Killing Without Dying: the Absence
of Suicide Missions
STATHIS N. KALYVAS IGNACIO SÁNCHEZ-CUENCA
Sean O'Callaghan was a member of the Irish Republican Army (IRA) who later became an informer for the Irish police. Having spent most of his life ighting against terrorism, O'Callaghan (1999: 308) concludes that 'the IRA is an organization that produces people prepared to starve themselves to death, people prepared to spend large part of their lives in prison. In short, it produces people who are prepared to inflict death, pain and suffering on themselves as well as on others in pursuit of a cause.' Yet, in spite of this, the IRA never launched a suicide mission (SM). The same holds for many other insurgent and terrorist organizations all over the world, no matter how violent their tactics, how virulent their ideology, or how extreme their members' preferences and commitment.1Euzkadi ta Azkatasuna (ETA) in Spain, the Red Brigades in Italy, the Baader Meinhof in Germany, the Shining Path in Peru, and the Kosovo Liberation Army, to name but a few, never engaged in SMs. Even the Algerian Groupe Islamique Armé (GIA), an extremely violent and militant Islamist organization that engaged in tens of massacres of civilians, resorted to SMs only once. In fact, according to a recent estimate only 113 out of 7,053 terrorist incidents (1.6 per cent) were suicide bombings and the great majority were attributed to a handful of organizations.2 Here, we turn the central question of this volume on its head and ask why so many organizations do not resort to SMs.
Any explanation of why some organizations do engage in such missions must also account for the fact that the great majority of insurgent and terrorist organizations refrain from using this method. In particular, if we assume that SMs are instrumentally planned to achieve a set of specific goals, we must also be able to say whether the absence of SMs results from altogether different goals or, rather, from differences in the costs associated with the parameters of the production of violence. In Becker's terminology (1996), the variation on SMs can be attributed either to variations in preferences or to variations in the relative costs of the inputs that are necessary to