I. Introduction
On February 24, 2022, Russia launched a military invasion of Ukraine. As of this writing, the first interstate war on European soil since 1945 is raging. World leaders assert that the conflict threatens international peace and security. However, while they have imposed a range of economic, financial, and cultural sanctions on Russia and supply weapons to Ukraine, they are determined to ensure that their armed forces should not intentionally and directly confront Russian forces, be it on land, in air, or at sea, unless their own country or one of their formal allies within the North Atlantic Treaty Organization (NATO) are under threat.
This essay is about the ethics of military intervention in interstate conflicts that threaten international peace and security (for short, intervention). I ask, “Is intervention in such cases morally justified?” In international public law, the answer to that question is “yes”; threats to international peace and security provide one of two exceptions to the legal and moral prohibition (as set out in Article 2[4] of the UN Charter) on using force as a means for resolving interstate disputes. The 1990–1991 Gulf War is the most recent illustration of this point, which is a cornerstone of our current collective security system.Footnote 1
Just war theorists are not as verdictive. Compared to the ethics of humanitarian intervention and the ethics of national self-defense, the ethics of third-party military involvement in interstate conflicts remains strikingly underdeveloped in contemporary just war theory. Michael Walzer’s discussion of the moral foundations of the law of neutrality in his Just and Unjust Wars and a handful of philosophical analyses of the 1990–1991 Gulf War are the exceptions to the rule.Footnote 2 This is a regrettable oversight.
It is true that most post-1945 military conflicts have taken place within the internationally recognized borders of sovereign states. Also, some scholars recently argue that, over time, the incidence and destructiveness of war are waning. Nevertheless, even if the “decline of war thesis,” as articulated by Steven Pinker, is correct (and there are reasons to doubt it), the world is witnessing persistent and rising military tensions.Footnote 3 I do not mean only the ongoing war between Russia and Ukraine, but also tensions between Israel and the United States, on the one hand, and Iran, on the other hand; between China, on the one hand, and, on the other hand, Indonesia, Taiwan, Vietnam, Brunei, Malaysia, and the Philippines, over the South China Sea; and between North Korea and South Korea.
One may wonder why this issue warrants philosophical investigation. After all (it might be thought), wars of self-defense against unjust aggression are widely regarded as the paradigmatic example of in-principle just war. Moreover (it might also be thought), an unjust aggression is in itself a threat to international peace and security. Interveners are not merely helping the victim of the aggression; they are also intervening on behalf of all of us so as to forestall what, for all intents and purposes, is or has the potential to turn into a global crisis. Once one has provided a justification for wars of self-defense against territorial and political aggression, one has ipso facto provided a justification for intervention for the sake of international peace and security, and thus gone a long way toward vindicating the normative foundations of our collective security system.
Matters are not so simple. I argue in Section II that to defend intervention requires defending preventive military force, deterrent military force, and the resort to force for the sake of rights the defense of which is not standardly regarded as just causes for military action. In the remainder of this essay, I focus on the deterrence argument. I show in Section III that deterrence is morally justified in relatively few cases.Footnote 4 I then examine in Section IV two sets of problems with the argument: those raised by deterrence failures and those raised by the level of uncertainty under which leaders who use deterrent force operate. With respect to deterrence failures, I claim that there still is scope for limited deterrence. With respect to uncertainty, I claim that that there is scope for mitigating its impact, by building on Allen Buchanan’s and Robert Keohane’s proposals for reforming the current collective security system.Footnote 5 I offer a brief conclusion in Section V. Notwithstanding its endorsement of intervention in principle and of relevant institutional reforms, the upshot of my essay is that, in the world as we know it, the most serious military threats to international peace and security call for not intervening by military means.
We need some preliminary remarks. First, I set aside cases in which a state has bound itself to provide military assistance to another by means of a treaty, a case of which Article 5 of NATO’s founding treaty is a paradigmatic example. Whether and when international treaties in general, and defensive treaties in particular, are morally binding is a separate question that I lack the space to address here.
Second, I focus on conflicts that are characterized by the reciprocal resort to kinetic, lethal military force, in which interveners actively participate. I set aside cyberattacks and interventions by proxy, such as funding or providing arms to belligerents, as well as alternatives to military force, such as economic sanctions, conditional aid, and diplomatic negotiations. Sometimes, those alternatives stand a better chance of protecting international peace and security, but the claim that military intervention would fail the necessity and effectiveness requirements in such cases is compatible with the view that I seek to explore here, namely, that it has a just cause. It is also worth exploring whether military force is justified when those measures fail.Footnote 6
Third, I focus on cases in which international peace and security are thought to be under threat as a result of a conflict between states. Whether my arguments apply to other kinds of conflicts (for example, to conflicts within state borders involving secessionist or revolutionary movements) is not a question I pursue here.
Fourth, this essay is about the set of norms that govern the resort to military force, that is, jus ad bellum. At the bar of jus ad bellum, states, coalitions thereof, and (on some views) nonstate actors may justifiably resort to military force only if they have a just cause, if force is a proportionate and necessary response, and if it is likely to succeed. My main aim is to explore the view that the deterrence of threats to international peace and security is a just cause for military intervention in interstate conflicts. Throughout, when I say that intervention is justified in such and such case, I mean that it satisfies the requirements of jus ad bellum. Footnote 7
Fifth, I assume for the sake of argument that the citizenries of defending and intervening states consent to military action. Some readers might take the view that if those citizenries or even a minority within them withhold consent, intervention is morally unjustified. Others might say, on the contrary, that when international peace and security are under threat, consent is not necessary. I want to show, however, that even on the more standard view that consent is a necessary condition for permissible military action and even if consent is forthcoming, intervention in interstate conflicts for the sake of international peace and security is much more difficult to justify than is usually thought.Footnote 8
Sixth and finally, I refer to the state that initiates the conflict as Aggressor; to the defending state as Defender; and to the intervening party, be it another state or a coalition thereof, as Intervener. I use those labels as convenient shortcuts for the citizens and officials of those states. When I speak of, for example, Intervener being justified in resorting to force, I mean that its citizens and officials acting on their behalf are justified in so doing. I also take for granted that all human beings wherever they reside in the world have rights to the freedoms and resources they need in order to lead a flourishing life; those rights impose pro tanto duties on all others, wherever they are in the world, to support the institutions—be they domestic or global—needed to secure those freedoms and to provide those resources.
