I. Introduction
Although there has been a move away from the comprehensive, multilateral economic sanctions of the sort that devastated Iraq, states still engage in unilateral or bilateral broad sanctions (e.g., the ongoing U.S. embargo on Cuba) and multilateral, “targeted” sanctions. The latter, sometimes called “smart sanctions,” have been developed since the 1990s and include measures such as asset freezes, travel bans, and smart trade sanctions. Economic sanctions are also widely seen as a tool to implement the responsibility to protect (R2P) doctrine. They are included in the report by the International Commission on Intervention and State Sovereignty as a measure to realize the “responsibility to react.” Footnote 1 As the R2P has evolved, sanctions have continued to be seen as an important measure to realize the goals of the tackling and prevention of mass atrocities. For instance, the United Nations Secretary-General’s report in 2012 on R2P sees sanctions as a central element of the toolbox to implement “pillar three,” which concerns coercive responses to mass atrocities. Footnote 2
Yet, in the literature on economic sanctions, they have been subject to extensive criticism. Their use has been widely criticized for failing to realize any change in policy in the political communities subject to the sanctions. Footnote 3 They are also often criticized for being indiscriminate and for leading to widespread suffering, as well as for being disastrous for civil and political rights in target states. Footnote 4 The comprehensive sanctions regime against Saddam Hussein’s Iraq, in particular, was widely seen as morally reprehensible, given the reported deaths of five hundred thousand children under the age of five. Footnote 5 Criticism of sanctions is so extensive that it is often suggested that war may be morally preferable to sanctions. The point is often made in the context of the discussion of the principle of last resort in Just War Theory (JWT). It is claimed that we should avoid a literal understanding of last resort, since war may be better than the alternatives and, in particular, economic sanctions. Footnote 6
According to what I will call the “Prevailing View” on the morality of economic sanctions, which can be found in much of the (relatively small) body of literature on the ethics of sanctions, broader sanctions (and potentially smart sanctions) are highly objectionable. Footnote 7 For proponents of the Prevailing View, this is largely because sanctions are seen as (i) indiscriminate, (ii) intending the harms that they inflict, and/or (iii) using the suffering of the innocent as a means to enact policy change. For some, sanctions are morally impermissible or highly problematic, regardless of the merits and demerits of the alternatives (and sometimes even if there are exceptions in the sanctions regime for the provision of humanitarian necessities). Footnote 8 Indeed, proponents of the Prevailing View sometimes reject outright the permissibility of broader sanctions. Footnote 9 Others argue that it is conceivable that sanctions may be permissible if they meet certain conditions (particularly when they take the form of smart sanctions), but generally hold that sanctions tend to be objectionable because of their alleged indiscriminateness, wrongful intentions, and/or use of innocents as a means. Footnote 10
By contrast, in this essay, I (partially) defend the case for economic sanctions. I argue that sanctions are not necessarily morally problematic and, in doing so, argue that sanctions are less morally problematic than is often claimed by proponents of the Prevailing View. This is because several of the noninstrumental objections to sanctions are largely unpersuasive. Thus, I argue that, although sanctions may still sometimes be morally impermissible, this is less likely to be the case than depicted by the Prevailing View. I go on to argue that, on the contrary, sanctions may sometimes be morally preferable to the leading alternatives and, in particular, to war and doing nothing.
More specifically, the essay will proceed as follows. In the early part of the essay, I will reply to (i) what I call the “Liability Objection,” which asserts that sanctions are impermissible because of their apparent indiscriminateness. I will then consider and largely repudiate the claims that sanctions are highly problematic because they (ii) involve means that intend harms — the “Intending-the-Harm Objection” — and (iii) wrongly use the suffering of innocents as a means — the “Instrumentalization Objection.” I will go on to argue that sanctions are more likely to distribute fairly the currently inevitable harms to innocents of tackling aggression and mass atrocities. In doing so, I will argue, more broadly, that we should often favor a “Harm-Distribution Approach” when considering issues, such as sanctions, in the ethics of war and peace.
Before beginning, some clarifications are necessary. First, as already noted, my focus will be on the noninstrumental reasons for and against sanctions. Instrumental considerations concern how efficacious sanctions are in tackling aggression and mass atrocities (measured by their effects on the promotion of the enjoyment of basic human rights) and the costs that they impose when doing so. Such considerations are likely to carry significant moral weight in an overall assessment of the case for sanctions, compared to the leading alternatives. Footnote 11 However, there is little grounding for strong conclusions in this regard. There is significant contestation about how effective sanctions are, often mired in methodological debates about measures of effectiveness, case selection, and time comparison. There is therefore little consensus on whether sanctions tend to address effectively aggression and mass atrocities. The empirical literature on the efficacy of sanctions also tends not to be comparative (or at least not comparative with robust supporting evidence). Footnote 12 As such, there is currently little empirical evidence on which to base comparative moral claims about the efficacy of sanctions and the leading alternatives, such as war. The lack of robust comparative evidence also affects the strength of my noninstrumental, comparative claims in defense of sanctions to the extent that these draw on contingent features of sanctions and war. These will therefore be tentative: I will present only presumptive, rather than definite, reasons in favor of sanctions, based on what we can reasonably expect of sanctions given their features and the nature of the international system.
Second, my account of the noninstrumental case for and against sanctions will be only partial. This is because a full account of the noninstrumental case for and against sanctions would need to consider in detail the various noninstrumental features of all of the alternatives as well, such as wars, international criminal prosecutions, peace operations, nonviolent resistance, and diplomatic criticism. There may be noninstrumental reasons particular to these alternatives to oppose or favor them; for reasons of space, I simply focus on the considerations relevant to sanctions and compare them to war and doing nothing — perhaps the leading alternatives. Similarly, I cannot consider the symbolic arguments for the importance of sanctions, which may also have noninstrumental import. Footnote 13 Space also precludes considering the potential for sanctions to be a form of punishment.
