American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 August 2003 The Strategie Logic of Suicide Terrorism ROBERT A. PAPE The University of Chicago uicide terrorism is rising around the world, but the most common explanations do not help us understand why. Religious fanaticism does not explain why the world leader in suicide terrorism is the Tamil Tigers in Sri Tanka, a group that adheres to a Marxist/Teninist ideology, while existing psychological explanations have been contradicted by the widening range of socio-economic backgrounds of suicide terrorists. To advance our understanding of this growing phenomenon, this study collects the universe of suicide terrorist attacks worldwide from 1980 to 2001,187 in all. In contrast to the existing explanations, this study shows that suicide terrorism follows a strategic logic, one specifically designed to coerce modern liberal democracies to make significant territorial concessions. Moreover, over the past two decades, suicide terrorism has been rising largely because terrorists have learned that it pays. Suicide terrorists sought to compel American and French military forces to abandon Tebanon in 1983, Israeli forces to leave Tebanon in 1985, Israeli forces to quit the Gaza Strip and the West Bank in 1994 and 1995, the Sri Tankan government to create an independent Tamil state from 1990 on, and the Turkish government to grant autonomy to the Kurds in the late 1990s. In all but the case of Turkey, the terrorist political cause made more gains after the resort to suicide operations than it had before. Thus, Western democracies should pursue policies that teach terrorists that the lesson of the 1980s and 1990s no longer holds, policies which in practice may have more to do with improving homeland security than with offensive military action. S Terrorist organizations are increasingly relying on suicide attacks to achieve major political objectives. For example, spectacular suicide terrorist attacks have recently been employed by Palestinian groups in attempts to force Israel to abandon the West Bank and Gaza, by the Liberation Tigers of Tamil Eelam to compel the Sri Lankan government to accept an independent Tamil homeland, and by Al Qaeda to pressure the United States to withdraw from the Saudi Arabian Peninsula. Moreover, such attacks are increasing both in tempo and location. Before the early 1980s, suicide terrorism was rare but not unknown (Lewis 1968; O'Neill 1981; Rapoport 1984). However, since the attack on the U.S. embassy in Beirut in April 1983, there have been at least 188 separate suicide terrorist attacks worldwide, in Lebanon, Israel, Sri Lanka, India, Pakistan, Afghanistan, Yemen, Turkey, Russia and the United States. The rate has increased from 31 in the 1980s, to 104 in the 1990s, to 53 in 2000-2001 alone (Pape 2002). The rise of suicide terrorism is especially remarkable, given that the total number of terrorist incidents worldwide fell during the period, from a peak of 666 in 1987 to a low of 274 in 1998, with 348 in 2001 (Department of State 2001). What accounts for the rise in suicide terrorism, especially, the sharp escalation from the 1990s onward? Robert A. Pape is Associate Professor, Department of Political Science, 5828 South University Avenue, The University of Chicago, Chicago, IL 60637 (r-pape@uchicago.edu). I thank Robert Art, Steven Cicala, Alex Downs, Daniel Drezner, Adria Lawrence, Sean Lynn-Jones, John Mearsheimer, Michael O'Connor, Sebastian Rosato, Lisa Weeden, the anonymous reviewers, and the members of the program on International Security Policy at the University of Chicago for their superb comments. I especially thank James K. Feldman and Chaim D. Kaufmann for their excellent comments on multiple drafts. I would also like to acknowledge encouragement from the committee for the Combating Political Violence paper competition sponsored by the Institute for War and Peace Studies at Columbia University, which selected an earlier version as a winning paper. Although terrorism has long been part of international politics, we do not have good explanations for the growing phenomenon of suicide terrorism. Traditional studies of terrorism tend to treat suicide attack as one of many tactics that terrorists use and so do not shed much light on the recent rise of this type of attack (e.g., Hoffman 1998; Jenkins 1985; Laqueur 1987). The small number of studies addressed explicitly to suicide terrorism tend to focus on the irrationality of the act of suicide from the perspective of the individual attacker. As a result, they focus on individual motives—either religious indoctrination (especially Islamic Fundamentalism) or psychological predispositions that might drive individual suicide bombers (Kramer 1990; Merari 1990; Post 1990). The first-wave explanations of suicide terrorism were developed during the 1980s and were consistent with the data from that period. However, as suicide attacks mounted from the 1990s onward, it has become increasingly evident that these initial explanations are insufficient to account for which individuals become suicide terrorists and, more importantly, why terrorist organizations are increasingly relying on this form of attack (Institute for Counter-Terrorism 2001). First, although religious motives may matter, modern suicide terrorism is not limited to Islamic Fundamentalism. Islamic groups receive the most attention in Western media, but the world's leader in suicide terrorism is actually the Liberation Tigers of Tamil Eelam (LTTE), a group who recruits from the predominantly Hindu Tamil population in northern and eastern Sri Lanka and whose ideology has Marxist/Leninist elements. The LTTE alone accounts for 75 of the 186 suicide terrorist attacks from 1980 to 2001. Even among Islamic suicide attacks, groups with secular orientations account for about a third of these attacks (Merari 1990; Sprinzak 2000). Second, although study of the personal characteristics of suicide attackers may someday help identify 1 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 individuals terrorist organizations are likely to recruit for this purpose, the vast spread of suicide terrorism over the last two decades suggests that there may not be a single profile. Until recently, the leading experts in psychological profiles of suicide terrorists characterized them as uneducated, unemployed, socially isolated, single men in their late teens and early 20s (Merari 1990; Post 1990). Now we know that suicide terrorists can be college educated or uneducated, married or single, men or women, socially isolated or integrated, from age 13 to age 47 (Sprinzak 2000). In other words, although only a tiny number of people become suicide terrorists, they come from a broad cross section of lifestyles, and it may be impossible to pick them out in advance. In contrast to the first-wave explanations, this article shows that suicide terrorism follows a strategic logic. Even if many suicide attackers are irrational or fanatical, the leadership groups that recruit and direct them are not. Viewed from the perspective of the terrorist organization, suicide attacks are designed to achieve specific political purposes: to coerce a target government to change policy, to mobilize additional recruits and financial support, or both. Crenshaw (1981) has shown that terrorism is best understood in terms of its strategic function; the same is true for suicide terrorism. In essence, suicide terrorism is an extreme form of what Thomas Schelling (1966) calls "the rationality of irrationality," in which an act that is irrational for individual attackers is meant to demonstrate credibility to a democratic audience that still more and greater attacks are sure to come. As such, modern suicide terrorism is analogous to instances of international coercion. For states, air power and economic sanctions are often the preferred coercive tools (George et al. 1972; Pape 1996, 1997). For terrorist groups, suicide attacks are becoming the coercive instrument of choice. To examine the strategic logic of suicide terrorism, this article collects the universe suicide terrorist attacks worldwide from 1980 to 2001, explains how terrorist organizations have assessed the effectiveness of these attacks, and evaluates the limits on their coercive utility. Five principal findings follow. First, suicide terrorism is strategic. The vast majority of suicide terrorist attacks are not isolated or random acts by individual fanatics but, rather, occur in clusters as part of a larger campaign by an organized group to achieve a specific political goal. Groups using suicide terrorism consistently announce specific political goals and stop suicide attacks when those goals have been fully or partially achieved. Second, the strategic logic of suicide terrorism is specifically designed to coerce modern democracies to make significant concessions to national self-determination. In general, suicide terrorist campaigns seek to achieve specific territorial goals, most often the withdrawal of the target state's military forces from what the terrorists see as national homeland. From Lebanon to Israel to Sri Lanka to Kashmir to Chechnya, every suicide terrorist campaign from 1980 to 2001 has been waged by terrorist groups whose main goal has been to establish or maintain self-determination for their community's homeland by compelling an enemy to withdraw. Further, every suicide terrorist campaign since 1980 has been targeted against a state that had a democratic form of government. Third, during the past 20 years, suicide terrorism has been steadily rising because terrorists have learned that it pays. Suicide terrorists sought to compel American and French military forces to abandon Lebanon in 1983, Israeli forces to leave Lebanon in 1985, Israeli forces to quit the Gaza Strip and the West Bank in 1994 and 1995, the Sri Lankan government to create an independent Tamil state from 1990 on, and the Turkish government to grant autonomy to the Kurds in the late 1990s. Terrorist groups did not achieve their full objectives in all these cases. However, in all but the case of Turkey, the terrorist political cause made more gains after the resort to suicide operations than it had before. Leaders of terrorist groups have consistently credited suicide operations with contributing to these gains. These assessments are hardly unreasonable given the timing and circumstances of many of the concessions and given that other observers within the terrorists' national community, neutral analysts, and target government leaders themselves often agreed that suicide operations accelerated or caused the concession. This pattern of making concessions to suicide terrorist organizations over the past two decades has probably encouraged terrorist groups to pursue even more ambitious suicide campaigns. Fourth, although moderate suicide terrorism led to moderate concessions, these more ambitious suicide terrorist campaigns are not likely to achieve still greater gains and may well fail completely. In general, suicide terrorism relies on the threat to inflict low to medium levels of punishment on civilians. In other circumstances, this level of punishment has rarely caused modern nation states to surrender significant political goals, partly because modern nation states are often willing to countenance high costs for high interests and partly because modern nation states are often able to mitigate civilian costs by making economic and other adjustments. Suicide terrorism does not change a nation's willingness to trade high interests for high costs, but suicide attacks can overcome a country's efforts to mitigate civilian costs. Accordingly, suicide terrorism may marginally increase the punishment that is inflicted and so make target nations somewhat more likely to surrender modest goals, but it is unlikely to compel states to abandon important interests related to the physical security or national wealth of the state. National governments have in fact responded aggressively to ambitious suicide terrorist campaigns in recent years, events which confirm these expectations. Finally, the most promising way to contain suicide terrorism is to reduce terrorists' confidence in their ability to carry out such attacks on the target society. States that face persistent suicide terrorism should recognize that neither offensive military action nor concessions alone are likely to do much good and should invest significant resources in border defenses and other means of homeland security. 