II. Interstate Conflicts and Global Crises
Suppose—and take as fixed throughout this essay—that state A attacks state D without warrant. State D (I assume) has a justification for defending itself by means of military force. A military conflict of this kind occasions severe direct and indirect harms to a number of people, such as loss of life, bereavement, life-changing injuries, and loss of homes and livelihood. By definition, it is a crisis. However, it does not necessarily threaten international peace and security, and so does not thereby trigger a global crisis.
In his discussion of the international legal order and its moral foundations, Walzer moots (though does not endorse) one possible defense of intervention that assumes, on the contrary, that an interstate conflict is by definition a threat to international peace and security. When Aggressor attacks Defender, it breaches the morally justified legal prohibition on aggression, a prohibition that states endorse by dint of their membership in the United Nations and their commitment to its Charter. In so doing, Aggressor threatens international peace and security and thus wrongs all other states.Footnote 9
On another view, as articulated by Yoram Dinstein, “an armed attack is like an infectious disease in the body politic of the family of nations. Every State has a demonstrable self-interest in the protection of international peace, for once the disease starts to spread, there is no telling if and where it will stop.”Footnote 10 Aggressor’s wrong is not just that it breaches the legal prohibition on aggression; it is also that it puts us all at risk of exactly this kind of harm either at its hands or at the hands of some other state that, should Aggressor be successful, would think itself licensed to resort to war to press its unjust ends.
The first defense of intervention suffers from two fatal weaknesses.Footnote 11 For a start, to justify the resort to military force is almost always to justify the resort to lethal force. However, the defense of a norm qua norm, as distinct from the interests that it protects, cannot on its own be a just cause for acts of killing. Moreover, Defender’s interests are not threatened in the same way as Intervener’s interests or indeed our own are, on whose behalf Intervener is acting. Its territory is wrongfully attacked, its ability to govern itself is wrongfully under threat, and the lives and limbs of its citizens and soldiers are at stake. We thus need to know what wrong exactly is incurred by Intervener and the rest of us and whether this wrong is severe enough to provide it with a justification for resorting to lethal force against Attacker.
The second defense of intervention plugs the gap, for it points to wrongful harms that justify the resort to force in individual self-defense. Yet, it too falls short of supporting intervention. After all, the “disease” does not always threaten to spread to the world at large. Likewise, a virus outbreak does not always turn into an epidemic; an epidemic does not always turn into a pandemic, even without interventions from outsiders. Suppose that Aggressor can quickly overpower Defender and that their dispute is of no strategic or economic importance to much larger powers outside the region. It is not clear at all that this conflict is a threat to international peace and security in any meaningful sense of those terms and is thus a global crisis, even if it is a regional crisis. The invasion of the U.K.-ruled Falklands Islands and the South Georgia and Sandwich Islands by Argentina in 1982 as well as the Congo Wars of the 1990s come to mind here. There was no suggestion at the time that Argentina would conduct further military aggressions on the United Kingdom or, indeed, on any of its allies, and that other countries would regard Britain’s failure to respond as encouragement to act on their own aggressive intentions. Devastating as the Congo Wars were for the region, it is not clear how destabilizing they have been to the world at large.
Contrast those examples with the COVID-19 pandemic. It clearly was a global crisis. The virus is lethal, spread quickly throughout the world, and has had severe direct and indirect impacts on the world’s population: death and long-COVID symptoms; millions of people throughout the world losing their job as a result of lockdown measures or of the economic recession consequent to the pandemic; millions of people having their non-COVID-related life-saving treatments delayed; and a legacy of ill health, long-term structural deficits, and irretrievably lost economic and social opportunities.
Suppose, then, that at time t1, Aggressor attacks Defender. The ensuing conflict is or threatens to morph into a global crisis if (for example) it is of such nature as to kill hundreds of thousands of people and/or lead to mass cross-border population displacements (as would happen in the case of a nuclear attack); if the conflict, past the initial attack, leads to growing cross-regional instability, large-scale disruption of supply chains, a dramatic decline in standards of living for millions of individuals, and a fortiori if it goes nuclear; and so on.Footnote 12
Such an outcome could happen in different ways. For example, Aggressor’s invasion of Defender itself occasions those harms or Aggressor’s initial attack against Defender, if successful, would be a prelude to its attacking other states with similarly devastating consequences globally. In the event that Aggressor should be successful, other states might be emboldened to resorting to force to pursue their ends, again with similarly devastating consequences globally. Defender’s response might needlessly escalate the conflicts. Third parties’ reaction to an initially localized conflict between Aggressor and Defender might also lead to an escalation of violence, with the same consequences. These are the kinds of scenarios that, in the light of World War II, the international community sought to forestall by setting up the United Nations and its collective security system.
