016: 263) argues that the growth in Iraqi war fighting capacity had been unclear to the Iranians, but when they ‘became aware of these Iraqi advantages and the results they brought on the battlefield, Iran quickly sought peace’. However, this study argues that private information did not play a decisive a role. Several battlefield events were very discouraging and clear to Tehran long before the final year of the war and still the Iranian leadership did not care to update their beliefs about the expected outcome of the war. The analysis will instead demonstrate that the length of the war was caused primarily by religiously based causal beliefs raising p(win). Religious ideas also increased the stakes, u(win), and diminished the costs, c, of warfare. All of p(win), u(win), and the costs of war, c, can change during a war and affect the prospects of creating a bargaining space. I will start the analysis by examining how battlefield events covaried with the belligerents’ war aims.

Early phases of the war

On 22 September 1980, the Iraqi Air Force launched sorties against Iranian air bases in an apparent effort to copy the Israeli offensive success against its Arab neighbors in 1967. It is possible that the temporary military weakness of Iran following the Islamic revolution had created a window of opportunity to initiate a preventive war (Amos, 1984: 58–60; Levy & Froelich, 1985: 137–139). Tibi (1998: 156) and Yapp (1996: 428) also argue that Saddam Hussein had the very limited war aims of deterring Khomeini from seeking to export the Islamic revolution, which might destabilize Iraq, and of occupying areas of strategic importance, so as to secure for Iraq better access to the Gulf. According to Hiro (1990: 40; see also King, 1987: 15–17), these objectives were the Shatt al-Arab waterway, the cities of Abadan and Khorramshahr on its eastern bank, and the cities of Ahvaz and Dezful deeper in Iranian territory. Taking over these areas with the help of offense-dominant weapon technology was not expected be a long process. Recovered documents indicate that Hussein believed that Iraq would be able to achieve its limited territorial key objectives in 10 to 14 days (Zabih, 1988: 169–170).

The air strikes were not as effective as expected because of the inefficiency of the Iraqi air force and the dispersion of well-bunkered targets, although the following land invasion fared somewhat better. The Iraqi army managed to penetrate some 15 km into Iranian territory in the south, and as far as 45 km in the north. Iraq soon stood a chance of occupying all the large cities in southwestern Iran (Cordesman & Wagner, 1990: 82–88). The Iraqi army was not proficient in conducting its Blitzkrieg, however, which gave Iran time to reorganize its defenses. Iraqi forces rarely risked bypassing an objective or opposing forces. Some combat units halted when they met relatively light opposition, and Iraq showed scant ability to maneuver its armor or use its air power (Cordesman & Wagner, 1990: 90; Hiro, 1990: 48). By late 1980, the Iraqi army had established a pattern of failing to conduct effective offensive combined-arms operations: ‘They moved too slowly, too little, and too late’ (Cordesman & Wagner, 1990: 97).

As its offensive progress was greatly reduced, and the Iraqi army failed to achieve its original war aims, Hussein announced that he was ready for a ceasefire on 28 September 1980. In accordance with the rational choice expectation of the theory of endogenous war termination, he lowered his war aims and merely demanded that Iran accept the territorial changes brought about by the initial Iraqi offensive, accept Iraq’s complete rights over the Shatt al-Arab waterway, and withdraw from three islands in the Gulf (Hiro, 1990: 42). However, the Iranians, who had not yet fully mustered the strength of their military against the Iraqis, had no reason to accept such terms. With inferior forces they had succeeded in slowing down the Iraqi onslaught and, with the mobilization of new forces, their expected offensive capacity, p(win), had not yet been tested. The plea for peace was left unreciprocated at this early stage of war because of expectations of increasing military capacity.

In January and March 1981, Iran launched unsuccessful counterattacks to regain the occupied territory. Similarly, Iraq sought in March to advance farther into Iran, but failed. From the point of view of the theory of endogenous war termination, by this time these battlefield failures ought to have revealed such negative information about Iran’s offensive capacity that Tehran also would lower its war aims, leading to a congruence in the belligerents’ minimum demands. Yet, at this critical point for war duration, quite the opposite happened.

On 20 June 1981, after a spring of failed counter-attacks, the more moderate President Bani Sadr was removed from office after being impeached by the Iranian Parliament for incompetence. The next president, Rajai, was the preferred choice of the clerics (Hiro, 1990: 52). Their uncompromising attitude toward negotiations was reflected in Khomeini’s statement that ‘there is no question of peace or compromise, and we shall never have any discussion with them’ (Hiro, 1990: 53). On 19 March 1982, Iran launched an offensive in the Dezful-Shush region that managed to push the frontier some 40 km back to the west. It is indicative of a marked decrease in offensive expectations, p(win), that Saddam Hussein allowed the withdrawal of forward units and the commitment of reserves after Iraq had begun to suffer serious setbacks (Cordesman & Wagner, 1990: 131). In accordance with this trend of diminishing offensive capacity, and as predicted by the theory of endogenous war termination, Iraq again lowered its war aims. In mid-March, Taha Yassin Ramadan, the Deputy Prime Minister, declared that Iraq was prepared to withdraw from occupied areas once peace negotiations had started and displayed signs of progress (Karsh, 1987: 23).

