Dr. Brad Roberts
Institute for Defense Analyses
26 February 2003
Among both policymakers and academics, interest in deterrence has waxed and waned over the decades. Especially since the end of the Cold War, it has mostly waned. But 9/11 brought it back to the fore. More precisely, in the wake of 9/11, as we gained an appreciation of the difficulties of actually destroying al Qaeda and of securing the homeland, we have had to consider what deterrence might be able to contribute to a war that we cannot win quickly by defeat and defense. At first glance, this seems an almost silly question. How, after all, might it be possible to deter suicidal terrorists, others for whom violence is a self-validating way of life, and a network that has no territory that can be put at risk? But might a second glance tell us something more about this subject?
A year ago the Defense Advanced Research Projects Agency (DARPA) chartered the Institute for Defense Analyses (IDA) and the RAND Corporation to jointly evaluate the potential role of deterrence in the war on terror. The presentation today represents some of the results of my participation on that study team. Please understand that this is a work in progress and that I am here to present some models, hypotheses, and preliminary conclusions as a way to catalyze broader discussion in our community about these questions. Please also understand that these comments are my personal views, and should not be construed to represent the views of IDA, DARPA, or the Department of Defense.
This presentation will have two basic parts. The first develops a simple analytical model of contemporary terrorism, it use of violence, and the possible contributions of deterrence to the effort to induce terrorist restraint. The second applies this model to al Qaeda.
Who are our adversaries in the "war on terror?" If it were only a handful of dedicated suicide bombers and a leadership circle that dispatches them, then deterrence seems largely irrelevant. Terrorist movements are to varying degrees organized and tightly knit or disorganized and loosely knit. As a generic type, they seem best described with the rubric, "complex, adaptive systems." These systems encompass many actors beyond the bomber and his leader. Who are they?
Typically there are also state sponsors of terrorism. The Department of States lists seven countries as currently sponsoring terrorism as a matter of policy. A significantly larger number of states (perhaps as many as 30) might better be characterized as enablers than as sponsors of terror. Such states appear not to have a policy to sponsor terrorism. Instead, rogue elements in the state structure, largely in the intelligence and military institutions, may be supporting terrorism independently of the state. Or enabling states may be providing sanctuary only because they don't have the means to expel the terrorists. Others forms of enabling behavior include turning a blind eye to the operations of criminal organizations, to money laundering, and to the arms merchants as they assist terrorists. Especially in the wake of the collapse of the USSR, we have obviously all grown increasingly concerned about the ways in which terrorist organizations may be "enabled" through their access to the legacy NBC weapons and infrastructures of collapsing states. There are also enablers other than states. These include the moral legitimizers of the movement, the weapons experts, the financiers, and the dupes who are played upon for various purposes at various times (for money, technology, political cover, etc.). Yet another type of relationship between terrorist and sponsor is the terrorism sponsored statewith the Taliban government of Afghanistan serving as a powerful example.
This "complex, adaptive terrorist system" also includes assets upon which the terrorists depend. These include dispersed terrorist cells and lone operators and the command and control systems for themas well as their financial resources. Other assets include sanctuary for the leadership, recruits, and training purposes. Money is needed too. Important assets often overlooked by our analytical community are the credibility of the movement in its own eyes and its image of itself as righteous and powerfulintangible assets but important nonetheless.
All of this surrounds the core organizational structurethe leaders, their inner circle of advisors, the lieutenants who run the activities of the organization, technical experts of various kinds (not just in military operations but also recruiting, planning, communicating, and financing), contacts in supportive states and other institutions, the trained killers, and the foot soldier recruits. In our experience of modern terrorism over the last few decades, such organizations may be small and highly structured or large and loosely networked.
These various pieces then define the components of a complex, adaptive system. Let us ask ourselves: who in this system is committed to giving their life for the cause? Not the sponsors, not the enablers, not the lieutenants, not most of the foot soldiers, definitely not the leaders. They're not all ready to die. They may not even be ready to sacrifice very many substantial interests, depending on how close they are to the core of the system. This suggests that we are wrong to frame the deterrence question only in terms of what it might contribute to preventing suicide bombers.
