Humanitarian Intervention and the Responsibility to Protect by James Pattison
Review of Pattison, J. (2010) Humanitarian Intervention and the Responsibility to Protect: Who Should Intervene?, Oxford: Oxford University Press, pp. 296.
Review by David Miller
There is much to admire in Pattison’s book. It gives us a systematic analysis of the problem of humanitarian intervention, moving easily from philosophical first principles to practical discussion of the logistics of intervention. Such breadth of approach seems to me essential if we are to gain a proper perspective on this most controversial of issues in international politics today. Although my task here is to explore the position Pattison develops critically, let me first single out three areas in which I found myself wholeheartedly agreeing with what he says. One is his excellent discussion of the motives and intentions of would-be humanitarian interveners in chapter 6, where he effectively demolishes the view that a justified intervention must be one that is undertaken for the right (i.e. purely humanitarian) motive. In the same chapter, he also dismisses the frequently-voiced opinion that the selectivity of intervention – the fact that it may occur in one case, but then not in others that are equally or more serious from a moral point of view – fatally damages it as an idea. Finally, I found his discussion of whether a doctrine of humanitarian intervention could be legally codified in such a way that an international body could pronounce on which agent or agents should intervene in any given set of circumstances subtle and persuasive. Without dismissing possible reforms of international law that would allow this to happen, he is properly sceptical of their significance. A strong note of political realism is sounded throughout the book, notwithstanding the fact that its main aim is to defend, on normative grounds, both the right and the duty of states and other agents to engage in such intervention.
Pattison presents this as a theory of legitimate humanitarian intervention. Given that a humanitarian disaster is unfolding, involving the large-scale violation of human rights, say at the hands of a brutal government or an insurgent ethnic group, who may legitimately intervene using coercive force to stops further rights violations and restore social order? In his answer to this question, legitimacy is treated as a scalar concept: different potential interveners may be more or less legitimate (pp. 32-4). In particular, Pattison draws a distinction between agents who are adequately legitimate, and therefore have the right to intervene, and the agent who is most legitimate, and therefore has the duty to intervene. (Agents here may be states, coalitions of states, international bodies such as NATO, etc). The main factor in determining an agent’s degree of legitimacy is their likely effectiveness in bringing the disaster to an end, although as we shall see ‘effectiveness’ turns out to be a more complicated matter than it might at first appear.
We may wonder whether legitimacy is the best concept to do the work that Pattison wants done, which is primarily to single out which among the parties who might undertake an intervention is the one that should. Usually it is treated as categorical rather than scalar: either it was legitimate for agent A to perform action x or it was not. However there are some contexts in which a scalar interpretation does make sense. For example, we can speak of a newly formed government increasing its legitimacy by virtue of the way it behaves over time. So let us grant Pattison’s assumption that a humanitarian intervener’s legitimacy can be a matter of degree. Does it help us in establishing either the right or the duty of intervention?
Consider the right to intervene first. This requires the intervening agent to achieve a threshold level of legitimacy, which according to Pattison is a function of several factors, including having sufficiently military and non-military resources to avert the humanitarian crisis, being sufficiently ‘representative’ of both the population under threat and the people in whose name the intervention is undertaken (I will explore this condition further below), conforming to jus in bello requirements, and so forth. Here, it seems to me, he is in danger of setting the bar too high. Consider a case where a small, weakly armed, state borders a larger one in which an ethnic minority faces a genocidal threat. The small state cannot avert the disaster by itself, but what it can do, since no other agent appears to be taking any action, is to establish an enclave close to its own borders in which some members of the threatened minority can seek refuge. Why wouldn’t this state have the right to intervene, even though it may seem not to satisfy most of Pattison’s conditions for ‘adequate legitimacy’? Reflection on this case suggests a simpler test: an agent has the right to intervene in a humanitarian crisis when it has a good prospect of making things better (by human rights standards) and little prospect of making them worse; and when its intervention does not pre-empt some better-equipped party’s taking action.
