Coalitions of the Willing and Responsibilities to Protect: Informal Associations, Enhanced Capacities, and Shared Moral Burdens

Abstract

“Coalition of the willing” is a phrase that we hear invoked with frequency in world politics. Significantly, it is generally accompanied by claims to moral responsibility. Yet the label commonly used to connote a temporary, purpose-driven, self-selected collection of states sits uneasily alongside these assertions of moral responsibility.

This article explores how the informal nature of such associations should inform judgments of moral responsibility. I begin by briefly recounting what I call a model of institutional moral agency in order to explain why it seems theoretically and practically problematic to talk about the moral responsibilities of informal associations. I then focus on coalitions of the willing as prominent, and challenging, examples of such associations, before raising misgivings about my own rather stark distinction if it means that accounts of moral responsibility must be reduced to the members—or potential members—of such coalitions in a way that neglects the moral significance of their acting together. Prompted by these concerns, I explore arguments by Virginia Held and Larry May about moral responsibility in relation to informal associations and identify insights that can be taken from these positions to refine our expectations and evaluations of the actions associated with such collectivities. Finally, I consider the particular implications of these insights in relation to the widely espoused duty to intervene to rescue vulnerable populations.

“Coalition of the willing” is a phrase that we hear invoked with frequency—and often urgency—in world politics. Significantly, it is generally accompanied by claims to moral responsibility. (Such appeals recently bolstered calls to establish a coalition of the willing to intervene in Syria.) Yet the label commonly used to connote a temporary, purpose-driven, self-selected collection of states (and sometimes nonstate and intergovernmental actors) sits uneasily alongside these assertions of moral responsibility. This unease might be attributed to its association with a particular case. For some, the label was tainted when the United States led a “coalition of the willing” into Iraq in ostensible response to the threat of weapons of mass destruction: the actual willingness of all of those states initially announced as members is as contested as the legitimacy of the 2003 offensive, the ensuing protracted conflict, and the subsequent occupation.¹ Nevertheless, the idea—and ideal—of a coalition of the willing has persisted beyond the infamy of that one iteration. The problem is, rather, that it is unclear how to speak coherently about assigning moral responsibilities—and apportioning blame—in relation to such ad hoc associations. Can coalitions of the willing be considered bearers of duties? Alternatively, should our calls to action—and our cries of condemnation in the wake of action that is stalled, ineffective, or deemed unjust—be directed toward their constituents? Or should our attention be redirected altogether, toward more formal, enduring and, arguably, legitimate organizations?

These are neglected points of inquiry, and they might valuably be examined in relation to a subject that has received an enormous amount of attention recently among both policymakers and scholars: military intervention in response to specific humanitarian crises as one means of discharging the proposed “responsibility to protect” (RtoP).² Indeed, it is in this context that coalitions of the willing have been variously invoked and ignored, summoned with the language of moral imperatives and dismissed as illegitimate.

There has been widespread support for the idea that the so-called international community has a remedial moral responsibility to protect vulnerable populations from grave human right violations when their own governments fail to do so, and that this protection may, when necessary, include military intervention. But where exactly is this responsibility located? In other words, which body, or bodies, can be expected to discharge a duty to safeguard those who lack the protection of—or, indeed, come under threat from—their own government? The question is particularly pressing when the United Nations is unwilling or unable to act and there is no one state to fill the breach—no “agent of last resort,” to invoke Michael Walzer's phrase (along with all of the controversy and potential risks that he acknowledges reliance on such a protector entails).³ This article examines coalitions of the willing as one (likely provocative) answer to this question, and explores how the informal nature of such associations should inform the judgments of moral responsibility that we make in relation to them.

I will begin by briefly recounting what I call a model of institutional moral agency in order to explain why it is both conceptually coherent and necessary in practice to maintain that moral responsibilities can be borne by formal organizations (such as states, multinational corporations, and intergovernmental organizations), but why it seems theoretically and practically problematic to talk about the moral responsibilities of informal associations.⁴ I will then focus on coalitions of the willing as prominent, and challenging, examples of the latter category, before raising misgivings about my own rather stark distinction if it means that accounts of moral responsibility must be reduced to the members—or potential members—of such coalitions in a way that neglects the moral significance of their acting together. Prompted by these concerns, I will explore important arguments by Virginia Held and Larry May about moral responsibility in relation to informal associations, and will identify insights that can be taken from these positions to refine our expectations and evaluations of the actions associated with such collectivities in world politics. Finally, I will consider the particular implications of these insights for addressing how the widely espoused duty to intervene to rescue vulnerable populations can be understood in relation to coalitions of the willing—and, perhaps most controversially, propose that, under certain circumstances, states and other entities have a duty to form such ad hoc associations.