To justify intervention, one must show not merely that the proposed military action would not itself turn a localized and contained conflict into a global crisis. One must also show, in the first instance, that preventing the commission of further armed attacks, deterring such attacks, or thwarting grievous global harms concomitant on such attacks are just causes for the resort to military force.Footnote 13
This is not a trivial task. On what one may call the orthodox view of the morality of war, the resort to military force in self- or other defense is morally permitted so long as it is a response to ongoing or imminent unjust force. Unless Aggressor’s or other states’ subsequent attacks on other states would be imminent in the event of Aggressor’s victory over Defender, the orthodox view endorses intervention only as a means to help Defender here and now; it prohibits both preventive and deterrent interventions. Moreover, on the orthodox view, an unwarranted armed attack is the only just cause for military intervention in an interstate conflict. Preventing population displacements, protecting a minimum standard of living, protecting supply chains, and forestalling outbreaks of violence in neighboring countries, are not just causes for such intervention.Footnote 14
III. The Deterrence Argument
Much work has been done since the 1990s on the ethics of preventive war and the ethics of waging war against threats that do not take the form of an armed attack. I do not tread these relatively familiar debates here. Instead, I assume that the fact that a wrongful harm has not yet materialized does not render it impermissible to thwart it by force. I also assume that the protection of fundamental rights other than rights to territorial integrity and political independence is a just cause for resorting to military force. In the remainder of this essay, I focus on the deterrence argument for intervention.Footnote 15
Deterrence is usually explored in the literature on nuclear deterrence and in the literature on punishment. There is comparatively little philosophical work on conventional (as opposed to nuclear) military deterrence. This is surprising because the post-1945 collective security system set out in the UN Charter—in particular, Chapter VII—is set up not merely as a means to stop attacks on international peace and security as they arise, but in large part as a deterrence mechanism. Indeed, in the two cases to date in which the United Nations Security Council (UNSC) authorized the use of military force in an interstate conflict—the 1950–1953 Korean War and the 1990–1991 Gulf War—deterrence seemed an important rationale for intervention.Footnote 16
Under Chapter VII so construed, the international community via the UNSC threatens at time t0 to resort to military force against aggressors at t2 if the latter carry out military attacks at t1. It is empowered to make good on its threat by resorting to military force at t2. Its use of force is not meant merely to stop Aggressor. It is also meant to signal, ex post, that its threat at t0 was credible and thus to give credibility ex ante to its further threat of resorting to (typically) greater force at t4 in response to Aggressor’s further breaches at t3. Furthermore, the use of deterrent force at t2 is meant to deter both Aggressor and other states—Aggressor + —from resorting to military force at t3. When Intervener seeks to deter Aggressor, it engages in what philosophers of punishment call special deterrence. When it seeks to deter Aggressor + , it engages in so-called general deterrence.
Thus framed, the question of conventional deterrence differs from the question of nuclear deterrence as the latter is standardly examined in the relevant literature. With nuclear deterrence, the main question concerns whether the mere threat of nuclear force is permitted as a deterrent. The question here is whether the use of military force is justified as a means to render credible a threat of typically greater force. In that respect, the question of whether the use of military force is morally justified as a deterrent is analogous to the question, at the heart of the literature on punishment, of whether the imposition of hard treatment is morally justified as a means to deter the commission of criminal wrongdoings. In both cases, the question arises because the mere threat of harm, at t0, has failed to deter and the credibility of the institution (respectively, state punishment and collective security) is now at stake.Footnote 17
A. Special deterrence
Suppose that Aggressor has long sought to retake a large part of Defender’s territory over which it does not have a rightful claim. At t1, it mounts repeated raids in Defender’s airspace and territorial waters and amasses thousands of infantry troops alongside their shared border. Defender, which is considerably weaker in military terms, has not so far resorted to defensive force. Unless Intervener comes to its help militarily at t2, Aggressor will mount a full-scale invasion at t3 that, if successful, would destabilize the entire region with severely harmful consequences for other parts of the world.
Aggressor’s breach provides a just cause for resorting to military force as a means of neutralizing it and thereby forestalling the more serious threat that it poses to international peace and security. The breach also provides a just cause for resorting to force at t2 as a means to deter Aggressor from mounting a renewed attempt at t3. At the bar of the just cause requirement, there is no morally salient difference between blocking Aggressor from pursuing its bellicose policy and changing its incentive structures by making the policy prohibitively costly. Subject to considerations of proportionality, necessity, and likelihood of success, deterrent intervention is, all things considered, justified.
That said, three remarks are in order. First, to claim that one may use force against another agent as a means to deter it from committing further wrongdoings is to imply that it is amenable to being deterred. By implication, the use of lethal force against a wrongdoer cannot be justified as a means to deter that wrongdoer from committing further wrongs. In the present context, legitimate targets for special-deterrent military force are those individuals who are liable to being harmed by dint of their participation in Aggressor’s unjust attack. Deterrence cannot justify killing them all, even if neutralization does. Suppose that both deterrence and neutralization would succeed. For deterrence to be (conceptually) possible, there have to be some agents left to be deterred. Deterrence thus results in lesser loss of life than does neutralization, which counts in deterrence’s favor.Footnote 18
Second, difficulties arise if the degree of force, F, that Intervener employs fails to deter Aggressor either because Aggressor does not believe that Intervener will employ greater force at t4 or because the degree of threatened force is not high enough to deter it from pursuing its unjust ends. Intervener is justified in resorting to F, then, only if it is effective not just in the sense that Aggressor believes that Intervener will use further force at t4, but also in the sense that Aggressor must be dissuaded from further attacks at t3. Footnote 19
Suppose that F is neither credible nor dissuasive, but that greater force F*—say, drone strikes on military targets combined with the bombing of a dual-use facility—would be. Suppose, however, that F* would be a disproportionate response to Aggressor’s breach at t1. Even so, if F* is a proportionate response to its putative breach at t3, Intervener might be justified in resorting to it as a way to render credible its threat of resorting to greater force F**—say, a full-scale invasion—at t4 should Aggressor pursue its bellicose policy at t3. This point holds even if F** would be a disproportionate response to Aggressor’s wrongdoing at t3. In this kind of case, Intervener’s resort to F* at t2 is not itself a disproportionate response to Aggressor’s future wrongs; but it does serve as a means to render credible a threat of ex hypothesi disproportionate force F** at t4. The question concerns whether Intervener may nevertheless so act.