Consider next the duty to intervene. Pattison claims that the most legitimate potential intervener has such a duty, which is in principle legally enforceable. I am doubtful whether there could be such a duty, although there will be cases in which such an agent has a strong reason to intervene. Three factors combine to defeat the proposed duty to intervene. First, humanitarian intervention by its very nature is a response to the rights-violating activities of other parties, who will continue to violate rights on a large scale if the intervention does not take place. This differentiates it both from natural disasters, and from cases where the rights-violators have left the scene, leaving others to pick up the pieces. It is important to emphasize, then, that A’s alleged duty is not the general duty to protect human rights, but the specific duty to stop B killing, maiming, ethnically, cleansing, etc C, and it is not clear that there can be such a duty, especially, to introduce the second factor, when performing it would be relatively costly. Humanitarian intervention, as Pattison emphasizes, is costly in material terms, since it requires large scale deployment of military and non-military resources if it is to be effective. It also, however, involves potential costs to the reputation and security of the agent who undertakes it, particularly if that agent is a single state. Pattison suggests, in his Conclusion, that successful intervention can enhance a state’s reputation as a good global citizen, but it is equally important to bear in mind that intervention almost always involves taking sides in a fight between a regime and its opponents, or between rival ethnic groups. It is virtually certain that the target of the intervention will have sympathisers and supporters abroad, including states with material interests at stake (consider China’s oil interests in Sudan, and its opposition to humanitarian action in Darfur). Thus the intervener may face retaliation, in the worst case in the form of terror attacks against its own citizens.
The third factor, applying specifically to democratic states (the most likely candidates for legitimate intervention according to Pattison’s criteria) is that the state owes its own citizens a duty of protection, which includes the duty not to expose them unnecessarily to external risk. Pattison accepts that legitimate humanitarian interventions cannot, in general, be undertaken by conscript forces (p. 110). He argues, however, that in the case of professional soldiers, their contracts of employment include their being required to undertake humanitarian missions. Even if this is so, it is not clear that initial consent overrides the state’s duty to treat all of its citizens, including those in uniform, with equal concern and respect, as I have argued elsewhere. So although Pattison accepts that the alleged duty to intervene is subject to a cost limitation, simply setting the human costs of intervention against its benefits – lives lost against lives saved – may not be the right way of thinking about the issue (pp. 77-8). Rather the state’s duty to its citizens should be seen as primary, and its acknowledgement of a responsibility to protect outsiders must be subject to that constraint. The language of responsibility, which implies that each political community must judge for itself when the costs of an intervention may reasonably be imposed on its own members, is therefore preferable to the language of duty.
We could nevertheless recast Pattison’s main question in the form ‘which agent bears the primary responsibility to intervene?’ without altering the substance of his answer. As noted above, his focus is on comparative ‘effectiveness’, but this is broken down into three categories ‘local external’, ‘global external’ and ‘internal’ (pp. 74-9). In each case the relevant currency is the net gain to human rights that will result from an intervention, but the focus falls respectively on human rights in the crisis area, human rights in the wider world (including possible knock-on effects in the form of intervention encouraged or deterred elsewhere), and human rights in the intervening state (allowing Pattison to include costs in the form of soldiers’ deaths and injuries into the calculus, as mentioned above). Overall, however, the main idea it to assess potential interveners by looking at the likely consequences of their intervention. But Pattison modifies this in various ways (he calls his position ‘Moderate Instrumentalism’). One issue is how those who will be the beneficiaries of an intervention by agent A view that prospect: what if intervention by the US carries the greatest chance of success, but the population to be rescued views all Americans as imperialists? Pattison introduces the idea of ‘local external legitimacy’ to deal with this problem, but expresses it rather confusingly. It depends, he says, ‘on whether an intervener is morally justifiable for those in the political community that is subject to its intervention’. But morally justifiable according to which standards? Consequentialist standards or standards that reflect local beliefs about who is entitled to intervene in their society? If the latter, then we have indeed driven a wedge between ‘effectiveness’ and ‘legitimacy’.