A Model of Institutional Moral Agency⁵

Claims to moral responsibility are ubiquitous in world politics. Such statements make use of two different but closely related understandings of responsibility. Prospective moral responsibility involves ex ante judgments regarding acts that ought to be performed, or forbearances that must be observed. Retrospective moral responsibility entails ex post facto assessments of a particular event or set of circumstances for which an agent's acts or omissions were such that the agent is the object of praise or blame. The forward-looking variation is heard in assertions of duty and obligation;⁶ the backward-looking variation emerges most often in charges of blame and accountability. Statements by two UN secretaries-general, both contemplating the consequences of inaction in the face of widespread human rights violations, provide respective examples of each understanding. “The international community has a moral responsibility,” Ban Ki-moon recently declared, “to stop the bloodbath and find peace for the people of Syria.”⁷ Apportioning moral responsibility for approximately 800,000 deaths in 1994, Kofi Annan lamented that “the international community” is “guilty of sins of omission” in the context of the Rwanda genocide.⁸ These statements exemplify the judgments of moral responsibility so commonly voiced in world politics. However, they also highlight the worrying ambiguity that often accompanies them—and is apparent in key articulations of RtoP—by suggesting that duties might somehow be assigned, and blame apportioned, to the international community—an entity that would seem incapable of unified, purposive action in the first place.

Importantly, to be intelligible, judgments of moral responsibility must be directed toward entities capable of responding to ethical imperatives. In other words, they must be directed toward moral agents, or those bodies that possess capacities to contemplate, recognize the significance of, and ultimately execute different courses of action. Overlooking those bodies that qualify as moral agents in world politics when considering how best to respond to crises, or mistakenly assuming that moral responsibilities can be borne by those bodies that do not qualify, can have detrimental consequences. Such missteps hamper attempts to consider, coordinate, and execute remedial action effectively and robustly. They also result in missed opportunities to determine what went wrong when crises are neither prevented nor mitigated. Avoiding such shortcomings is particularly pressing in just those cases that prompt calls to protect vulnerable populations—when inaction can have far-reaching and tragic consequences, and failing to learn from past mistakes deprives us of the knowledge to make meaningful pledges of “never again.”

Yet identifying moral agents in world politics is far from straightforward. Most philosophers understand moral agency exclusively in terms of individual human beings. Such an assumption seems unnecessarily limiting. In practice, it severely restricts assessments of moral responsibility in the context of some of the most pressing problems in world politics: when the relevant actions or outcomes cannot possibly be (or could not possibly have been) performed or produced by individual human beings acting individually. In theory, such an assumption flies in the face of the observation that the norms, rules, procedures, practices, and cultures of formal organizations frame and channel the decisions and actions of those individual human beings within them—with the consequence that these organizations can reach decisions and act in ways not adequately described in terms of the sum of decisions and actions of their constituents. Duty and blame would seem to be most accurately attached, in some circumstances, to the organizations themselves.

Given these objections to the idea that individual human beings exhaust the class of moral agent—along with the point that bodies such as financial institutions, states, and the United Nations, for example, are often described as bearers of duties and appropriate objects of blame in practical discourses in world politics—it makes sense to consider the possibility that such collectivities also qualify. However, this potential has, until recently, been neglected in the discipline of International Relations (IR).⁹ In fact, the moral agency of collectivities is often precluded by the field's narrow conception and application of agency.¹⁰ Fortunately, the potential moral agency of collectivities has received sustained attention elsewhere.

Focusing primarily on business firms, Peter French challenges what he identifies as an “anthropocentric bias” in our moral reasoning and aims to illustrate that corporations can be “moral persons.” Borrowing terminology from geology, he distinguishes between what he calls an “aggregate collectivity” (“merely a collection of people”) and a “conglomerate collectivity” (“an organization of individuals such that its identity is not exhausted by the conjunction of the identities of the persons in the organization”).¹¹ He concludes that the latter are “full-fledged members of the moral community, of equal standing with the traditionally-acknowledged residents: human beings.”¹² Highlighting parallels between the capacities of individual human agents and those of institutions, Onora O'Neill argues that institutions can also be agents for whom ethical reasoning is both accessible and action-guiding.¹³

Adding to and elaborating upon French's account of “conglomerate collectivities,” and inspired by O'Neill's “thin theory of institutional agency,” I have proposed that a collectivity qualifies as a moral agent if it possesses five characteristics: (1) an identity that is more than the sum of the identities of its constitutive parts, or what might be called a “corporate identity”; (2) a decision-making structure that can commit the group to a policy or course of action that is different from the individual positions of some (or all) of its members;¹⁴ (3) mechanisms by which group decisions can be translated into actions (thereby establishing, with the previous characteristic, a capacity for purposive action); (4) an identity over time; and (5) a conception of itself as a unit (meaning simply that it cannot be merely externally defined).¹⁵ I refer to collectivities that have these qualifying features as “institutional moral agents.” They can be subject to the assignment of duties and the apportioning of blame in the context of particular acts and omissions in a way that is not reducible to their individual constituents—as long as they enjoy the (limited) independence from other agents and structural constraints necessary to perform the requisite actions.¹⁶ To avoid misunderstanding, this proposed model in no way precludes or undermines the moral agency of those individual human actors or subgroups that constitute the institutional moral agent. Rather, moral agency exists simultaneously at different levels, and moral agents at all levels can be responsible for concurrent, complementary, or even coordinated acts and omissions.

In light of this model, one might ask which bodies in world politics would be able to respond to the proposed moral imperative to protect vulnerable populations. Although each case warrants detailed examination beyond the scope of this article, one might argue that most states and many intergovernmental organizations (including the United Nations and, perhaps, some regional alliances of states such as NATO and the Arab League) possess the sophisticated, integrated capacities for deliberation and action that allow them to qualify as institutional moral agents.¹⁷ As moral agents, they could reasonably be expected to discharge such a duty in accordance with established moral guidelines and in the context of enabling conditions.¹⁸ They could also, then, be blamed for the acts or omissions that derogate from this duty.