Third, I define economic sanctions as “coercive economic measures by one state or several states in order to realize political change in the target.” This definition coheres with the widely held view in the literature that sanctions are authorized by states and statist organizations, such as by the United Nations Security Council, but not by nonstate actors. Of course, nonstate actors may also instigate economic harms, such as by boycotting the produce of various regimes, but these should not be viewed as economic sanctions. It is also worth noting that, on this definition, the target of the sanctions regime need not be states; for instance, smart sanctions may target particular individuals through asset freezes, travel bans, and smart trade sanctions.
II. Objections to Sanctions
A. The Liability Objection
The central objection to sanctions by those who endorse the Prevailing View is what I will call the “Liability Objection.” Footnote 14 In short, this holds that sanctions are impermissible because they are indiscriminate in that they impose significant harms on those who are not liable. Accordingly, on this objection, sanctions are morally problematic because, as Joy Gordon argues, they “are inconsistent with the principle of discrimination from just war doctrine.” Footnote 15 More specifically, they appear to violate the principle of noncombatant immunity, which asserts that civilians should not be intentionally targeted in war. This is most obvious for comprehensive, multilateral sanctions, such as those against Iraq, where many of those who bore the brunt of the sanctions regime were children and the vulnerable. In fact, such is the apparent lack of discrimination in sanctions that various scholars claim that they are similar or analogous to sieges, since they undermine the economy of a society and, in doing so, prevent the production or importation of necessities. Footnote 16
In what follows, I will not completely reject the Liability Objection. Rather, I will argue that, although it may sometimes give us reason to oppose sanctions, it does not establish that sanctions are necessarily impermissible. This is because any of the potential responses to aggression and mass atrocities, even doing nothing, may involve harms to those who are not liable. In the case of doing nothing, innocents who are currently subject to the threat of aggression and mass atrocities will have to continue to bear harms for which they are not liable. Similarly, war also almost always involves the imposition of harms on those who are not liable to them. To start with, combatants may not be liable to attack. On “revisionist” JWT, combatants’ liability is not determined by their status as combatants, but rather by their individual moral responsibility and the justifiability of the war in which they participate. Footnote 17 Accordingly, many combatants fighting on the just side may be morally innocent, since insofar as they are helping in the pursuit of a just cause, they have done nothing morally wrong that abrogates their right not to be subject to potentially lethal force. Moreover, combatants fighting on the unjust side may also not be liable to attack. This is because they may not be morally responsible for their being combatants (for example, because they are conscripted) or because they are pursuing just causes within an all-things-considered unjust war and, in doing so, are not contributing to the overall unjust war. In addition, war also typically harms many noncombatants, who may also not be liable to attack. To that extent, even if one rejects revisionist JWT, war would still typically be subject to the Liability Objection. This is because almost all accounts of JWT hold that noncombatants are generally not liable to the harms of war. Furthermore, there are huge practical difficulties in determining who is liable to attack, and, even when this can be reliably determined, there are problems in ensuring that only liable agents are harmed.
It might be thought that smart sanctions avoid the Liability Objection since they target only those liable. But smart sanctions can also impose harms on those who are innocent. Footnote 18 First, travel bans and asset freezes can be mistakenly imposed on those who are innocent. Second, as Gordon argues, aviation bans “can undermine a major component of a nation’s transportation sector, adversely affecting the civilian population generally.” Footnote 19 An example was the ban on aviation on Gaddafi’s Libya, which Gordon argues reduced the availability of food since it jeopardized crop dusting and the importation of agricultural and veterinary supplies. Footnote 20 Third, arms embargoes may prevent the victims of aggression from being able to defend themselves. For instance, the arms embargo on the former Yugoslavia in 1991 denied Bosnia the means to defend itself against much better equipped forces. Footnote 21 Fourth, smart trade sanctions on particular industries, such as on timber or oil (which are often viewed as benefiting political elites), can damage an important sector of the economy and undermine the exports of the state and, as a result, weaken its ability to fulfill basic governmental functions. Footnote 22 Moreover, smart sanctions are widely held to be even more ineffective than broader sanctions. They may therefore be little better than doing nothing; innocents currently facing the threat of aggression or mass atrocities will still have to continue to face this threat.
The problem, then, is that, when responding to aggression and mass atrocities, all the various options, such as war, sanctions, and doing nothing, will typically lead to harm to innocents, either directly if the agent harms innocents when responding to the situation or indirectly if there is an insufficient response to the situation which means that those currently bearing the harms must still bear them. As such, all options are subject to the Liability Objection. It follows that there may be a lesser-evil justification for sanctions: given that no option can avoid the Liability Objection, we need to look to the option that will impose the least harm on innocents, and sanctions may sometimes impose less harm than the alternatives. This depends, of course, on sanctions being sometimes the lesser evil, that is, on whether sanctions are in fact sometimes likely to impose less harm on innocents. This will depend in turn on the efficacy of sanctions at addressing the situation, compared to all the alternatives, including doing nothing, and the incidental harms that sanctions will cause. I have already indicated that there is little consensus in the empirical literature on such matters. Notwithstanding, both broader and smart sanctions seem likely at least sometimes to be better than the alternatives in reducing the overall amount of harm that innocents face. To reiterate, this is not to deny that the Liability Objection may still apply to economic sanctions or to hold that sanctions are often likely to be justified as a lesser evil. The point, rather, is that the Liability Objection does not provide an indefeasible objection to economic sanctions; it depends on the efficacy of sanctions.