2 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 THE LOGIC OF SUICIDE TERRORISM Most suicide terrorism is undertaken as a strategic effort directed toward achieving particular political goals; it is not simply the product of irrational individuals or an expression of fanatical hatreds. The main purpose of suicide terrorism is to use the threat of punishment to coerce a target government to change policy, especially to cause democratic states to withdraw forces from territory terrorists view as their homeland. The record of suicide terrorism from 1980 to 2001 exhibits tendencies in the timing, goals, and targets of attack that are consistent with this strategic logic but not with irrational or fanatical behavior: (1) timing— nearly all suicide attacks occur in organized, coherent campaigns, not as isolated or randomly timed incidents; (2) nationalist goals—suicide terrorist campaigns are directed at gaining control of what the terrorists see as their national homeland territory, specifically at ejecting foreign forces from that territory; and (3) target selection—all suicide terrorist campaigns in the last two decades have been aimed at democracies, which make more suitable targets from the terrorists' point of view. Defining Suicide Terrorism Terrorism involves the use of violence by an organization other than a national government to cause intimidation or fear among a target audience (Department of State 1983-2001; Reich 1990; Schmid and Jongman 1988). Although one could broaden the definition of terrorism so as to include the actions of a national government to cause terror among an opposing population, adopting such a broad definition would distract attention from what policy makers would most like to know: how to combat the threat posed by subnational groups to state security. Further, it could also create analytic confusion. Terrorist organizations and state governments have different levels of resources, face different kinds of incentives, and are susceptible to different types of pressures. Accordingly, the determinants of their behavior are not likely to be the same and, thus, require separate theoretical investigations. In general, terrorism has two purposes—to gain supporters and to coerce opponents. Most terrorism seeks both goals to some extent, often aiming to affect enemy calculations while simultaneously mobilizing support or the terrorists cause and, in some cases, even gaining an edge over rival groups in the same social movement (Bloom 2002). However, there are trade-offs between these objectives and terrorists can strike various balances between them. These choices represent different forms of terrorism, the most important of which are demonstrative, destructive, and suicide terrorism. Demonstrative terrorism is directed mainly at gaining publicity, for any or all of three reasons: to recruit more activists, to gain attention to grievances from soft-liners on the other side, and to gain attention from third parties who might exert pressure on the other side. Groups that emphasize ordinary, demonstrative terrorism include the Orange Volunteers (Northern Ireland), National Liberation Army (Columbia), and Red Brigades (Italy) (Clutterbuck 1975; Edler Baumann 1973; St. John 1991). Hostage taking, airline hijacking, and explosions announced in advance are generally intended to use the possibility of harm to bring issues to the attention of the target audience. In these cases, terrorists often avoid doing serious harm so as not to undermine sympathy for the political cause. Brian Jenkins (1975,4) captures the essence of demonstrative terrorism with his well-known remark, "Terrorists want a lot of people watching, not a lot of people dead." Destructive terrorism is more aggressive, seeking to coerce opponents as well as mobilize support for the cause. Destructive terrorists seek to inflict real harm on members of the target audience at the risk of losing sympathy for their cause. Exactly how groups strike the balance between harm and sympathy depends on the nature of the political goal. For instance, the Baader-Meinhoft group selectively assassinated rich German industrialists, which alienated certain segments of German society but not others. Palestinian terrorists in the 1970s often sought to kill as many Israelis as possible, fully alienating Jewish society but still evoking sympathy from Muslim communities. Other groups that emphasize destructive terrorism include the Irish Republican Army, the Revolutionary Armed Forces of Colombia (FARC), and the nineteenth-century Anarchists (Elliott 1998; Rapoport 1971; Tuchman 1966). Suicide terrorism is the most aggressive form of terrorism, pursuing coercion even at the expense of losing support among the terrorists' own community. What distinguishes a suicide terrorist is that the attacker does not expect to survive a mission and often employs a method of attack that requires the attacker's death in order to succeed (such as planting a car bomb, wearing a suicide vest, or ramming an airplane into a building). In essence, a suicide terrorist kills others at the same time that he kills himself.1 In principle, suicide terrorists could be used for demonstrative purposes or could be limited to targeted assassinations. In practice, however, suicide terrorists often seek simply to kill the largest number of people. Although this maximizes the coercive leverage that can be gained from terrorism, it does so at the greatest cost to the basis of support for the terrorist cause. Maximizing the number of enemy killed alienates those in the target audience who might be sympathetic to the terrorists cause, while the act of 1 A suicide attack can be defined in two ways, a narrow definition limited to situations in which an attacker kills himself and a broad definition that includes any instance when an attacker fully expects to be killed by others during an attack. An example that fits the broad definition is Baruch Goldstein, who continued killing Palestinians at the February 1994 Hebron Massacre until he himself was killed, who had no plan for escape, and who left a note for his family indicating that he did not expect to return. My research relies on the narrow definition, partly because this is the common practice in the literature and partly because there are so few instances in which it is clear that an attacker expected to be killed by others that adding this category of events would not change my findings. 2 Hunger strikes and self-immolation are not ordinarily considered acts of terrorism, because their main purpose is to evoke understanding and sympathy from the target audience, and not to cause terror (Niebuhr 1960). 3 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 suicide creates a debate and often loss of support among moderate segments of the terrorists' community, even if also attracting support among radical elements. Thus, while coercion is an element in all terrorism, coercion is the paramount objective of suicide terrorism. The Coercive Logic of Suicide Terrorism At its core, suicide terrorism is a strategy of coercion, a means to compel a target government to change policy. The central logic of this strategy is simple: Suicide terrorism attempts to inflict enough pain on the opposing society to overwhelm their interest in resisting the terrorists demands and, so, to cause either the government to concede or the population to revolt against the government. The common feature of all suicide terrorist campaigns is that they inflict punishment on the opposing society, either directly by killing civilians or indirectly by killing military personnel in circumstances that cannot lead to meaningful battlefield victory. As we shall see, suicide terrorism is rarely a one time event but often occurs in a series of suicide attacks. As such, suicide terrorism generates coercive leverage both from the immediate panic associated with each attack and from the risk of civilian punishment in the future. Suicide terrorism does not occur in the same circumstances as military coercion used by states, and these structural differences help to explain the logic of the strategy. In virtually all instances of international military coercion, the coercer is the stronger state and the target is the weaker state; otherwise, the coercer would likely be deterred or simply unable to execute the threatened military operations (Pape 1996). In these circumstances, coercers have a choice between two main coercive strategies, punishment and denial. Punishment seeks to coerce by raising the costs or risks to the target society to a level that overwhelms the value of the interests in dispute. Denial seeks to coerce by demonstrating to the target state that it simply cannot win the dispute regardless of its level of effort, and therefore fighting to a finish is pointless—for example, because the coercer has the ability to conquer the disputed territory. Hence, although coercers may initially rely on punishment, they often have the resources to create a formidable threat to deny the opponent victory in battle and, if necessary, to achieve a brute force military victory if the target government refuses to change its behavior. The Allied bombing of Germany in World War II, American bombing of North Vietnam in 1972, and Coalition attacks against Iraq in 1991 all fit this pattern. Suicide terrorism (and terrorism in general) occurs under the reverse structural conditions. In suicide terrorism, the coercer is the weaker actor and the target is the stronger. Although some elements of the situation remain the same, flipping the stronger and weaker sides in a coercive dispute has a dramatic change on the relative feasibility of punishment and denial. In these circumstances, denial is impossible, because military conquest is ruled out by relative weakness. Even though some groups using suicide terrorism have received important support from states and some have been strong enough to wage guerrilla military campaigns as well as terrorism, none have been strong enough to have serious prospects of achieving their political goals by conquest. The suicide terrorist group with the most significant military capacity has been the LTTE, but it has not had a real prospect of controlling the whole of the homeland that it claims, including Eastern and Northern Provinces of Sri Lanka. As a result, the only coercive strategy available to suicide terrorists is punishment. Although the element of "suicide" is novel and the pain inflicted on civilians is often spectacular and gruesome, the heart of the strategy of suicide terrorism is the same as the coercive logic used by states when they employ air power or economic sanctions to punish an adversary: to cause mounting civilian costs to overwhelm the target state's interest in the issue in dispute and so to cause it to concede the terrorists' political demands. What creates the coercive leverage is not so much actual damage as the expectation of future damage. Targets may be economic or political, military or civilian, but in all cases the main task is less to destroy the specific targets than to convince the opposing society that they are vulnerable to more attacks in the future. These features also make suicide terrorism convenient for retaliation, a tit-for-tat interaction that generally occurs between terrorists and the defending government (Crenshaw 1981). The rhetoric of major suicide terrorist groups reflects the logic of coercive punishment. Abdel Karim, a leader of Al Aksa Martyrs Brigades, a militant group linked to Yasir Arafat's Fatah movement, said the goal of his group was "to increase losses in Israel to a point at which the Israeli public would demand a withdrawal from the West Bank and Gaza Strip" (Greenberg 2002). The infamous fatwa signed by Osama Bin Laden and others against the United States reads, "The ruling to kill the Americans and their allies—civilians and military—is an individual duty for every Muslim who can do it in any country in which it is possible to do it, in order to liberate the al-Aqsa Mosque and the holy mosque [Mecca] from their grip, and in order for their armies to move out of all the lands of Islam, defeated and unable to threaten any Muslim" (World Islamic Front 1998). Suicide terrorists' willingness to die magnifies the coercive effects of punishment in three ways. First, suicide attacks are generally more destructive than other terrorist attacks. An attacker who is willing to die is much more likely to accomplish the mission and to cause maximum damage to the target. Suicide attackers can conceal weapons on their own bodies and make last-minute adjustments more easily than ordinary terrorists. They are also better able to infiltrate heavily guarded targets because they do not need escape plans or rescue teams. Suicide attackers are also able to use certain especially destructive tactics such as wearing "suicide vests" and ramming vehicles into targets. The 188 suicide terrorist attacks from 1980 to 2001 killed an average of 13 people each, not counting the unusually 4 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 large number of fatalities on September 11 and also not counting the attackers themselves. During the same period, there were about 4,155 total terrorist incidents worldwide, which killed 3,207 people (also excluding September 11), or less than one person per incident. Overall, from 1980 to 2001, suicide attacks amount to 3% of all terrorist attacks but account for 48% of total deaths due to terrorism, again excluding September 11 (Department of State 1983-2001). Second, suicide attacks are an especially convincing way to signal the likelihood of more pain to come, because suicide itself is a costly signal, one that suggests that the attackers could not have been deterred by a threat of costly retaliation. Organizations that sponsor suicide attacks can also deliberately orchestrate the circumstances around the death of a suicide attacker to increase further expectations of future attacks. This can be called the "art of martyrdom" (Schalk 1997). The more suicide terrorists justify their actions on the basis of religious or ideological motives that match the beliefs of a broader national community, the more the status of terrorist martyrs is elevated, and the more plausible it becomes that others will follow in their footsteps. Suicide terrorist organizations commonly cultivate "sacrificial myths" that include elaborate sets of symbols and