Coalitions of the Willing as a “Hard Case”

Coalitions of the willing, however, belong to a broad category of unlikely candidates for institutional moral agency: informal associations, or those collectivities that lack formal organizational structures and decision-making procedures. Informal associations are prominent in both the scholarly domain of IR and the practical world of international politics. “International society,” “epistemic communities,” “transnational advocacy networks,” and “communities of practice” are all informal associations that are notable, and fruitful, objects of analysis in IR.¹⁹ Terrorist networks, protest movements, and coalitions of the willing—as well as the more amorphous collectivities known respectively as “the international community,” “the (global) market,” “the media,” and “the Internet”—are examples of informal associations regularly invoked in practical discourses and debates.²⁰ Each would prima facie struggle to be considered a duty-bearer in its own right, and this matters to how we interpret, judge, and respond to the assertions of moral responsibility made in relation to them. Of this diverse range of examples, coalitions of the willing are a particularly interesting case because they have been regularly ushered into (and occasionally conspicuously excluded from) recent prominent discussions of moral responsibility in world politics, including discussions of RtoP. They also provide an exceptionally challenging case because they do not straighforwardly fail to meet every criterion for institutional moral agency, but, rather, demand careful consideration of the degree to which they might satisfy at least some.

Coalitions of the willing are common phenomena in world politics. The label is most often used for associations that are summoned and established in cases of military intervention—with or without UN authorization, and frequently on proposed humanitarian grounds—but is also applied in the context of single-issue campaigns involving norm promotion.²¹ The term has achieved currency relatively recently. It was reportedly used for the first time in 1990–1991, when a U.S.-led, UN-authorized coalition of the willing responded to Iraq's invasion of Kuwait,²² and was employed again in the late 1990s in the context of the campaign to prohibit anti-personnel landmines.²³ A coalition of the willing established around NATO, which included some non-NATO states, intervened in Kosovo in 1999 on humanitarian grounds without UN Security Council backing. In the same year, an Australia-led, UN-endorsed multinational coalition of the willing intervened in East Timor. The 2003 U.S.-led group of states that launched a “preventive” war against Iraq—one that was also subsequently, and rather improbably, justified on humanitarian grounds—has arguably been the most prominent, and infamous, example of such a coalition. In 2011 a multinational coalition of the willing led by the United States took military action to protect civilians in Libya (followed shortly after by NATO control of the military effort) in response to UN Security Council Resolution 1973. Most recently, calls have been made to establish a coalition of the willing to advance the climate change agenda,²⁴ and, in the absence of Security Council authorization, to put a halt to the violence in Syria.²⁵

Despite the differences among these various examples, an important set of shared features ties them together. Coalitions of the willing are self-selected (and often self-authorized) constellations of states and sometimes nonstate and intergovernmental actors (including, for example, nongovernmental organizations and regional alliances of states, respectively) that come together to respond to a specific crisis and, in responding, act outside the control of any formal, overarching organization to which they might already belong. The members of a coalition of the willing are thereby temporarily united in pursuit of a common purpose, but the coalition itself lacks an established organizational and decision-making structure. For those coalitions of the willing that either convene or are implored to materialize in order to engage in military intervention on humanitarian grounds, the issue of their authority to act invariably arises. The circumstances under which a coalition of the willing might intervene militarily on humanitarian grounds include both those in which the United Nations has effectively “subcontracted” such an ad hoc association to act (to use Thomas Weiss' term),²⁶ and those in which the members of a coalition of the willing put themselves forward as agents of intervention when the United Nations is unwilling or unable to act at all.²⁷ I will refer to the associations in each case as “subcontracted” and “vigilante” coalitions, respectively.²⁸ Questions of authority—including whether the UN should have a monopoly on authorizing intervention in cases of humanitarian crises—are, of course, important.²⁹ Yet, whether or not vigilante coalitions of the willing should have the legal authority to intervene in cases of mass atrocity and large-scale loss of life is distinct from a possibility that will be explored in this article: that both subcontracted and vigilante coalitions (or at least their constituent states and regional organizations) have a moral obligation to intervene in certain urgent circumstances.

Coalitions of the willing receive little attention as subjects for academic study in IR and have been ignored within philosophical discussions of collective responsibility (perhaps unsurprisingly, given the hitherto largely domestic focus of such analyses).³⁰ Moreover, despite their concrete presence in military interventions on humanitarian grounds, they are conspicuously absent from RtoP reports and policy documents. Significantly, the seminal 2001 report The Responsibility to Protect, drafted by the International Commission on Intervention and State Sovereignty (ICISS), acknowledges the potential necessity of vigilante coalitions (“ad hoc coalitions . . . acting without the approval of the Security Council”) engaging in military intervention for human protection purposes if the Security Council “fails to discharge . . . its responsibility in conscience-shocking situations crying out for action.”³¹ However, when the UN member states unanimously endorsed RtoP at the 2005 World Summit, it was made explicit that the responsibility to protect must be discharged exclusively through the Security Council (with, perhaps predictably, no mention of coalitions of the willing).³²

If we accept that there is a responsibility to protect vulnerable populations from serious human rights abuses, then considering who should, and indeed can, discharge this responsibility when the UN fails to act is fundamental. The coalition of the willing is an obvious candidate for consideration. The question of how we understand such coalitions—whether as entities that we can reasonably expect to bear moral burdens, or as ad hoc associations of individual duty-bearers—necessarily follows. In order to respond, it might prove useful to return to the criteria proposed above and interrogate the preliminary judgment that coalitions of the willing fail to qualify as institutional moral agents.