If Hussein had by now interpreted the battlefield events as indicating a marked decline in Iraq’s current and expected offensive capacity, Iran had opposite expectations of increasing p(win). In early May 1982, Iran succeeded in further regaining some of the occupied territory, and by 23 May, the city of Khorramshahr had been recaptured without much resistance from the withdrawing Iraqis (Karsh, 1987: 25). As a result of poor battlefield performance and lack of offensive expectations, on 20 June 1982, Hussein announced that Iraq would finally enact a unilateral withdrawal to the old border (Hiro, 1990: 64). The Iraqi war aims reflected the expectations of the theory of endogenous war termination and suggested that a bargaining space could be created unless Iran raised its war aims. The Iranian Supreme Defense Council now had a bitter debate over whether continuing the war into Iraq would be useful.

The idea of an invasion was categorically opposed by the military leadership, which doubted the army’s ability to carry it out. The military was supported in their judgment by some moderate politicians such as the Premier, Mir Hossein Moussavi, and the President, Sayyed Ali Khamanei, who opposed an invasion on the grounds of its high human, material and political costs. (Karsh, 1987: 42; see also Hiro, 1990: 86)

However, Rafsanjani, the Parliament speaker, related from the meeting that ‘the Imam commanded that no one should speak of cease-fire’ (Takeyh, 2010: 371). Thus, by the summer of 1982, the conservative factions of the Iranian leadership around Khomeini had crushed the more moderate opposition and were even coming under pressure from a still more conservative group of mullahs (Kifner, 1982). The hardliners with a different cost–benefit analysis won the debate and ignored the regular army’s and the moderates’ expectations of low offensive capacity, p(win), and rising costs of warfare, c.

On 9 July 1982, Rafsanjani announced the new higher Iranian goals: retaining the old Algiers treaty that gave Iran the right of navigation in Shatt al-Arab; repatriation of over 100,000 Iraqi citizens expelled from Iraq; USD 100 billion in war reparations; and punishing Saddam Hussein as a war criminal (Hiro, 1990: 86). As Iraq categorically refused to consider such terms, a bargaining space did not materialize and several offensives against Iraqi territory were launched during the summer. The attacks were thwarted, however, by a solid Iraqi line of defense at a heavy human cost to the attacker. The Iranian Chief of Staff threatened to resign if unqualified people continued to meddle with the conduct of the war (Karsh, 1987: 26).

Indeed, the rift between the regular army and the religiously inspired Pasdaran (Revolutionary Guards) and Basij (a volunteer-based paramilitary force) widened as the army played a small role in the failed attacks that relied on massive frontal infantry assaults without close air support or armor (Karsh, 1987: 27). Tehran then launched the largest land operations to that time: during the first, in mid-February 1984, over half a million men were engaged in a battle that brought no gains to Iran; during the second, only the Majnun Island oilfields were captured (Karsh, 1987: 28). The balance of force in air power and artillery tilted in favor of Iraq (Cordesman, 1984: 684). However, Iran did not learn from its poor battlefield performance and lower its war aims. Why did Iran invade Iraq and why did it not lower its war aims as the war went awry?

Iranian religious beliefs

One possible reason why Iran continued the war was its fear of Iraqi commitment problems. Kamla Kharrazi, the head of the War Propaganda Office, asked, ‘Who can assure us that the regime of apostate Saddam after ending the war, will not regain its strength and attack us once more’ (Takeyh, 2010: 272). However, as the Iranian leadership believed that they would win the war with a high p(win) and waged the war to spread the revolution with a high u(win), commitment problems became a secondary reason for continuing the war. Religiously based causal beliefs diminish the role of commitment problems, as such beliefs make even weaker states believe they are strong.

The crucial explanatory factor for the lack of learning from the battlefield events can be found within the realm of religion. The clerics propagated ‘Shi’i generated epic aspects of the war, mourning, opposition to existing values in the city, martyrdom, action as opposed to words, purity and devotion, and spiritual rewards in the afterlife’ (Farhi, 2004: 104). Religion not only managed to mobilize the masses to fight for a holy cause but also reinforced the leadership’s expectations of Iran’s offensive capacity, as the negative events on the battlefield became increasingly irrelevant to the decisionmaking process of the leadership.

