In the lead article of this symposium, Allen Buchanan and Robert Keohane propose an informal regime wherein the use of weaponized unmanned aerial vehicles (“lethal drones,” in their shorthand) could be regulated at the international and transnational level. It is a rare ambition of scholarship to propose new institutional structures, and professors Buchanan and Keohane go about the task in their usual bold and innovative fashion. Indeed, this article will be a tremendous encouragement to scholars in international relations, reminding them not to lose sight of their subject's formidable normative implications, and to scholars in ethics, challenging them to embrace questions of political practicality as a central part of their research.
The authors' proposed Drone Accountability Regime (DAR) has significant implications for a number of debates prominent on the international-relations research agenda, including those about effective institutional design, norm diffusion, the role of soft law in the spread and fragmentation of international law, and the significance of reputation in states' decision-making. In this short comment, I will not be able to do justice to the contributions the article makes to these debates. Instead, I will focus on the following question: How does the informal DAR relate to, and how would it interact with, existing international treaty and customary (that is, binding) law that currently governs the use of force, including the use of lethal drones? The ethical implications of the regime's potential implementation would largely depend on its relationship with existing law.
Here the architects of the DAR face a threefold challenge. First, any proposed informal norm that is weaker than existing laws, based on the rationale that those laws are often disregarded or too out of touch with reality, threatens whatever formal compliance may exist. It also undermines the very idea of binding international legal obligations as setting a nonnegotiable minimum standard. However, second, any informal norm that is more stringent than established law is for that reason less likely to meet with acceptance by states. It also raises the question why it is only these specific means of warfare (that is, drones) that should be subject to this more demanding normative standard. Third, at the same time, any proposed informal norm that simply reflects existing legal obligations appears, prima facie, redundant.
How can an innovative informal regime rise to this threefold challenge? The considerable cost of undercutting international law is unlikely to be outweighed by the potential benefits of a weaker informal regime. However, positing informal norms that reframe or exceed legal obligations might be sensible if drone use as a means of waging war really does raise unique challenges. In that case, it could be worth institutionalizing existing obligations in a manner more conducive to eliciting compliance by drone users. If drones are unique, it might even make sense for their use to be subject to informal restrictions that are tighter than current international law. Moreover, if it is likely to have positive spillover effects onto the regulation of other means of waging war, such a demanding drone regime may have broader ethical value. Alternatively, however, a technology-specific regime might close off paths for a more inclusive development of informal norms related to the use of force, and possibly even for new laws themselves. The relation between the ethical value of an informal regime that exceeds the stringency of international law and the persuasiveness of the argument that drones are unique is dialectical. The more we are convinced of the uniqueness of drones, the easier it will be to justify a separate, stricter regime. The more we are convinced of that uniqueness, however, the less beneficial and more potentially detrimental a specific drone regime is for the progressive regulation of warfare in general.
The DAR and International Law Regulating the Conduct in War
The DAR asks for three eminently sensible measures on the part of drone-using states that concern conduct in war: publicizing targeting criteria ex ante; keeping records on the consequences of drone strikes; and explaining targeting decisions ex post. At first glance, this looks like a restatement of international law, as these measures are essential for a belligerent's ability to follow in good faith the central rules of international law regulating military operations. How could a belligerent ensure that forces comply with the law without issuing targeting guidelines? How could a belligerent exercise constant care and minimize collateral damage as much as possible without a clear grasp and systematic assessment of the consequences of past attacks? However, one innovation of the DAR proposal is Buchanan and Keohane's transparency requirement. Providing transparency in targeting by subjecting this data to public scrutiny is currently not required by the laws of war.
The DAR also asks states to adopt a policy according to which drone operators may, under certain conditions, refuse to carry out an attack. When precisely a combatant may refuse an order from higher up in the chain of command is currently beyond the purview of international law, with the caveat that combatants have an obligation not to follow a manifestly unlawful order. In fact, they may be criminally liable if they do. Here, the DAR again sets the informal standard higher than the current legal standard. The proposal of a rule of conscientious objection on a strike-by-strike basis is even more of a departure from current practice than in the case of the three rules mentioned above, which build on existing legal obligations.
Certainly, we have reason to believe that both additions to the canon of belligerents' obligations would not be acceptable to states without further justification. Records of treaty negotiations suggest that states tend to object to external scrutiny of their militaries and to international norms that might interfere with military discipline. Similarly, states routinely shroud their targeting criteria and air campaign statistics in as much secrecy as possible. Does the uniqueness of drones mean there is an increased urgency to providing transparency in targeting and permitting conscientious objection to individual attacks? In other words, is there a good reason why states might accept these additional (informal) obligations?