Other wedges appear when Pattison imposes jus in bello requirements on interveners, since these are partly deontological in form and therefore cannot be included in a simple human rights calculus. So we begin to see a pluralist answer to the responsibility question emerging, where deciding which potential intervener bears primary responsibility in a given case will involve weighing various factors alongside effectiveness; how legitimate will the intervention be by the standards of the receiving community, how likely is it that the intervening force will adhere to jus in bello, is the intervention ‘internally representative’, meaning is it acceptable to the citizens of the intervening state? Two other factors that we might think relevant to the question are, however, given short shrift. One is whether a potential intervener bears some historic responsibility for the crisis itself. Pattison thinks that this will often be indeterminate (p. 193), but it is hard to believe that it is more indeterminate than the issue of future effectiveness. The past may need interpretation, but at least the record is open to inspection, whereas when we are trying to judge who will be most effective in bringing the human rights violations to an end, we are at best in the realm of informed conjecture. Pattison also sets aside bonds of cultural similarity as a relevant factor. But these will often be closely related to the issue of ‘representativeness’. When referring briefly to intervention in a Muslim state (p. 193), he appears to overlook the resistance many Muslims feel to the military presence of non-Muslims in territory regarded as ‘Muslim lands’. This does not of course single out any one Islamic state as the ‘most legitimate’ intervener, but it certainly narrows the field.
A pluralist theory of responsibility is open to the objection that the relevant factors may be weighed differently by different agents, and therefore no one agent emerges consensually as having the strongest reason to intervene. This might seem to point towards creating a single agent whose role is precisely to carry the primary responsibility for humanitarian interventions. Pattison explores this in chapter 8, asking what would be required for an effective UN military force capable of engaging in several interventions at once. He concedes, however that this could only happen under the auspices of ‘cosmopolitan democratic institutions’ and that these are unlikely to emerge in the foreseeable future (pp. 234-6). One might say something stronger here, namely that a world in which cosmopolitan institutions were democratically legitimate would also be a world in which occasions for humanitarian intervention were rare indeed, since it would be a world made up almost entirely of democratic states. The problem Pattison is grappling with in the rest of the book is our problem, and to that problem effectiveness alone, important though that is, cannot provide the whole solution.
 D. Miller, ‘The Responsibility to Protect Human Rights’ in L. Meyer (ed.) Legitimacy, Justice and Public International Law (Cambridge: Cambridge University Press, 2009).
Review by Graham Long
James Pattison should be congratulated on writing a comprehensive, coherent and rigorous account of the rights and wrongs of humanitarian intervention. Pattison sets out to answer the question, “who possesses the right and duty to intervene?” Pattison’s answer is that, in order to possess the right to undertake intervention, the appropriate actor must be legitimate. The duty to intervene falls on the most legitimate actor, subject to a set of further concerns about burden-sharing. Legitimacy as a political concept, especially in the liberal tradition, has been taken as being centrally about consent. However, Pattison’s conception is much wider and less specific. Legitimate power is “morally justifiable power”. Pattison’s central idea, that only morally justifiable interventions can be permissible and obligatory is surely correct, and he does valuable work in applying his concept of legitimacy to a range of contemporary actors and examples.
On Pattison’s account, an intervention’s moral justifiability is largely but not wholly determined by its effectiveness (in a sense I discuss shortly). This claim forms the basis of Pattison’s “Moderate Instrumental” approach, which is a central component of his account. But in addition to these avowedly consequential considerations, there are relevant “non-consequentialist values” (185) We assess (1) the intervener’s adherence to ius in bello restrictions, and (2) their representativeness, both internally and externally. Ultimately, on Pattison’s approach, this range of consequentialist and non-consequentialist assessments are commensurable, and can be summed and assessed against a threshold for justifiable intervention. As Pattison describes his account, our assessment is “a matter of degree” across all these respects. Thus, in unusual circumstances, effectiveness alone “can be sufficient for an adequate degree of legitimacy” (183).
Pattison’s moderate instrumental approach is consequentialist: it judges interventions to be morally justifiable or not on the basis of an assessment of likely consequences. However, it is a ‘rights-consequentialism’, as I term it here: that is to say, the consequences are measured (as far as I can tell, solely) in their effects on human rights – “whether an intervener is likely to promote or harm the enjoyment of human rights (and primarily the basic rights)” (182). I agree with Pattison that an account of human rights should be central to our moral assessment of humanitarian intervention, but this feature also prompts my line of criticism here. Since the non-consequential aspects of Pattison’s assessment can themselves itself be put in terms of the promotion or harm of important rights, it is unclear what this separation aims to achieve, and how successfully it achieves it.