Some opponents of nuclear deterrence would aver, on the following grounds, that Intervener may not so act. Intervener’s threat at t2 is not credible unless Intervener actually intends to resort to F** at t4, but if resorting to F** at t4 is morally wrong, then so is intending at t2 to do so. Given that intending to do so is wrong, threatening to do so is also wrong. Although those opponents of nuclear deterrence target mere threats to use nuclear weapons, they would by implication condemn the resort to conventional force as a means to render a wrongful threat credible.Footnote 20
This objection invites two responses. The first response concedes that intending to resort to F** at t4 is impermissible, but it denies that Intervener must necessarily form such conditional intention at t2; all that it needs is for Aggressor to be uncertain as to whether Intervener will so act. If Intervener has strong reasons to believe that Aggressor is uncertain and will be deterred from pursuing its policy when faced with threats of F**, it is difficult to see why it may not so threaten at t2, even though it does not intend to resort to F** at t4.
This response assumes that bluffing is morally permissible and that it is possible for a regime to bluff or, at the very least, to engineer doubts as to its intentions. Those who reject either assumption might be tempted by the second response to the objection. This response accepts that Intervener must form at t2 the conditional intention to resort to F** at t4, but it denies that the fact that F** is impermissible entails that threatening to resort to it is impermissible. Threats, the response holds, have effects—here, the beneficial effect of forestalling future wrongful harms—that must be taken into account when ascertaining whether it is permissible to issue them. Sometimes, those effects render permissible a threat to do the impermissible. Either way, if Intervener may threaten Aggressor, then it may act in such a way as to make its threat credible—in this case, by resorting to F* at t2. Footnote 21
The second response is this. Suppose that Intervener impermissibly issues a threat to Aggressor at t2. Given that the threat itself is impermissible, so is the resort to force as a means to render it credible. However, now that the threat has been issued, Intervener’s credibility is on the line. Suppose that Aggressor rides roughshod over Intervener and wrongfully attacks Defender at t3. If Aggressor would have attacked Defender anyway, Intervener’s use of countervailing force at t4 is not morally troubling. However, if Intervener’s ex hypothesi wrongful threat triggers Aggressor’s move, then we may wonder whether Intervener is morally permitted to respond. To say that it is raises a moral hazard, since Intervener could deliberately issue a threat that it knows is impermissible, so as to create a situation in which it is then morally permitted to intervene. More broadly, it also raises the interesting question, already familiar in the literatures on nuclear deterrence and the endings of wars, of whether political actors must desist here and now from pursuing a course of action that they started unjustly.
Neither point undermines my limited case for intervention on grounds of special deterrence. The response to the concern about moral hazard is that, even if Intervener is justified in using force against Aggressor at t4, the fact remains that it did act unjustly at t2, a fact which will have to be taken into account in any justified post-conflict settlement. The response to the question of extrication is that the grounds which rendered the initiation of a course of action impermissible may shift, such that persisting is morally permissible.Footnote 22
B. General deterrence
So far, I have assumed that the resort to force is meant to deter Aggressor from continuing with its policy. In general deterrence, by contrast, it is meant to deter other states from threatening international peace and security. Suppose that Aggressor’s success would embolden Aggressor + into resorting to force to pursue its ends against Defender + , with devastating consequences globally. Does this make a difference to the moral permissibility of intervention?
Assume for the sake of argument that resorting to F is a necessary, likely effective, and proportionate means to deter Aggressor. We need to distinguish between two cases. In the first case, resorting to F against Aggressor is also a necessary and likely effective means to deter Aggressor + as well as a proportionate response to its future breach. If so, the fact that F deters Aggressor + provides Intervener with a further reason so to act.
In the second case, F suffices against Aggressor, but not against Aggressor + ; Aggressor + will renounce going to war at t3 only if Intervener subjects Aggressor to F*. Let us quantify the difference between F and F* as f. To say that Intervener may resort to F* against Aggressor is to say that it is justified in subjecting Aggressor to f in addition to F, for the sake of deterring Aggressor + . For example, it is to say that it may launch drone strikes (F) and bomb a range of dual-use facilities (f), albeit at the cost of more lives, limbs, and livelihood than if it had only done the former.
A familiar objection to intervention in such cases appeals to the Kantian prohibition on using persons as mere means to an end.Footnote 23 Aggressor is innocent of Aggressor + ’s future wrongdoings and is in the same position vis-à-vis Defender + as is the international community. To say that Intervener is justified in subjecting it to f (in addition to F) is to say that it may be used as a mere means to protect Aggressor + ’s future victims. However, we ought not generally to use, let alone harm, the innocent as mere means to our or someone else’s ends, however valuable those ends. Subjecting an innocent person to a high risk of losing her livelihood, sustaining life-changing injuries, or being killed for the sake of protecting another person is to use her as a mere means to the latter’s ends. By implication, then, general deterrence is not a morally justified response to attacks on international peace and security.