For Buchanan and Keohane, the unique risk of drones is that they make it “more difficult to identify violations of constraints against targeting noncombatants.” But if we compare drone strikes to regular air operations with manned aircraft this claim appears contestable. The ability of drones to loiter over targets does not necessarily influence the “visibility” of violations of international law. Of course, if the link between a drone and its command center were cut entirely so that drone strikes were executed without a human operator taking part in the decision-making process, this would remove one potential whistle-blower who could report unlawful killings. However, the authors explicitly bracket the problem of what they call “fully autonomous drones” from their proposal.
While the acceptability of these innovative informal obligations to states might be enhanced by the argument that drones raise unique challenges to the ex post facto assessment of combat operations, the ethical value of the specific provisions the DAR proposes certainly does not depend on whether drones make it more difficult to identify violations of the laws of war. As it is, identifying violations of the laws of war in the proverbial fog is difficult enough to render any added requirements of transparency extremely beneficial. In addition, for conscientious belligerents, transparency would facilitate the sharing of best practices, as the authors point out. For less conscientious belligerents it would create an incentive not to let military considerations trump humanitarian considerations when devising these policies. The implementation of the transparency measures would also provide much-needed data on how states actually operationalize legal concepts, such as constant care, feasibility, and proportionality. Such data would, in turn, enhance the action guidance of law and its usefulness as a tool to judge conduct in war in hindsight.
But if these provisions are sensible and urgent regardless of whether or not drones are unique, the question that arises is how, then, would the DAR influence the regulation of warfare in general? Is there a danger that singling out drones suggests that transparency is dispensable in the use of other means and methods of warfare? I doubt it. If transparency in drone use were to become the norm, this would weaken or even destroy the argument that opening up targeting policies to external scrutiny is militarily prohibitive in other cases. The DAR might also demonstrate that military considerations do not forbid treating the individual combatant as a fully capable moral agent. That positive spillover effect is all the more likely the less we perceive drones as a unique means of waging war.
The DAR and International Law Regulating the Resort to Force
Under the proposed DAR, “nonconsensual strikes are permissible within the territory of failed states when transnational violence is being organized within their borders and valid state consent is impossible to attain.” Buchanan and Keohane consider this to be in line with current international law based on two assumptions. First, they consider the legal criterion for self-defense against a nonstate actor without the consent of the state on whose territory the attack will take place to be the test of whether that state proved “unwilling or unable” to itself control the threatening nonstate actor. Second, they argue that the source of being “unwilling and unable” is, in colloquial terms, “state failure.” Both assumptions are problematic.
State failure is a poorly defined concept. But do we really want to declare a state as failed based purely on its toleration of a nonstate actor (such as a terrorist organization) on its territory that another state perceives as threatening? If a state tolerates the terrorist organization because it is unable to control it, the state is weak. If it is unwilling to reign in the terrorist organization, however, it is possible that the state is perfectly stable as far as its internal sovereignty is concerned but that the regime in power may be politically or morally corrupt. Not every pariah state is also a failed state.
Even if state failure was synonymous with a state's inability or unwillingness to neutralize a terrorist organization on its territory, this “unwilling or unable test” represents a long-standing U.S. interpretation of international law, but is not itself established international law. The International Court of Justice, for instance, sets the bar much higher. It only allows self-defense against a nonstate actor without the consent of the state on whose territory the attack would take place if that state itself “sent” the nonstate actor.Footnote 1 In other words, the threat against which the drone-using state would defend itself would have to be attributable to the state whose consent is dispensed with. The Court has been widely criticized for this injunction. Legal scholarship is neatly divided on whether a state that is merely “unable and unwilling” to neutralize a threatening nonstate actor, but does not itself support it, is a legitimate target of nonconsensual intervention.Footnote 2
Bottom line, the application of states' legal right to self-defense against nonstate actors, such as transnational terrorist groups operating from the territory of another state, is “one of the most contentious issues facing the international community today.”Footnote 3 It may be too much to ask of an article that integrates international-relations scholarship and philosophical reasoning to do justice to contestation in a third field. However, the authors' ambition not to “carve out an exception to the laws of war [but to] help ensure better compliance” means they either need to err on the side of uncontroversial lex lata as expressed in the text of the UN Charter and expounded by the International Court of Justice or make sure that the regime can evolve alongside the law, rather than charge ahead of it.