In fact, there is reason to hold that the Liability Objection is likely to deem sanctions preferable to the alternatives. My reasoning is as follows. First, numerous individuals are culpable for performing small wrongs that in part cause aggression and mass atrocities. Footnote 23 For instance, they recklessly or negligently contribute to a foreseeable harm, such as by negligently participating in a manufacturing sector that helps to uphold an aggressive regime or to an economy that foreseeably upholds the unjust global economic order. They may therefore be liable to small amounts of harm to address aggression and mass atrocities, in proportion to their small degree of wrongdoing. Second, broader sanctions seem likely to be preferable since they are more likely to involve the imposition of small harms on a large number of people (even if the overall amount of harm is the same or currently unknowable). By contrast, war and doing nothing seem more likely to involve a significant concentration of harms whereby few individuals (such as soldiers and those currently suffering in the face of mass atrocities) bear larger harms that are disproportionate to their wrongdoing. To help see this, suppose that the United Kingdom (UK) is trying to prevent Indonesia from obtaining nuclear weapons. Suppose further that British citizens generally make it politically impossible for the UK government to disarm their own nuclear weapons. The British possession of nuclear weapons foreseeably increases the desire of other states, such as Indonesia, to possess nuclear weapons. British citizens generally act with reckless disregard for the interests of those affected by their strong support for maintaining British nuclear weapons. Suppose further that Indonesians widely support, both materially and politically, their government’s attempt to obtain nuclear weapons, again with reckless disregard for those affected. Many of those in both Britain and in Indonesia are liable to some harm, given their reckless but minor contribution to an unjust nuclear threat (let us assume that nuclear threats are unjust). If the options chosen in response to this threat are to go to war or to do nothing, there will be a significant concentration of harms. In the case of war, this will be on British and Indonesian combatants and noncombatants. In the case of doing nothing, it will be on those threatened, such as the citizens in Indonesia’s neighboring states who are threatened by its nuclear weapons. With these two options, it seems very likely that some individuals will have to bear disproportionate costs — much greater costs than those for which they are liable. By contrast, if economic sanctions are chosen, the costs on the sanctioning state (the United Kingdom) and the sanctioned state (Indonesia) in terms of loss of trade and material wealth might still be proportionate to the degree to which many British and Indonesians are liable. Thus, the overall amount of harm may be the same, but sanctions are more likely to impose smaller, proportionate harms on a much greater number of individuals. Footnote 24
This is, of course, not to adopt a collectivized view of responsibility. Nor is it to deny outright the Liability Objection. Certain sanctions regimes, such as the sanctions on Iraq, impose disproportionate costs on a large number of individuals. Even less egregious sanctions are still likely to impose costs on those who are not liable for any harm (for example, young children). My point, then, is not that sanctions are exempt from the Liability Objection because they always impose smaller costs on a larger number of individuals. Rather, it is that there is a reason to hold that, other things being equal, sanctions are likely to be preferable to the alternatives of war and doing nothing because they seem more likely (but not certainly) to impose smaller, proportionate costs on a larger number of individuals.
A. The Intending-the-Harm Objection
A second major objection to sanctions is that, as a means, they intentionally target innocents. I will call this the “Intending-the-Harm Objection." That is, it is alleged by critics of sanctions that sanctions intentionally target innocents so that their suffering will persuade the target state to change its policy. Footnote 25 This may appear to contravene a central premise of the Doctrine of Double Effect that, when promoting a good end, any bad effects must be incidental side effects rather than intended. For instance, in Michael Walzer’s formulation, it is a condition of innocence in the face of double effect that “[t]he intention of the actor is good, that is, he aims only at the acceptable effect; the evil effect is not one of his ends, nor is it a means to his ends.” Footnote 26
This objection is mistaken. To start with, it should be noted that most sanctions regimes do not rely on the suffering of innocents in the target state to push the government to change its policy (as proponents of this objection may well admit). For example, some sanctions instead involve bans of transfers of weapons (i.e., the arms embargo on the Democratic Republic of Congo) or material used for the production of nuclear weapons (e.g., sanctions on Iran). The objection, then, at best applies only to certain types of sanctions regimes; most sanctions regimes do not follow this logic.
Furthermore, the objection also applies to both wars and doing nothing. Wars may use the suffering of innocent soldiers or of civilians in cases such as those involving terror bombing, where innocent soldiers or civilians are targeted in order to demoralize the opposing forces and to persuade the opposing side to cease its efforts. Perhaps less obvious is that doing nothing may also use the suffering of innocents as a means. States may make the calculation that letting innocent civilians in another state suffer will lead to a rebellion or to severe international opprobrium of the aggressor, and so be the means by which the aggressor changes its policy.
Moreover, the objection does not even have any purchase against sanctions regimes that do rely on the suffering of innocents in the target state to push the government to change its policy. The central problem with the objection is that the suffering of the innocents should not be viewed as wrong because it is intended. In other words, I am skeptical of the DDE on the issue of means (the ensuing discussion will, in effect, briefly present an objection to the DDE in the context of sanctions). Footnote 27 Consider the following cases:
Angola 1: Angola has been sending its troops into the civil war in the Democratic Republic of Congo in order to gain favorable access to natural resources. The United States imposes sanctions on Angola, knowing that the harm inflicted will cause a certain amount of suffering of innocent Angolans. But it also knows that the pressure of this suffering will (e.g., by internal uprisings) lead to the Angolan government ceasing its interference in the Democratic Republic of Congo, which is the central objective of the United States. Overall, much more good will be achieved.
Angola 2: Again, the United States imposes sanctions on Angola, knowing that the harm inflicted will cause a certain amount of suffering on innocent Angolans. This time it knows that the embarrassment of being sanctioned will mean that the Angolan government ceases its interference in the Democratic Republic of Congo (again, the main reason for the action of the United States). The policy will be as effective and lead to the same amount of suffering as in Angola 1.