Are Coalitions of the Willing Institutional Moral Agents?

A coalition of the willing struggles to meet the first criterion for institutional moral agency: that it possess an identity that is more than the sum of the identities of its constitutive parts. Indeed, the importance that is often placed on the precise membership of such a coalition is indicative of this difficulty.³³ Moreover, a considerable obstacle to a consistent and convincing corporate identity is the coalition's failure to satisfy the second criterion. The absence of an established decision-making mechanism at the level of the coalition prevents the diverse perspectives, preferences, and policies of its members (along with their potentially very different motivations for association) from being channeled into a position that is more than an agglomeration of discrete individual stances. Any independent identity that the coalition as a whole might achieve therefore remains partial and precarious at best.

Deliberation does, of course, take place between the members of the coalition, and is necessary for the coordinated action that its constituents come together to achieve. Yet, rather than the formal decision-making of so-called structured institutions such as the United Nations, NATO, and (all but failed) states, which entails codified rules and established practices for arriving at policies, the coalition of the willing relies on what might be called informal decision-making.³⁴ This involves negotiation, bargaining, and consensus-building among the various constituents in the absence of existing organizational structures and decision-making procedures.³⁵ In the case of both subcontracted and vigilante coalitions assembled for military interventions, one member often takes the lead (whether a state such as the United States or a regional alliance such as NATO).³⁶ Other coalitions of the willing, most notably those committed to norm promotion, tend to be less hierarchical in negotiating courses of action.³⁷ Yet both cases display a further feature frequently associated with informal decision-making: the inability to translate calls for collective action into decisions that are binding upon the group as a whole if such proposals depart from the positions of some of its members.³⁸

In sum, informal decision-making establishes and holds together the loose association of agents (or, indeed, fails to do so), achieves some consensus on a general course of action, and coordinates individual contributions (often imperfectly), with the specifically military command-and-control function of coalitions in the former case commonly provided through a single chain within the lead organization.³⁹ Crucially, there is no overarching decision-making apparatus that can be said to both direct the actions and represent the intentions of the collectivity as a whole.⁴⁰ With respect to the third criterion—that the collectivity have mechanisms by which decisions can be translated into actions and policies can be implemented—although the individual members of the coalition can carry out their respective decisions in order to coordinate their actions, and can even choose to follow the instructions of a lead agent, the lack of a decision-making structure at the level of the coalition impedes corporate purposive action.

As an association that is by definition temporary, a coalition of the willing also stumbles at the fourth, identity-over-time criterion. A degree of continuity is necessary for it to make sense to talk about the moral expectations that we have of a collectivity performing certain actions at some point in the future. Logically, a coalition of the willing cannot be assigned duties prospectively when it does not exist prior to what is often taken to be the duty-generating event or set of circumstances that its constituents band together to confront. Moreover, it follows that, as a group, it cannot incur blame retrospectively for failing to discharge duties that it could not have been assigned in the first place.⁴¹ This judgment is crucial for many of the cases being addressed: that is, when an ad hoc coalition is called to action to respond to the gross violation of human rights, but no such entity yet exists. It also highlights one of the vital questions that this article seeks to address: to whom or to what can duties be assigned in such a case?

However, it might be wise to pause on this criterion and consider another set of circumstances. Coalitions of the willing enjoy some continuity once they have been established, even if this is unlikely to represent the persistence of a corporate identity. Some are surprisingly eclectic (the 1990–1991 Gulf War coalition, for example, comprised liberal democracies, traditional Arab states, and more radical Arab states, including Syria) and render an enduring association highly unlikely. Yet others are relatively homogeneous, with many of their members bound together by overlapping regional and international associations that both predate and outlive the particular coalition in question (such as the 1999 Kosovo coalition).⁴² In each case, once a coalition of the willing has been established, questions of prospective and retrospective moral responsibility arise in relation to the specific timeframe in which it exists.⁴³ The proposed responsibility to react, which the members come together to uphold, does not exhaust the range of possible duties to be discharged in responding militarily to protect vulnerable populations. Rather, other duties are also recognized in the context of that action: that military intervention be conducted discriminately and proportionately, for example. Questions remain of where these duties rest and where blame for their breach is to be apportioned. Convincing answers depend on the moral status of the coalition of the willing. Insight into this status is, in turn, revealed in the difficulty that such coalitions have in meeting the previous criteria—even if they can have a (limited) identity over time. Finally, the criterion that the collectivity be self-asserting is easily met by a coalition of the willing, given that its constituents are self-selected and explicitly committed to both a common goal and collective action.⁴⁴

Even acknowledging the diversity within the category of coalition of the willing, and understanding each criterion for institutional moral agency as being able to be satisfied by degree (rather than constituting an all-or-nothing test), coalitions of the willing do not possess the sophisticated, integrated capacities for deliberation and action that would allow them to qualify as institutional moral agents.⁴⁵ According to this analysis, coalitions of the willing are not moral agents; the moral responsibility to protect vulnerable populations must be borne elsewhere. Individual states and other institutional agents might have duties to intervene, and can be blamed for inaction (or acting disproportionately or indiscriminately). However, such prescriptions and evaluations must remain exclusively at the level of those agents that make up the temporary association. We cannot coherently talk about assigning duties or apportioning blame to the coalition itself.