A good place to begin to think about the strategic purposes of a counterterrorism strategy is this systems perspective. Strategy should seek to break apart the system. Strategy should be aimed at breaking apart the network at its nodes and isolating and somehow attacking the individual pieces. Those nodes are often quite vulnerableperhaps one or two key individuals. Moreover, their effective operation is heavily dependent on trust. Sowing distrust in such organizations is an old technique of counterterrorism but one with new potential in an era of advanced information technologies.
History seems to confirm the notion that the connections among these pieces of the system are weak. Criminal organizations, for instance, have proven themselves to be quite resistant to the efforts of terrorist organizations to enlist their services. This should not surprise us: after all, as parasites such organizations must take care not to damage the health of the organism on which they dependin this case, a state or society. Further, terror cells do not demonstrate a strong record of effective, autonomous action.
How is violence used to advance the interests of this complex, adaptive terrorist system? Because the American experience is so limited, we tend of think in terms of the discrete attacks inflicted on us by terrorists. In contrast, terrorist leaders seem to think in terms of a campaign of violence, orchestrated over an extended period of time. How should we understand how those leaders might understand the purposes of such violence? What might this suggest about the potential contributions of deterrence? It is useful to think in terms of the "lifecycle" of a terrorist group or movement. In simple terms, there are five generic steps in this "lifecycle."
The first is the formative stage, in which the core organization is formed, a vision of a "better world" takes shape along with a strategy of violence to bring it into being, and lieutenant are hired and trained and the rest of the movement is assembled. In this phase, violence is used to probe the vulnerabilities of targets and also to enforce internal discipline. This is where there seem to be some important failures of deterrence, in the sense that the probing violence of terrorists is not met with the type of reply that convinces them that they cannot go further.
The second stage involves the initial acts of strategic violence. In other words, they unleash their plan. Here, the history of terrorism suggests that violence is used for three different purposes. One is instrumentalto generate fear in order to extract concessions from a targeted political entity. The violence of the PLO or IRA is exemplary of this. Another is catalyticto awaken others to the potential of violence and the need to act. The Oklahoma City bombing is exemplary here, reflecting as it did Timothy McVeigh's belief that he could unleash a great race war in America. A third possible purpose is apocalypticto transform large social, political, or other realities in the service of some transcendent goal. The violence of Aum Shinrikyo in Japan best fits this notion, as its leaders believed that they could bring about the end of the secular period of Japanese history and usher in a new world with Aum members at the helm. Deterrence would seem to have no role to play in dissuading these initial acts of strategic violence if it has played no useful role in the formative stage.
The third stage is best thought of as the adaptive phase. What happens when the instrumentalists don't get their goal, the catalytic acts don't change the world, or the apocalyptic ones don't change a thing? At this point, terrorist groups seem either to adapt or collapse. Happily, our experience has not so far included an apocalyptic group that has successfully adapted to its failures. The groups with a catalytic purpose seem often to bludgeon on for a while before splintering. The instrumentalists seem to keep at it the longest, often with intense internal debates about the means and ends of their effort. Indeed, much of the history of terrorism is in this adaptive phase, as many groups struggle while some succeed but many fail. From a deterrence perspective, a central question here is whether such groups might respond to strategic frustration by escalating. Happily again, the history of terrorism is not rich in such examples. To be sure, over the last decade there has been a noticeable increase in the willingness to conduct indiscriminate attacks. But with the exception of Aum, none has sought to reap the full lethal potential of weapons of mass destruction.
The fourth stage is simply endgame: Here the purpose of the use of violence seems to depend directly on whether the group has won or lost. In winning, some groups turn the violence inwardly again, to clean up their act as they prepare for some legitimate political role. In losing, there is some getting even, as will as the inevitable formation of a splinter group, that not only gets even but also reenters the stage with acts of violence calculated to send a message that the risks have gone up. Deterrence considerations would seem not to be relevant, especially for the splinter group.
In a final stage, groups either fade away or that splinter group begins the cycle anew.
For sophisticated movements that fit the "complex, adaptive model," the key phase is that third one: adapt or collapse. Historical experience suggests that terrorist groups are not effective innovators. Over the last couple of decades many have formed and disappeared. Their difficulty in innovating has something to do with the price of failure. Terrorist organizations are not known for their ability to tolerate failureindeed, as argued above, their reputation for power and success is one of their most important assets. Those who fail are embarrassments and often are punished brutally. Aum Shinrikyo occasionally punished failure by death. This is not an environment in which innovation can affect big changes.