It seems that both cases are equally permissible. It would be odd to hold that Angola 2 is permissible, but Angola 1 is not because the harm in Angola 1 is intentional. To see this, it helps to consider the objection in terms of motives and intentions. Agents’ intentions are their objectives or the purposes of their more immediate action, whereas their motives are their underlying reasons for acting (or their ultimate goal). Although I cannot argue for this fully here, it is agents’ motives, and not their intentions, that seem to matter noninstrumentally. Footnote 28 This is because what seems morally important are agents’ ultimate goals; as far as is possible, these are what should be assessed noninstrumentally, given that they ultimately determine agents’ actions, rather than their intentions, which are simply the objectives or purposes of agents’ immediate action, designed to bring about their motives. In the case of sanctions, agents need to be motivated, for instance, by concern to reduce the suffering of innocents (such as in a humanitarian crisis); the sanctioning agents should be assessed according to whether they have this concern. If the agents possess such motives, it is unclear why it matters that some of their intentions are nefarious when considered in isolation (such as when they intentionally target innocents so that the suffering of innocents will persuade the target state to change its policy), but ultimately are formed in order to realize their rightful motives (such as tackling the crisis). Footnote 29 For instance, in Angola 1, the motive is good: to help those in the Democratic Republic of Congo. The central intent is designed to achieve this motive: a sanctions regime to help those in the Democratic Republic of Congo. The imposition of the harms on the innocent is a subsidiary intent in that it is used in order to achieve the main intent and ultimately to fulfill the motive. But why should this subsidiary intent mean that the overall policy is impermissible? This becomes even clearer if we add that there is no other reasonable option to achieve the desired end. If this objective could be reasonably achieved without such harms, it might well be. When this is the case, Angola 1 seems permissible.
We may think differently when there is a variety of means and an agent chooses a suboptimal means. For instance, suppose that the imposition of sanctions by the United States in Angola 1 leads to a much higher number of deaths than other feasible means. Even then, it still is odd to object that the United States is wrong since it intends to harm those affected by its choice. Rather, it would be acting wrongly because it acts recklessly by failing to adopt a reasonable option that would have reduced the harm. Footnote 30 (Another way of framing this point is to argue that the agent’s action would fail to meet the requirement of necessity, that is, it would fail to avoid inflicting harms that are not necessary to achieve the proportionate end.) What seems wrong with the reckless failure to adopt a reasonable option is not the desire or will to inflict harm on innocent others, but rather the failure to be sufficiently motivated to give proper weight to the interests of others in one’s deliberations by avoiding unnecessary harm.
This is not to deny that sanctions may be wrong because of the recklessness or negligence of decision-makers. On the contrary, I suspect that this will often be the case. Yet, war and doing nothing also often involve reckless or negligent decision-making, when leaders fail to be sufficiently motivated by the interests of those beyond their borders (or within their borders). Thus, although sanctions sometimes might be wrong since they are often reckless or negligent, this is not unique to sanctions, and, contra the Intending-the-Harm Objection, they are not wrong because the intent that underlies them is to inflict the harm they bring about. Not only does it seem unlikely that sanctions would be intended to harm, but even if they were, sanctions should not be objected to for this reason.
C. The Instrumentalization Objection
The third, related major objection to sanctions on the Prevailing View is what I will call the “Instrumentalization Objection”: they involve the instrumental use of the suffering of nonliable agents to achieve the greater good. Footnote 31 For instance, Gordon claims that “sanctions reduce individuals to nothing more than means to an end by using the suffering of innocents as a means of persuasion, thereby violating the Kantian principle that human beings are ‘ends in themselves’.” Footnote 32 The Instrumentalization Objection is separate from (or in addition to) the concern that the sanctioning agent intends harm to the sanctioned party. It applies even when the sanctioning agent foreseeably harms sanctioned parties, using them as means to an end, regardless of the intent of the sanctioning agent.
This objection is subject to notable limitations. First, as for the Intending-the-Harm Objection, it will apply, at most, to only certain types of sanctions regimes — where the suffering of innocents in the target state is used to push the government to change its policy. For instance, the harms of the measures of smart sanctions such as asset freezes and travel bans are often unintended side effects of the coercive use of force against liable agents (with the aim of affecting the agents’ actions), rather than the harms being a means to affect the agents’ actions. The objection may also often not apply to broader sanctions. The broader sanctions policy may clearly not use the suffering imposed as a means. For instance, the sanctioning agent may launch sanctions because it does not want to be complicit in the target state’s behavior.
Second, as for the Intending-the-Harm Objection, it will also apply to war and doing nothing. These options may involve the unintentional but foreseeable use of harm as a means, such as when doing nothing will harm innocents in a way that foreseeably leads to an insurrection and a subsequent shift in an aggressive state’s policy.
Third, the instrumental use of the suffering of nonliable agents to achieve the greater good per se does not appear to be wrong. This is because what appears to be wrong when agents are used instrumentally is that they are shown disrespect — they are not viewed in the agent’s deliberations as separate persons worthy of respect. In other words, they are treated as mere means, since their interests do not figure sufficiently in agents’ deliberations. But the instrumental use of the suffering of innocents may not always treat innocents as mere means. For this, there must be a denial of respect. And sanctions may often not involve such a denial of respect; that is, they will not use innocents as mere means. To start with, the sanctioning agent may be concerned with the interests of the agent and the interests of others; the import of the agent’s interests — although figuring in the agent’s deliberations — may be outweighed by the greater weight it gives to the interests of the others (e.g., it is motivated to save the greater number who will be saved by a sanctions regime). Indeed, the sanctioning agent may lament that there is no way to avoid imposing the costs of sanctions on the innocent party in order to help others. In other words, the sanctions policy may not be reckless or negligent, since there is no reasonable alternative that would achieve the same end without causing such harm. Moreover, it is also conceivable — and perhaps sometimes likely — that a sanctions regime uses the suffering of those harmed as a means in order to assist these same individuals. For example, a state may inflict some suffering on the people of an authoritarian state in order to pressure the government to change its policy toward these people. In such cases, the sanctions policy that uses the citizens as means may in fact be motivated by respect for these citizens.