rituals to mark an individual attacker's death as a contribution to the nation. Suicide attackers' families also often receive material rewards both from the terrorist organizations and from other supporters. As a result, the art of martyrdom elicits popular support from the terrorists' community, reducing the moral backlash that suicide attacks might otherwise produce, and so establishes the foundation for credible signals of more attacks to come. Third, suicide terrorist organizations are better positioned than other terrorists to increase expectations about escalating future costs by deliberately violating norms in the use of violence. They can do this by crossing thresholds of damage, by breaching taboos concerning legitimate targets, and by broadening recruitment to confound expectations about limits on the number of possible terrorists. The element of suicide itself helps increase the credibility of future attacks, because it suggests that attackers cannot be deterred. Although the capture and conviction of Timothy McVeigh gave reason for some confidence that others with similar political views might be deterred, the deaths of the September 11 hijackers did not, because Americans would have to expect that future Al Qaeda attackers would be equally willing to die. The Record of Suicide Terrorism, 1980 to 2001 To characterize the nature of suicide terrorism, this study identified every suicide terrorist attack from 1980 to 2001 that could be found in Lexis Nexis's on-line database of world news media (Pape 2002) .3 Examination of the universe shows that suicide terrorism 3 This survey sought to include every instance of a suicide attack in which the attacker killed himself except those explicitly authorized has three properties that are consistent with the above strategic logic but not with irrational or fanatical behavior: (1) timing—nearly all suicide attacks occur in organized, coherent campaigns, not as isolated or randomly timed incidents; (2) nationalist goals—suicide terrorist campaigns are directed at gaining control of what the terrorists see as their national homeland territory, specifically at ejecting foreign forces from that territory; and (3) target selection—all suicide terrorist campaigns in the last two decades have been aimed at democracies, which make more suitable targets from the terrorists' point of view. Nationalist movements that face nondemocratic opponents have not resorted to suicide attack as a means of coercion. Timing. As Table 1 indicates, there have been 188 separate suicide terrorist attacks between 1980 and 2001. Of these, 179, or 95%, were parts of organized, coherent campaigns, while only nine were isolated or random events. Seven separate disputes have led to suicide terrorist campaigns: the presence of American and French forces in Lebanon, Israeli occupation of West Bank and Gaza, the independence of the Tamil regions of Sri Lanka, the independence of the Kurdish region of Turkey, Russian occupation of Chechnya, Indian occupation of Kashmir, and the presence of American forces on the Saudi Arabian Peninsula. Overall, however, there have been 16 distinct campaigns, because in certain disputes the terrorists elected to suspend operations one or more times either in response to concessions or for other reasons. Eleven of the campaigns have ended and five were ongoing as of the end of 2001. The attacks comprising each campaign were organized by the same terrorist group (or, sometimes, a set of cooperating groups as in the ongoing "second intifada" in Israel/Palestine), clustered in time, publically justified in terms of a specified political goal, and directed against targets related to that goal. The most important indicator of the strategic orientation of suicide terrorists is the timing of the suspension of campaigns, which most often occurs based on a strategic decision by leaders of the terrorist organizations that further attacks would be counterproductive to their coercive purposes—for instance, in response to full or partial concessions by the target state to the terrorists' political goals. Such suspensions are often accompanied by public explanations that justify the by a state and carried out by the state government apparatus (e.g., Iranian human wave attacks in the Iran-Iraq war were not counted). The survey is probably quite reliable, because a majority of the incidents were openly claimed by the sponsoring terrorist organizations. Even those that were not were, in nearly all cases, reported multiple times in regional news media, even if not always in the U.S. media. To probe for additional cases, I interviewed experts and officials involved in what some might consider conflicts especially prone to suicide attacks, such as Afghanistan in the 1980s, but this did not yield more incidents. According to the CIA station chief for Pakistan from 1986 to 1988 (Bearden 2002), "I cannot recall a single incident where an Afghan launched himself against a Soviet target with the intention of dying in the process. I don't think these things ever happened, though some of their attacks were a little hare-brained and could have been considered suicidal. I think it's important that Afghans never even took their war outside their borders—for example they never tried to blow up the Soviet Embassy in Pakistan." 5 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 TABLE 1. Suicide Terrorist Campaigns, 1980-2001 No. of No. Date Terrorist Group Terrorists' Goal Attacks Killed Target Behavior Completed Campaigns 1. Apr-Dec 1983 Hezbollah U.S./France out of Lebanon 6 384 Complete withdrawal 2. Nov 1983-Apr 1985 Hezbollah Israel out of Lebanon 6 96 Partial withdrawal 3. June 1985-June 1986 Hezbollah Israel out of Lebanon security zone 16 179 No change 4. July 1990-Nov 1994 LTTE Sri Lanka accept Tamil state 14 164 Negotiations 5. Apr 1995-Oct 2000 LTTE Sri Lanka accept Tamil state 54 629 No change 6. Apr 1994 Hamas Israel out of Palestine 2 15 Partial withdrawal from Gaza 7. Oct 1994-Aug 1995 Hamas Israel out of Palestine 7 65 Partial withdrawal from West Bank 8. Feb-Mar 1996 Hamas Retaliation for Israeli assassination 4 58 No change 9. Mar-Sept 1997 Hamas Israel out of Palestine 3 24 Hamas leader released 10. June-Oct1996 PKK Turkey accept Kurd autonomy 3 17 No change 11. Mar-Aug 1999 PKK Turkey release jailed leader 6 0 No change Ongoing Campaigns, as of December 2001 12. 1996- Al Qaeda U.S. out of Saudi Peninsula 5 3,329 TBDa 13. 2000- Chechnen Rebels Russia out of Chechnya 4 53 TBD 14. 2000- Kashmir Rebels India out of Kashmir 3 45 TBD 15. 2001- LTTE Sri Lanka accept Tamil state 6 51 TBD 16. 2000- Several Israel out of Palestine 39 177 TBD Total incidents 188 No. in campaigns 179 No. isolated 9 Source: Pape (2002). aTo be determined. decision to opt for a "cease-fire." Further, the terrorist organizations' discipline is usually fairly good; although there are exceptions, such announced cease-fires usually do stick for a period of months at least, normally until the terrorist leaders take a new strategic decision to resume in pursuit of goals not achieved in the earlier campaign. This pattern indicates that both terrorist leaders and their recruits are sensitive to the coercive value of the attacks. As an example of a suicide campaign, consider Hamas's suicide attacks in 1995 to compel Israel to withdraw from towns in the West Bank Hamas leaders deliberately withheld attacking during the spring and early summer in order to give PLO negotiations with Israel an opportunity to finalize a withdrawal. However, when in early July, Hamas leaders came to believe that Israel was backsliding and delaying withdrawal, Hamas launched a series of suicide attacks. Israel accelerated the pace of its withdrawal, after which Hamas ended the campaign. Mahmud al-Zahar, a Hamas leader in Gaza, announced, following the cessation of suicide attacks in October 1995: We must calculate the benefit and cost of continued armed operations. If we can fulfill our goals without violence, we will do so. Violence is a means, not a goal. Hamas's decision to adopt self-restraint does not contradict our aims, which include the establishment of an Islamic state instead of Israel.... We will never recognize Israel, but it is possible that a truce could prevail between us for days, months, or years. (Mishal and Sela 2000, 71) If suicide terrorism were mainly irrational or even disorganized, we would expect a much different pattern in which either political goals were not articulated (e.g., references in news reports to "rogue" attacks) or the stated goals varied considerably even within the same conflict. We would also expect the timing to be either random or, perhaps, event-driven, in response to particularly provocative or infuriating actions by the other side, but little if at all related to the progress of negotiations over issues in dispute that the terrorists want to influence. Nationalist Goals. Suicide terrorism is a high-cost strategy, one that would only make strategic sense for a group when high interests are at stake and, even then, as a last resort. The reason is that suicide terrorism maximizes coercive leverage at the expense of support among the terrorists' own community and so can be sustained over time only when there already exists a high degree of commitment among the potential pool of recruits. The most important goal that a community can have is the independence of its homeland (population, property, and way of life) from foreign influence or control. As a result, a strategy of suicide terrorism is most likely to be used to achieve nationalist goals, such as gaining control of what the terrorists see as 6 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 TABLE 2. Motivation and Targets of Suicide Terrorist Campaigns, 1980-2001 Region Dispute Homeland Status Terrorist Goal Target a Democracy? Lebanon, 1983-86 West Bank/Gaza, 1994-Tamils in Sri Lanka, 1990-Kurds in Turkey, 1990s Chechnya, 2000-Kashmir, 2000-Saudi Peninsula, 1996- U.S./F/IDF military presence IDF military presence SL military presence Turkey military presence Russia military presence Indian military presence U.S. military presence______ U.S./F/IDF withdrawal IDF withdrawal SL withdrawal Turkey withdrawal Russian withdrawal Indian withdrawal U.S. withdrawal Yes Yes Yes(1950)a Yes (1983)a Yes (1993)a Yes Yes Sources: Pape (2002). Przeworski et al. 2000 identifies four simple rules for determining regime type: (1) The chief executive must be elected, (2) the legislature must be elected, (3) there must be more than one party, and (4) there must be at least one peaceful transfer of power. By these criteria all the targets of suicide terrorism were and are democracies. Przeworski et al. codes only from 1950 to 1990 and is updated to 1999 by Boix and Rosato 2001. Freedom House also rates countries as "free," "partly free," and "not free," using criteria for degree of political rights and civil liberties. According to Freedom House's measures, Sri Lanka, Turkey, and Russia were all partly free when they were the targets of suicide terrorism, which puts them approximately in the middle of all countries, a score that is actually biased against this study since terrorism itself lowers a country's civil liberties rating (freedomhouse.org). aDate established as a democracy (if not always a democracy). their national homeland territory and expelling foreign military forces from that territory. In fact, every suicide campaign from 1980 to 2001 has had as a major objective—or as its central objective— coercing a foreign government that has military forces in what they see as their homeland to take those forces out. Table 2 summarizes the disputes that have engendered suicide terrorist campaigns. Since 1980, there has not been a suicide terrorist campaign directed mainly against domestic opponents or against foreign opponents who did not have military forces in the terrorists homeland. Although attacks against civilians are often the most salient to Western observers, actually every suicide terrorist campaign in the past two decades has included attacks directly against the foreign military forces in the country, and most have been waged by guerrilla organizations that also use more conventional methods of attack against those forces. Even Al Qaeda fits this pattern. Although Saudi Arabia is not under American military occupation per se and the terrorists have political objectives against the Saudi regime and others, one major objective of Al Qaeda is the expulsion of U.S. troops from the Saudi Peninsula and there have been attacks by terrorists loyal to Osama Bin Laden against American troops in Saudi Arabia. To be sure, there is a major debate among Islamists over the morality of suicide attacks, but within Saudi Arabia there is little debate over Al Qaeda's objection to American forces in the region and over 95% of Saudi society reportedly agrees with Bin Laden on this matter (Sciolino 2002). Still, e ven if suicide terrorism follows a strategic logic, could some suicide terrorist campaigns be irrational in the sense that they are being waged for unrealistic goals? The answer is that some suicide terrorist groups have not been realistic in expecting the full concessions demanded of the target, but this is normal for disputes involving overlapping nationalist claims and even for coercive attempts in general. Rather, the ambitions of terrorist leaders are realistic in two other senses. First, suicide terrorists' political aims, if not their