The reliance on technical capabilities is not easily
criticized by ethicists, since the development
of these weapons is made in the name of ethics.
John Kaag and Whitley Kaufman1
Drones, more formally known as “unmanned aerial vehicles” (UAVs), are remotely controlled military aircraft that can be used for surveillance or attack. They are often controlled by operators thousands of miles from the site of the attack. Considerable controversy has arisen recently, much of it ethical, about the use of attack drones for the “targeted killings” of individuals outside the context of a conventional war, for example, as they are used in Pakistan. (By conventional war, I mean an armed conflict that at least starts as a struggle between two states.2) Drones are ethically interesting, in part, because the case for drones is often made in moral terms.3 The paper will begin by following this recent controversy, focusing on the use of drones for targeted killings outside of a conventional war. Later in the paper, I will consider the ethical issues raised by drone use more generally.
The development of military technology often poses new and difficult moral problems, and drones are no exception.4 In investigating the ethics of new military technology, it is helpful to distinguish between two sorts of moral problems it may create. The first sort, ordinary problems, may be addressed or resolved by a modification in the way in which the technology is configured or the military activities involving it are conducted. The second sort, extraordinary problems, are problems so severe that they may require as a response not that the technology or its use be altered, but that the technology not be used at all. This is obviously a rough distinction, if only for the fact that there is no sharp line between engaging in an activity in a modified way and not engaging in it at all. Some of the moral problems raised by drones are ordinary, but some are extraordinary.
The United States, and the West in general, is in a global struggle with Islamic insurgents, a struggle often referred to in the US as the global war on terror (GWOT). Conventional wars, such as those in Iraq and Afghanistan, are part of the GWOT, but much of the GWOT, the more controversial parts, takes place outside of conventional war. I will use the term GWOT idiosyncratically to refer to those parts of this struggle that are outside of a conventional war. One question we need to consider is whether the GWOT, understood in this way, is a war at all. The GWOT is an asymmetric conflict. Asymmetric conflicts, in general, are those in which one of the sides has at a great relative advantage over the other (the other being at a corresponding disadvantage). There are many forms of asymmetry, but the kind I am concerned with is one in which one side has a great military advantage in terms of its technology and/or the size of its military forces.5
This form of asymmetry characterizes most of the wars fought by the US and other Western powers in the past few decades. This is even more the case with the GWOT, where the asymmetry is extreme. Drones are, in fact, especially useful in fighting such asymmetric wars. Because of their surveillance capabilities, they are good at tracking and attacking individuals. Other advanced technologies, such as cruise missiles, are good at attacking hitting fixed targets, mainly infrastructure, but insurgent groups often have little in the way of infrastructure. Their ability to inflict harm resides in their personal, which is what drones are good at targeting. As the technology of drones has matured during the GWOT, the US has come to rely increasingly on them to fight its asymmetric battles. An increasing portion of US military aviation is devoted to drones. The cruise missile is a paradigmatic weapon of the cold war, while the drone is a paradigmatic weapon of the post-cold war world.
In the first section, to clear the ground a bit, I mention two strategic objections to drones, then in the second section present three moral benefits that supporters of drones claim for the technology. In the remainder of the paper, I present and develop five moral objections to drones; some of these apply to drones used for targeted killings and others apply also to drone use more generally. Finally, in the last section, I provide an overall assessment of the ethical status of drones.
1. Two Strategic Objections to Drones
Two of the most commonly heard objections to drones are more strategic than moral. They are primarily prudential because they concern not what drones do to their victims, but what they do to the interests of their users. Each of these objections entails that the use of drones is counterproductive.
The first objection is that the use of drones by the US against Islamic insurgents is counterproductive because, as it is crudely put, drones produce more terrorists than they kill, due to the animosity they generate in local populations.6 Concerning the civilians killed by drones, “every one of these dead noncombatants represents an alienated family, a new desire for revenge, and more recruits for a militant movement that has grown exponentially even as drone strikes have increased.”7 If defeat of the insurgents in Pakistan and elsewhere requires that their numbers be reduced and that the local populations be turned against them, drone attacks will not succeed.
The second strategic objection is that the development and use of drones by the US is counterproductive because it encourages the development of military capabilities and attitudes among other states that are, in the long run, inimical to US interests. First, it will lead to the development and deployment of drones by other states, some of whom will be opponents of the US. As two journalists note: “With Russia and China watching, the United States has set an international precedent for sending drones over borders to kill enemies.”8 Moreover, the deployment of drones by other states, whether friendly to US interests or not, will lead to a harmful climate of military instability. In addition, the use of drones by the US for target killings is widely perceived in the rest of the world as a flouting of international law.9 I will discuss later the role of international law in the criticism of drone use, but the point here is that respect for international law depends on a sense of reciprocity among states, and the perception by the rest of the world that the greatest military power is ignoring international law will lead other states to do so as well, resulting in a sort of international lawlessness that harms the interests of everyone. The use of drones may make everyone worse off. We have seen this dynamic before in the development of nuclear weapons technology during the cold war.