This is not to deny that states do sometimes use innocents as mere means. They sometimes use the suffering of innocents recklessly or negligently by failing to put in place an option that reduces the suffering of innocents because, for instance, sanctions are more politically convenient for the government of the state and they use the suffering of the sanctions regime in order to attempt to achieve policy change in the target government. In such cases, the state’s reckless or negligent, instrumental imposition of harm shows insufficient attention to the interests of those affected. The Instrumentalization Objection may therefore still sometimes apply. My point is that (1) sanctions will not always instrumentalize others; (2) the Instrumentalization Objection will also apply to other options; and (3) using someone as a means per se is not wrong — what is wrong is using someone as a mere means — and sanctions may not involve the use of others as a mere means. As such, the force of the Instrumentalization Objection to economic sanctions is heavily circumscribed.
III. In Favor of Sanctions
Thus far, I have largely rejected three potential objections to sanctions and, so, have rejected much of the case for the Prevailing View. In this section, I will present a reason to favor sanctions. I argue that sanctions are likely to distribute harms more fairly compared to both war and doing nothing.
Before arguing to this effect, however, it is worth replying to an immediate worry: considering fairness in detail is redundant in the ethics of war and peace since liable agents should bear all the costs. More precisely, one might think that a fair distribution of the costs has those who are liable bear all of the costs. Their wrongful action means that they should bear the costs. In reply (and as argued above), with all of the various options, there are likely to be some who must bear costs for which they are not liable. Thus, although an option that places all the burdens on liable agents should be chosen first, this is not currently feasible. With all the options, there will still be some currently unavoidable harm to those not liable. How should we distribute these harms — the remainder?
I have already suggested that we should look at the option that will be the lesser evil. I have also suggested that, even when sanctions are not the lesser evil and the overall amount of harm may be the same (or unknowable), sanctions are preferable to the alternatives since they are likely to involve more-widespread imposition of smaller, proportionate costs, rather than disproportionate costs on the few. I will now present a further consideration in favor of sanctions: fairness in the distribution of costs to innocents. That is to say, in addition to concerns about minimizing overall harm and the infliction of proportionate harms, fairness in the distribution of costs to innocents should also influence which policy is favored.
In what follows, I will consider three leading ways of distributing fairly the remainder, that is, of the unavoidable harms to innocents of aggression and mass atrocities. My aim is not to defend one of these ways over the others in general; I think that all are plausible (all are widely cited in the relevant literatures as ways of distributing costs to innocents). My focus is on the relevance of these ways of assigning the harms of tackling mass atrocities and unjust aggression and their relevance for the case for and against sanctions vis-à-vis the alternatives. I will suggest that some — but not all — favor economic sanctions.
A. Consent
The first way of distributing costs is to assign them to those innocents who consent to them. Indeed, consent is sometimes held to be relevant in the context of the permissibility of placing the burdens of sanctions on individuals. For instance, Lori Damrosch argues that, when the objective is the empowerment of the civilian population, the international community should defer to the views of authentic judgments of leaders about the degree of hardship that the population is willing to suffer in a sanctions policy. Footnote 33 If the leaders judge that the population is willing to suffer the burdens of sanctions, they can be permissibly assigned these burdens.
Consent to burdens is relevant in general. It matters that, for instance, in the choice of whether to distribute unavoidable costs to Anna (who gives her informed, express, and free consent) or to Brenda (who does not), we should distribute the costs to Anna. Anna’s preference to take on the costs should be respected by others. However, consent of the innocent is not typically relevant in the context of economic sanctions. This is because, in practice, those in the sanctioned state do not typically give their informed, express, and free consent to the sanctions regime. Footnote 34
George Lopez replies that there need not be express consent: there would be hypothetical consent, he argues, since we can reasonably expect that repressed populations would consent to take on the burdens of sanctions. Footnote 35 In his words, “[w]e can morally argue for sanctions on some countries because history supports the contention that repressed people will consent to such.” Footnote 36 However, this reply falters. Even if there were express consent to the policy, it is doubtful whether innocents who consent should have the burdens distributed to them. It is important here to consider to what the individuals are (or would be) consenting. In the case of sanctions, it is consent to the sanctions regime, compared to not having the sanctions regime and continuing in the prevailing situation (e.g., state aggression). Yet, given the sorts of situations that sanctions typically address — such as state repression — the consent of individuals is not fully free. They are faced with a Hobson’s choice: (1) a potentially very burdensome sanctions regime or (2) the suffering involved in the status quo. As such, their express or hypothetical consent lacks a reasonable alternative. It therefore does not provide a strong enough reason to burden them. Although when in the position of having to choose between two egregious harms, they may choose the lesser harm (or can be expected to), this does not mean that they should be the ones to bear the burdens of the situation. There may be further options or policies that would not burden them at all (or burden them far less), such as a peace operation, and to which they would be likely to give their express, tacit, or hypothetical consent, instead of the two much more burdensome alternatives.
On the contrary, it may appear that the import of distributing harms to innocents who consent to them provides a major reason in favor of war instead of economic sanctions. This is because when they sign up, soldiers consent to the burdens of tackling aggression and mass atrocities as part of their military covenant whereby they agree to take on risks. Yet, there are three worries with this suggestion. First, and most obvious, many soldiers are conscripted and so do not sign up to take on these burdens. Second, even if soldiers do consent, it is questionable whether this consent meets the standards required for free consent. Volunteer soldiers may have been subjected to significant socioeconomic pressures that mean that their consent has been coerced. Moreover, it is unclear that consent given by soldiers is adequately informed; recruitment officers can manipulate and mislead potential recruits. For example, the U.S. Army has allegedly targeted children in a way that resembles “predatory grooming” and lied to student recruits. Footnote 37 Third, noncombatants (who do not clearly consent to burdens) typically make up the majority of casualties in wars and so war is still likely to impose significant costs on those who do not consent to them. Consent, therefore, does not seem to provide a reason to favor war either.