Let us consider now how this model applies to al Qaeda. Al Qaeda's formative stage lasted for about ten years. Over that period, it gathered members, enforced discipline on the group, and began to test the United States. One of its authors has written about American failures of deterrence in this period. U.S. responses to the embassy and Cole bombings appeared to only reinforce al Qaeda's view of America as a wimpy country that would back down once its nose were bloodied.
Obviously we are now beyond the phase of al Qaeda's initial acts of strategic violence and into the responsive phase. A core question here is simply: does al Qaeda's leadership perceive a need to adapt? Were the attacks of 9/11 a success, both operationally and strategically? Has what's happened since in South Asia, the Middle East, East Asia, and the United States played to or against al Qaeda's interests? We cannot know its answers to these questions but there are arguments on all sides. The central deterrence question is this: if the leadership of al Qaeda wishes to exploit weapons of mass destruction for their full lethal potential, is it willing to put at risk what remains of its complex, adaptive network? Is it willing to lose even more of its assets, enablers, sponsors, and moral legitimizers in exchange for further bludgeoning of America? Here as in the Cold War, perceptions of credibility are central: are U.S. threats to put those items at further risk likely to be seen as credible by the al Qaeda leadership? Does al Qaeda believe that the United States can do more damage or not? And does it believe that killing Americans in far larger numbers would not come at the cost of perceived legitimacy amongst those Muslims whose interests it claims to represent?
Moreover, if the al Qaeda leadership desires to adapt strategically, is the organization capable of doing so? American analysts are typically greatly impressed by the tactical innovations employed on 9/11 and in other al Qaeda operations. But let us put this in perspective. Since 9/11 we have seen no operations not already scripted in the handbook issued in the training camps (and available on the web). The "terrorist spectaculars" preferred by the leadership have apparently been attempted but without success (in Europe and in America). Can bin Laden and his immediate inner circle rethink a strategy and agenda that they have elaborated with such considerable attention, detail, and personal investment? Would they see escalation to the exploitation of NBC weapons for their full lethal potential as morally defensible in their own terms, especially if repeated over an extended campaign? And would they see it prudent if the perceived righteousness of their cause is part of what provides them a flow of recruits and the support of their various enablers and perhaps sponsors? We have opinions aplenty on these questions in America but precious few facts.
As a closing aside, let me note a problem we encountered in seeking answers to these various questions in our study. Our literature includes many studies on the emergence and initial acts of terrorist organizations. In contrast, there are quite few studies on why terrorist organizations go away and on how they adapt. We understand more about the groups still with us today than the ones that went away. We understand also much more about the potential capabilities of these groups than about their actual operational skills. We understand their incentives but not their disincentives, the sources of their desire to kill but not the way they think about the red-lines in front of them (and they do). We understand more about organizations than about networks and movements. Finally, the American expert community is largely ignorant of the experience of other countries in confronting and thinking through these issues. Experts from France, Israel, and especially the UK were helpful to us but only because we went and searched them out.
In sum, deterrence does have something to contribute to the war on terror. Its contributions may be modest but they are not irrelevant. As a complement to strategies emphasizing offense and defense, it has a role to play. But for deterrence to work, we must know what to hold at risk. The constellation of sponsors, enablers, and assets may provide opportunities on this countand the nodes that connect them. Attacking their trust of one another seems particularly promising. Putting their state sponsors and enablers at risk is an obvious pointand a focus of U.S. policy.
Dr. Brad Roberts is a member of the research staff at the Institute for Defense Analyses (IDA) in Alexandria, Virginia. The views expressed here are his own and should not be attributed to IDA or any of its sponsors. Prior to joining IDA, he served as editor of The Washington Quarterly and as a research fellow at the Center for Strategic and International Studies. His recent publications include "Hype or Reality: The New Terrorism and Mass Casualty Attacks" (2000) and "China, Nuclear Weapons, and Arms Control" in Foreign Affairs (1999).
Rapporteur, Oliver Fritz
back to Seminar Schedule, Spring 2003