Even if we accepted the “unwilling or unable” test as lex lata and the diagnosis of state failure as equivalent to the finding of unwillingness or inability—in other words, even if the criteria for legitimate resort to drone strikes according to the informal DAR did not undercut the legal standard—the DAR would bear a risk of weakening compliance with international law. The effect of restating formal laws in an informal regime depends first and foremost on whether that restatement addresses the unique challenges of the technology in question. The two risks created or exacerbated by the availability of drones, according to the authors, are “(1) violations of sovereignty by unauthorized incursions into another state's territory; and (2) overuse of the military option, encouraged by lower costs, especially in terms of reduced risk to the user's military personnel and the avoidance of negative reactions to ground troops.”
While violation of sovereignty is a risk, if not an immediate consequence, of virtually every cross-border military action, the argument regarding the risk of overuse seems convincing. But is it plausible that a United States with no drones at its disposal would limit itself to diplomatic endeavors to neutralize the threat stemming from terrorist suspects in Yemen and Somalia rather than conduct “normal” air strikes or special operations missions, which have the same potential as drone strikes to violate a state's sovereignty? It is beyond the scope of this comment to thoroughly investigate this complex counterfactual claim. Nonetheless, if we agree with Professor Buchanan and Professor Keohane that drones indeed lower states' threshold to resort to force, the next question is: does the proposed regime effectively address the danger of overuse?
The DAR asks for consent or a waiver of consent of the state on whose territory drone strikes are to be carried out. If the state is failed and such consent or a waiver cannot be obtained, states have to deliver a public justification for their drone strikes ex ante and allow the publication of records ex post. States are also enjoined to participate in good faith in an inquiry into the validity of their justification. The “war paradigm” the authors embrace suggests that they consider it possible that the resort to drones alone can constitute an act of war, and thus violate the legal prohibition on the use of force in international relations as enshrined in Article 2(4) of the UN Charter. Such violations tend to be accompanied by public justifications, usually in terms of more or less plausible appeals to self-defense. The more transparent the grounds for those justifications are, the better. But how much legitimizing force will these procedural steps have and how much should they have? As I understand the proposal, the justification, publication of records, and inquiry have two aims: to establish the plausibility that consent was really impossible to obtain, and to show that the threat was imminent and grave enough to warrant forcible self-defense. The authors do not explore a scenario in which a drone-using state might comply perfectly with the suggested procedures, but consent would still have been possible to obtain; or a case where a threat did not rise to the level of an armed attack. In the absence of a neutral assessment and pronouncement on the objective merit of the justification, the act of justifying may seem enough to legitimize the resort to force. If the DAR provides potential belligerents with access to procedural legitimacy, it might encourage states to resort to force even if they cannot avail themselves of the legitimacy shield of substantive legal standards.
Even if the DAR's ex post inquiry were equipped to uncover violations of international law—for instance, that the supposed attack was not imminent—the informal establishment of a violation of the most fundamental rule of international law would be a poor consolation, given that the attack was initially endowed with legitimacy. The crucial point is that this extra source of procedural legitimacy is not available for the justification of “normal wars.” If the singular risk of drones is the encouragement of their overuse, making it any easier to legitimize resort to their use seems the entirely wrong course of action. It does not help that the authors propose that the Assembly of States of the DAR could “authorize action to eliminate drones.” The authors emphasize the informal nature of the rules of the DAR and that the regime's bodies would “not have mandatory power.” But what kind of authorization powers would the DAR then possess: the kind that justifies a course of action as legitimate even though, or whether or not, it is legal?
Concluding Thoughts
Any proposed new regime, whether formal or informal, needs to find the space in which it reflects a progressive ideal while also being acceptable to states. I think the DAR gets that balance exactly right regarding the use of drones as a means of conducting hostilities. It creates (informal) obligations not yet enshrined in law and reaffirms obligations that are implicit in the law by requesting transparency in compliance. The contestability of the claim that drones are unique means the progressive contributions of the DAR will likely have a positive influence on the standards perceived to apply to warfare in general.
That said, this likely spillover effect could be highly problematic for international law on the resort to force. Even though it is informal and tailored to drones, a regime that undercuts established law in order to endorse how states may want to behave when resorting to drones may erode existing legal standards regarding the resort to force. Even if we believe that state failure means a given state can neither consent to intervention nor neutralize nonstate actors on its territory, and that this finding triggers a permission of forcible intervention under international law, the DAR risks reducing the reputational costs of unlawful resorts to force by offering procedures that legitimize actions that may not ultimately fulfill the legal criteria for self-defense. If, on the other hand, we believe that drones are unique in that they specifically tempt drone-wielding states to dispense with state consent and use military means before other options have fully been explored, then a regime that suggests that consent can be replaced with public justifications may be even more problematic.