The first reason why Tehran kept its high war aims high was that the Iranian leadership had religiously based causal beliefs. They increased the estimated probability of victory, p(win), and slowed down the process of updating beliefs based on battlefield events, which were discredited as a reliable source of information. While Hussein based his assessment of justified war aims on the battlefield events, as expected by the theory of endogenous war termination, Tehran held on to its high war aims as it deemed its expected future offensive capacity to be high because Iran had divine help on its side: ‘According to Khomeini, victory in the war ultimately depended on the creator’s fadl (favour) and lutf (benevolence)’ (Gieling, 1999: 60). In the end, the outcome of the war would be affected by Iranians’ level of belief and striving in the way of God: ‘God would help the Iranians only if they helped Him’ and in return render them victorious (Gieling, 1999: 60). However, frequent mention was made of the belief that God would aid the believers because they were fighting for a ‘divine cause’ (Associated Press, 1987) to defeat Iraq. Khomeini argued in several speeches that victory would ultimately belong to the believers (Gieling, 1999: 149). For example, he declared that ‘with the dispensation of the Supreme Lord, the banner of Islam is likely to be hoisted throughout the globe in the not-too-distant future’ (Brumberg, 2001: 132–133). Rafsanjani further said that as ‘we are sure of victory […] we refuse to negotiate’ (Khadduri, 1988: 95). The leadership therefore had causal beliefs that led them to neglect the battlefield events and expect the Iranian offensive capacity, p(win), to increase because of the believers’ reliance on God.

Not only expectations of God’s help but also religious fervor explain why strategic information was not interpreted as expected by the theory of endogenous war termination: ‘The Mullahs […] continued to try to substitute ideological fervor for strategy, tactics, and training’ (Cordesman & Wagner, 1990: 169–170). The hardliners argued against the doubts of the military professionals that any limitations ‘could be overcome by deploying large numbers of fighters imbued with revolutionary Islamic zeal’ (Hiro, 1990: 86). As Prime Minister Moussavi would later explain, ‘the power of faith can outmaneuver a complicated war machine’ (Takeyh, 2010: 369). In effect, religious fervor and God’s help were expected to raise Iran’s offensive capacity, p(win), and the probability of achieving even higher war aims. When leaders believe that fervor and divine intervention can substitute for military training, it is reasonable that their perceived p(win) will be higher than when training is believed to be crucial.

The second reason why Tehran kept its war aims high was that religious beliefs raised the offensive stakes. As the offensive stakes, that is, the utility of winning, u(win), were high, bearing the soaring costs of war, c, was possible without lowering the expected utility of continuing the war. The offensive stakes were high for religious reasons, as agreeing to a religiously unjust peace agreement made negotiations impossible. The issue was not about a traditional commitment problem emanating from fear of a future breach of a peace treaty, but was based on reasoning in terms of religious values.

The leaders made it clear that, in the case of this war, ‘peace was not in conformity with Islam’ (Gieling, 1999: 165). This meant that Hussein’s peace proposals were treated as un-Islamic. ‘All peace negotiations and calls for a settlement of the war were worthless in the eyes of the Iranian leaders unless these were accompanied by divine justice (adâla). Fighting had to continue until ‘adâla had been achieved’ (Gieling, 1999: 167). Owing to a religious striving for divine justice, any peace treaty would have to include the punishment of Hussein as a war criminal and his removal from power. Prime Minister Muhammad Rizâ Mahdawî Kanî referred to the Quran (49:9) in arguing that to compromise with an oppressor would be morally wrong (Gieling, 1999: 165). Anecdotal evidence from early Islamic history was also used to justify not entering into peace negotiations when doing so implied reconciliation with unbelief, heresy, and aggression (Gieling, 1999: 112). In a broadcast on 4 April 1985, Khomeini further stressed the extremely negative consequences of defeat, u(loss), from a religious respective:

It is our belief that Saddam wishes to return Islam to blasphemy and polytheism […] Islam will receive such a blow that it will not be able to raise its head for a long time […] The issue is one of Islam versus blasphemy, and not of Iran versus Iraq. (Brumberg, 2001: 133)

Furthermore, the religious sentiments also raised the offensive stakes, u(win), by increasing the value of Iraqi territory in the eyes of the clergy now controlling the Iranian state. Khomeini said that Islamic revolution ‘is being exported’ (Brumberg, 2001: 132–133), and Rafsanjani argued that ‘we have to strengthen the revolution […] and spread its ideas’ (Takeyh, 2010: 371). The road to Jerusalem was seen as going through the occupation of Iraq (Abrahamian, 2008: 175; Takeyh, 2010: 371). As Hiro writes, ‘They saw the march either to the holy cities of Najaf and Karbala in southern Iraq, or Kadhimain near Baghdad, as part of the advance to Jerusalem, the third holiest city in Islam’ (Hiro, 1990: 106). As early as 1982, when the decision to invade Iraq was made, Khomeini argued that the ‘path to Jerusalem was through Karbala’ (Takeyh, 2010: 371). These sentiments did not wane with later battlefield failures and were also reflected in a speech that Khomeini gave to Muslims making the annual pilgrimage to Mecca in 1987:

I declare my own as well as the unreserved support of the Iranian nation, Government and authorities for all Islamic struggles of nations and courageous and Moslem young people toward the liberation of Jerusalem […] We will export our experiences to the whole world and present the outcome of our struggles against tyrants to those who are struggling along the path of God. ( New York Times, 1987)

However, the prospects of spreading revolution were dependent on expectations of high offensive capacity, as the hardliners expected the Iraqi Shiite population to join the revolution if the invasion was successful (Walt, 1996: 261; Takeyh, 2009: 94).