I think this is plainest in the case of ius in bello. Pattison maintains that adherence to ius in bello is non-consequentially important. This might be true, but still, the importance of adherence to ius in bello can be captured by a rights-consequentialism. Importantly, our account of ius in bello norms is often spelt out in terms of human rights. Ius in bello’s restriction on targeting civilians is founded in, or expressed by, the idea that this is a rights-violation. Unlike combatants, bystanders have done nothing to lose their relevant human rights, and so are relevantly innocent.
To illustrate the implications of this, consider Alf and Betty. Alf is an innocent, whose life is unjustifiably threatened by his regime. Betty is an innocent, whose life is unjustifiably threatened by the intervener’s military action to save Alf. The promotion of, or harm to, Alf’s right, on Pattison’s account, is a matter for his central rights-consequentialist calculus. It is necessary that intervention prevents harm to Alf in order for intervention to be justified: this is Pattison’s measure of effectiveness. The right of Betty, however, is (a) a non-necessary matter – an intervention can be justified even where Betty’s right is not upheld – and (b) a non-instrumental consideration – it matters in its own right. This, to my mind, is a puzzling position. It seems rather arbitrary to treat one right within a calculus of effective rights-fulfilment, and the other outside such a calculus. Why, in the logic of Pattison’s account, is every human rights violation not assessed in the same terms? Certainly, it is not clear why the effect on Betty’s right is not a matter of effectiveness central to our assessment of the intervention (as expressed by the necessity of effectiveness). Suppose a second external power could intervene to prevent Betty’s unjust death as a result of the first intervention to save Alf. For this putative second intervener, upholding Betty’s right is necessary: a central issue of effectiveness. It is unclear why Betty’s right should matter in such different ways to the two interveners.
A deeper consideration of the content of Pattison’s ius in bello rules – his position that ius in bello for interventions should be separate from, and more restrictive than, “traditional” ius in bello – seems to lends support to my contention above, though it might also be problematic on its own terms. His argument for a distinctive ius in bello seems grounded in two considerations. The first is that an intervener could not kill civilians indiscriminately yet be called humanitarian – we demand a “consistency of means and ends” (107). This consistency requirement suggests that killing civilians in bello is not so different from the kind of killing of civilians that prompts intervention. Pattison’s second consideration is that humanitarian operations tend to be more limited in scope than wars. Neither of these, however, seem the kind of substantive difference that should generate separate rules of ius in bello particular to this kind of conflict. The second is purely contingent: it is easy to imagine a more limited kind of war and a less limited kind of intervention – perhaps the current intervention over Libya lies in this territory. More fundamentally, since the just cause for most just wars could – or must – be expressed in terms of rights-violations, it is not clear that there is a conceptual distinction here to underpin these two considerations. I am not convinced that the intervener who kills civilians is more morally inconsistent than the just war-fighter who does so. And even if this is so, it is unclear whether internal inconsistency should make any principled difference to the rules of war.
I wonder if Pattison’s second non-instrumental concern, that of representativeness, can again be folded into a rights-consequentialism. Pattison argues that intervention is a sufficiently special issue of policy that it requires extra efforts by a government to attain the express consent of its populace, and in this he may well be right. Pattison’s thought is that individual self-government seems to possess non-consequentialist value. But it is precisely this principle or value that animates or finds expression in the idea of a right to take part in the government of one’s country. Pattison seems to overlook this link. In considering an example of putative intervention by the Vietnamese government, he writes “The views of the Vietnamese people matter… not because taking account of those views will best serve international law and order, nor because doing so is the best for the Vietnamese people’s enjoyment of human rights, but because it is their state” (Pattison, 139). But, surely, the Vietnamese people have a human right to have their views heard – as for example, laid out in the UN charter. The importance of self-government justifies enshrining this as a right, or else self-government is important because it is a right. Either way, these considerations appear inseparable – that Vietnam is “their state” is the content of the human right here, precisely the idea that the right protects. As with the case of Betty, this rights-violation could in principle form the basis for a suitably ‘light-touch’ intervention by a third party (though Pattison is reluctant to consider violations of less basic rights as grounds for intervention).