methods, are often more mainstream than observers realize; they generally reflect quite common, straight- forward nationalist self-determination claims of their community. Second, these groups often have significant support for their policy goals versus the target state, goals that are typically much the same as those of other nationalists within their community. Differences between the terrorists and more "moderate" leaders usually concern the usefulness of a certain level of violence and—sometimes—the legitimacy of attacking additional targets besides foreign troops in the country, such as attacks in other countries or against third parties and civilians. Thus, it is not that the terrorists pursue radical goals and then seek others' support. Rather, the terrorists are simply the members of their societies who are the most optimistic about the usefulness of violence for achieving goals that many, and often most, support. The behavior of Hamas illustrates the point. Hamas terrorism has provoked Israeli retaliation that has been costly for Palestinians, while pursuing the—apparently unrealistic—goal of abolishing the state of Israel. Although prospects of establishing an Arab state in all of "historic Palestine" may be poor, most Palestinians agree that it would be desirable if possible. Hamas's terrorist violence was in fact carefully calculated and controlled. In April 1994, as its first suicide campaign was beginning, Hamas leaders explained that "martyrdom operations" would be used to achieve intermediate objectives, such as Israeli withdrawal from the West Bank and Gaza, while the final objective of creating an Islamic state from the Jordan River to the Mediterranean may require other forms of armed resistance (Shiqaqi 2002; Hroub 2000; Nusse 1998). Democracies as the Targets. Suicide terrorism is more likely to be employed against states with democratic political systems than authoritarian governments for several reasons. First, democracies are often thought to be especially vulnerable to coercive punishment. Domestic critics and international rivals, as well as terrorists, often view democracies as "soft," usually on the grounds that their publics have low thresholds of cost tolerance and high ability to affect state policy. Even if there is little evidence that democracies are easier to coerce than other regime types (Horowitz 7 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 and Reiter 2001), this image of democracy matters. Since terrorists can inflict only moderate damage in comparison to even small interstate wars, terrorism can be expected to coerce only if the target state is viewed as especially vulnerable to punishment. Second, suicide terrorism is a tool of the weak, which means that, regardless of how much punishment the terrorists inflict, the target state almost always has the capacity to retaliate with far more extreme punishment or even by exterminating the terrorists' community. Accordingly, suicide terrorists must not only have high interests at stake, they must also be confident that their opponent will be at least somewhat restrained. While there are infamous exceptions, democracies have generally been more restrained in their use of force against civilians, at least since World War II. Finally, suicide attacks may also be harder to organize or publicize in authoritarian police states, although these possibilities are weakened by the fact that weak authoritarian states are also not targets. In fact, the target state of every modern suicide campaign has been a democracy. The United States, France, Israel, India, Sri Lanka, Turkey, and Russia were all democracies when they were attacked by suicide terrorist campaigns, even though the last three became democracies more recently than the others. To be sure, these states vary in the degree to which they share "liberal" norms that respect minority rights; Freedom House rates Sri Lanka, Turkey, and Russia as "partly free" (3.5-4.5 on a seven-point scale) rather than "free" during the relevant years, partly for this reason and partly because terrorism and civil violence themselves lowers the freedom rating of these states. Still, all these states elect their chief executives and legislatures in multiparty elections and have seen at least one peaceful transfer of power, making them solidly democratic by standard criteria (Boix and Rosato 2001; Huntington 1991; Przeworski et al. 2000). The Kurds, which straddle Turkey and Iraq, illustrate the point that suicide terrorist campaigns are more likely to be targeted against democracies than authoritarian regimes. Although Iraq has been far more brutal toward its Kurdish population than has Turkey, violent Kurdish groups have used suicide attacks exclusively against democratic Turkey and not against the authoritarian regime in Iraq. There are plenty of national groups living under authoritarian regimes with grievances that could possibly inspire suicide terrorism, but none have. Thus, the fact that rebels have resorted to this strategy only when they face the more suitable type of target counts against arguments that suicide terrorism is a nonstrategic response, motivated mainly by fanaticism or irrational hatreds. TERRORISTS' ASSESSMENTS OF SUICIDE TERRORISM The main reason that suicide terrorism is growing is that terrorists have learned that it works. Even more troubling, the encouraging lessons that terrorists have learned from the experience of 1980s and 1990s are not, for the most part, products of wild-eyed interpretations or wishful thinking. They are, rather, quite reasonable assessments of the outcomes of suicide terrorist campaigns during this period. To understand how terrorists groups have assessed the effectiveness of suicide terrorism requires three tasks: (1) explanation of appropriate standards for evaluating the effectiveness of coercion from the standpoint of coercers; (2) analysis of the 11 suicide terrorist campaigns that have ended as of 2001 to determine how frequently target states made concessions that were, or at least could have been, interpreted as due to suicide attack; and (3) close analysis of terrorists' learning from particular campaigns. Because some analysts see suicide terrorism as fundamentally irrational (Kramer 1990; Merari 1990; Post 1990), it is important to assess whether the lessons that the terrorists drew were reasonable conclusions from the record. The crucial cases are the Hamas and Islamic Jihad campaigns against Israel during the 1990s, because they are most frequently cited as aimed at unrealistic goals and therefore as basically irrational. Standards of Assessment Terrorists, like other people, learn from experience. Since the main purpose of suicide terrorism is coercion, the learning that is likely to have the greatest impact on terrorists' future behavior is the lessons that they have drawn from past campaigns about the coercive effectiveness of suicide attack. Most analyses of coercion focus on the decision making of target states, largely to determine their vulnerability to various coercive pressures (George 1972; Pape 1996). The analysis here, however, seeks to determine why terrorist coercers are increasingly attracted to a specific coercive strategy. For this purpose, we must develop a new set of standards, because assessing the value of coercive pressure for the coercer is not the same problem as assessing its impact on the target. From the perspective of a target state, the key question is whether the value of the concession that the coercer is demanding is greater than the costs imposed by the coercive pressure, regardless of whether that pressure is in the form of lives at risk, economic hardship, or other types of costs. However, from the perspective of the coercer, the key question is whether a particular coercive strategy promises to be more effective than alternative methods of influence and, so, warrants continued (or increased) effort. This is especially true for terrorists who are highly committed to a particular goal and so willing to exhaust virtually any alternative rather than abandoning it. In this search for an effective strategy, coercers' assessments are likely to be largely a function of estimates of the success of past efforts; for suicide terrorists, this means assessments of whether past suicide campaigns produced significant concessions. A glance at the behavior of suicide terrorists reveals that such trade-offs between alternative methods are important in their calculations. All of the organizations 8 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 that have resorted to suicide terrorism began their coercive efforts with more conventional guerrilla operations, nonsuicide terrorism, or both. Hezbollah, Hamas, Islamic Jihad, the PKK, the LTTE, and Al Qaeda all used demonstrative and destructive means of violence long before resorting to suicide attack. Indeed, looking at the trajectory of terrorist groups over time, there is a distinct element of experimentation in the techniques and strategies used by these groups and distinct movement toward those techniques and strategies that produce the most effect. Al Qaeda actually prides itself for a commitment to even tactical learning over time—the infamous "terrorist manual" stresses at numerous points the importance of writing "lessons learned" memoranda that can be shared with other members to improve the effectiveness of future attacks. The most important analytical difficulty in assessing outcomes of coercive efforts is that successes are more ambiguous than failures. Whenever a suicide terrorist campaign, or any coercive effort, ends without obtaining significant concessions, presumably the co-ercers must judge the effort as a failure. If, however, the target state does make policy changes in the direction of the terrorists' political goals, this may or may not represent a coercive success for suicide attack in the calculations of the terrorists. The target government's decision could have been mainly or partly a response to the punishment inflicted by the suicide attacks, but it also could be a response to another type of pressure (such as an ongoing guerrilla campaign), or to pressure from a different actor (such as one of the target state's allies) or a different country, or the target's policy decision may not even have been intended as a concession but could have been made for other reasons that only coincidently moved in a direction desired by the terrorists. Different judgments among these alternatives yield different lessons for future usefulness of suicide attack. Standard principles from social psychology suggest how terrorists are likely to resolve these ambiguities. Under normal conditions, most people tend to interpret ambiguous information in ways that are consistent with their prior beliefs, as well as in ways that justify their past actions (Jervis 1976; Lebow 1981). Suicide terrorists, of course, are likely to have at least some initial confidence in the efficacy of suicide attack or else they would not resort to it, and of course, the fact of having carried out such attacks gives them an interest in justifying that choice. Thus, whenever targets of suicide terrorism make a real or apparent concession and it is a plausible interpretation that it was due to the coercive pressure of the suicide campaign, we would expect terrorists to favor that interpretation even if other interpretations are also plausible. This does not mean that we should simply expect terrorists to interpret virtually all outcomes, regardless of evidence, as encouraging further terrorism; that would not constitute learning and would make sense only if the terrorists were deeply irrational. To control for this possibility, it is crucial to consider the assessments of the same events by other well-informed people. If we find that when suicide terrorist leaders claim credit for coercing potential concessions, their claims are unique (or nearly so), then it would be appropriate to dismiss them as irrational. If, on the other hand, we find that their interpretations are shared by a significant portion of other observers, across a range of circumstances and interests—from target state leaders, to others in the terrorists' community, to neutral analysts—then we should assume that their assessments are as rational as anyone else's and should take the lessons they draw seriously. In making these judgments, the testimony of target state leaders is often especially telling; although states like the United States and Israel virtually never officially admit making concessions to terrorism, leaders such as Ronald Reagan and Yitzhak Rabin have at times been quite open about the impact of suicide terrorism on their own policy decisions, as we see below. Finally, understanding how terrorists' assess the effectiveness of suicide terrorism should also be influenced by our prior understanding of the fanatical nature of the specific terrorists at issue. If the most fanatical groups also make what appear to be reasonable assessments, then this would increase our confidence in the finding that most terrorists would make similar calculations. Hamas and Islamic Jihad are the most crucial case, because these groups have been considered to be fanatical extremists even among terrorists (Kramer 1996). Thus, detailed examination of how Hamas and Islamic Jihad leaders assessed the coercive value of suicide attacks during the 1990s is especially important. The Apparent Success of Suicide Terrorism