Third, Iran’s expected utility of continuing the war was also high because the perceived costs of war, c, were low from a religious perspective. As the war dragged on, the content of Khomeini’s speeches suggested his growing conviction that ‘martyrdom was the supreme form of mystical experience’ (Brumberg, 2001: 128). In a speech in October 1980, Khomeini said that ‘the natural world is the lowest part of creation […] The true arena is the divine world which is inexhaustible’; the martyrs were not a cost but an asset, as they were helping Iran to become ‘a divine country’ (Brumberg, 2001: 128). Thus, Khomeini argued that ‘Iran’s recent setbacks on the battlefront represented gain’, as ‘the nation that goes for martyrdom […] can hardly think of anything else. As for its economy, it does not matter’ (Brumberg, 2001: 132). Moreover, as Khomeini suggested that ‘the commander in chief of the forces is God’ (Williamson & Woods, 2014: 83), observations of battlefield events became disconnected from strategic thinking:

God has ordered us to fight those who are against Islam […] We either win or we don’t. If we do, blessed be God, we have accomplished the mission and have been successful. If we die, we have done what we have been told. We do not have defeat, defeat is not for us. For us, we are either victorious or we don’t win but we have saved our honor [before] God. (Williamson & Woods, 2014: 83–84)

Bargaining space opens up

The ground war had come to a virtual halt by late 1984. Iran’s general performance improved, however, as the direction of the war was placed in the hands of the regular army (Karsh 1987: 36). In two large operations in mid-February 1985, the Iranians succeeded in advancing to the Fao Peninsula in the south and to the outskirts of the city of Sulaymaniyah in the north. As the war neared its sixth anniversary, senior Iranian spokesmen again began to stress the need to deliver a ‘final blow’ (Karsh, 1987: 33; see also Hiro, 1990: 171).

Between September 1986 and February 1987, Iran sought to capitalize on its previous offensive gains by launching a series of massive attacks on Iraq. The Iranian efforts led to scant gains, however, despite the government’s committing large numbers of troops in human wave attacks against the Iraqi defenses. These attacks exacted a heavy human toll of at least 200,000 casualties (Cordesman & Wagner, 1990: 254). However, this battlefield information was not used for updating Tehran’s beliefs and the strategy was not changed.

In March 1987, the Iranians made some gains in the north, but even then, the battle was a blood-bath. Still, despite having lost as many as 600,000 to 700,000 soldiers since the start of the war, Rafsanjani claimed that Iran would emerge victorious in the coming year (Cordesman & Wagner, 1990: 260–261). During April and May, an Iranian assault failed to achieve significant territorial changes at the war front, however. In June, Rafsanjani announced that Iran would give up its human wave tactics due to their costliness in terms of casualties and concentrate instead on surprise attacks (Cordesman & Wagner, 1990: 282, 302). The perceived costs of war, c, were now increasing.

While the religious leadership continued to favor religious zeal rather than the professional advice of the regular army and Khomeini still believed in November that an Iraqi defeat is ‘inevitable’ (Cordesman & Wagner, 1990: 324), Iran was now facing mobilization problems, the economy was failing, and the troops were running low on ammunition (Chubin, 1989: 23; Cordesman & Wagner 1990: 324; Willett, 2004: 53; Takeyh, 2010: 380). However, against the expectations of the theory of endogenous war termination, the war aims were still not lowered.

During the course of the war the Iraqi army had improved both its professionalism and military materiel. In particular, the Revolutionary Guards of about 100,000 men had received extensive training in offensive operations (Cordesman & Wagner, 1990: 353–355). On 17 April 1988, the Iraqi forces made quick gains against the defenders in the Fao Peninsula with the help of nerve gas (Cordesman & Wagner, 1990: 374; Willett, 2004: 53). The level of Iranian religious fervor had waned as ‘few units showed any sign of the willingness to die that had characterized Iranian forces in previous campaigns’ (Cordesman & Wagner, 1990: 374).