What I am suggesting, then, is that the work done by keeping some rights-based considerations outside of the central rights-based effectiveness calculation is unclear. This is not to say that no account of this work could be supplied. It might be that these non-instrumental principles represent an attempt to ensure special consideration for the rights of particular groups and the duties correlative to such rights. Or it might be that the keeping of some rights-questions separate and immune to consequential considerations is intended to honour something in the nature of rights themselves: it is sometimes thought, for example, that rights embody the “separateness of persons”. However, the presence of a larger, scalar, assessment of these components seems to undermine this sense of specialness. Recall that on Pattison’s account, it is possible for an intervention to be morally justified by virtue of its effectiveness, even whilst it is highly imperfect in these non-instrumental respects. Thus, it is not clear what kinds of trade-offs between rights Pattison’s structure actually prevents.
The purest kind of consequentialism appears simple in term of its calculus. For example, it might be thought that nothing hinges on how the good consequences were brought about or distributed (except insofar as this is instrumental in maximising the good consequences): however, measuring the consequences in terms of rights changes this, since notions of rights carry within them complex moral content: they track precisely these things. Rights are entitlements and trumps; they can be upheld individually or institutionally, waived or suspended, justifiably infringed or violated, and some are more important than others. This is helpful for Pattison’s argument: human rights need only prompt intervention where they have not been waived or forfeited. However, I want to suggest that Pattison’s account does not fully recognise the duties that accompany human rights.
Pattison’s account begins with a defence of a general duty to intervene to prevent rights-violations. However, the later focus on how this duty should be allocated seems at odds with the – surely plausible – idea that that anything less than the fulfilment of our duty is a moral wrong. If I understand Pattison’s account correctly, it sanctions as legitimate and permissible interventions which fall far short of our duties in respect of human rights. Now, it would be no objection, I take it, that interventions might unavoidably kill or fail to save civilians (either in bello, or regarding the targets of the intervention). Rights-fulfilment, after all, admits of tragic shortfalls and tradeoffs. But on Pattison’s account, any actor who can prevent (some sufficient number of?) rights-violations may well meet this standard of moral justifiability, even whilst falling short of what duty requires. Pattison writes “it will often be the case that several of the current interveners will possess an adequate degree of legitimacy, since they will make an improvement in the enjoyment of the human rights of those suffering the crisis” (210, my italics). But merely making an improvement is not enough to make their action morally permissible, when it falls short of what is owed to the victims, and there is another alternative scenario in which the duty is fulfilled. For Pattison, just as a legitimate intervener can massacre a number of civilians and still possess sufficient scalar legitimacy, an agent can have a right to intervene, by virtue of possessing adequate moral justification, even though it fails to costlessly rescue several parties – perhaps, as long as the intervention “makes an improvement” overall. If a number of individuals each have a human right that generates a corresponding duty of intervention, it cannot be permissible to fulfil only some of these duties and not others where fulfilment of all these duties is co-possible and not unduly costly. The permissibility of such an intervention gives the intervener the false sense that they have done enough. Of course, it is better that something be done than nothing be done, but this generates only a remedial account of duties to undertake imperfect interventions where perfect interventions are not possible, rather than the guaranteed justifiability of failing in what we owe the victims. Pattison is right that the international community must build the capacity to be able to fulfil this duty – as he says, capacity-building and reform is a duty in itself. But perhaps the objection here is to the use of bald effectiveness to specify the permissibility of intervention: I am suggesting that rather than effectiveness being understood in terms of rights promotion, it should be assessed in terms of duty-fulfilment. This might mean, for example, that the threshold for the moral permissibility of intervention should map on to the human lives that must be saved, or the efforts that must be made, if we are to fulfil our duty, rather than the requirement of merely “making an improvement”.
Pattison, then, gives us a strong and plausible account of the duty of intervention and how it might be allocated. But in the ways I have discussed, perhaps my objection is that he fails to take rights – and their corresponding duties – quite seriously enough as the currency of his consequentialist account. This need not be an issue for Pattison alone so much as a challenge for rights-consequentialism generally: It may be that as a general matter of morality, individual human rights and duties sit deeply at odds with the idea of a consequentialist calculus.