A Gap in the Analysis? Moral Responsibilities and Informal Associations

This might seem a logical end-point to this analysis. Yet three related concerns remain, each of which cautions against treating coalitions of the willing as mere aggregates of individual agents acting independently for the purposes of addressing questions of moral responsibility. First, if moral responsibility must be attributed to individual states (and possibly nonstate and intergovernmental actors), rather than the ad hoc associations that they might come together to form, there will be some cases in which no state (or other actor) can be expected to discharge a particular duty (to protect a population from egregious human right violations, for example) simply because each, as a discrete agent, lacks either the capacities or enabling conditions to do so.⁴⁶ If the United Nations is also unable to respond, then we seem to have no choice but to concede that no agent can reasonably be burdened with a duty to act in response to the crisis. This is eminently unsatisfactory. Second, accounts of moral responsibility with respect to some informal associations, such as coalitions of the willing, seem to resist being reduced without remainder to their component parts. If their members can accomplish things acting in concert that they cannot achieve when acting individually, this must somehow affect our judgments of moral responsibility in relation to them, even though the coalition that they come together to form cannot be a moral agent in itself. Third, coalitions of the willing are not amorphous collectivities in the way that the international community is. They are not purposive actors in themselves, but they nevertheless seem to possess something that at least resembles a capacity for purposive action.⁴⁷ Meeting some of the criteria addressed above to a limited degree does not take the coalition of the willing far enough to qualify as an institutional moral agent. Nevertheless, its informal deliberative capacity, the continued (albeit temporary) association of its members, and its constitution by actors who have come together to participate in a common project are features that, combined, seem morally significant. Perhaps informal associations like coalitions of the willing should not be so easily dismissed in assessments of moral responsibility.

Philosophers such as Virginia Held and Larry May have focused on questions of moral responsibility in the context of just those types of collectivity that I have argued cannot bear moral responsibilities at the corporate level: associations that lack formal organizational structures and decision-making procedures. Insights from their work provide a more nuanced and demanding account of moral responsibility in relation to informal associations in world politics than what has been alluded to so far, and one that will complement the analysis of institutional moral agency already offered.

Held's “Random Collection”

The question that Held set out to answer in an oft-cited 1970 article is reflected in its title: “Can a Random Collection of Individuals be Morally Responsible?”⁴⁸ The types of groups that she wishes to hold up to scrutiny are those that contingently share a time and place, but lack any specific decision-making procedures. We might, she suggests, think of unacquainted passengers riding together on a train or pedestrians sharing a sidewalk as thereby belonging to this category. Neither would have specific methods for deciding to act. Indeed, the groups that she addresses fail to qualify as institutional moral agents on each and every criterion offered above. They also, I will suggest, stand in a potentially instructive relationship to the types of ad hoc coalition overlooked in recent articulations of RtoP.

Held makes two claims. First she argues that, in some circumstances, a random collection—or, more accurately, those that make up such a collection—may be held responsible for not acting.⁴⁹ The specific circumstances in which they may be held responsible are those in which the following conditions are met: (a) the individuals constituting the random collection were faced with a stranger in dire need of rescue and the gravity of the crisis demanded a response; (b) these individuals could have rescued the stranger by acting together, even though no one in the group could have done so acting independently; (c) the required action was “obvious” and its foreseeable outcome “clearly favorable”; and (d) carrying out this action would have been possible without prior deliberation or special coordination between the individuals and would not have been open to disagreement.⁵⁰ As my intention is to apply Held's argument at the international level, it is worth noting that the extent to which the fourth condition can be met when we are talking about discrete states, rather than individual human beings, is questionable. It is difficult to imagine a scenario in which collective action involving states would require neither prior deliberation nor special coordination.⁵¹ This condition would therefore prima facie shield individual states from any expectation that they engage in remedial collective action outside a preexisting organizational structure. An exemption of this sort would have potentially tragic consequences in those cases in which no formal intergovernmental organization were willing or able to act to discharge a moral responsibility, such as the proposed imperative to protect vulnerable populations. The duty would thereby go unmet without any agent having derogated from its moral responsibility to respond. However, before lamenting that states would be let off the moral hook given Held's analysis, this application to international relations might be valuably pursued—and the apparent leniency of this condition qualified—in the context of her subsequent assertion.

According to Held's second claim, even in cases in which a random collection cannot be held responsible for failing to perform an action, the individuals who constitute it may be held responsible for not “forming [themselves] into an organized group capable of deciding which action to take.”⁵² The specific example that she constructs to convey this argument is worth relating. Held describes three pedestrians, who are strangers to each other, walking down an isolated street when a small building collapses.