To put it briefly, drone use creates “blowback” that does more harm than good to US security. These objections do have some moral import because, if the use of drones is counterproductive, it will be a great waste of lives and resources. Were the GWOT a war in the morally relevant sense, the moral objections to this waste would be represented by its failure to satisfy the jus ad bellum principle that a war is just only if it has a reasonable chance of success.10 But complex conceptual and empirical issues lie behind these objections, and I will not consider them further.
2. Three Moral Virtues of Drones
Supporters claim that drones are a morally valuable military technology, and three virtues of drones are often cited. First, drones have a greater capacity for precision, thus potentially reducing “collateral damage” and better adhering, when used in war, to the important in bello principles of discrimination and proportionality. Bradley Strawser claims that “UAV technology actually increases a pilot’s ability to discriminate.”11 Drone operators can observe a potential target for hours before deciding whether to attack, making them better able to avoid civilian casualties. Kenneth Anderson makes a similar point, claiming that drone use allows us to avoid the “proportionality trap,” a trade-off characteristic of military actions where attacks that provide greater military advantage also put civilians increasingly at risk and attacks providing more protection to civilians achieve less military advantage.12 Drones break this linkage, achieving increased military advantage while also providing greater protection for civilians. They are a win-win.
Second, the use of drones poses no risk to the pilots, allowing military operations to better realize the principle of force protection. Force protection is not generally recognized as an in bello principle, but, other things being equal, it certainly is morally preferable to reduce one’s own combatant casualties. Strawser also argues that drones have this virtue. He casts it in terms of what he calls the “principle of unnecessary risk,” according to which military leaders have a moral obligation when pursing a legitimate military goal to choose a means, if available and just, that does not pose a risk to their combatants.13 Strawser’s principle is a specification of the principle of force protection, which counsels reducing risk to one’s combatants, though not necessarily eliminating it. Strawser’s specification of the principle of force protection is the principle at its limits.
Third, the use of drones may avoid the need to wage a full-scale war.14 This is related to the first virtue. Just as it would be preferable to use drones instead of troops for a particular mission in a conventional war, so it would be preferable to use drones instead of fighting a full-scale war involving boots on the ground, assuming these are exhaustive alternatives. Speaking of “force-short-of-war,” a category of military force in which he would presumably include drone use outside of a full-scale war, Michael Walzer notes that the use of such force avoids the “unpredictable and often catastrophic consequences” of war.15 If a state can avoid a need to go to war by using drones to obviate the threat it faces, it certainly is a moral benefit to have drone technology available for that purpose. One might think of the use of drones as a step that needs to be tried in order that a prospective war satisfy the jus ad bellum criterion of last resort.16
These three moral claims about drones represent the idea that the use of drones is morally preferable to other methods to the same purpose, such as traditional air strikes or the use of combatants, whether in special operations of a full-scale war. First, air strikes are less precise and the use of combatants on the ground generally leads to a greater number of civilian casualties. “Many military experts support the government’s claim that using conventional airstrikes or troops on the ground to attack terrorist compounds would be likely to kill far more civilians than drones have.”17 Second, of course, boots on the ground, and to a lesser extent traditional airstrikes, puts one’s combatants at risk.
3. Six Moral Objections and Two Paradigms
I will consider six moral objections to the use of drones. While all of these objections apply to the policy of drone use in the GWOT for targeted killings, some also apply to the use of drones more generally.
Drone use for targeted killings occurs outside the context of war, which is morally objectionable.
Drone strikes cause an excessive number of civilians casualties and otherwise violate principles of jus in bello.
Drone use is a rejection of traditional military virtues and affords its victims no opportunity to surrender.
Drone use removes political constraints on war making, undermining democratic controls, and creating a risk of perpetual military violence.
Because of the extreme asymmetry in a one-sided drone war, the use of drones is fundamentally unfair.
As a preliminary to the discussion of these objections, note that there are two ways to morally assess harmful behavior and a government’s response to it: they may be considered acts of war, or crime and law enforcement. There is the paradigm of warring and the paradigm of policing, each represented by a different set of rules limiting the coercive and violent actions involved in the government’s response.18 The rules are moral as well as legal. Warring is governed morally by the part of just war theory concerning conduct in war, jus in bello, and legally by International Humanitarian Law (IHL). In contrast, policing is governed morally by respect for individual human rights and legally by International Human Rights Law (IHRL). The contrasting pairs of rules are significantly different, with the moral and legal focus on human rights being, in general, more restrictive. One relevant difference between them is that a human rights focus concerns the conduct of those against whom force may be used. Force can be used only against those who have committed some wrong. In contrast, jus in bello and IHL allow or prohibit force against individuals due to their status, specifically, whether they are combatants or civilians.19 A related difference is that, special circumstances aside, a human rights regime allows the taking of life only after there is a something like a judicial finding of guilt (regarding conduct), while a regime under jus in bello or IHL allows the taking of life merely on an individual determination of status (for example, as shown by the wearing of a uniform).