The Rights and Duties of Humanitarian Intervention: A Reply to Two Critics by James Pattison
Graham Long and David Miller present careful analyses of my book, Humanitarian Intervention and the Responsibility to Protect: Who Should Intervene? In doing so, they offer some thoughtful criticisms of my Moderate Instrumentalist Approach. In what follows, I will defend this approach against their objections and highlight points where I think that it is in fact in agreement with their analyses.
Rights-Consequentialism and the Moderate Instrumentalist Approach
On the Moderate Instrumentalist Approach, the enjoyment of human rights is the measure of whether an intervener is effective. Effectiveness is the primary factor in the legitimacy of an intervener, but there are also morally important, noninstrumental factors, most notably (i) fidelity to the principles of jus in bello and (ii) (internal and local external) representativeness. Long suggests that these two noninstrumentalist factors could also be framed in terms of human rights. If they could, then all of the main moral values of the Moderate Instrumentalist Approach may be seen in human rights terms. Accordingly, Long is worried that is unclear what sorts of rights trade-offs this approach would sanction. For, I argue that, unlike absolutism, the Moderate Instrumentalist Approach does not prohibit trade-offs and, unlike consequentialism, it does not sanction all trade-offs when overall rights compliance would be increased. But if all the values that I am concerned with can be framed in terms of human rights, how can I make this move and avoid one of the two extremes of absolutism or pure rights consequentialism?
To answer this, it helps to consider in turn the two central types of noninstrumentalist value that I defend. First, the central reason why the fidelity to the principles of jus in bello does not collapse into either absolutism or consequentialism is the difference between ‘doing’ and ‘allowing’ (see pp. 115–17). This difference largely grounds the import of the fidelity to the principles of jus in bello and, in particular, the principle of noncombatant immunity. That is, it is important that an intervener avoids doing harm itself (e.g., by killing innocent civilians), even if it were to allow more harm in the long run (perhaps by being less effective). The difference between doing and allowing is not absolute, however. It can sometimes be permissible to do harm oneself rather than allowing a greater harm—and, in particular, when extremely beneficial consequences (in terms of the mass violation of basic human rights) are likely. This nonabsolute difference between doing and allowing goes some way to explain why I can adopt a midway approach that sanctions some, but not all, trade-offs between rights. To put it in terms of rights, it is sometimes better that an intervener does not violate rights to X, even if this will allow more rights violations of X overall. But when there would otherwise be a much larger number of rights violations of X, it is preferable that the intervener does infringe the rights to X.
The importance of representativeness might seem to be more easily captured by a purely rights-consequentialist approach. On such an approach, the right to individual self-government (which, to some extent, underlies the value of representativeness) is one right to be maximised, along with the right to life, the right not to be tortured, and so on. (The values captured by the ‘Resources Argument’ and the ‘Burdens Argument’ would also need to be put in terms of rights.)
However, this may have little practical import. My aim in the defence of the noninstrumental importance of representativeness is to show that the value of individual self-government (as well as the values captured by the Resources and Burdens Arguments) should be taken into account in the morality of intervention (representativeness has often been overlooked). Even if this value is conceived of in terms of rights and included in the overall rights calculus, it will still play the same role in the morality of humanitarian intervention. That is, any potential intervener would need to take into account the rights of individual self-government, in addition to the effects on other rights, such as the right to life. Suppose, for instance, intervention by France in Togo (in response to a hypothetical crisis) would not be legitimate because it would violate a much larger number of the rights to self-government of the French and Togolese citizens than the number of lives that it would save. If we include the value of representativeness within the rights calculus, France would not be effective at promoting the overall enjoyment of human rights and so would be likely to be illegitimate. Likewise, if we exclude the value of representativeness from the rights calculus, France would not be representative and so would be likely to be illegitimate.
Jus in Bello and Humanitarian Intervention
Long questions my claim that the principles of jus in bello should be more restrictive than the traditional accounts of these principles. He suggests that one of the arguments that I present for this view—that humanitarian intervention should be more limited in scope than wars—is purely contingent because wars can also be limited in scope. More generally, he suggests that since just wars are ultimately justified in terms of rights, there is little to demarcate them from humanitarian intervention. Both should follow the stricter account of the principles of jus in bello that I endorse.
