Center for Strategic Decision Research


Dealing with Non-State Actors

Mr. Robert Nurick
Director, Carnegie Moscow Center

Deterrence of non-state actors is in some respects a curious subject because there is a widespread presumption that many of the non-state actors are in some fundamental sense non-deterrable. On the face of it, this appears to be so. We know that such groups are very different from state actors, that they pose a very different kind of security problem, and that there is good reason to think that the traditional concepts of deterrence are not appropriate. 

Nevertheless, it is useful to look more closely at why we believe the old concepts of deterrence are at best in need of revision. We may then be able to eliminate some of the policy challenges that have been posed by the new security agenda-an agenda in which non-state actors, especially those with potential access to weapons of mass destruction, are very much at the center. 


Let us go back for a moment and remember what we thought was meant by deterrence. In the Cold War days, the notion of deterrence was focused on the threat of retaliation, the threat of punishment. This notion was very much based on state-to-state interactions and it was believed that three sets of conditions were critical for deterrence to work. One was that you could identify your adversary, the person or state; the second was that the adversary had highly valued assets, and if you could attack them, you could affect your adversary's decisions, intentions, and actions; and the third was that you had the means to deter your adversary and the ability to credibly threaten the valued assets. 


Now, arguably, none of these conditions apply, or at least apply in the same way to catastrophic terrorism as they did during the Cold War. We are not always sure who the perpetrator is; we do not know what the different terrorist groups and other non-state actors value highly; and because we often cannot identify the physical objects that threaten, it is hard to know what the real threats are in different circumstances. 

Sometimes, it is argued, deterrence can still operate under these new and different conditions and it may eventually be an effective way to deal with terrorism, or to reduce access to weapons of mass destruction. I agree that if terrorist groups are unsuccessful long enough and often enough, eventually deterrence will start to work, and terrorists will turn to other methods to achieve their aims. But this is a longer-term hope; it is not one, I think, that should inform policy response, certainly not in the United States following the attacks of September 11. In fact, some hold the view that today's threats are existential, that there are groups out there so implacably hostile to the United States that the only thing that can be done is to reduce their ability to hurt the country. There is no chance of changing their intentions, which is what traditional deterrence sought to do. 

Since the Cold War days, however, there has been another kind of deterrence, which in the language of political science is called deterrence by denial. This in some ways sounds like the recipe for homeland security. But I think it is a bit misleading to call the important efforts to beef up homeland security as reflections of some kind of underlying belief in deterrence, even of this sort. I do not believe it is a strong rationale for current policy. Of course there is also the hope that over time, if non-state groups find it more and more difficult to carry out attacks, they will eventually give up. But as I said earlier, this is a possibility for the very long term. 


What does all this suggest? What are some of the policy issues that these differences raise? I would like to briefly note three consequences that flow from these differences. 

The first is that we have tried to find links to states and to recreate some of the conditions traditional to deterrence, because we are more familiar with these conditions and because they give us guidelines about how to act. The assumption here is that non-state actors do not exist in a vacuum, that they are often supported, harbored, or tolerated by states that we can identify; sometimes they simply exploit state weakness in the hope that state sovereignty will somehow provide them some protection. This is one of the reasons that the Bush administration has made a point of saying that they will hold states responsible for the actions of non-state groups that operate from their territory or with their support. 

Links do exist between state actors, who presumably are deterrable, and non-state actors, whose deterrability is more questionable; however the links are controversial. Iraq is one area in which controversy rages. But even there, where a charge was made that a link was essentially meant to provide additional political justification for a course that was taken for other reasons, there is still the fundamental question of whether the need to look for states to hold responsible for non-state actors we can't find will increase or decrease the problem of terrorism over the longer term. This issue, which started to be discussed in the run-up to Iraq, deserves and will get a lot more attention in the future. 

The second point is an aspect of traditional deterrence that has got little attention but probably deserves some: the assumption that it is important to understand your adversary, and that your very ability to deter depends on it. You need to know what your adversary values; you need to know your adversary's vulnerabilities-hence the enormous support of Soviet studies in the United States and the enormous support of American studies in Russia. This aspect of deterrence failed to attract much attention because of the assumption that non-state actors are in fact existentially non-deterrable. This assumption is worth reviewing, and may in fact be true at least in many cases. I think it is very important to look at this quite carefully and see if we really believe it and, if so, why. 

The third issue-the difference between the conditions for traditional deterrence and those for current policy problems-has something to do with military instruments. I just want to note one aspect of this very complicated topic, which is preemption, a subject that became intertwined with broader policy questions about how to respond to terrorism. My point here is that the conditions that make deterring non-state actors difficult also make preemption of non-state actors difficult. If it is hard to know how to punish or threaten to punish non-state actors in retaliation for a terrorist act they committed, it is also difficult to know how to credibly threaten to punish them before they act. This is a dilemma we will face over and over again when thinking about policy responses, particularly military responses to catastrophic terrorism. 


Finally, I would like to discuss an institutional issue that many speakers have raised. Given that terrorism is a multifaceted phenomenon requiring multifaceted responses, we need to determine how to coordinate these responses and how institutions should adapt to deal with them. This issue has been at the heart of recent debates about the future of NATO. For example, should NATO adapt itself to the new agenda and, if so, how should it relate to other institutions? It has been very hard to think about NATO's future without thinking about other institutions, starting with the EU. 

One solution, which implies a kind of division of labor, has been reflected more and more in policy documents both from NATO and the EU. NATO, as a traditional military institution, should deal primarily with the high-end military aspects of terrorism while the EU, given its mandate, should focus more on justice and home affairs, the primarily non-military and long-term policy responses to non-state actors, particularly those involved in terrorism.   

This division does make a great deal of sense, but it also raises some problems, as we learned in the U.S. after September 11, or as the French also learned dealing with terrorism of a different sort in the 1980s. Indeed, trouble will arise if there is a strict institutional divide between organizations responsible for military and defense functions and those dealing with domestic intelligence, home affairs, and health aspects raised by biological weapons and the like. There will be great difficulties if the two sets of institutions dealing with terrorist issues do not coordinate well and do not talk often, especially institutions on the international level. 

However, I am not arguing against a division of labor-a division of labor is necessary. But NATO and the EU must start coordinating more than ever before, which we are starting to see some signs of; and the point of these interactions should not be to simply refine a division of labor but to keep coordinating processes in which justice and home affairs, interior ministries, domestic intelligence, health services, and other departments work transparently with those who deal with defense, and vice versa. 


Top of page | Home | ©2003 Center for Strategic Decision Research