The objection fails if there is a morally salient connection between Aggressor’s use of force against Defender at t1 and Aggressor + ’s use of force against Defender + at t3, such that Aggressor is not in fact innocent of Aggressor + ’s wrongdoings. Suppose that even though Aggressor is not attacking Defender + , it cannot but foresee that its attack on Defender would, if successful, embolden Aggressor + . Aggressor’s wrongdoing is not just the wrongdoing of attacking Defender; it is also the wrongdoing of attacking Defender foreseeing that this will embolden Aggressor + to attack Defender + . Suppose, more strongly still, that Aggressor attacks Defender with the intention to embolden Aggressor + into attacking Defender + ; it commits both the wrong of an unjust aggression and the wrong of incitement. In both cases, Aggressor’s additional wrongdoing provides Intervener with a just cause for resorting to f in itself as a means to deter Aggressor + ; subject to the requirements of necessity and proportionality, combined with the wrongdoing of attacking Defender, it provides Intervener with a justification for resorting to F*. Footnote 24
In this case, Aggressor’s attack on Defender is causally related to Aggressor + ’s attack on Defender + and thereby contributes to undermining international peace and security. It is plausible that most cases of military aggression will be of that kind. There are few such conflicts that the emboldening impact of a failure to intervene is likely to reverberate beyond Aggressor’s borders.Footnote 25 Nevertheless, suppose that it is not related. Aggressor + had plans to attack Defender + anyway, irrespective of Aggressor’s attack on Defender. The deterrence argument says that Intervener is justified in resorting to military force against Aggressor as a means to deter Aggressor + , even though it is ex hypothesi innocent of the latter’s wrongdoings. Intervener’s resort to F* against Aggressor does seem vulnerable to the Kantian objection.
A classic reply to this objection, developed by Victor Tadros in defense of criminal punishment, says that wrongdoers have lost their claim not to be used as a mere means for the sake of others. Moreover, not only are they under remedial duties to their own victims to protect them from further wrongful harms, they are also under protective duties to the victims of other wrongdoers. Harming them at t2 to deter those wrongdoers at t3 is one way to enforce their protective duty.Footnote 26 In the context at hand, then, Aggressor owes it to Aggressor + ’s victims to protect them from Aggressor + ’s attacks and its concomitant grievous wrongful harms. By resorting to military force against it as a means to deter Aggressor + , Intervener is simply enforcing Aggressor’s duty.
This reply does not work, though. We all are under duties to victims of wrongdoing, whether or not we have contributed to those wrongdoings. Those duties flow from a general obligation of assistance to those in need and are subject to a no-undue costs proviso: Consistent with the prohibition on using the innocent as mere means to other persons’ ends, there are limits to the harms that we are under a duty to incur and that it is permissible deliberately to inflict on us for the sake of those in need. (Tadros endorses this point.) The claim that Aggressor is under a duty of assistance to Aggressor + ’s future victims and thus to shoulder the burdens (within limits) of thwarting the commission of future wrongs is incompatible with the Kantian prohibition. For it comes at the considerable moral cost of relaxing the prohibition on acts of military aggression against states—their citizens and leaders—who have not forfeited their rights to political independence and territorial integrity. Even if one accepts that the innocent are under duties to incur some harms for the sake of victims of wrongdoings, it is doubtful that the harms attendant on acts of aggression—which must be severe enough, remember, in order effectively to deter Aggressor + —are compatible with the no-undue costs proviso and, by implication, with the injunction against using the innocent as mere means.
If proponents of the Duty View wish to hold on to the prohibition on military aggression against the innocent while endorsing Intervener’s resort to deterrent force f (in addition to F) against Aggressor, they need to show that the latter is under a more stringent duty to deter Aggressor + than are other parties—such as states that were not part of the initial conflict—precisely by dint of its wrongdoing at t1. I cannot rehearse all the possible moves a proponent of the Duty View might deploy. Here is one, however, drawn from Tadros’s work.Footnote 27 In the context of punishment, wrongdoers are under an impersonal duty to redeem themselves. They can do so by showing that they are committed to the moral values that they impaired by acting as they did. One way to do that is to incur the costs of protecting future victims of other wrongdoers. Punishment enforces that duty. In the geopolitical context at hand, then, Aggressor is under an impersonal duty to redeem itself for having violated Defender’s rights to territorial integrity and political independence and, in so doing, subjecting its population and the world at large to a range of harms. It can discharge that duty by incurring the costs of being subjected to F* rather than F as a means to deter Aggressor + . Intervener enforces that duty.
Let us assume that wrongdoers are under an obligation to redeem themselves. While this argument draws a bright moral line between wrongdoers and the innocent—between Aggressor and others—it runs against an insuperable difficulty. Redemption is inherently expressive. As Tadros himself implies, to redeem oneself in the eyes of the victim of one’s wrongdoing and of third parties is not merely to act in such a way as to further the moral norms one has violated; it is also to acknowledge that one has committed a wrong. Whereas one can be coerced into conducting oneself in the required way, one cannot be coerced into forming and manifesting the requisite sincere belief. A wrongdoer in general and Aggressor in particular thus cannot redeem itself by doing x if it has to be coerced into doing so rather than because it believes that it has done wrong and that doing x is the right way to redress the wrong. To say that Intervener is justified in resorting to military force as a means to enforce Aggressor’s duty to redeem itself is self-defeating.
IV. Two Problems
To recapitulate, subject to meeting the requirements of necessity and effectiveness, Intervener is justified in resorting to military force F* against Aggressor at t2 to deter it from carrying out further attacks at t3 and thereby threatening international peace and security, so long as F* is a proportionate response to Aggressor’s future wrongs and even if the degree of force that Intervener thus threatens to use at t4 should Aggressor persist is a disproportionate response. Subject to the aforementioned requirements, Intervener is justified in resorting to F* against Aggressor at t2 as a means to deter Aggressor + from attacking Defender + and thereby threatening international peace and security at t3 in the following two cases: (a) Aggressor’s wrongful attack on Defender in itself warrants F* and (b) Aggressor’s wrongful attack only warrants F, but it is connected to Aggressor + ’s wrongdoing or to the state of affairs resulting from that wrongdoing in such a way as to warrant subjecting it to additional force f. Pending further defense of the Duty View or other arguments in favor of general deterrence, Intervener is not justified in other cases. In this section, I address two further, serious concerns about deterrent intervention in interstate conflicts: deterrence failures and the problem of uncertainty.