I agree with Long’s argument that just wars may ultimately be justified in terms of rights and, as such, the case for the stricter principles of jus in bello would apply to them too. Indeed, I suggest (p. 125 n. 10) that the reason that I offer for the increased stringency and importance of jus in bello could apply to other uses of force, especially those that occupy enemy territory. But, although I agree with Long, this view of just war is contentious. Many maintain that war can be justified in response to values that are not necessarily reducible to individual rights, such as the threat to communal integrity. If one subscribes to such a view, then there are differences between the rights focus of humanitarian intervention and other just wars. My aim is simply to make the case for strict fidelity to the principles of jus in bello for humanitarian intervention in particular, in contrast to those who maintain that wars can be justified for reasons not reducible to rights and, consequently, assert less strict principles of jus in bello for such wars.
The Duty to Intervene
Long suggests that rather than measuring humanitarian intervention in terms of the promotion of the enjoyment of human rights, it should be measured by duty fulfilment. This is because, on the Moderate Instrumentalist Approach, interveners that fall short in their fulfilment of their duties can nevertheless be viewed as acting permissibly if they make an improvement in the overall enjoyment of basic human rights. What worries Long is that, on my approach, interveners seem to be viewed as acting justifiably, even if they fail to do what they should.
I agree that such interveners are morally required to do more. Indeed, on the Moderate Instrumentalist Approach, an intervener is fully legitimate only if it meets all the requirements of this approach. So, for instance, (hypothetical) intervention by NATO in Uzbekistan that fails to consider the opinions of the Uzbeks would be morally lacking. Nevertheless, I also want to maintain that interveners that fall short of meeting all the requirements of the Moderate Instrumentalist Approach (which may not always be their fault) can, if they meet enough of them to pass the threshold of legitimacy I identify, still act permissibly in one sense at least, that is, permissibly compared to not intervening. (I call this ‘an adequate degree of legitimacy’.) We should of course prefer to have an intervener that meets all of the requirements of the Moderate Instrumentalist Approach—that is why they are not fully legitimate. The fact that they do not do so is morally problematic.
However, I do not think that Long’s suggestion of measuring the effectiveness of humanitarian intervention in terms of duty compliance would be beneficial. There is a degree of consensus (although, of course, not full agreement) surrounding human rights, which is most apparent in the political endorsements of the rights of the Universal Declaration of Human Rights and the defence of this list of rights by human rights theorists. As such, there is likely to be a degree of shared understanding about how to measure whether an intervener will or will not be effective in terms of human rights. The duties that interveners possess, however, are more contested. Many deny that there is a duty to intervene. Since, on this view, there are no relevant duties to be fulfilled by interveners, if effectiveness is measured by duty fulfilment, there could be no measure of effectiveness. Furthermore, even if the duty to intervene is accepted, there are likely to be differences about what is required for duties to be fulfilled and so how to measure effectiveness. This is because amongst those who accept that there exists a duty to intervene, there are widely differing views on its strength and on what the duty entails.
By contrast, Miller is worried that, by requiring an intervener to possess several factors in order to have an adequate degree of legitimacy, I set the bar too high for the right to intervene. He suggests instead looking to whether an agent has a good prospect of making things better in terms of human rights and little prospect of making them worse (and when it does not pre-empt a more effective intervention by another intervener). Like the ‘Extreme Instrumentalist Approach’ that I consider (pp. 93–4), the problem with such a test is that it disregards the noninstrumental importance of fidelity to the principles of jus in bello and representativeness that are also significant for an intervener’s legitimacy and for who has the right to intervene.
Miller also doubts whether exists a duty to intervene. First, he argues that humanitarian intervention, by its very nature, is a response to rights-violating activities and it is unclear whether there exist duties on third parties to halt the violation of others’ human rights. Second, he claims that the duty to intervene is unclear when intervention would be costly. Third, he argues that the state’s duty to its citizens should be seen as primary.