A man inside is trapped; he calls to the three for help. He is bleeding from a lower-leg injury and needs immediate assistance. All four persons know that a tourniquet should be applied to his thigh, but this cannot be done until various beams are removed, and removing any would require the strength of all three. The three observers do not agree on how to proceed . . . .⁵³

Each person makes a different proposal, they argue, do not act, and the man slowly bleeds to death. Held observes that any of their proposed actions would have saved the man, and a “reasonable person” should have known that any action would have been better than none; “the problem was deciding which to take.” She concludes that we cannot hold the members of the random collection responsible for the nonperformance of the action (given the fourth condition outlined above), but that they “can be held responsible for failing . . . to adopt a decision method” and thereby transforming themselves into the sort of group that might have been capable of the deliberation and coordination required for an effective response.⁵⁴

Intuitively, Held's judgment appears flawed in two respects—at least in the context of the domestic case that she illustrates. First, she seems too forgiving in concluding that the individuals are not responsible for failing to act when they are unable to reach an agreement (but are, rather, responsible for not forming themselves into the sort of group that would have been capable of acting), if, as she maintains, it is abundantly clear that any action would have been better than no action at all. Surely, then, each individual should have accepted the gravity of the situation, suppressed his or her own desire to orchestrate the rescue, and simply gone along with the others.⁵⁵ A more appropriate scenario for the individuals to have been spared responsibility would have been one in which it were not, in fact, clear that any action would have been better than none. Perhaps a false move of one beam, for example, would have conceivably resulted in the death of the trapped man—or indeed of all four individuals. Such a scenario would have provided good reason for not contributing immediately to any proposed action, and would have genuinely required a group of individuals capable of some degree of deliberation over the consequences of possible actions and able to coordinate themselves to execute the considered course. Second, if transformation into a group capable of this sort of deliberation had been necessary, it does not seem plausible to suggest that the individuals could have achieved this in the time they had to move the beams without the victim bleeding to death.

Yet, when an analogy is drawn with a possible set of circumstances at the international level, and states (or even regional organizations of states) represent the pedestrians, Held's judgment becomes more compelling in both respects. First, whether the crisis is a malfunctioning nuclear power plant (such as the one involved in the Fukushima Daiichi nuclear disaster of 2011) or the mass violation of human rights in another state (such as Rwanda in 1994 or Syria as of 2012), the situation is going to be infinitely more complex than simply moving three beams. Indeed, in such complex crises, it is not immediately obvious that just any action would be better than none. Some actions will worsen rather than improve the situation, or cause more overall harm than good. Second, in many (although not all) cases of international crisis, there is time for relevant agents to transform themselves from a random collection into a group capable of accomplishing objectives that would be beyond their reach in isolation or as part of a completely nonintegrated aggregate. (This potential is particularly important in the absence of a formal, permanent organization willing and able to respond, and given that either establishing or reforming such an organization would require a prohibitively longer time frame.) Genocidal intent, for example, is often clearly articulated before it is acted upon: weapons might be stockpiled, hatred incited, human beings rhetorically reduced to “cockroaches.”⁵⁶ In short, crises—or at least their escalation—may be anticipated, leaving time to establish some capacity for deliberation and special coordination among relevant agents.

Held's argument provides purchase in addressing certain, apparently intractable, cases in world politics in which the gravity and imminence of a crisis demands a response, and yet no agent seems able to act effectively on its own. Inspired by Held's position, one might propose that even if remedial action requires both a multilateral effort on the part of available agents (because of the limited capacities or constraining conditions faced by each) and prior deliberation and special coordination between them (because the requisite action is neither immediately obvious nor beyond disagreement), and even if these agents are not part of some preexisting organizational structure that would facilitate such cooperation, there nevertheless remains a moral imperative for them to do something. Namely, they each have a duty to contribute to establishing the type of collectivity capable of the requisite collective action. It is important to emphasize that Held actually suggests that members of a random collective be held responsible for failing to transform themselves into an “organized group”—comparable to those formal organizations that I argue qualify as institutional moral agents.⁵⁷ (For Held, an organized group has “a method for deciding to act: it has officials who can act in its name, or a voting procedure . . . or customary procedures to guide its actions.”)⁵⁸ Yet her argument and the particular time-sensitive, life-and-death examples that she invokes actually point toward an intermediary type of association that the members of her random collection have an obligation to form themselves into: an informal association, loosely organized and capable of some deliberation leading to coordinated action, but lacking the organizational or decision-making trappings of structured institutions.

Applying Held's argument to the international level—but with the crucial clarification that loose organizational structures may be sufficient for the deliberation and coordination needed to support some forms of collective action—leads to a compelling, and potentially controversial, suggestion. One might propose that individual states (and nonstate and intergovernmental actors) each have an obligation, in the absence of a viable alternative, to contribute to the establishment, and then functioning, of an informal temporary association for the purposes of responding to a particular crisis effectively, robustly, and in time to mitigate disaster. This seems especially persuasive in the worrying case where there is no formal organization that can be expected to discharge a duty to respond and no hope of reforming or creating one in the limited time needed for an effective response.⁵⁹

Of course, once such an ad hoc coalition has been established, we are talking about neither a “random collection of individuals” in Held's sense, nor what I have called an institutional moral agent. Rather, we are addressing what was formerly a random collection of individuals who have together transformed themselves into something between the two: that is, an informal association made up of actors who have come together in pursuit of a common goal. For consideration of such an intermediary category of collectivity, Larry May's work is invaluable.