4. Drones and the “Global War on Terror”
Consider objections (1), that drone use for targeted killings occurs outside the context of conventional war, which is morally unacceptable. This is a claim about the paradigm under which drone use should be considered, that it should be considered as policing rather than warring. The GWOT is not a war.
Of course, the response to this objection is that the GWOT is in fact a war, a “war on terror,” fought on a global battlefield, so that any drone strike against combatants in this war is permissible.20 The government asserts that the GWOT is a war: “The United States is in an armed conflict with al-Qa’ida and its associated forces.”21 But the question is whether the GWOT is a war at all. The question of the justice of the war is a different and secondary question. The appropriate way to answer whether the GWOT is a war is to investigate whether it sufficiently resembles a standard case of war that the rules of the warring paradigm can be applied to it. The assumption here is that, unless the GWOT is sufficiently like a standard case of war, it is not a war in the morally relevant sense, that is, a conflict to which jus in bello applies.
The GWOT departs from the standard case of war to such an extent that it is implausible to claim that it is anything other than war in an analogical sense (as in the “war on crime”). A number of points support this. First, it is a self-proclaimed fight not primarily against an organization, as in the standard case of war, but against a tactic, that is, terror.22 Perhaps in response to this concern, the Obama administration has replaced “GWOT” with the phrase “armed conflict with al-Qaeda, the Taliban, and associated forces.”23 But the new phrase makes the point that the US sees itself as battling not a single organization or even a set of organizations, but, given the elasticity in the term “associated,” an open-ended list of organizations, effectively, all alleged practitioners of terror. As Steve Coll notes: “A war against a name is a war in name only.”24 Second, the GWOT, unlike a standard case of war, has no temporal and spatial bounds, as terror will never be completely eliminated and the struggle against it has no spatial boundaries, no fixed geographical location. In all these respects, the GWOT is war only in an analogical sense, only in the way that the “war on crime” is a war. Crime is not an organization, and it is without spatial or temporal bounds, as it occurs everywhere and presumably will never come to an end. Indeed, there is more than a comparison between terrorism and crime, as terrorism is a form of crime to which the policing paradigm should be applied.
One other difference between the GWOT and the standard case of war is that each relates differently to the notion of state sovereignty. The standard case of war in conceptualized in terms of sovereignty—it is fought to attack sovereignty or to defend sovereignty. But the GWOT, conceived as a global struggle, considers sovereignty only from a practical perspective. It sees itself as attacking terrorists wherever they are, which is always in some sovereign state, but the sovereignty of that state is immaterial to the justification for the strike.25 The US may seek permission from a state for a strike, but it does so merely as a practical matter to facilitate its activities, not because it has an obligation to do so, not because respect for sovereignty is part of the justification for the strike.
Drone strikes have occurred in conventional war zones, for example, in Afghanistan, Iraq, and Libya. But many drone strikes for targeted killings, over three hundred in the past ten years, have occurred elsewhere, in Pakistan, Yemen, and Somalia, where the US is not a belligerent in a conventional war.26 This difference is marked in US policy by the fact that, while drone strikes in recognized war zones are conducted by the Air Force, strikes outside recognized war zones are conducted by the Central Intelligence Agency, a civilian organization.27 A commentator notes “just how radical it is [that] for the first time in history, a civilian intelligence agency is using robots to carry out a military mission, selecting people for killing in a country where the United States is not officially at war.”28 It is odd that, while the US claims that the GWOT is a war, it assigns the operations of that war to a civilian organization. As Mary Ellen O’Connell notes: “Only members of the United States armed forces have the combatant’s privilege to use lethal force without facing prosecution. CIA operatives are not trained in the law of armed conflict.”29 Indeed, in this regard, the US policy faces a dilemma. Either a targeted killing by drone in Pakistan, say, is an act of war (as the US claims) or it is not. If it is an act of war, it is not morally acceptable because the rules of war require that acts of war be undertaken by military rather than civilian personnel. If it is not an act of war, then it is morally unacceptable because it falls under the policing rather the warring paradigm and does not show the proper respect for human rights.