A. Deterrence failures
The claim that Intervener is justified in resorting to military force in the aforementioned cases is subject to its intervention meeting the effectiveness condition. Aggressor and Aggressor + must form the belief at t3 that there is a risk that Intervener will make good on its threat of further harm at t4 if they carry out further attacks. Furthermore, they must desist from so doing precisely on those grounds; otherwise, deterrence qua deterrence will have failed.
Ex hypothesi, however, deterrence has already failed, since Aggressor attacked Defender at t1 notwithstanding Intervener’s threat at t0. In order to deter Aggressor (and quite possibly Aggressor + as well) from resorting to wrongful force at t3, Intervener must at t2 overcome the credibility deficit it suffered at t1, in the knowledge that it may well fail again. Intervener’s decision of which quantum of force to use must rest on an assessment of the probability that it will succeed factored by the magnitude of the harms that would ensue should it fail, relative to the harms that would accrue if it does nothing at all. Intervener, thus, faces a dilemma: (a) either it resorts to, say, F* at t2 as a means credibly to signal that it will resort to F** at t4 if need be, in which case it risks locking itself into an escalating conflict and rendering the crisis worse than it is, or (b) it desists here and now, in which case it risks allowing the crisis to become worse than it is. Other things being equal, contributing to causing harm is worse than allowing harm to happen. Other things being equal, then, Intervener should refrain from using deterrent force at t2, a fortiori so if escalation occasions greater harms and risks thereof than does nonintervention. The difficulty for deterrence as a strategy, however, is that these are precisely the cases in which deterrence is most needed. Put differently, the greater an aggressor’s capacity and willingness to make light of interveners’ threats, the greater the need for intervention and yet, at the same time, the greater the case against it. Deterrent force is most effective against states that, by dint of their comparatively low capacity for threatening international peace and security, are the least likely to warrant it; it is least effective against those that, by dint of their immense capacity for harm, are the most likely to warrant it.
The difficulty is particularly acute in cases in which Aggressor (or indeed Aggressor + ) and Intervener both have nuclear capacities to the point of being able to destroy one another. Nuclear deterrence is usually meant to deter not just the resort to nuclear force, but also the resort to conventional force on the understanding that the latter might be met with a nuclear response and on the assumption that neither party will risk annihilation by escalating a nascent conflict. Ex hypothesi, nuclear deterrence so construed has failed in the scenarios at issue here. Russia’s ongoing war of aggression against Ukraine is a depressingly perfect illustration of the point. Here is another, not so hypothetical example. In the autumn of 2021 and the Spring of 2022, China launched waves of fighter jets and bombers into Taiwan’s airspace, following a long-standing policy of naval build up in the South China sea. It is thought that China will have the wherewithal to mount a full-scale invasion of the island within a few years. If the United States’ possession of nuclear weapons is meant in part to deter China from threatening Taiwan as it has done so far, deterrence clearly has failed. The question then is whether the next time China conducts such an exercise, the United States (and its regional allies such as Australia) may justifiably resort to conventional force as a means to deter China from invading Taiwan. Their resort to force at t2, recall, is meant to signal that should China nevertheless press ahead at t3, they will employ (presumably greater) force again at t4. If the force that they threaten at t2 to use at t4 is nuclear, they have to instill in China’s leaders the belief that there is a likelihood that they will so act despite the fact that China could retaliate in kind, with catastrophic consequences for the world at large, including the U.S. Given those consequences, it is unlikely that the U.S. could credibly threaten such response merely by using conventional force. Conventional deterrence, then, would fail. The most likely way to render it credible would be to resort to nuclear force at t2, but given that this would in all likelihood invite a similar response from China, deterrence would have failed in this case, too.Footnote 28
Suppose, contrastingly, that the United States threatens to use conventional force at t4 should China invade at t3. Its use of force at t2 will help render its threat credible. Should China then desist on that basis, conventional deterrence will have succeeded, thus rendering nuclear deterrence and its concomitant risks unnecessary in this case. However, the United States must gamble on China forming the belief that they have taken the nuclear option off the table and nevertheless deciding not to respond with conventional force. It is a hugely risky gamble. Moreover, even if one can uncouple the resort to nonnuclear force from the threat of nuclear force (by no means a foregone conclusion), the fact remains that countries with nuclear capacities also have large nonnuclear capacities and can inflict serious damage on one another. The difficulty highlighted two paragraphs ago thus remains.
Is conventional deterrence a moral nonstarter, then? Not necessarily. Even if using force against Russia and China would not succeed at deterring them, it might succeed at deterring other putative aggressors from pursuing their ends by force. Thus, even if using deterrent force against Russia and China would be, all things considered, impermissible, using deterrent force against a lesser foe might be, all things considered, permissible, subject to considerations of proportionality, necessity, and likelihood of success. The claim that the more necessary deterrence is, the less effective against the initial wrongdoer, and vice versa, while plausible, does not show that deterrence is never morally justified.Footnote 29
B. Uncertainty
At t1, Aggressor subjects Defender to military force. Intervener’s leaders believe and argue that leaving the ensuing conflict unchecked will threaten international peace and security; on the basis of this, they employ deterrent force against Aggressor. For all they know, however, the conflict would remain localized. In that spirit, Walzer, who is the only just war theorist properly to engage with the issue of intervention in interstate conflicts, offers an uncompromising defense of states’ right, indeed duty, not to intervene, even in cases in which an aggression is or is likely to morph into a global crisis.Footnote 30
Although Walzer’s argument targets a decision to wage war, his argument has purchase against decisions to resort to force short of war that is likely to escalate into a full-blown military intervention. Crucially—and precisely because this puts pressure on the judgment that an interstate conflict is or promises to turn into a global crisis—it applies not only to deterrent force, but also to preventive force as well as against force to neutralize ongoing threats whenever those who wish to resort to force rely on that judgment. This must be taken seriously. Global crises call for extraordinary measures. The risk is that unscrupulous governments will claim that we are in the grip of a global crisis as a means to justify taking such measures when, in fact, their proposed course of action is a morally unwarranted response to the crisis, is deeply unpopular (even if morally warranted per se), or both. However, the converse is also true; precisely because global crises in general and threats to international peace and security in particular are thought to require extraordinary, costly, and often unpopular measures, governments that are unwilling so to act might avoid labeling a set of events as such. The problem is particularly acute for conventional deterrence. In order to deter putative aggressors, intervening states have to be credible; in order to be credible, they have to be willing to expose their own armed forces and populations to retaliatory harms on the part of aggressors.Footnote 31 Opportunistic mislabeling goes both ways—not just in the direction of predatory military intervention under the ill-fitting cloak of self-defense, but also in the direction of blind refusal to see where dangers lie. Either way, it is particularly likely to occur the more uncertain actors are about the facts of the case.