Perhaps the most striking aspect of recent suicide terrorist campaigns is that they are associated with gains for the terrorists' political cause about half the time. As Table 1 shows, of the 11 suicide terrorist campaigns that were completed during 1980-2001, six closely correlate with significant policy changes by the target state toward the terrorists' major political goals. In one case, the terrorists' territorial goals were fully achieved (Hezbollah v. US/F, 1983); in three cases, the terrorists territorial aims were partly achieved (Hezbollah v. Israel, 1983-85; Hamas v. Israel, 1994; and Hamas v. Israel, 1994-95); in one case, the target government to entered into sovereignty negotiations with the terrorists (LTTE v. Sri Lanka, 1993-94); and in one case, the terrorist organization's top leader was released from prison (Hamas v. Israel, 1997). Five campaigns did not lead to noticeable concessions (Hezbollah's second effort against Israel in Lebanon, 1985-86; a Hamas campaign in 1996 retaliating for an Israeli assassination; the LTTE v. Sri Lanka, 1995-2002; and both PKK campaigns). Coercive success is so rare that even a 50% success rate is significant, because international military and economic coercion, using the same standards as above, generally works less than a third of the time (Art and Cronin 2003). There were limits to what suicide terrorism appeared to gain in the 1980s and 1990s. Most of the gains for the terrorists' cause were modest, not involving interests 9 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 central to the target countries' security or wealth, and most were potential revocable. For the United States and France, Lebanon was a relatively minor foreign policy interest. Israel's apparent concessions to the Palestinians from 1994 to 1997 were more modest than they might appear. Although Israel withdrew its forces from parts of Gaza and the West Bank and released Sheikh Yassin, during the same period Israeli settlement in the occupied territories almost doubled, and recent events have shown that the Israel is not deterred from sending force back in when necessary. In two disputes, the terrorists achieved initial success but failed to reach greater goals. Although Israel withdrew from much of Lebanon in June 1985, it retained a six-mile security buffer zone along the southern edge of the country for another 15 years from which a second Hezbollah suicide terrorist campaign failed to dislodge it. The Sri Lankan government did conduct apparently serious negotiations with the LTTE from November 1994 to April 1995, but did not concede the Tamil's main demand, for independence, and since 1995, the government has preferred to prosecute the war rather than consider permitting Tamil secession. Still, these six concessions, or at least apparent concessions, help to explain why suicide terrorism is on the rise. In three of the cases, the target government policy changes are clearly due to coercive pressure from the terrorist group. The American and French withdrawal was perhaps the most clear-cut coercive success for suicide terrorism. In his memoirs, President Ronald Reagan (1990,465) explained the U.S. decision to withdraw from Lebanon: The price we had to pay in Beirut was so great, the tragedy at the barracks was so enormous.... We had to pull out.... We couldn't stay there and run the risk of another suicide attack on the Marines. The IDF withdrawal from most of southern Lebanon in 1985 and the Sri Lankan government decision to hold negotiations with the LTTE were also widely understood to be a direct result of the coercive punishment imposed by Hezbollah and LTTE respectively. In both cases, the concessions followed periods in which the terrorists had turned more and more to suicide attacks, but since Hezbollah and the LTTE employed a combination of suicide attack and conventional attack on their opponents, one can question the relative weight of suicide attack in coercing these target states. However, there is little question in either case that punishment pressures inflicted by these terrorist organizations were decisive in the outcomes. For instance, as a candidate in the November 9, 1994, presidential election of Sri Lanka, Mrs. Chandrika Kumaratunga explicitly asked for a mandate to redraw boundaries so as to appease the Tamils in their demand for a separate homeland in the island's northeast provinces, often saying, "We definitely hope to begin discussions with the Tamil people, with their representatives—including the Tigers—and offer them political solutions to end the war... [involving] extensive devolution." This would, Kumaratunga said, "create an environment in which people could live without fear." (Sauvagnargues 1994; "Sri Lanka" 1994). The other three concessions, or arguable concessions, are less clear-cut. All three involve Hamas campaigns against Israel. Not counting the ongoing second intifada, Hamas waged four separate suicide attack campaigns against Israel, in 1994,1995,1996, and 1997. One, in 1996, did not correspond with Israeli concessions. This campaign was announced as retaliation for Israel's assassination of a Hamas leader; no particular coercive goal was announced, and it was suspended by Hamas after four attacks in two weeks. The other three all do correspond with Israeli concessions. In April 1994, Hamas begin a series of suicide bombings in relation for the Hebron Massacre. After two attacks, Israel decided to accelerate its withdrawal from Gaza, which was required under the Oslo Agreement but which had been delayed. Hamas then suspended attacks for five months. From October 1994 to August 1995, Hamas (and Islamic Jihad) carried out a total of seven suicide attacks against Israel. In September 1995, Israel agreed to withdraw from certain West Bank towns that December, which it earlier had claimed could not be done before April 1996 at the soonest. Hamas then suspended attacks until its retaliation campaign during the last week of February and first week of March 1996. Finally, in March 1997, Hamas began a suicide attack campaign that included an attack about every two months until September 1997. In response Israeli Prime Minister Netanyahu authorized the assassination of a Hamas leader. The attempt, in Amman, Jordan, failed and the Israeli agents were captured. To get them back Israel agreed to release Sheikh Ahmed Yassin, spiritual leader of Hamas. While this was not a concession to the terrorists' territorial goals, there is no evidence that Hamas interpreted this in anyway different from the standard view that this release was the product of American and Jordanian pressure. Accordingly the key Hamas campaigns that might have encouraged the view that suicide terrorism pays were the 1994 and 1995 campaigns that were associated with Israel's military withdrawals from Gaza and the West Banks. Terrorists' assessments of these events are evaluated in detail. The Crucial Case of Hamas The Hamas and Islamic Jihad suicide campaigns against Israel in 1994 and 1995 are crucial tests of the reasonableness of terrorists' assessments. In each case, Israel made significant concessions in the direction of the terrorists' cause and terrorist leaders report that these Israeli concessions increased their confidence in the coercive effectiveness of suicide attack. However, there is an important alternative explanation for Israel's concessions in these cases—the Israeli government's obligations under the Oslo Accords. Accordingly, evaluating the reasonableness of the terrorists' assessments of these cases is crucial because many observers characterize Hamas and Islamic Jihad as fanatical, irrational groups, extreme both within Palestinian society and among terrorists groups in general (Kramer 1996). Further, these campaigns are also of special interest because they helped to encourage the most intense 10 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review ongoing campaign, the second intifada against Israel, and may also have helped to encourage Al Qaeda's campaign against the United States. Examination of these crucial cases demonstrates that the terrorist groups came to the conclusion that suicide attack accelerated Israeli's withdrawal in both cases. Although the Oslo Accords formally committed to withdrawing the IDF from Gaza and the West Bank, Israel routinely missed key deadlines, often by many months, and the terrorists came to believe that Israel would not have withdrawn when it did, and perhaps not at all, had it not been for the coercive leverage of suicide attack. Moreover, this interpretation of events was hardly unique. Numerous other observers and key Israeli government leaders themselves came to the same conclusion. To be clear, Hamas may well have had motives other than coercion for launching particular attacks, such as retaliation (De Figueredo and Weingast 1998), gaining local support (Bloom 2002), or disrupting negotiated outcomes it considered insufficient (Kydd and Walter 2002). However, the experience of observing how the target reacted to the suicide campaigns appears to have convinced terrorist leaders of the coercive effectiveness of this strategy. To evaluate these cases, we need to know (1) the facts of each case, (2) how others interpreted the events, and (3) how the terrorists interpreted these events. Each campaign is discussed in turn. Israel's Withdrawal from Gaza, May 1994. The Facts. Israel and the Palestinian Liberation Organization signed the Oslo Accords on September 13, 1993. These obligated Israel to withdraw its military forces from the Gaza Strip and West Bank town of Jericho beginning on December 13 and ending on April 13,1994. In fact, Israel missed both deadlines. The major sticking points during the implementation negotiations in Fall and Winter of 1993-94 were the size of the Palestinian police force (Israel proposed a limit of 1,800, while the Palestinians demanded 9,000) and jurisdiction for certain criminal prosecutions, especially whether Israel could retain a right of hot pursuit to prosecute Palestinian attackers who might flee into Palestinian ruled zones. As of April 5,1994, these issues were unresolved. Hamas then launched two suicide attacks, one on April 6 and another on April 13, killing 15 Israeli civilians. On April 18, the Israeli Knesset voted to withdraw, effectively accepting the Palestinian positions on both disputed issues. The suicide attacks then stopped and the withdrawal was actually conducted in a few weeks starting on May 4,1994.4 These two suicide attacks may not originally have been intended as coercive, since Hamas leaders had announced them in March 1994 as part of a planned series of five attacks in retaliation for the February 24th Hebron massacre in which an Israeli settler killed 29 Palestinians and had strong reservations about negotiating a compromise settlement with Israel (Kydd and Walter 2002). However, when Israel agreed to withdraw more promptly than expected, Hamas decided to 4 There were no suicide attacks from April to October 1994. Vol. 97, No. 3 forgo the remaining three planned attacks. There is thus a circumstantial case that these attacks had the effect of coercing the Israelis into being more forthcoming in the withdrawal negotiations and both Israeli government leaders and Hamas leaders publically drew this conclusion. Israeli and Other Assessments. There are two main reasons to doubt that terrorist pressure accelerated Israel's decision to withdraw. First, one might think that Israel would have withdrawn in any case, as it had promised to do in the Oslo Accords of September 1993. Second, one might argue that Hamas was opposed to a negotiated settlement with Israel. Taking both points together, therefore, Hamas' attacks could not have contributed to Israel's withdrawal. The first of these arguments, however, ignores the facts that Israel had already missed the originally agreed deadline and, as of early April 1994, did not appear ready to withdraw at all if that meant surrendering on the size of the Palestinian police force and legal jurisdiction over terrorists. The second argument is simply illogical. Although Hamas objected to surrendering claims to all of historic Palestine, it did value the West Bank and Gaza as an intermediate goal, and certainly had no objection to obtaining this goal sooner rather than later. Most important, other observers took explanations based on terrorist pressure far more seriously, including the person whose testimony must count most, Israeli Prime Minister Yitzhak Rabin. On April 13,1994, Rabin said, I can't recall in the past any suicidal terror acts by the PLO. We have seen by now at least six acts of this type by Hamas and Islamic Jihad.... The only response to them and to the enemies of peace on the part of Israel is to accelerate the negotiations. (Makovsky and Pinkas 1994). On April 18,1994, Rabin went further, giving a major speech in the Knesset explaining why the withdrawal was necessary: Members of the Knessett: I want to tell the truth. For 27 years we have been dominating another people against its will. For 27 years Palestinians in the territories ... get up in the morning harboring a fierce hatred for us, as Israelis and Jews. Each morning they get up to a hard life, for which we are also, but not solely responsible. We cannot deny that our continuing control over a foreign people who do not want us exacts a painful price.... For two or three years we have been facing a phenomenon of extremist Islamic terrorism, which