May's “Middle Position”

May observes that the structures of informal associations (such as amateur sports teams and mobs) “enable their members to perform actions that they could not have performed on their own.”⁶⁰ Moreover, he argues that the “in-between status” of the mob in particular—“a status somewhere between a random collection of individuals and an organized group”—should prompt us to rethink the requisite features for coherently talking about a group acting.⁶¹ In constructing his argument, May focuses on the significance of relationships and, specifically, the notion of solidarity. He maintains that collective action and responsibility can be predicated on individuals in groups, insofar as one refers to both individual persons and the relations among them. He thereby adopts an alternative to both my position as outlined above (that some collectivities can be moral agents, and therefore the bearers of duties, in their own right) and the “individualist” insistence that accounts of group action and responsibility are always reducible to descriptions of the actions and responsibilities of their individual human constituents. May explains his “middle position” as follows:

My own view is that social groups should be analyzed as individuals in relationships. Groups themselves do not exist in their own right; but the individuals who compose groups also are often not understandable as acting in isolation from one another.⁶²

In distinguishing his position from that of Peter French, May argues that French is wrong both to assert that a “conglomerate” somehow entails a group that is ontologically independent of its individual members in relationships and to assume that an “aggregate” is straightforwardly reducible to its individual members.⁶³ While I agree with French that certain collectivities have identities that are more than the sum of their constitutive parts—and in a way that is morally significant—May's point that French's aggregates are not always reducible to descriptions of individual action is a compelling one. In other words, I adopt the more radical “minority position” (explicitly rejected by May) that certain types of collectivity can qualify as moral agents in their own right,⁶⁴ but nevertheless maintain that there is something in May's more moderate stance that is important for certain collectivities that neither meet the criteria for institutional moral agency outlined above nor can be adequately described in terms of Held's random collections. In short, May's argument is important for collectivities such as coalitions of the willing.

May reminds us that French actually, albeit briefly, distinguishes between two different types of aggregate collectivity: “the sort whose definition explicitly involves spatial/temporal contiguity” and “the sort defined in terms of a common characteristic or feature.”⁶⁵ (Neither has formal organizational structures or decision-making procedures.) Descriptions of the actions of both, French maintains, are reducible to the actions of their individual constituents, with consequences for questions of moral responsibility. May challenges this and suggests that only the former (what he calls “random aggregate groups”) actually supports French's position that they are reducible to the actions of their individual constituents.⁶⁶ As for the latter, May proposes that this “subclass of aggregates, those defined in terms of a common feature, is surely not one which can be understood as a summation of the parts of persons who compose it.” Offering mobs and teams as examples of this subclass, he maintains that “both groups are defined by reference to the solidarity which allows the members of the group to engage in joint purposive behaviour.”⁶⁷ This middle ground between groups that I have argued are capable of corporate purposive action and groups such as Held's “random collectives” that are reducible to the actions of their constituents promises to tell us something important about moral responsibility in world politics in general, and about the reasonable expectations that we might have of coalitions of the willing in particular.

Two points of clarification should, however, be offered about how I propose to use May's work in this respect. First, although I agree with May's important point that there is more than one type of aggregate collectivity, I disagree with his claim that individuals and relationships are all that there is. Some collectivities can achieve a status independent of their constituents. Second, while I am drawing on May's insight that individuals in informal associations act in a way that cannot be simply reduced to the actions and intentions of individuals acting in isolation, it is important to emphasize that by applying this insight to an analysis of coalitions of the willing I am not using May's work in a way that he would—or could, given the totality of his argument—endorse. This is because, going back to the previous point, I am assuming that the sorts of collectivities to which he denies independent ontological standing are actually the individuals in the informal associations being addressed. States, and possibly other institutional agents, are to my coalitions of the willing what individual human actors are to May's mobs and amateur sports teams. This is not a problem for me given my arguments surrounding institutional moral agency, but it would be a problem for May's very different “middle position.”

Importantly, May effectively introduces another category of collectivity: those entities that lack formal organizational structures, but that “can facilitate joint action or common interest” nonetheless. Within this category May includes both collectivities that have “informal organizational structures” and those “such as mobs” that “can attain unity even though they are not organized at all.”⁶⁸ Both types of collectivity, May argues, are distinct from those random collections of persons “which do not have the ability to engage in joint action.”⁶⁹ Although the latter variation (which May understands to be exemplified by mobs) is less useful to the discussion here, the former would seem logically to be the type of collectivity that the individuals within Held's random collection have an obligation to form themselves into under certain circumstances—at least in the short term, given the time constraints that she assumes.⁷⁰ Indeed, May's account of the often-overlooked characteristics of such informally organized collectivities provides the foundation for a potentially valuable supplement to analyses of moral responsibility made possible by the model of institutional moral agency outlined above.

The Moral Significance of Acting in Concert: Coalitions of the Willing and Responsibilities to Protect

The proposal that I have taken from Held's article is the following: If there is a duty to perform a particular action, and if individual agents can come together to perform this action when they could not have performed it individually, then they each have an obligation to contribute to establishing the type of group necessary for this duty to be discharged. The examples that she invokes in constructing this argument gesture toward a category of informally-organized associations whose members are nevertheless capable of some degree of deliberation and special coordination. May's work is valuable in giving substance to this category. From May, I have taken the point that social relationships and their participants' shared interests and desire to work toward a common goal can make possible “joint purposive action.”