In addition to these conceptual and moral objections to the claim that GWOT is a war in the proper sense, there are also objections coming from international law.30 These are developed by Philip Alston.31 He argues that according to IHL, there are two conditions that must be satisfied for hostilities to constitute an armed conflict (that is, a war). One is that an armed conflict is a struggle between sufficiently organized and structured “parties.” But “al-Qaeda and other alleged ‘associated’ groups are often only loosely linked, if at all,” so they “cannot constitute a ‘party’ as required by IHL—although they can be criminals.”32 The other condition is that “there must be a minimal threshold of intensity and duration” to the violence perpetrated by each party, and it is questionable whether the violence perpetrated by al-Qaeda and associates “rises to the level necessary for an armed conflict to exist.” Alston concludes: “Taken cumulatively, these factors make it problematic for the US to show that” it is at war with the “terrorists.”
Because the GWOT is not a war in the morally relevant sense, I refer to it simply as a use of military violence. GWOT-style targeted killings and war are distinct forms of military violence and should be morally judged on separate bases. Daniel Brunstetter and Megan Braun make this point:
Clearly, drone strikes are acts of violence, but is their use an act of war? In war zones, such as Afghanistan and Iraq their use suggest that they are acts of an ongoing war against insurgents. However, in such states as Yemen and Pakistan, with which the United States is not at war, they are acts of violence carried out against targeted individuals in noncombatant zones with the tacit consent of the state government.33
Using an idea of Michael Walzer’s, these authors suggest that such acts can be regarded as acts of “force-short-of-war” and may be subsumed under the moral category jus ad vim, justice in the use of force, analogous to, but more permissive than, jus ad bellum.34 But it would be a mistake to see targeted killings by drone as in a moral category close to but not quite war. Walzer sees jus ad vim as applying to military actions taken by one state against another, and one reason that the GWOT is not a war is that the military violence is directed not against states but against individuals alleged to be members of a trans-national group.
The other part of the objection is that the use of drones for targeted killings outside of war is morally objectionable. Alston defines targeted killings as “the intentional, premeditated and deliberate use of lethal force, by States or their agents acting under colour of law, or by an organized armed group in armed conflict, against a specific individual who is not in the physical custody of the perpetrator.”35 He notes that targeted killing may occur inside or outside of war. While targeted killings in war may be acceptable, those outside of war are not. “Outside the context of armed conflict, the use of drones for targeted killing is almost never likely to be legal.”36 This because it falls under the policing paradigm and requires respect for human rights. O’Connell makes this point:
Drones are not lawful for use outside of combat zones. Outside such zones, police are the proper law enforcement agents and police are generally required to warn before using lethal force. Restricting drones to the battlefield is the most important single rule governing their use.37
The victims of targeted killings are dispatched without any judicial determination of guilt for alleged harmful conduct. These strikes “must be governed by international human rights law.”38 If targeted killings by drone were considered under the warring paradigm, they might be justified by the status of the victims as combatants under jus in bello or IHL. But without war, just war theory can do no justificatory work. War may legitimate military violence, but outside the context of a war, military violence is simply violence, and lacking in moral justification. Terrorism is not warfare but crime, and its perpetrators should be treated as criminals, not combatants. This means that they cannot be subject to targeted killing. Under the human rights paradigm, these killings are extrajudicial and amount to assassination.
5. National Self-Defense
Defenders of drone strikes outside conventional wars have one other argument to justify these strikes. They claim that the strikes are an exercise of a right of national self-defense. For example, the US claims: “Targeting a member of an enemy force who poses an imminent threat of violent attack to the United States is not unlawful. It is a lawful act of self-defense.”39 This claim is interesting because it offers what is in effect a third way to justify drones strikes, an effort to deny that the paradigms of policing and warring are exhaustive. If targeted killings in places like Pakistan can be justified under neither IHL (because there is no war there) nor IHRL (because they are extrajudicial executions), maybe there is a third possible mode of justification. In this spirit, defenders of drones have appealed to a principle of national self-defense. This principle can be seen as a hybrid between the two paradigms. Like the warring paradigm, it permits the killing of individuals without judicial process, and like the policing paradigm, it is not inconsistent with human rights.
The national self-defense justification is based on a domestic analogy with individual self-defense, but the government’s understanding of it may be based on confusion between the just war categories of jus ad bellum and jus in bello.40 The initiation of war under jus ad bellum can be justified on the basis of national self-defense. But once war is underway and jus in bello becomes relevant, the killing of enemy combatants is justified not on the grounds of self-defense (which would be a matter of conduct), but simply because they are combatants (a matter of status). This confusion can be seen when the government justifies targeted killings as based on “the inherent right to national self-defense recognized in international law (see, e.g., U.N. Charter art. 51).” But Article 51 in a matter of jus ad bellum concerning the justification of going to war, not an in bello justification of killing individuals. If the national self-defense justification is to succeed, it must be seen as independent of an in bello justification, since the military struggle with the insurgents is not a war. Were it not independent, it would not be a third way. But there is some indication that the government perceives the two justifications as independent, as when it speaks of a drone strike as “a lethal operation against an enemy leader undertaken in national self-defense or during an armed conflict.”41