recalls Hezbollah, which surfaced in Lebanon and perpetrated attacks, including suicide missions.... There is no end to the targets Hamas and other terrorist organizations have among us. Each Israeli, in the territories and inside sovereign Israel, including united Jerusalem, each bus, each home, is a target for their murderous plans. Since there is no separation between the two populations, the current situation creates endless possibilities for Hamas and the other organizations. Independent Israeli observers also credited suicide terrorism with considerable coercive effectiveness. The most detailed assessment is by Efraim Inbar (1999,141- 42): 11 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 A significant change occurred in Rabin's assessment of the importance of terrorist activities.... Reacting to the April 1994 suicide attack in Afula, Rabin recognized that terrorists activities by Hamas and other Islamic radicals were "a form of terrorism different from what we once knew from the PLO terrorist organizations.... "Rabin admitted that there was no "hermitic" solution available to protect Israeli citizens against such terrorist attacks.... He also understood that such incidents intensified the domestic pressure to freeze the Palestinian track of the peace process. Islamic terrorism thus initially contributed to the pressure for accelerating the negotiations on his part. Arab writers also attributed Israeli accommodation to the suicide attacks. Mazin Hammad wrote in an editorial in a Jordanian newspaper: It is unprecedented for an Israeli official like Y. Rabin to clearly state that there is no future for the settlements in the occupied territories.... He would not have said this [yesterday] if it was not for the collapse of the security Israel.... The martyrdom operation in Hadera shook the faith of the settlers in the possibility of staying in the West Bank and Gaza and increased their motivation to pack their belongings and dismantle their settlements. ("Hamas Operations" 1994) Terrorists' Assessments. Even though the favorable result was apparently unexpected by Hamas leaders, given the circumstances and the assessments voiced by Rabin and others, it certainly would have been reasonable for them to conclude that suicide terrorism had helped accelerate Israeli withdrawal, and they did. Hamas leader Ahmed Bakr (1995) said that "what forced the Israelis to withdraw from Gaza was the intifada and not the Oslo agreement," while Imad al-Faluji judged that all that has been achieved so far is the consequence of our military actions. Without the so-called peace process, we would have gotten even more.... We would have got Gaza and the West Bank without this agreement.... Israel can beat all Arab Armies. However, it can do nothing against a youth with a knife or an explosive charge on his body. Since it was unable to guarantee security within its borders, Israel entered into negotiations with the PLO.... If the Israelis want security, they will have to abandon their settlements ... in Gaza, the West Bank, and Jerusalem. ("Hamas Leader" 1995) Further, these events appear to have persuaded terrorists that future suicide attacks could eventually produce still greater concessions. Fathi al-Shaqaqi (1995), leader of Islamic Jihad, said, Our jihad action has exposed the enemy weakness, confusion, and hysteria. It has become clear that the enemy can be defeated, for if a small faithful group was able to instill all this horror and panic in the enemy through confronting it in Palestine and southern Lebanon, what will happen when the nation confronts it with all its potential.... Martyrdom actions will escalate in the face of all pressures ... [they] are a realistic option in confronting the unequal balance of power. If we are unable to effect a balance of power now, we can achieve a balance of horror. Israel's Withdrawal from West Bank Towns, December 1995. Tlie second Hamas case, in 1995, tells es- sentially the same story as the first. Again, a series of suicide attacks was associated with Israeli territorial concessions to the Palestinians, and again, a significant fraction of outside observers attributed the concessions to the coercive pressure of suicide terrorism, as did the terrorist leaders themselves. The Facts. Tlie original Oslo Accords scheduled Israel to withdraw from the Palestinian populated areas of the West Bank by July 13, 1994, but after the delays over Gaza and Jericho all sides recognized that this could not be met. From October 1994 to April 1995, Hamas, along with Islamic Jihad, carried out a series of seven suicide terrorist attacks that were intended to compel Israel to make further withdrawals and suspended attacks temporarily at the request of the Palestinian Authority after Israel agreed on March 29, 1995 to begin withdrawals by July 1. Later, however, the Israelis announced that withdrawals could not begin before April 1996 because bypass roads needed for the security of Israeli settlements were not ready. Hamas and Islamic Jihad then mounted new suicide attacks on July 24 and August 21,1995, killing 11 Israeli civilians. In September, Israel agreed to withdraw from the West Bank towns in December (Oslo II) even though the roads were not finished. The suicide attacks then stopped and the withdrawal was actually carried out in a few weeks starting on December 12,1995.5 Israeli and Other Assessments. Although Israeli government spokesmen frequently claimed that suicide terrorism was delaying withdrawal, this claim was contradicted by, among others, Prime Minister Rabin. Rabin (1995) explained that the decision for the second withdrawal was, like the first in 1994, motivated in part by the goal of reducing suicide terrorism: Interviewer: Mr Rabin, what is the logic of withdrawing from towns and villages when you know that terror might continue to strike at us from there? Rabin: What is the alternative, to have double the amount of terror? As for the issue of terror, take the suicide bombings. Some 119 Israelis ... have been killed or murdered since 1st January 1994, 77 of them in suicide bombings perpetrated by Islamic radical fanatics.... All the bombers were Palestinians who came from areas under our control. Similarly, an editorial in the Israeli daily Yediot Aharonot ("Bus Attack" 1995) explained, If the planners of yesterday's attack intended to get Israel to back away from the Oslo accord, they apparently failed. In fact, Prime Minister Y. Rabin is leaning toward expediting the talks with the Palestinians.... The immediate conclusion from this line of thinking on Rabin's part— whose results we will witness in the coming days—will be to instruct the negotiators to expedite the talks with the Palestinians with the aim of completing them in the very near future. Terrorists' Assessments. As in 1994, Hamas and Islamic Jihad came to the conclusion that suicide 5 There were no suicide attacks from August 1995 to February 1996. There were four suicide attacks in response to an Israeli assassination from February 25 to March 4,1996, and then none until March 1997. 12 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 terrorism was working. Hamas's spokesman in Jordan explained that new attacks were necessary to change Israel's behavior: Hamas, leader Muhammad Nazzal said, needed military muscle in order to negotiate with Israel from a position of strength. Arafat started from a position of weakness, he said, which is how the Israelis managed to push on him the solution and get recognition of their state and settlements without getting anything in return. (Theodoulou 1995) After the agreement was signed, Hamas leaders also argued that suicide operations contributed to the Israeli withdrawal. Mahmud al-Zahhar (1996), a spokesman for Hamas, said, The Authority told us that military action embarrasses the PA because it obstructs the redeployment of the Israeli's forces and implementation of the agreement.... We offered many martyrs to attain freedom.... Any fair person knows that the military action was useful for the Authority during negotiations. Moreover, the terrorists also stressed that stopping the attacks only discouraged Israel from withdrawing. An early August Hamas communique (No. 125,1995) read, They said that the stragglers' operations have been the cause of the delay in widening the autonomous rule in the West Bank, and that they have been the reason for the deterioration of the living and economic conditions of our people. Now the days have come to debunk their false claims ... and to affirm that July 1 [a promised date for IDF withdrawal] was no more than yet another of the "unholy" Zionist dates.... Hamas has shown an utmost degree of self-restraint throughout the past period.... but matters have gone far enough and the criminals will reap what their hands have sown. Recent Impact of Lessons Learned. In addition to the 1994 and 1995 campaigns, Palestinian terrorist leaders have also cited Hezbollah experience in Lebanon as a source of the lesson that suicide terrorism is an effective way of coercing Israel. Islamic Jihad leader Ramadan Shallah (2001) argued that: The shameful defeat that Israel suffered in southern Lebanon and which caused its army to flee it in terror was not made on the negotiations table but on the battlefield and through jihad and martyrdom, which achieved a great victory for the Islamic resistance and Lebanese People.... We would not exaggerate if we said that the chances of achieving victory in Palestine are greater than in Lebanon.... If the enemy could not bear the losses of the war on the border strip with Lebanon, will it be able to withstand a long war of attrition in the heart of its security dimension and major cities? Palestinian terrorists are now applying the lessons they have learned. In November 2000, Khalid Mish'al explained Hamas's strategy for the second intifada, which was then in its early stages: Like the intifada in 1987, the current intifada has taught us that we should move forward normally from popular confrontation to the rifle to suicide operations. This is the normal development.... We always have the Lebanese ex- periment before our eyes. It was a great model of which we are proud. Even before the second intifada began, other Hamas statements similarly expressed, The Zionist enemy ... only understands the language of Jihad, resistance and martyrdom, that was the language that led to its blatant defeat in South Lebanon and it will be the language that will defeat it on the land of Palestine. (Hamas Statement 2000) The bottom line is that the ferocious escalation of the pace of suicide terrorism that we have witnessed in the past several years cannot be considered irrational or even surprising. Rather, it is simply the result of the lesson that terrorists have quite reasonably learned from their experience of the previous two decades: Suicide terrorism pays. THE LIMITS OF SUICIDE TERRORISM Despite suicide terrorists' reasons for confidence in the coercive effectiveness of this strategy, there are sharp limits to what suicide terrorism is likely to accomplish in the future. During the 1980s and 1990s, terrorist leaders learned that moderate punishment often leads to moderate concessions and so concluded that more ambitious suicide campaigns would lead to greater political gains. However, today's more ambitious suicide terrorist campaigns are likely to fail. Although suicide terrorism is somewhat more effective than ordinary coercive punishment using air power or economic sanctions, it is not drastically so. Suicide Terrorism Is Unlikely to Achieve Ambitious Goals In international military coercion, threats to inflict military defeat often generate more coercive leverage than punishment. Punishment, using anything short of nuclear weapons, is a relatively weak coercive strategy because modern nation states generally will accept high costs rather than abandon important national goals, while modern administrative techniques and economic adjustments over time often allow states to minimize civilian costs. The most punishing air attacks with conventional munitions in history were the American B-29 raids against Japan's 62 largest cities from March to August 1945. Although these raids killed nearly 800,000 Japanese civilians—almost 10% died on the first day, the March 9, 1945, fire-bombing of Tokyo, which killed over 85,000—the conventional bombing did not compel the Japanese to surrender. Suicide terrorism makes adjustment to reduce damage more difficult than for states faced with military coercion or economic sanctions. However, it does not affect the target state's interests in the issues at stake. As a result, suicide terrorism can coerce states to abandon limited or modest goals, such as withdrawal from territory of low strategic importance or, as in Israel's case in 1994 and 1995, a temporary and partial withdrawal from a more important area. However, suicide 13 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 Strategic Logic of Suicide Terrorism terrorism is unlikely to cause targets to abandon goals central to their wealth or security, such as a loss of territory that would weaken the economic prospects of the state or strengthen the rivals of the state. Suicide terrorism makes punishment more effective than in international military coercion. Targets remain willing to countenance high costs for important goals, but administrative, economic, or military adjustments to prevent suicide attack are harder, while