States are thus caught between the risk of intervening when international peace and security are not in fact under threat, thereby wrongfully causing people to incur grievous harms, and the risk of not intervening even though international peace and security are in fact under threat, thereby wrongfully allowing some people to incur grievous harms. Other things being roughly equal, under conditions of uncertainty, it is better to err on the side of not harming than on the side of allowing harm to happen. However, the problem of uncertainty and the concomitant risk of wrongful military action are not reasons for rejecting intervention out of hand. As Buchanan argues, ascertaining whether a wrongdoing is objectively, in itself, a just cause for war is only one of the tasks (albeit a crucially important one) that just war theory should set itself. Another task consists (a) in ascertaining whether a wrongdoing provides a justification for military force given the institutional framework within which we operate and, if not, (b) in reflecting on and building institutional frameworks that would mitigate the aforementioned epistemic and motivational risks and, in so doing, enhance our chances of doing the morally right thing, objectively speaking.Footnote 32
Our current institutional framework for addressing threats to international peace and security is not equal to the task of thwarting such threats as they arise from interstate military conflicts. Chapter VI of the UN Charter mandates member states to seek peaceful resolutions to their disputes. Chapter VII states that the UNSC “shall determine the existence of any threats to the peace, breach of the peace, or act of aggression and shall make recommendations, or decide what measures shall be taken in accordance with articles 41 and 42, to protect international peace and security” (Art. 39). It may authorize measures short of war as well as the deployment of armed forces, should it deem it necessary, and can delegate enforcement to states and/or regional organizations (Arts. 43, 51–52). The five permanent members of the UNSC (China, France, Russia, U.K., and the U.S.) each have a right to veto any substantive course of action put to the UNSC, though they must abstain from voting on matters pertaining to disputes to which they are a party (Art. 27). Once the UNSC has determined that international peace and security are under threat and decides to take enforcement action, member states are under a duty to provide the required assistance (Arts. 48–49).Footnote 33
Due to the right to veto and to the fact that the UN is unwilling to enforce Article 27, the UNSC is not able to pass a resolution describing military actions on the part of any of the five permanent members as a threat to international peace and security. Indeed, a draft resolution condemning Russia’s invasion was rejected on February 25, 2022, Russia having used its veto power. Granted, the General Assembly is able to pass such a resolution. In 1950, largely to remedy paralysis at the UNSC, it passed the “Uniting for Peace” Resolution. Under the terms of the resolution, if the Council “fails to exercise its primary responsibility for the protection of international peace and security” when there appears to be a threat to it, the General Assembly “shall consider the matter immediately with a view to making appropriate recommendations to Members for collective measures, including in the case of a breach of the peace or act of aggression the use of armed force when necessary, to protect international peace and security.”Footnote 34 On March 2, 2022, the Assembly voted in favor of a resolution condemning the war in Ukraine. That success aside, however, in matters pertaining to international peace and security, the General Assembly’s resolutions are not binding. Furthermore, due to the high number of authoritarian states, the Assembly is not adequately representative of the latter’s citizenries. In any event, even if it were representative, the UNSC alone is legally empowered to take or delegate enforcement action for the sake of international peace and security. While it has authorized the use of force in intrastate conflicts, particularly since the end of the Cold War, it is toothless in the face of interstate conflicts in which one of its permanent members is directly involved, has a stake, and/or has an ally in another veto-yielding permanent member.Footnote 35
If this relatively uncontroversial diagnosis is correct, this leaves us with the following options, all of which seek to minimize the occurrence of wrongful harms under conditions of uncertainty: (1) reform the current system, (2) bypass it altogether, or (3) do nothing. If there are realistic prospects for options (1) and (2) such as to mitigate both the risks of unwarranted intervention or unwarranted failures to intervene, then doing nothing is not a morally acceptable option. Which of the two remaining options we should advocate does not depend merely on whether, in the world as we know it, it stands a realistic chance of being adopted. It also depends on whether the resulting institutional setup would meet basic conditions for legitimacy such as representativeness and impartiality, the establishment of adequate procedures for resolving disagreements between states, and the existence of a fit between the institution’s goals (namely, the protection of international peace and security) and its processes and performance.Footnote 36
Bypassing the current system altogether would take the form of endorsing unilateral intervention or of acceding to states’ refusal to act, for self-interested reasons and even though action is called for. Alternatively, it may take the form of building a new and competing set of institutions from scratch. None of this is attractive. Imperfect as the United Nations (by which I mean both UN institutions and its member states) are, notably when it comes to keeping major powers in check, the decisions they make are less unrepresentative and less likely to be partial than decisions made unilaterally by a subset of their members. They also have a history of relative success in some cases, albeit mostly involving humanitarian interventions against weak states, and thus relevant experience that institutions set up wholly de novo would not have.