B. Benefiting
A second way of distributing costs is to look to those who are seemingly benefited by the action. Jeff McMahan defends a form of this principle when considering the just distribution of harms between combatants and noncombatants in cases of humanitarian intervention. He argues that, other things being equal, those who are the subjects of humanitarian intervention should bear greater costs than bystanders or just intervening soldiers, since they benefit from intervention. Footnote 38 Similarly, it might seem that this provides a reason to favor sanctions since those in the sanctioned political community will be benefited by the removal of their authoritarian leader or the halting of the mass atrocities. By contrast, the costs of war will also be significant for those who do not clearly benefit (i.e., the intervener). The costs of doing nothing will be borne by those facing the current situation (e.g., those facing mass atrocities) rather than those who benefit from doing nothing (e.g., bystander states not fulfilling their duties to assist).
Of course, it is questionable whether sanctions do in fact improve the situation of most of those in the political community. Yet, even if they do, being benefited in this way is not typically morally relevant, at least in the context of economic sanctions. The problem is that those being “benefited” might still be significantly disadvantaged. The sanctioned party is often in a very bad situation to start with, such as being under an authoritarian leader or the threat of aggression by another state. Although their situation may improve, it will have often started at a very low level. Those who have not been “benefited” by the sanctions regime may often be far better off. Consider, for instance, the case of those who have in fact had treatment to tackle a genetic disorder. They may seem to have been benefited by the treatment, but they would have been in a much worse situation to start with. Their treatment only brings them back up to the level of equality. They should not be the ones to pay for all of their treatment. In fact, the converse may be true: if the treatment is itself burdensome (e.g., requires chemotherapy), it may be that others should bear the burdens of paying for the treatment since those “benefited” have already had the significant misfortune of having to bear the costs of having a genetic disorder. Similarly, if we hold that many of those who must bear the burdens of mass atrocities or an authoritarian leader are not morally responsible for this situation, then, given that these individuals have already had to bear the costs of the mass atrocities or their authoritarian leader, others should bear the costs of tackling the situation.
To be sure, being benefited by a potential policy is morally relevant to the distribution of costs when the starting point is equality of opportunity or when those who are benefited are those who are already in an unequal, better position. In such cases, when those who are benefited bear the costs, this may decrease the unequal effects of their being benefited. However, it does not seem that benefiting will generally be relevant to the context of economic sanctions since the sanctioned party will typically be relatively badly off, compared to other states. For instance, at the time of writing, states subject to financial sanctions by the United Kingdom include Belarus, the Democratic Republic of Congo, Egypt, Eritrea, Guinea, Guinea-Bissau, Iran, Ivory Coast, Liberia, North Korea, Somalia, Sudan, Syria, Tunisia, and Zimbabwe.
That said, although being benefited generally does not seem to be relevant to sanctions, there are some instances when it may be. First, being benefited may be morally relevant when considering the internal costs to the economies of rich sanctioning states that benefit from, for instance, the removal of an aggressive regime in a neighboring state. Second, benefiting may also be relevant for smart sanctions. As noted above, measures such as asset freezes and restrictions on travel potentially harm innocents, since those directly targeted may be in fact innocent and, even if not, their families will be affected. Nevertheless, smart sanctions may still be preferable because those burdened, even if innocent, are typically those (i.e., the elite) who do very well out of the current system. This is providing that the measures are efficacious and proportionate to the degree to which the affected individuals are benefited by their elite position.
C. Sharing the Costs
A third way of distributing the remainder is the equal sharing of reasonable and unreasonable costs among nonliable agents. (By costs, I mean the currently inevitable harms of the situation). Why should reasonable costs be shared equally between nonliable agents?39 Most obviously, costs should be shared so that no one must bear unreasonable costs, that is, costs that they are not duty-bound to bear (nor liable to bear). By sharing costs, the amount that each has to bear will be lowered and, so, potentially each must bear only the costs that agents can be reasonably asked to bear in their general duty to tackle aggression and mass atrocities. In addition, by sharing even otherwise reasonable costs (i.e., the costs that they are not liable to bear because of their previous wrongdoing but can be reasonably asked to bear in the fulfillment of their duties), the overall burden that one individual has to bear will be lower. This may mean that their lives are much less likely to be affected by taking on the currently inevitable (but reasonable) costs of the situation. If one individual is required to bear all of the (otherwise reasonable) costs of the situation, he or she will be potentially disadvantaged compared to others, who do not have to pay these costs. To see this, suppose that, in order to save a child drowning in a pond, it would cost an individual one thousand dollars, which would be within the parameters of a reasonable cost for the individual to bear. If others cannot share in these costs, the individual can be reasonably asked to pay the entire amount. But now suppose that it is possible to share the costs between a thousand agents so that each pays one dollar. The costs should be shared, even though it would be otherwise reasonable to require of the one individual to pay one thousand dollars. This is because, by sharing in the costs, the effect on one particular agent is massively reduced; the particular agent is not disadvantaged relative to others by, for instance, his or her random proximity to the pond. The brute bad luck of being in this situation is, in effect, shared across all parties, so that the agent is not unluckily a thousand dollars worse off. For instance, assume that bearing the thousand dollar cost means that the agent would have to go to a less prestigious university compared to the others, with the result that his or her future opportunities, although still good, are not as good as those who do not have to pay. By everyone paying a dollar, everyone’s future opportunities, relative to each other, are unaffected.
There may be cases where the overall burden of the situation is such that some innocents will inevitably have to bear unreasonable costs (i.e., costs that they are not liable to bear because of previous wrongdoing and cannot be reasonably asked to bear in the fulfillment of their duties). Like reasonable costs, unreasonable costs should also be shared in order to reduce the burdens on particular individuals. For instance, suppose that, to defeat an aggressor, it is necessary to resort to the compulsory acquisition of private property. It is better that the unreasonable burdens of the compulsory acquisition (such as the amount seized) be shared, as far as possible (such as by increasing the number of those who are subject to the compulsory acquisition of private property and by reducing the amount seized from each individual). The reasoning is the same: spreading the costs will mean that the lives of those impacted are less negatively affected by taking on the currently inevitable costs of the awful situation. If one individual is required to bear all of the unreasonable costs of the situation, that individual will potentially be disadvantaged compared to others.