suicide attackers themselves are unlikely to be deterred by the threat of retaliation. Accordingly, suicide attack is likely to present a threat of continuing limited civilian punishment that the target government cannot completely eliminate, and the upper bound on what punishment can gain for coercers is recognizably higher in suicidal terrorism than in international military coercion. The data on suicide terrorism from 1980 to 2001 support this conclusion. While suicide terrorism has achieved modest or very limited goals, it has so far failed to compel target democracies to abandon goals central to national wealth or security. When the United States withdrew from Lebanon in 1984, it had no important security, economic, or even ideological interests at stake. Lebanon was largely a humanitarian mission and not viewed as central to the national welfare of the United States. Israel withdrew from most of Lebanon in June 1985 but remained in a security buffer on the edge of southern Lebanon for more than a decade afterward, despite the fact that 17 of 22 suicide attacks occurred in 1985 and 1986. Israel's withdrawals from Gaza and the West Bank in 1994 and 1995 occurred at the same time that settlements increased and did little to hinder the IDF's return, and so these concessions were more modest than they may appear. Sri Lanka has suffered more casualties from suicide attack than Israel but has not acceded to demands that it surrender part of its national territory. Thus, the logic of punishment and the record of suicide terrorism suggests that, unless suicide terrorists acquire far more destructive technologies, suicide attacks for more ambitious goals are likely to fail and will continue to provoke more aggressive military responses. Policy Implications for Containing Suicide Terrorism While the rise in suicide terrorism and the reasons behind it seem daunting, there are important policy lessons to learn. The current policy debate is misguided. Offensive military action or concessions alone rarely work for long. For over 20 years, the governments of Israel and other states targeted by suicide terrorism have engaged in extensive military efforts to kill, isolate, and jail suicide terrorist leaders and operatives, sometimes with the help of quite good surveillance of the terrorists' communities. Thus far, they have met with meager success. Although decapitation of suicide terrorist organizations can disrupt their operations temporarily, it rarely yields long-term gains. Of the 11 major suicide terrorist campaigns that had ended as of 2001, only one—the PKK versus Turkey—did so as a result of leadership decapitation, when the leader, 14 20:32 August 2003 in Turkish custody, asked his followers to stop. So far, leadership decapitation has also not ended Al Qaeda's campaign. Although the United States successfully toppled the Taliban in Afghanistan in December 2001, Al Qaeda launched seven successful suicide terrorist attacks from April to December 2002, killing some 250 Western civilians, more than in the three years before September 11,2001, combined. Concessions are also not a simple answer. Concessions to nationalist grievances that are widely held in the terrorists' community can reduce popular support for further terrorism, making it more difficult to recruit new suicide attackers and improving the standing of more moderate nationalist elites who are in competition with the terrorists. Such benefits can be realized, however, only if the concessions really do substantially satisfy the nationalist or self-determination aspirations of a large fraction of the community. Partial, incremental, or deliberately staggered concessions that are dragged out over a substantial period of time are likely to become the worst of both worlds. Incremental compromise may appear—or easily be portrayed—to the terrorists' community as simply delaying tactics and, thus, may fail to reduce, or actually increase, their distrust that their main concerns will ever be met. Further, incrementalism provides time and opportunity for the terrorists to intentionally provoke the target state in hopes of derailing the smooth progress of negotiated compromise in the short term, so that they can reradicalize their own community and actually escalate their efforts toward even greater gains in the long term.6 Thus, states that are willing to make concessions should do so in a single step if at all possible. Advocates of concessions should also recognize that, even if they are successful in undermining the terrorist leaders' base of support, almost any concession at all will tend to encourage the terrorist leaders further about their own coercive effectiveness. Thus, even in the aftermath of a real settlement with the opposing community, some terrorists will remain motivated to continue attacks and, for the medium term, may be able to do so, which in term would put a premium on combining concessions with other solutions. Given the limits of offense and of concessions, homeland security and defensive efforts generally must be a core part of any solution. Undermining the feasibility of suicide terrorism is a difficult task. After all, a major advantage of suicide attack is that it is more difficult to prevent than other types of attack. However, the difficulty of achieving perfect security should not keep us from taking serious measures to prevent would-be terrorists from easily entering their target society. As Chaim Kaufmann (1996) has shown, even intense 6 The Bush administration's decision in May 2003 to withdraw most U.S. troops from Saudi Arabia is the kind of partial concession likely to backfire. Al Qaeda may well view this as evidence that the United States is vulnerable to coercive pressure, but the concession does not satisfy Al Qaeda's core demand to reduce American military control over the holy areas on the Arab peninsula. With the conquest and long term military occupation of Iraq, American military capabilities to control Saudi Arabia have substantially increased even if there are no American troops on Saudi soil itself. PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 ethnic civil wars can often be stopped by demographic separation because it greatly reduces both means and incentives for the sides to attack each other. This logic may apply with even more force to the related problem of suicide terrorism, since, for suicide attackers, gaining physical access to the general area of the target is the only genuinely demanding part of an operation, and as we have seen, resentment of foreign occupation of their national homeland is a key part of the motive for suicide terrorism. The requirements for demographic separation depend on geographic and other circumstances that may not be attainable in all cases. For example, much of Israel's difficulty in containing suicide terrorism derives from the deeply intermixed settlement patterns of the West Bank and Gaza, which make the effective length of the border between Palestinian and Jewish settled areas practically infinite and have rendered even very intensive Israeli border control efforts ineffective (Kaufmann 1998). As a result, territorial concessions could well encourage terrorists leaders to strive for still greater gains while greater repression may only exacerbate the conditions of occupation that cultivate more recruits for terrorist organizations. Instead, the best course to improve Israel's security may well be a combined strategy: abandoning territory on the West Bank along with an actual wall that physically separates the populations. Similarly, if Al Qaeda proves able to continue suicide attacks against the American homeland, the United States should emphasize improving its domestic security. In the short term, the United States should adopt stronger border controls to make it more difficult for suicide attackers to enter the United States. In the long term, the United States should work toward energy independence and, thus, reduce the need for American troops in the Persian Gulf countries where their presence has helped recruit suicide terrorists to attack America. These measures will not provide a perfect solution, but they may make it far more difficult for Al Qaeda to continue attacks in the United States, especially spectacular attacks that require elaborate coordination. Perhaps most important, the close association between foreign military occupations and the growth of suicide terrorist movements in the occupied regions should give pause to those who favor solutions that involve conquering countries in order to transform their political systems. Conquering countries may disrupt terrorist operations in the short term, but it is important to recognize that occupation of more countries may well increase the number of terrorists coming at us. Appendix: Suicide Terrorist Campaigns 1980-2001 Weapon Target Killed* Completed Campaigns Campaign #1: Hezbollah vs. US, France I.April 18, 1983 car bomb US embassy, Beirut 63 2. Oct 23, 1983 car bomb US Marine barracks 241 3. Oct 23, 1983 car bomb French barracks 58 4. Dec 12, 1983 grenades US Embassy, Kuwait 7 5. Dec 21, 1983 car bomb French HQ, Beirut 1 6. Sept?, 1984 truck bomb US Beirut embassy 14 Campaign #2: Hezbollah vs. Israel 1. Nov 4, 1983 car bomb IDF post in Tyre, Lebanon 50 2. Jun 16, 1984 car bomb IDF post, southern Lebanon 5 3. Mar 8, 1985: truck bomb IDF post 12 4. Apr 9, 1985: car bomb IDF post 4 5. May 9, 1985: suitcase bomb SLA checkpoint 2 6. June 15, 1985: car bomb IDF post in Beirut 23 Campaign #3: Hezbollah vs. Israel and South Lebanon Army I.July 9, 1985 car bombs 2 SLA outposts 22 2. July 15, 1985 car bomb SLA outpost 10 3. July 31, 1985 car bomb IDF patrol, south Lebanon 2 4. Aug 6, 1985 mule bomb SLA outpost 0 5. Aug 29, 1985 car bomb SLA outpost 15 6. Sept 3, 1985 car bomb SLA outpost 37 7. Sept 12, 1985 car bomb SLA outpost 21 8. Sept 17, 1985 car bomb SLA outpost 30 9. Sept 18, 1985 car bomb SLA outpost 0 10. Oct 17, 1985 grenades SLA radio station 6 11. Nov 4, 1985 car bomb SLA outpost 0 12. Nov 12, 1985 car bomb Christ, militia leaders, Beirut 5 13. Nov 26, 1985 car bomb SLA outpost 20 14. April 7, 1986 car bomb SLA outpost 1 15. July 17, 1986 car bomb Jezzine, S. Lebanon 7 16. Nov 20, 1986: car bomb SLA outpost 3 15 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 Appendix—Continued Weapon Target Killed* Campaign #4: Liberation Tigers of Tamil Eelam vs. Sri Lanka 1. Jul 12, 1990 boat bomb naval vessel, Trincomalee 6 2. Nov 23, 1990 ??? army camp, Manakulam 0 3. Mar 2, 1991 car bomb defense minister, Colombo 18" 4. Mar 19, 1991 truck bomb army camp, Silavathurai 5 5. May 5, 1991 boat bomb naval vessel 5 6. May 21, 1991 belt bomb Rajiv Gandhi, Madras, India 1** 7. June 22, 1991 car bomb defense ministry, Colombo 27 8. Nov 16, 1992 motorcycle bomb navy commander, Colombo 1** 9. May 1, 1993 belt bomb President Premadasa, Colombo 23« 10. Nov 11, 1993 boat bomb naval base, Jaffna Lagoon 0 11. Aug 2, 1994 grenades air force helicopter, Palali 0 12. Sept 19, 1994 mines naval vessel, Sagarawardene 25 13. Oct 24, 1994 belt bomb Presidential candidate, Colombo 53" 14. Nov 8, 1994 mines naval vessel, Vettilaikerny 0 Campaign #5: LTTE vs. Sri La ika 1. Apr 18, 1995 scuba divers naval vessel, Trincomalee 11 2. Jul 16, 1995 scuba divers naval vessel, Jaffna peninsula 0 3. Aug 7, 1995 belt bomb government bldg, Colombo 22 4. Sep 3, 1995 scuba divers naval vessel, Trincomalee 0 5. Sep 10, 1995 scuba divers naval vessel, Kankesanthurai 0 6. Sep 20, 1995 scuba divers naval vessel, Kankesanthurai 0 7. Oct 2, 1995 scuba divers Naval vessel, Kankesanthurai 0 8. Oct 17, 1995 scuba divers naval vessel, Trincomalee 9 9. Oct 20, 1995 mines 2 oil depots, Colombo 23 10. Nov 11, 1995 belt bombs army HQ, crowd, Colombo 23 11. Dec 5, 1995 truck bomb police camp, Batticaloa 23 12. Jan 8, 1996 belt bomb market, Batticaloa 0 13. Jan 31, 1996 truck bomb bank, Colombo 91 14. Apr 1, 1996 boat bomb navy vessel, Vettilaikerni 10 15. Apr 12, 1996 scuba divers port building, Colombo 0 16. Jul 3, 1996 belt bomb gov motorcade, Jaffna 37 17. Jul 18,1996 mines naval gunboat, Mullaittivu 35 18. Aug 6, 1996 boat bomb naval ship, north coast 0 19. Aug 14, 1996 bicycle bomb public rally, Kalmunai 0 20. Oct 25, 1996 boat bomb gunboat, Trincomalee 12 21. Nov 25, 1996 belt bomb police chief vehicle, Trincomalee 0 22. Dec 17, 1996 motorcycle bomb police unit jeep, Ampara 1 23. Mar 6, 1997 grenades China Bay air base 0 24. Oct 15, 1997 truck bomb World Trade Centre, Colombo 18 25. Oct 19, 1997 boat bomb naval gunboat, northeastern coast 7 26. Dec 28, 1997 truck bomb chief Cecil Tissera, south Sri Lanka 0 27. Jan 25, 1998 truck bomb Buddhist shrine, Kandy 11 28. Feb 5, 1998 belt bomb Air Force headquarters, Colombo 8 29. Feb 23, 1998 boat bombs 2 landing ships off Point Pedru 47 30. Mar 5, 1998 bus bomb train station, Colombo 38 31. May 15, 1998 belt bomb army brigadier, Jaffna peninsula 1 32. Sep 11,1998 belt bomb new mayor of Jaffna 20" 33. Mar 15, 1999 belt bomb police station, Colombo 5 34. May 29, 1999 belt bomb Tamil rival leader, Batticaloa 2 35. Jul 25, 1999 belt bomb passenger ferry, Trincomalee 1 36. Jul 29, 1999 belt bomb Tamil politician, Colombo 1 37. Aug 4, 1999 bicycle bomb police vehicle, Vavuniya 12 38. Aug 9, 1999 belt bomb military commander, Vakarai 1 39. Sep 2, 1999 belt bomb Tamil rival, Vavuniya, 3 40. Dec 18, 1999 2 belt bombs Pres C. Kumaratunga, Colombo 38*" 41. Jan 5, 2000 belt bomb PM S. Bandaranaike, Colombo 11*** 42. Feb 