This leaves us with reforming the UN. Obviously, I cannot offer a detailed blueprint for reform here. I only have a tentative two-pronged proposal that draws on Buchanan and Keohane’s work.Footnote 37 Return to our recurrent case: Aggressor launches an attack on Defender. The international community must take the following three steps. First, it must determine whether the conflict is or would, if left unchecked, morph into a global crisis, such as to warrant military intervention. Second, it must decide whether to intervene. Third, it must hold itself or a subset of its members accountable ex post for a wrongful intervention or, as the case may be, a wrongful failure to intervene.
Consider the first step. To declare that an interstate conflict threatens international peace and security is to put states and their citizenries on notice and to direct them to take the steps necessary to contain it. Contrariwise, to declare that the world is not facing a global crisis exempts states and their citizenries from having to act. It stands to reason that such declaration ought to be made by a multilateral, impartial, and representative body. Furthermore, the determination that military force is a necessary, proportionate, and effective response—or, on the contrary, that nonviolent alternatives are warranted—carries serious material and moral risks, either way, for the world at large. It also stands to reason that such determination should be made by a multilateral, impartial, and representative body. Whether the UN General Assembly and a fortiori the UNSC could be such a body is doubtful, for reasons set out above. Hence, the first prong of the proposal: establish a separate body with the sole function of evaluating calls for the use of force. Such a body would comprise state delegations, failing which states would not support its establishment or recognize its decisions as authoritative. However, partly to help compensate for authoritarian states’ failures of representativeness vis-à-vis their citizenries, it would include officials of the UN’s major agencies and of nongovernmental human rights organizations, such as Amnesty International and Human Rights Watch, and (more controversially) major charities, such as the Red Cross or Médicins Sans Frontières. While these organizations and charities are not representative of citizenries in the sense in which we tend to think of representation, they routinely deal with the humanitarian costs of military conflicts; have the expertise to determine whether the use of force or, on the contrary, a decision not to intervene, would trigger or worsen a global crisis; and can and do speak on behalf of some of the most vulnerable individuals in the world. It is partly for that reason, in fact, that they are referred to as the “Third UN,” alongside UN institutions based in Geneva and New York and UN member states. This suggestion does not radically reshape the international order. As we saw, neither the General Assembly nor the UNSC has monopoly over the mere (as distinct from binding) determination of what counts as a threat to international peace and security. Moreover, the UN has long included civil society organizations in many of its operations.Footnote 38
Second, consider the decision to intervene by force and, if so, the question of to whom the intervention will be entrusted. Calls for permanent members to renounce their veto right over the use of force itself are wholly unrealistic. It is also unrealistic to suppose that the international community can do anything other than rely on a coalition of the willing, ranging from ad hoc coalitions to regional organizations such as NATO or the African Union. At the same time, it is worth noting that a decision to use force must be reached by a majority of the Council’s permanent and nonpermanent members. If the institution I described above determines that international peace and security are not under threat or that they are, but intervention would worsen the crisis, the UNSC’s members who take its judgment seriously have the means to withhold authorization. Of course, this does not guarantee that a wrongful intervention will not take place, as shown by the example of the U.S.-led coalition’s unauthorized invasion of Iraq in 2003. However, in the cases at hand, UNSC authorization, with warts and all, is (I fear) the best that we can hope for.
That said, third, as Buchanan and Keohane persuasively argue, states that seek authorization to resort to deterrent force ought to be willing to subject themselves to an ex post evaluation of their decisions by an impartial and representative body and to accept that body’s determination as to what they owe the victims of their wrongful decisions. By the same token, I submit that states that withhold authorization should also be willing to go through the same process. Again, there is no guarantee that states that have been found derelict, be it for triggering or worsening a global crisis by dint of their use of force or by allowing such a crisis to unfold by dint of refusing to use force, will “pay up.” Nevertheless, a system that affords opportunities for scrutiny and allocation of remedial responsibilities is better than none at all.
Is it realistic that states—not least the most powerful of all—would agree to this, here and now? Writing in the mid-2000s to late-2010s, Buchanan and Keohane seemed optimistic that they would, on the grounds that acceding to scrutiny would help states wanting to resort to force and seeking allies overcome suspicions that they are acting in a purely self-interested way. Writing in the early 2020s at a time of growing international military tensions—indeed, outright war—and during a pandemic in which the most powerful states have honored multilateralism in the breach more than the observance (to put it mildly), I am pessimistic. If such pessimism is warranted, we must reconcile ourselves to the fact that conventional deterrence, that pillar of the collective security system, will remain powerless in precisely the kind of conflicts that are most likely to trigger a global crisis. In the event that there are hopes for reform, they lie in reforming the system from within rather than without.
V. Conclusion
I began by noting that just war theorists have paid scant attention to the ethics of intervention in interstate conflicts. For that matter, they have paid similarly scant attention to the ethics of conventional deterrence, notwithstanding the fact that the latter is a pillar of the international collective security system. In this essay, I argued that the resort to deterrent military force as a means to forestall a global crisis is morally justified, but only in very few cases. While some moves to reform the international collective security system might help state actors deal with the problem of uncertainty, they will only take us so far.
I thus end on a somewhat deflationary note. Deterrent military force is morally justified, objectively speaking, in some of the cases in which international peace and security are at stake. In the world as we know it, however, I doubt that it is justified—at least, not against nuclear powers. This does not mean that nothing can be done. We should strive toward reform without illusions. But it does mean that, for now at least and paradoxically, we should take seriously the possibility that nonintervention, construed as the rejection of the direct use of military force, is the morally correct response to the most serious threats to international peace and security.