On 25 May, similar offensive efforts led to the capture of the city of Salamcheh, but Khomeini still refused to negotiate. However, at the beginning of June, fears of a collapse emerged among the clerics. The gap between the causal beliefs of the highest religious leadership and the battlefield performance of the Iranian armed forces was by now becoming too obvious to ignore, as on 25 June the Iraqi army succeeded in driving 30 km into Iran without serious opposition (Cordesman & Wagner, 1990: 383–389). ‘Some Iranian clerics began lobbying Ayatollah Khomeini to end the war. If it went on much longer, they argued, it might endanger Khomeini’s Islamic revolution itself’ (Willett, 2004: 53–54). Rather than hoping to export the revolution, there were now increasing fears that the entire religious project would be in jeopardy if the war continued longer and ended in a military collapse. Causal beliefs that had initially assumed that God would be on Iran’s side successfully pursuing the war into Iraq were now replaced with an interpretation of God’s will that instead emphasized defending the local basis of the Islamic revolution.

By 12 July 1988, Iraq was in control of virtually all of its old territory. Reminiscent of the previous problems when occupying Iranian territory, Hussein did not raise his war aim of recovering the old borders. On 17 July, he gave a speech in which he repeated his call for a ceasefire and a return to the international borders (Hiro, 1990: 241). The same day, President Khamanei sent a letter to the UN Secretary-General Perez de Cuellar requesting a ceasefire. On 6 August, Iraq reciprocated Iran’s peace offer, and the next day the belligerents agreed to enter into direct negotiations (Cordesman & Wagner, 1990: 395–398; Willett, 2004: 55). Both sides accepted UN Resolution 598 calling for a return to the old borders.

Khomeini now said that he first ‘had promised to fight to the last drop of my blood and to my last breath’ but then ‘submitted to God’s will’ to make peace (Pear, 1988). ‘Based only on the interest of the Islamic republic’ at the urging of ‘all the high-ranking political and military experts’ (Pear, 1988; see also Gieling, 1999: 169) he changed his causal beliefs in God’s help and lowered his offensive expectations, p(win). Learning from the battlefield events was finally possible.

A letter published by the office of Rafsanjani in October 2006 reveals that Khomeini had been advised by Rezai, the commander of the Iranian Revolutionary Guards, that the war was not winnable: ‘No victories are in sight for the next five years’ (Nafisi, 2006; see also Takeyh, 2010: 377). Already during the first years of fighting, the regular army had observed the clearly negative battlefield events and updated its beliefs about the expected outcome of the war. Their criticism of the leadership over the conduct of the war had gone unheeded, but by this time the troops lacked religious fervor and even the religious Revolutionary Guards doubted Iran’s expected offensive capacity, which was essential for changing the leadership’s religiously based causal beliefs. The process of updating expectations took a long time because religious causal beliefs had created expectations of offensive capacity, that is, different priors, and battlefield events were not regarded as a credible source of information for updating these expectations till they were too obvious to ignore. Facing a potential military collapse, causal beliefs relying on God’s help to defeat Iraq on the battlefield were finally changed, which made the creation of a bargaining space and peacemaking possible.

Final analysis

The case of the Iran–Iraq War illustrates a lack of learning from the information provided by the battlefield events as they did not covary with the war aims. The theory of endogenous war termination, representing a simple rational choice perspective, therefore does not suffice to explain the chain of events during the war. Shirkey (2016) argues that private information was the key to understanding this long war. However, several battlefield events were very discouraging throughout the long war and revealed crucial information that was clear to the professional military, and still the Iranian leadership did not update their beliefs. In the end, a combination of religiously based high offensive stakes, u(win), low perceived cost of war, c, and religious causal beliefs of high expected offensive capacity, p(win), contributed to the length of the war.

The demands of the two sides were visibly discrepant at the outset of the war in 1980. Saddam Hussein initially sought territorial gains that Iran was not prepared to give him. After grasping its window of opportunity, Baghdad achieved some of its war aims but realized that Iraq’s offensive war capacity was lacking. In accordance with the rational choice expectation of the theory of endogenous war termination, Hussein demanded that Iraq should keep the occupied territories, but gave up the rest. Iran, in turn, had not yet unleashed its total war-making capacity and was confident it could expel the aggressor, so it refused to negotiate unless Iraq withdrew to the old borders. A bargaining space was still non-existent.

Following the successful Iranian counter-attacks in 1982, Baghdad lowered its expected offensive capacity and war aims by unilaterally withdrawing to the old borders. The war could have ended at this time with the creation of a bargaining space. However, with the removal of President Bani Sadr, the more fundamentalist factions of Iranian society consolidated their power in decisionmaking circles. As a result, against the judgment of the professional military and the expectations of the theory of endogenous war termination, the offensive expectations, p(win), and the offensive stakes, u(win), were raised based on religious reasoning. This increased Iran’s war aims and excluded negotiations because Iraq did not consider the enemy’s war aims justifiable based on the current battlefield events.