4, 2000 sea diver naval vessel, Trincomalee 0 43. Mar 2, 2000 belt bomb military commander, Trincomalee 1*** 44. Mar 10, 2000 belt bomb government motorcade Colombo 23 45. Jun 5, 2000 scuba diver ammunition ship, northeast coast 5 46. Jun 7, 2000 belt bomb Industries Minister Colombo 26** 47. Jun 14,2000 bicycle bomb air force bus, Wattala town 2 16 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 Appendix—Continued Group Weapon Target Killed 48. Jun 26, 2000 49. Aug 16,2000 50. Sep 15,2000 51. Oct 2, 2000 52. Oct 5, 2000 53. Oct 19, 2000 54. Oct 23, 2000 boat bomb belt bomb belt bomb belt bomb belt bomb belt bomb boat bombs merchant vessel, north coast military vehicle, Vavuniya hospital, Colombo SLMC leader, Trincomalee People's Alliance rally, Medawachchiya Cabinet ceremony, Colombo gunboat/troop carrier, Trincomalee 7 1 7 22« 12 0 2 Campaign #6: Hamas vs. Israel 1. Apr 6, 1994 2. Apr 13, 1994 Hamas Hamas car bomb belt bomb Afula, Israel Hadera, Israel 9 6 Campaign #7: Hamas/lslamic Jihad vs. Israel 1. Oct 19, 1994 2. Nov 11, 1994 3. Dec 25, 1994 4. Jan 22, 1995 5. Apr 9, 1995 6. July 24, 1995 7. Aug 21, 1995 Hamas Islamic Jihad Hamas Islamic Jihad Islamic Jihad Hamas Hamas belt bomb belt bomb belt bomb belt bomb belt bomb Tel Aviv Netzarim, Gaza Jerusalem Tel Aviv Netzarim, Gaza Tel Aviv Jerusalem 22 3 0 21 8 6 5 Campaign #8: Hamas vs. Israel 1. Feb 25, 1996 2. Feb 25, 1996 3. Mar 3, 1996 4. Mar 4, 1996 Hamas Hamas Hamas Hamas belt bomb belt bomb belt bomb belt bomb Jerusalem Ashkelon Jerusalem Tel Aviv 25 1 19 13 Campaign #9: Hamas vs. Israel 1. Mar 21, 1997 2. Jul 30, 1997 3. Sept 7, 1997 Hamas Hamas Hamas belt bomb belt bomb belt bomb café, Tel Aviv Jerusalem Jerusalem 3 14 7 Campaign #10: Kurdistan Workers Party (PKK) vs. Turkey 1. Jun 30, 1996 2. Oct 25, 1996 3. Oct 29, 1996 belt bomb belt bomb belt bombs Tunceli, Turkey Adana, Turkey Sivas, Turkey 9 4 4 Campaign #11: PKK vs. Turkey 1. Mar 4, 1999 2. Mar 27, 1999 3. Apr 5, 1999 4. Jul 5, 1999 5. Jul 7, 1999 6. Aug 28, 1999 belt bomb grenade belt bomb belt bomb grenades bomb Batman, Turkey Istanbul governor, Bingol, Turkey Adana, Turkey Huh, Turkey Tunceli, Turkey 0 0 0 0 0 0 Ongoing Compaigns Campaign #12: Al Qaeda vs. United States 1. Nov 13, 1995 2. Jun 25, 1996 3. Aug 7, 1998 4. Oct 12, 2000 5. Sep 9, 2001 6. Sep 11,2001 car bomb truck bomb truck bombs boat bomb camera bomb hijacked airplanes US military base, Riyadh, SA US military base, Dhahran SA US embassies, Kenya/Tanzania USS Cole, Yemen Ahmed Shah Masood, Afghanistan WTC/Pentagon 5 19 250 17 1 3037 Campaign #13: Chechen Separatists vs . Russia 1. Jun 7, 2000 2. Jul 3, 2000 3. Mar 24, 2001 4. Nov 29, 2001 truck bomb truck bomb car bomb belt bomb Russian police, Chech. Argun, Russia Chechnya military commander, Chechnya 2 30 20 1 Campaign #14: Kashr mir Separatists vs. India 1. Dec 25, 2000 2. Oct 1, 2001 3. Dec 13,2001 car bomb car bomb gunmen Srinagar, Kashmir Legislative assembly Parliament, New Delhi 8 30 7 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 Strategic Logic of Suicide Terrorism August 2003 Appendix—Continued Group Weapon Target Killed Campaign #15: LTTE vs Sri Lanka 1. Jul 24, 2001 belt bomb international airport, Colombo 12 2. Sep 16,2001 boat bomb navel vessel, north 29 3. Oct 29, 2001 belt bomb PM Wickremanayake, Colombo 3*** 4. Oct 30, 2001 boat bomb oil tanker, northern coast 4 5. Nov 9, 2001 belt bomb police jeep, Batticaloa 0 6. Nov 15, 2001 belt bomb crowd, Batticaloa 3 Compaign #16: Hamas/I slamic Jihad vs. Israel 1. Oct 30, 2000 Hamas belt bomb Jerusalem 15 2. Nov 2, 2000 Al Aqsa 2 3. Nov 20, 2000 Hamas car bomb Hadera, Israel 2 4. Dec 22, 2000 Al Aqsa belt bomb West Bank 3 5. Jan 1,2001 Hamas belt bomb Netanya, Israel 10 6. Feb 8, 2001 Hamas 4 7. Feb 14,2001 Hamas bus driver Tel Aviv 8 8. Mar 1,2001 Hamas 1 9. Mar 4, 2001 Hamas belt bomb Netanya, Israel 3 10. Mar 27, 2001 Hamas belt bomb Jerusalem 1 11. Mar 27, 2001 Hamas belt bomb Jerusalem (2nd attack) 0 12. Mar 28, 2001 Hamas belt bomb Kfar Saba, Israel 3 13. Apr 22, 2001 Hamas belt bomb Kfar Saba, Israel 3 14. Apr 23, 2001 PFLP car bomb Yehuda, Israel 8 15. Apr 29, 2001 Hamas belt bomb West Bank 0 16. May 18,2001 Hamas belt bomb Netanya, Israel 5 17. May 25, 2001 Islamic Jihad truck bomb Netzarim, Gaza 2 18. May 27, 2001 Hamas car bomb Netanya, Israel 1 19. May 30, 2001 Islamic Jihad 8 20. Jun 1,2001 Hamas belt bomb nightclub, Tel Aviv 22 21. Jun 22, 2001 Hamas belt bomb Gaza 2 22. Jul 2, 2001 Hamas car bomb IDF checkpt, Gaza 0 23. Jul 9, 2001 Hamas car bomb Gaza 0 24. Jul 16,2001 Islamic Jihad belt bomb Jerusalem 5 25. Aug 8, 2001 Al Aqsa car bomb Jerusalem 8 26. Aug 9, 2001 Islamic Jihad belt bomb Haifa, Israel 15 27. Aug 12,2001 Islamic Jihad belt bomb Haifa, Israel 0 28. Aug 21, 2001 Al Aqsa Jerusalem 0 29. Sept 4, 2001 Hamas belt bomb Jerusalem 0 30. Sept 9, 2001 Hamas belt bomb Nahariya, Israel 3 31. Oct 1,2001 ??? 1 32. Oct 7, 2001 Islamic Jihad car bomb Israel 2 33. Nov 26, 2001 Hamas car bomb Gaza 0 34. Nov 29, 2001 Islamic Jihad Gaza 3 35. Dec 1,2001 Hamas belt bomb Haifa, Israel 11 36. Dec 2, 2001 Hamas Jerusalem 15 37. Dec 5, 2001 Islamic Jihad belt bomb Jerusalem 0 38. Dec 9, 2001 ??? belt bomb Haifa, Israel 0 39. Dec 12,2001 ??? Isolated Attacks Gaza 0 1. Dec 15, 1981 ??? Car bomb Iraqi embassy, Beirut 30 2. May 25, 1985 Hezbollah car bomb ruler, Kuwait 0*** 3. Jul 5, 1987 LTTE truck bomb army camp, Jaffna Penin 20 4. Aug 15, 1993 ??? motorcycle bomb Egyptian Interior Minister 3 5. Jan 30, 1995 AIG™ truck bomb crowd, Algiers 42 6. Nov 19, 1995 Islamic Group truck bomb Egyptian embassy, Pakistan 16 7. Oct 29, 1998 Hamas belt bomb Gaza 1 8. Nov 17, 1998 ??? belt bomb Yuksekova, Turkey 0 9. Dec 29, 1999 Hezbollah grenades Lebanon 1 Note: Several reports of PKK suicide in May and June 1997 during fighting between PKK and Kurdish militias in Iraq, but coverage insufficient to distinguish suicide attack from suicide to avoid capture. * Not including attacker. ** Assassination target killed. *** Assassination target survived. ??? = unclaimed. **** Armed Isalamic Group. 18 PI: HAU CJ206E-01 American Political Science Review APSR.cls July 14,2003 20:32 American Political Science Review Vol. 97, No. 3 REFERENCES al-Shaqaqi, Fathi. 1995. "Interview with Secretary General of Islamic Jihad." Al-Quds, 11 April. FBIS-NES-95-70,12 April 1995. al-Zahhar, Mahmud. 1996. "Interview." Al-Dustur (Amman), 19 February. FBIS-NES-96-034, 20 February 1996. Art, Robert X, and Patrick M. Cronin. 2003. The United States and Coercive Diplomacy. Washington, DC: United States Institute of Peace. Bakr, Ahmed. 1995. "Interview." The Independent (London), 14 March. FBIS-NES-95-086, 4 May 1995. Bearden, Milton. 2002. Personal correspondence. University of Chicago, March 26. Bloom, Mia. 2002. "Rational Interpretations of Palestinian Suicide Bombing." Paper presented at the Program on International Security Policy, University of Chicago. Boix, Carlos, and Sebastian Rosato. 2001. "A Complete Dataset of Regimes, 1850-1999." University of Chicago. Typescript. "Bus Attack Said to Spur Rabin to Speed Talks." 1995. Yediot Aharonot, July 25. FBIS-NES-94-142, 25 July 1995. Clutterbuck, Richard. 1975. Living with Terrorism. London: Faber & Faber. Crenshaw, Martha. 1981. "The Causes of Terrorism." Comparative Politics 13 (July): 397-99. De Figueiredo, Rui, and Barry R. Weingast. 1998. "Vicious Cycles: Endogenous Political Extremism and Political Violence." Paper presented at the annual meeting of the American Political Science Association. Department of State. 1983-2001. Patterns of Global Terrorism. Washington, DC: DOS. Edler Baumann, Carol. 1973. Diplomatic Kidnapings: A Revolutionary Tactic of Urban Terrorism. The Hague: Nijhoff. Elliott, Paul. 1998. Brotherhoods of Fear. London: Blandford. George, Alexander, et al. 1972. Limits of Coercive Diplomacy. Boston: Little, Brown. Greenberg, Joel. 2002. "Suicide Planner Expresses Joy Over His Missions," New York Times, 9 May. Hamas Communique No. 125.1995. Filastin al-Muslimah (London), August. FBIS-NES-95-152, 8 August 1995. "Hamas Leader Discusses Goals." 1995. Frankfurter Runschau, 3 May. FBIS-NES-95-086, 4 May 1995. "Hamas Operations Against Israel Said to Continue." 1994. Al-Dustur (Amman, Jordan), 14 April. FBIS-NES-94-072, 14 April 1994. Hamas Statement. 2000. BBC Summary o f World Broadcasts, 23 July. Hoffman, Bruce. 1998. Inside Terrorism. New York: Columbia University Press. Horowitz, Michael, and Dan Reiter. 2001. "When Does Aerial Bombing Work? Quantitative Empirical Tests, 1917-1999." Iournal of Conflict Resolution 45 (April): 147-73. Hroub, Khaled. 2000. Hamas: Political Thought and Practice. Washington, DC: Institute for Palestine Studies. Huntington, Samuel P. 1991. The Third Wave: Democratization in the Twentieth Century. Norman: University of Oklahoma Press. Inbar, Efraim. 1999. Rabin and Israel's National Security. Baltimore: John's Hopkins University Press. Institute for Counter-Terrorism (ICT). 2001. Countering Suicide Terrorism. Herzliya, Israel: International Policy Institute for Counter-Terrorism. Jenkins, Brian N. 1975. "Will Terrorists Go Nuclear?" Rand Report P-5541. Santa Monica, CA: Rand Corp. Jenkins, Brian N 1985. International Terrorism. Washington, DC: Rand Corp. Jervis, Robert. 1976. Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press. Kaufmann, Chaim D 1996. "Possible and Impossible Solutions to Ethnic Civil Wars." International Security 20 (Spring): 136-75. Kaufmann, Chaim D 1998. "When All Else Fails: Ethnic Population Transfers and Partitions in the Twentieth Century." International Security 23 (Fall): 120-56. Kramer, Martin. 1990. "The Moral Logic of Hizballah." In Origins of Terrorism, ed. Walter Reich. New York: Cambridge University Press. Kramer, Martin. 1996. "Fundamentalist Islam at Large: Drive for Power." Middle Fast Quarterly 3 (June): 37-49. Kydd, Andrew, and Barbara E Walter. 2002. "Sabotaging the Peace: The Politics of Extremist Violence." International Organization 56 (2): 263-96. Laqueur, Walter. 1987. The Age of Terrorism. Boston: Little, Brown. Lebo w, Richard Ned. 1981. Between Peace and War: The Nature of International Crisis. Baltimore, MD: Johns Hopkins University Press. Lewis, Bernard. 1968. The Assassins. New York: Basic Books. Makovsky, David, and Alon Pinkas. 1994. "Rabin: Killing Civilians Won't Kill the Negotiations." lerusalem Post, 13 April. Merari, Ariel. 1990. "The Readiness to Kill and Die: Suicidal Terrorism in the Middle East." In Origins of Terrorism, ed. Walter Reich. New York: Cambridge University Press. Mish'al, Khalid. 2000. "Interview." BBC Summary of World Broadcasts, 17 November. Mishal, Shaul, and Avraham Sela. 2000. The Palestinian Hamas. New York: Columbia University Press. Niebuhr, Reinhold. 1960. Moral Man and Immoral Society. New York: Scribner. Nusse, Andrea. 1998. Muslim Palestine: The Ideology of Hamas. Amsterdam: Harwood Academic. O'Neill, Richard. 1981. Suicide Squads. New York: Ballantine Books. Pape, Robert A. 1996. Bombing to Win: Air Power and Coercion in War. Ithaca, NY: Cornell University Press. Pape, Robert A. 1997. "Why Economic Sanctions Do Not Work." International Security 22 (Fall): 90-136. Pape, Robert A. 2002. "The Universe of Suicide Terrorist Attacks Worldwide, 1980-2001." University of Chicago. Typescript. Post, Jerrold M. 1990. "Terrorist Psycho-Logic: Terrorist Behavior as a Product of Psychological Forces." In Origins of Terrorism, ed. Walter Reich. New York: Cambridge University Press. Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development: Political Institutions and Weil-Being in the World, 1950-1990. Cambridge, UK: Cambridge University Press. Rabin, Yitzhaq. 1994. "Speech to Knessett." BBC Summary of World Broadcasts, 20 April. Rabin, Yitzhaq. 1995. "Interview." BBC Summary of World Broadcasts, 8 September. Rapoport, David C. 1971. Assassination and Terrorism. Toronto: CBC Merchandising. Rapoport, David C. 1984. "Fear and Trembling: Terrorism in Three Religious Traditions." American Political Science Review 78 (September): 655-77. Reagan, Ronald. 1990. An American Life. New York: Simon and Schuster. Reich, Walter, ed. 1990. Origins of Terrorism. New York: Cambridge University Press. Sauvagnargues, Philippe. 1994. "Opposition Candidate." Agence France Presse, 14 August. Schalk, Peter. 1997. "Resistance and Martyrdom in the Process of State Formation of Tamililam." In Martyrdom and Political Resistance, ed. Joyed Pettigerw. Amsterdam: VU University Press, 61-83. Schelling, Thomas. 1966. Arms and Influence. New Haven, CT: Yale University Press. Schmid, Alex P., and Albert J. Jongman. 1988. Political Terrorism. New Brunswick, NJ: Transaction Books. Sciolino, Elaine. 2002. "Saudi Warns Bush." New York Times, 27 January. Shallah, Ramadan. 2001. "Interview." BBC Summary of World Broadcasts, 3 November. Shiqaqi, Khalil, et al. 2002. The Israeli-Palestinian Peace Process. Portland, OR: Sussex Academic Press. Sprinzak, Ehud. 2000. "Rational Fanatics." Foreign Policy, No. 120 (September/October): 66-73. "Sri Lanka Opposition Leader Promises Talk with Rebels." 1994. lapan Economic Newswire, 11 August. St. John, Peter. 1991. Air Piracy, Airport Security, and International Terrrorism. New York: Quorum Books. Theodoulou, Michael. 1995. "New Attacks Feared." The Times (London), 21 August. FBIS-NES-95-165, 25 August 1995. Tuchman, Barbara W 1966. The Proud Tower. New York: Macmillan. World Islamic Front. 1998. "Jihad Against Jews and Crusaders." Statement, 23 February. 19