Sanctions are likely to be preferable to both war and intervention and doing nothing since they better spread harm, other things being equal (i.e., the aggregate harm being of equal magnitude). More precisely, sanctions are more likely to spread the costs of the policy so that no one is asked to bear more than reasonable costs, or the degree of unreasonable costs that nonliable individuals have to bear is lower.
First, broader sanctions are preferable partly because they seem to share the internal costs of the response to the situation, that is, the costs that the party responding to the situation bears. Footnote 40 In the cases of military intervention and war, these costs are often borne almost fully by soldiers. By contrast, as far as the sanctioning state goes, the harm of banning trade in certain goods with another state, for instance, would often not seem to be unreasonable for those in the sanctioning state. And if the burdens do seem to be unreasonable for a particular sector of society (e.g., those who work in the arms industry), they could be compensated, and the costs of compensation shared amongst those in the society. This is obviously much harder for soldiers who will suffer significant injury or even death tackling the situation. Footnote 41
Broader sanctions are also likely to better spread costs externally, that is, among those in the target state and those in third parties. (It is important to reiterate that this is other things being equal: when the overall amount of harm of war or sanctions is the same (i.e., not when wars will be less costly aggregately than sanctions).) As noted above, wars often involve the concentration of harms, with particular individuals and groups having to bear much greater costs (often physical injury or death) than others do. Similarly, doing nothing will typically lead to a concentration of harms, such as the persecution of those from certain ethnic, religious or socioeconomic groups. By contrast, again as noted above, the more diffuse nature of sanctions, particularly broader sanctions — given that they involve general economic burdens — means that they are more likely to (although will not always) spread the costs among the whole society. Of course, those working in particular parts in the economy (e.g., in the extraction of natural resources) may be more affected but, again, compensation may also be more easily realizable than wars or doing nothing in such scenarios, and the costs of such compensation spread.
It is worth linking this argument to the related argument in Section II. I argued above that sanctions should be favored because they are likely to impose more but smaller costs that are proportionate to the liability of many. My point here is that, even when these costs will be imposed on those who are not liable, or these costs will be disproportionate to the amount to which the burdened are liable, the fact that sanctions seem likely to impose more but smaller costs on innocents provides reason to favor them.
IV. The Liability Objection Revisited and the Harm-Distribution Approach
Against my defense of the import of sanctions for reasons of fairness, it might be argued that we should not redistribute harms to those who are not liable to these harms since this involves doing harm. The standard response to cases in mass atrocities and aggression should therefore be to fail to react, since at least then the state is not doing the harm itself, even if it allows a much greater harm.
This view implies that the distinction between doing and allowing has overwhelming weight, which seems counterintuitive and out of kilter with the mainstream view in moral philosophy that the doing and allowing distinction is not of absolute moral import. That is, it can still sometimes be permissible to do harm, in particular when countervailing considerations in favor of doing harm (e.g., achieving very beneficial consequences) outweigh the import of the distinction. Notwithstanding, a less extreme version of the objection runs as follows: it is permissible to do harm to nonliable agents only in exceptional cases, where there are hugely beneficial consequences at stake (e.g., the tackling of one hundred thousand impending deaths by doing harm to ten innocents). In other cases, where doing harm will be likely to achieve only very beneficial consequences (e.g., the tackling of a thousand impending deaths by doing harm to ten innocents), it is better to allow harm.
This less extreme version of the objection is also implausible. This is because, first, the distinction between doing and allowing does, I think, matter, but it can be outweighed not simply by hugely beneficial consequences, but also by very beneficial consequences, other things being equal. In other words, a significant decrease (and not only by huge degrees) in the overall amount of harm can mean that considerations of doing harm are outweighed. In the case above, it seems more important to tackle the thousand impending deaths, even if this involves doing harm to ten innocents, other things being equal. This is simply because achieving very beneficial consequences in terms of the overall reduction in unjust harm is of huge moral significance. Footnote 42
Second, in addition to decreasing the overall amount of harm, securing a fairer distribution of harm may also outweigh the import of the difference between doing and allowing. That is to say, on occasion it may be justifiable for an agent to bring about a fairer distribution of harms even if, when doing so, it does some harm itself (but still achieves the same overall amount of harm). For instance, suppose that ten thousand nonliable citizens in the south of Gabon are being repressed by the Gabonese government. France launches a successful sanctions policy that halts this repression but, in doing so, harms ten thousand nonliable citizens in the north of Gabon. Unlike those in the south, those in the north clearly and freely consent to such harms. Footnote 43 Such a redistribution — the doing of harm to those in the north — is permissible because it is a fairer distribution, even though the overall amount of harm to nonliable agents is the same in both cases.
This is relevant for thinking more broadly about the ethics of war and peace, and in particular lesser-evil justifications. Most admit that lesser-evil justifications can play a role in the justification of the imposition of harms on those who are not liable in certain cases. Nevertheless, the bar for lesser-evil justifications is typically held to be very high, largely because they involve doing harm. Footnote 44 They are often implied (1) to involve only a straightforward consequentialist judgment (for example, the minimization of overall harm) and (2) to apply only when there is an alternative, terrible catastrophe at stake (for example, supreme emergency). Footnote 45 By contrast, it seems to me that the bar for lesser-evil justifications to apply should be lowered. This is because, first, reducing overall harm is, I think, intuitively more weighty than often thought and, so, reduces some of the concern about doing harm and, second, the worry about doing harm is reduced further by countervailing considerations of fairness. That is to say, if lesser-evil justifications consider not only the amount of harm, but also the fairness of the distribution of harm of the various options, they may be much more likely to be relevant, since there need not be extreme bads that will be achieved for lesser-evil justifications to apply. There need be only a significant decrease in unjust harm and a better distribution of harm.