In 1986, Iran managed to occupy more Iraqi territory, but the massive losses that Iran suffered during failed offensive actions later the same year could again have ended the war. However, the same religiously based reasoning kept the expected utility of continuing the war high and prevented a lowering of the war aims. It was not until mobilization problems began to emerge and Iraq delivered a serious blow to Iran in 1988 that Iranian offensive expectations began to wane, matching the actual Iranian offensive capacity and finally leading to a lowering of the war aims. The realities of the battlefield and the recruitment problems had finally weakened the causal beliefs relying on God’s help and religious fervor to pursue the war into Iraq.

Reiter (2009) and Weisiger (2013) see the underlying reason for the prolonged war as commitment problems. However, as the Iranian leadership had religiously based causal beliefs that led them to assume that they would win the war with a high p(win), and ultimately waged the war to spread the revolution with a high u(win), commitment problems were only a secondary reason for continuing the war. The primary mechanism for lowering the Iranian war aims was a change in the causal beliefs that had relied on God’s help to win the war. As Smith & Stam (2004: 787) suggest, the key to analyzing war duration is to ‘examine how beliefs change’ during the war. However, in this case, the belligerents had a more fundamental disagreement about causality, as compared to divergent theoretical perspectives on the role of military technology. As long as Tehran was adamant about God’s help, negative battlefield events were discredited as a reliable source of information for updating beliefs.

In the end, poor battlefield performance was too obvious to ignore and a change in causal beliefs not only lowered Iran’s offensive expectations, p(win), but also gave rise to the idea that the revolution could be spread by other means than open war, which lowered the offensive stakes in conquering Iraqi territory, u(win). Thus, in 1988 a bargaining space finally opened.

If both sides in a militarized conflict change their war aims according to their battlefield successes or failures, a bargaining space quickly emerges. Battlefield events can therefore signal relative strength and resolve so that agreement on acceptable terms of peace becomes possible. Yet this is often only an ideal situation. A long war easily results if confounding factors intervene to end this convenient connection between battlefield events and the expected utility of warfare. The case of the Iran–Iraq War illustrates how religion, especially when it creates causal beliefs raising p(win), increases the stakes, u(win), and lowers the cost of warfare, c, is a factor that can raise the expected utility of continued warfare and raise the war aims.

Religious causal beliefs constitute a real word mechanism that not only creates different priors about expected military capacity, but also slows down the process of updating beliefs relating to the p(win), as battlefield events are not seen as credible information. Thus, incorporating causal beliefs that are based on an interpretation of the will of God is a complement to how rational choice theories have expected actors to form and update their beliefs about the expected outcome of a war based on observable evidence. The general lesson to be learned is that immaterial factors such as religion hold the potential to prolong the duration of armed conflicts and a change in these beliefs is required for a peaceful resolution to materialize.

Although the prevalence of religious conflicts has increased over time (Fox, 2002; Toft, 2007), the formation of beliefs remains understudied when applying rational choice to real world conflicts. The results of this study, pointing to the difficulties of reaching a bargaining solution, should caution states against getting involved in conflicts in which the advice of professional military is neglected and religious beliefs are central components, unless they are sure to decisively overrun their enemy. While this study has illustrated the role of religiously based causal beliefs as a plausible explanation of very long wars, to further test the hypothesis about such delays in information updating, future studies should seek to measure variation in religious causal beliefs and how it is associated with war duration.

Footnotes

Notes

References

Abrahamian

Ervand

(2008) History of Modern Iran. Cambridge: Cambridge University Press.

Amos

John W

(1984) The Iran–Iraq war: Conflict linkage, and spillover in the Middle East. In: Darius

Robert G

Amos

John W

Magnus

Ralph H

(eds) Gulf Security in the 1980s: Perceptual and Strategic Dimensions. Stanford, CA: Hoover Institute Press, 49–81.

Associated Press (1987) Khomeini, in first speech in 3 months, pledges war until victory. New York Times 11 February (http://www.nytimes.com/1987/02/11/world/khomeini-in-first-speech-in-3-months-pledges-war-until-victory.html).

Aumann

Robert J

(1976) Agreeing to disagree. Annals of Statistics 4(6): 1236–1239.

Bennett, D Scott & Allan C Stam (1996) The duration of interstate wars, 1816–1985. American Political Science Review 90(2): 647–662.

Bennett, D Scott & Allan C Stam (2004) The Behavioral Origins of War. Ann Arbor, MI: University of Michigan Press.

Blattman

Christopher

Miguel

Edward

(2010) Civil war. Journal of Economic Literature 48(1): 3–57.

Brumberg

Daniel

(2001) Reinventing Khomeini: The Struggle for Reform in Iran. Chicago, IL: University of Chicago Press.

Bueno de Mesquita, Bruce (1981) The War Trap. New Haven, CT: Yale University Press.

10.

Chiozza

Giacomo

Goemans

Hein E

(2003) Peace through insecurity: Tenure and international conflict. Journal of Conflict Resolution 47(4): 443–467.

11.