However, these three considerations do not undermine the case for the existence of a duty to intervene. First, there is no conceptual or principled reason to insist that ‘humanitarian intervention’ can occur only in response to the active violations of human rights. On the contrary, humanitarian intervention can be undertaken—both conceptually and justifiably—in response to situations where individuals’ human rights are going unfulfilled. It seems to me that a state whose population are starving or the subject of a natural disaster could be subject to ‘humanitarian intervention’, if their state is unwilling or unable to act, even if it is not originally responsible for the plight of its people. Indeed, there was a sizable public debate about whether there should be humanitarian intervention in response to the failure of the Burmese Junta to allow assistance to its people in the wake of Cyclone Nargis. (To be sure, intervention in response to a natural disaster would be unlikely to be included under the ‘responsibility to protect’ umbrella as this doctrine has developed since the UN World Summit in 2005). As such, even if one holds that third parties do not have duties to halt violations of others’ human rights, it might still be the case that there exist duties to respond to situations where individuals’ basic human rights are going unmet, such as in natural disasters.
Moreover, I think that third parties do have duties to halt the violations of others’ human rights. When this is of little cost, the existence of this duty seems patent. Suppose that Alan uses a knife to attack Bob, who is innocent. Without the risk of harm to himself, Chris could disarm Alan (say that Chris has a gun). It seems to me that Chris has a duty to disarm Alan, given that he can prevent an egregious violation of Alan’s rights without risk to himself. Just as we should save a child drowning in a pond when the costs are insignificant, so too we should save a child being drowned in a pond when the costs are insignificant. Of course, the existence of this duty becomes more complex when there are potentially significant costs involved. If Alan also has a gun, we may think differently about whether Chris has a duty to intervene. To that extent, I agree with Miller when I argue (p. 17) that when it is excessively costly to potential interveners, humanitarian intervention is not a duty and, on the contrary, is impermissible.
I also agree with Miller that the state’s duty to its citizens should be seen as primary. Indeed, I accept (pp. 17–19, 132–3) that states have fiduciary obligations to promote their citizens’ interests. However, we can admit this point and still assert that humanitarian intervention is a duty. First, humanitarian intervention may be of very little cost and of much benefit to the state, so does not contravene fiduciary obligations. Second, although a state’s duty to its citizens may be primary, this is not to say that it should occupy itself exclusively with the interests of its citizens. When fiduciary obligations are seem as primary, rather than exclusive, a state can still have certain obligations to those beyond its borders, such as to provide humanitarian aid and, on occasion, to intervene militarily in response to a serious humanitarian crisis. But when intervention will be costly or the crisis is less serious, the fiduciary obligations that states owe to their citizens may outweigh the duty to intervene.
Assigning the Duties of Intervention
Miller argues that I am too quick in my rejections of the importance of historical responsibility for the crisis and communal bonds in the assessment of the alternative ways of assigning the duties of intervention (pp. 191–4). These alternatives, he suggests, may be determined on occasions (e.g., Muslims may prefer intervention by an Islamic state) and potentially more easily determined than effectiveness. Miller may be correct about this. The potential indeterminateness of historical responsibility and communal bonds might not provide a strong reason to favour effectiveness as the central way of assigning the duty to intervene. But nor is the potential indeterminateness of effectiveness a strong reason to prefer the other alternatives. This is because there can be a stronger sense of an intervener’s likely effectiveness than is often assumed. A series of potential factors (e.g., sufficient military and nonmilitary resources, a suitable strategy, and international support) are likely to influence significantly the success of any intervention. The absence or presence of these factors will provide us with a good sense of the likely effectiveness of a particular intervener, in addition to looking at past cases and assessing the circumstances in the political community subject to the intervention.
Beyond the issue of determinateness, there is a strong reason for rejecting historical responsibility and communal bonds as alternative ways of assigning the duties of intervention. When using the potentially blunt tool of military force to end the mass violation of basic human rights, what matters most is that the intervention tackles this violation. In the context of large-scale human suffering, an intervener’s making up for its past misdeeds or its assisting those with which it has bonds seems to be far less important. Thus, when deciding who should intervene, we should primarily look to the intervener that will be effective at tackling the mass violation of basic human rights.
 See, for instance, James Nickel (2007) Making Sense of Human Rights, Second Edition (London: Blackwell).