In fact, rather than focusing on whether it is permissible to launch a response to mass atrocities and aggression that will do harm to nonliable agents, I think it often would be preferable to focus much more on matters of fairness in distribution (as well as on decreasing the overall aggregate harm). That is, when there is bound to be a remainder to be distributed, we should focus much more on how to distribute this harm as fairly as possible (and to decrease its overall size), rather than the fact that the agent will impose harms on those who are not liable. To that extent, we should adopt what I will call a “Harm-Distribution Approach.”
The Harm-Distribution Approach holds that rather than viewing cases, such as the imposition of sanctions, as primarily about the permissibility of launching a particular action, we should instead see them more as matters of distribution and, in particular, the distribution of currently inevitable burdens. On this approach, the central issue is this: How should the currently inevitable burdens of the situation (for instance, state repression or mass atrocities) be distributed to be fair and to minimize the overall harm? And which policy best enacts this distribution? Would sanctions be best or would an alternative be better? This differs from the framing of such issues in terms of the permissibility of launching a response (for instance, “Is it permissible to undertake sanctions in order to avoid the likely harms of the current situation?”). Such framings are widely used in the discussions of the ethics of war, humanitarian intervention, and self-defense. For instance, one of the most frequently discussed issues in the literature on the ethics of humanitarian intervention is whether it is permissible to undertake military intervention in order to tackle mass atrocities.
What the Harm-Distribution Approach takes seriously, then, is the current inevitability of someone facing burdens for which they are not liable and the need to distribute the remainder as fairly as possible (and to decrease the overall amount of harm). As we have seen, even doing nothing will mean that the burdens will necessarily be distributed to a certain group or individual. To that extent, the Harm-Distribution Approach can be seen as often appropriate for matters of Nonideal Theory since it is concerned with the distribution of the burdens of noncompliance with moral principles on the part of other agents. (It may also be relevant for Ideal Theory, where there is an unavoidable harm, for example, a natural disaster or an accident such as an innocent man falling down a well on another innocent person who is stuck down there.) Given mass atrocities, state aggression, and so on, which involve certain actors creating burdens that someone has to bear, who should bear them?46
Many discussions in the morality of self-defense at least implicitly consider the distribution of unavoidable harms (for instance, trolley problems where it is impossible not to do or allow harm to some innocents). My point is that we should adopt this distributive approach more widely and more explicitly. This is in order to emphasize the issue of fairness in the distribution of harms and in order not to overemphasize the import of avoiding doing harm. Adopting the Harm-Distribution Approach would bring the ethics of war and peace more in line with other fields in international political philosophy, such as the ethics of climate change. For instance, on the literature on the burdens of responding to climate change (adapting or mitigating), there is much less of a concern that those deciding how to allocate the costs of climate change might do harm in that they will impose some costs on nonliable agents. The focus is much more on the fact that (1) there are bound to be costs that some nonliable agents will have to bear (because the huge costs of tackling climate change are greater than that which could be borne by liable agents since, for instance, the previous polluters have died or were reasonably unaware of climate change) and, given (1), we need (2) to distribute the remainder — the costs of tackling climate change — as fairly as possible. In response to this issue, there is a vibrant debate on how the costs should be distributed. Footnote 47 The issue is not framed as “Is it permissible for State A to take on the burdens of tackling climate change and, if so, why?” but rather “How should the inevitable burdens of tackling climate change be distributed?”
Similarly, in cases of aggression and mass atrocities — the sorts of situations that war and sanctions respond to — we should be concerned more with how to deal with the remainder by focusing on how to minimize and to distribute fairly the overall harms of aggression and mass atrocities, rather than on whether it is permissible to launch a policy that may do harm to nonliable agents. These burdens are currently unavoidable or inevitable, to the extent that these issues are likely to continue to arise for the foreseeable future while there is widespread noncompliance with the demands of Ideal Theory.
Could the import of fairness in the distribution of harms to nonliable agents be captured by the standard permissibility frameworks? In short, yes. One could simply add a condition of fair distribution to the standard list of Just War conditions. Footnote 48 Or, one could reject a narrow account of discrimination, that is, an account that is too focused on (individual) responsibility. One could instead have an account of discrimination that does not appeal simply to responsibility, but rather includes a variety of considerations, including distributive ones. Indeed, in its most fundamental sense, discrimination requires us to discriminate between those to whom the costs should be distributed and those to whom they should not.
The worry, though, with these solutions is that the import of fairness will be missed, as the additional distributive condition is overlooked or a standard account of discrimination is adopted. This is because this concern can involve complex moral judgments, requiring a detailed account of distributive considerations. The assessment of the permissibility of launching a response would, then, become much more involved, and the issue of distribution often overlooked. By contrast, the concerns of this principle will be taken into account if the Harm-Distribution Approach is adopted, given its focus on how we should distribute costs (such as between civilians and soldiers) in response to currently inevitable harms. Thus, the Harm-Distribution Approach is preferable since it highlights an important consideration not easily captured and potentially missed by standard framings.
V. Conclusion
I have argued in defense of sanctions and, in doing so, rejected much of the Prevailing View. First, I argued (i), against the Liability Objection, that sanctions may still sometimes be permissible as a lesser evil and that there is reason to view them as likely to be preferable in this regard, (ii) that the Intending-the-Harm Objection is implausible, and (iii) that the apparent instrumentalization of sanctions is of little moral concern. Second, I argued that sanctions are potentially better than the alternatives because they more fairly distribute costs. In doing so, I have suggested that the Harm-Distribution Approach is a useful approach when thinking about responses to currently inevitable harms. It is worth reiterating that my defense of sanctions is contingent since some of the objections may apply on occasion, and tentative since I have not considered the instrumental case for and against sanctions, which is important for an overall assessment. Indeed, nothing that I have said is meant to offer a blanket defense of sanctions; sometimes they may be impermissible. Notwithstanding, sanctions may be less morally bad than is often claimed.