Chubin

Shahram

(1989) Iran and the war: From stalemate to ceasefire. In: Karsh

Efraim

(ed.) The Iran–Iraq War: Impact and Implications. Basingstoke: Palgrave Macmillan, 13–25.

12.

Chubin

Shahram

Tripp

Charles

(1988) Iran and Iraq at War. Boulder, CO: Westview.

13.

Cordesman

Anthony H

(1984) The Gulf and the Search for Strategic Stability: Saudi Arabia, the Military Balance in the Gulf, and Trends in the Arab–Israeli Military Balance. Boulder, CO: Westview.

14.

Cordesman

Anthony H

Wagner

Abraham R

(1990) The Lessons of Modern War. Volume II: The Iran–Iraq War. Boulder, CO: Westview.

15.

Farhi

Farideh

(2004) The antinomies of Iran’s war generation. In Potter

Lawrence

Sick

Gary G

(eds) Iran, Iraq, and the Legacies of War. Gordonsville, VA: Palgrave Macmillan, 101–120.

16.

Fearon

James D

(1995) Rationalist explanations for war. International Organization 49(3): 379–414.

17.

Fearon

James D

(1998) Bargaining, enforcement, and international cooperation. International Organization 52(2): 269–305.

18.

Fearon

James D

(2013) Fighting rather than bargaining. Revised version of paper presented at the Annual Meetings of the American Political Science Association, Chicago, IL, 30 August–2 September 2007 (https://web.stanford.edu/group/fearon-research/cgi-bin/wordpress/wp-content/uploads/2013/10/frtb6.pdf).

19.

Filson

Darren

Werner

Suzanne

(2002) A bargaining model of war and peace: Anticipating the onset, duration, and outcome of war. American Journal of Political Science 46(4): 819–837.

20.

Fox

Jonathan

(2002) Ethnoreligious Conflict in the Late 20th Century. Lanham, MD: Lexington.

21.

Fox

William TR

(1970) The causes of peace and conditions of war. Annals of the American Academy of Political and Social Science 392(1): 1–13.

22.

Gelpi

Christopher

(1997) Democratic diversions: Governmental structure and the externalization of domestic conflict. Journal of Conflict Resolution 41(2): 255–282.

23.

Gieling

Saskia

(1999) Religion and War in Revolutionary Iran. London: IB Tauris.

24.

Goemans

Hein E

(2000) War and Punishment: The Causes of War Termination and the First World War. Princeton, NJ: Princeton University Press.

25.

Hassner

Ron E

(2003) ‘To have and to hold’: Conflicts over sacred space and the problem of indivisibility. Security Studies 12(4): 1–33.

26.

Hiro

Dilip

(1990) The Longest War: The Iran–Iraq Military Conflict. London: Paladin.

27.

Horowitz

Michael C

(2009) Long time going: Religion and the duration of crusading. International Security 34(2): 162–193.

28.

Iklé

Fred C

(1991) Every War Must End. New York: Columbia University Press.

29.

Jackson

Mathew O

Morelli

Massimo

(2007) Political bias and war. American Economic Review 97(4): 1353–1373.

30.

Jervis

Robert

(1976) Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press.

31.

Karsh

Efraim

(1987) The Iran–Iraq War: A Military Analysis. London: International Institute for Strategic Studies.

32.

Kecskemeti

Paul

(1958) Strategic Surrender: The Politics of Victory and Defeat. Stanford, CA: Stanford University Press.

33.

Khadduri

Majid

(1988) The Gulf War: The Origins and Implications of the Iran–Iraq Conflict. New York: Oxford University Press.

34.

Kifner

John

(1982) Khomeini faction challenged by more conservative group. New York Times 13 April (http://www.nytimes.com/1982/04/13/world/khomeini-faction-challenged-by-more-conservative-group.html?%20pagewanted=all).

35.

King

Ralph

(1987) The United Nations and the Iran–Iraq War. New York: Ford Foundation.

36.

Leventoglu

Bahar

Slantchev

Branislav

(2007) The armed peace: A punctuated equilibrium theory of war. American Journal of Political Science 51(4): 755–771.

37.

Levy

Jack

(1989) The diversionary theory of war: A critique. In: Midlarsky

Manus I

(ed.) Handbook of War Studies. Boston, MA: Unwin Hyman, 259–288.

38.

Levy

Jack

(1998) The causes of war and the conditions of peace. Annual Review of Political Science 1(1): 139–165.

39.

Levy

Jack

Froelich

Mike

(1985) Causes of the Iran–Iraq War. In: Brown

James

Snyder

William P

(eds) The Regionalization of Warfare. New Brunswick, NJ: Transaction, 127–143.

40.

Moshiri

Farrokh

(1991) Iran: Islamic revolution against Westernization. In: Goldstone

Jack

Gurr

Ted R

Moshiri

Farrokh

(eds) Revolutions of the Late Twentieth Century. Boulder, CO: Westview, 116–135.

41.

Nafisi

Rasool

(2006) The Khomeini letter: Is Rafsanjani warning the hardliners? Iranian.com 11 October (http://www.iranian.com/RasoolNafisi/2006/October/Nuclear/index.html).

42.

New York Times (1987) Excerpts from Khomeini’s speeches. New York Times 4 August (http://www.nytimes.com/1987/08/04/world/excerpts-from-khomeini-speeches.html).

43.

Nilsson

Marco

(2012) Offense–defense balance, war duration, and the security dilemma. Journal of Conflict Resolution 56(3): 467–489.

44.

Pear

Robert

(1988) Khomeini accepts the ‘poison’ of ending the war with Iraq; UN sending mission. New York Times 21 July (http://www.nytimes.com/1988/07/21/us/khomeini-accepts-poison-of-ending-the-war-with-iraq-un-sending-mission.html).

45.

Pillar

Paul R

(1983) Negotiating Peace. Princeton, NJ: Princeton University Press.

46.

Porsholt

Lars

(1966) On methods of conflict prevention. Journal of Peace Research 3(2): 178–193.

47.

Powell

Robert

(1999) In the Shadow of Power. Princeton, NJ: Princeton University Press.

48.

Powell

Robert

(2006) War as a commitment problem. International Organization 60(1): 169–203.

49.

Powell

Robert

(2012) Persistent fighting and shifting power. American Journal of Political Science 56(3): 620–637.

50.

Quester

George H

(1970) Wars prolonged by misunderstood signals. Annals of the American Academy of Political and Social Science 392(1): 30–39.

51.

Ramsay

Kristopher W

(2008) Settling it on the field: Battlefield events and war termination. Journal of Conflict Resolution 52(6): 850–879.

52.

Reiter

Dan

(2009) How Wars End. Princeton, NJ: Princeton University Press.

53.

Reiter

Dan

Stam

Allan C

(1998) Democracy, war initiation, and victory. American Political Science Review 92(2): 377–389.

54.

Shirkey

Zachary

(2016) Uncertainty and war duration. International Studies Review 18(2): 244–267.

55.

Slantchev

Stanislav L

(2004) How initiators end their wars: The duration of warfare and the terms of peace. American Journal of Political Science 48(4): 813–829.

56.

Smith

Alastair

Stam

Allan C

(2004) Bargaining and the nature of war. Journal of Conflict Resolution 48(6): 783–813.

57.

Stanley

Elizabeth A

Sawyer

John P

(2009) The equifinality of war termination: Multiple paths to ending war. Journal of Conflict Resolution 53(5): 651–676.

58.

Svensson

Isak

Harding

Emily

(2011) How holy wars end: Exploring the termination patterns of conflicts with religious dimensions in Asia. Terrorism and Political Violence 23(2): 133–149.

59.

Takeyh

Ray

(2009) Guardians of the Revolution: Iran and the World in the Age of the Ayatollahs. Oxford: Oxford University Press.

60.

Takeyh

Ray

(2010) The Iran–Iraq War: A reassessment. Middle East Journal 64(3): 365–383.

61.

Tannenwald

Nina

(2005) Ideas and explanation: Advancing the theoretical agenda. Journal of Cold War Studies 7(2): 13–42.

62.

Tibi

Bassam

(1998) Conflict and War in the Middle East. Basingstoke: Macmillan.

63.

Toft

Monica D

(2007) Getting religion? The puzzling case of Islam and civil war. International Security 31(4): 97–131.

64.

Wagner

R Harrison

(2000) Bargaining and war. American Journal of Political Science 44(3): 469–484.

65.

Walt

Stephen

(1996) Revolution and War. Ithaca, NY: Cornell University Press.

66.

Weisiger

Alex

(2013) Logics of War: Explanations for Limited and Unlimited Conflicts. Ithaca, NY: Cornell University Press.

67.

Werner

Suzanne

(1998) Negotiating the terms of settlement: War aims and bargaining leverage. Journal of Conflict Resolution 42(3): 321–343.

68.

Willett

Edward

(2004) The Iran–Iraq War. New York: Rosen.

69.

Williamson

Murray

Woods

Kevin M

(2014) The Iran–Iraq War: A Military and Strategic History. Cambridge: Cambridge University Press.

70.

Wittman

Donald

(1979) How a war ends. Journal of Conflict Resolution 23(4): 743–763.

71.

Wolford

Scott

Reiter

Dan

Carrubba

Clifford J

(2011) Information, commitment, and war. Journal of Conflict Resolution 55(4): 556–579.

72.

Wright

Quincy

(1965) The escalation of international conflicts. Journal of Conflict Resolution 9(4): 434–449.

73.

Yapp

Malcolm E

(1996) The Near East Since the First World War. Harlow: Longman.

74.

Zabih

Sepehr

(1988) The Iranian Military in Revolution and War. London: Routledge.