It is important to clarify here that I hesitate to take this further and accept that group action and intention are possible in the case of informal associations; I do not think that solidarity can replace a formal organizational structure and decision-making procedures and thereby allow for purposive action in a way that is not ultimately distributive among the members of the association.⁷¹ The thought of somehow melding together the actions and intentions of these constituents seems unrealistic—and worrying, particularly if one then treats each constituent as somehow complicit in the ostensible actions of the informal association as a whole. Despite my defending the view that formal organizations that meet specific criteria are capable of purposive action, and can thereby coherently be understood to bear moral responsibilities at the corporate level, I do not think that the same is true for informal associations. When it comes to this category of collectivity, descriptions of actions and intentions, and corresponding ascriptions of moral agency and responsibility, are reducible to their individual constituents as long as one recognizes that such accounts are different in important ways from accounts of individuals acting independently or in isolation.⁷²

This difference is encompassed by what I take to be the defining feature of “joint action,” “shared activity,” or “collective action”: an outcome or effect that could not have resulted from the mere summation of individual acts.⁷³ Importantly, the necessarily unified, concerted endeavors of individual agents that characterize such action vary in both sophistication and the capacities necessary to perform them—a point that is particularly important if we are considering joint action at the international level. Some types of joint action—including that which Held suggests the members of her “random collectives” can be blamed for not performing given certain conditions—require only limited cooperation between individual agents. Others demand some degree of prior deliberation and special coordination. Following May (to the point that draws on his analysis of groups with informal organizational structures, but not as far as to embrace his analysis of those that lack them, such as mobs), I will refer to the latter, more sophisticated, variation as joint purposive action.⁷⁴ To be able to exercise joint purposive action, the members of a collectivity must have: (1) compatible interests (although not common or even necessarily complementary motivations); (2) a concomitant willingness to cooperate (something that might be called “participatory intention”);⁷⁵ and (3) the capacity to deliberate (however informally) in order to coordinate their actions (even imperfectly) in circumstances in which the required collective action is not obvious but, rather, open to disagreement.⁷⁶ Joint purposive action is thereby distinct from action that is best described at the level of the organization as a whole, and that generates coherent accounts of moral responsibility that cannot be reduced to its individual constituents. The model of institutional moral agency, presented above, aims to explore the latter (and to define the collectivities capable of it), but it overlooks the moral significance of the former. Coalitions of the willing are capable of—and, indeed, created to exercise—joint purposive action.

With the potential for joint purposive action comes shared responsibility. By this I mean responsibility that is necessarily distributive among the individual members of a collectivity for outcomes that can only be achieved when they act in concert.⁷⁷ I am not, however, suggesting that joint purposive action thereby entails diminished responsibility on the part of each member of the group. From such a perspective, the fact that multiple actors are required to discharge a duty would mean that the responsibility of each to act is somehow reduced, and that each could only be blamed in fraction for failing to do so. Moral responsibility, by this view, is diluted when it is shared. I do not mean this at all. Rather, the position I am proposing involves the individual constituents of those informally organized groups capable of this more sophisticated type of collective endeavor bearing greater individual responsibility than they would bear if they were acting independently. It also, by extension, suggests that we should have greater expectations of the individuals within random collectives that have the potential to contribute to establishing such groups than we should have of individual agents who do not have this opportunity (because they find themselves either in isolation or as part of a random collection whose other constituents lack a willingness to cooperate).

Joint Purposive Action, Enhanced Capacities, and Redefined Individual Resonsibilities to Protect

The arguments of both Held and May allow me to build on a relatively simple, but crucial, insight: that agents who come together, even in an informal association, to work toward a shared goal are able to achieve things by cooperating that they would not be able to achieve independently. This potential for enhanced capacities prompts me to suggest that individual agents (including states, nonstate actors, and intergovernmental organizations) should come together to form such an ad hoc group in the absence of a viable alternative and when confronted with a moral imperative that would otherwise go unmet.⁷⁸ This also leads me to propose that there is reason to have greater expectations of these agents—both when they are already members of informal associations with the capacity for joint purposive action and when they find themselves in “random collections” with the opportunity to form such associations. In other words, the enhanced capacities with which individual agents can be imbued as part of an informal association (existing or potential) lead to magnified individual responsibilities.

This argument regarding capacities for joint purposive action and attendant shared responsibility is directly relevant to considerations of who can—and should—discharge the proposed remedial responsibility to protect vulnerable populations. Indeed, if we accept the proposed moral responsibility to protect vulnerable populations as a starting point, and maintain that the “just cause” criterion was met in the case of the 2011 intervention in Libya, this argument has implications not only for how we judge the actions of those states that formed a coalition in order to carry out UN Resolution 1973 (these actions were required, and not merely due to states' Charter-defined responsibilities to implement Security Council decisions) but also for what would have been required in a counterfactual case of no resolution. In such a case, either a state capable of acting independently would have had a duty to intervene (Walzer's “agent of last resort”) or, more plausibly (given both the diverse range of capacities needed for an effective humanitarian intervention and the constraining perception of illegitimacy that threatens to accompany unilateral action), members of a “random collection” of institutional moral agents would have each had an obligation to contribute to a collectivization process, and then, as part of a joint endeavor, to respond to the crisis. Importantly, when such agents have enhanced capacities as members of an informal association, or the opportunity to achieve enhanced capacities by contributing to the establishment of an informal association, these greater expectations placed on them also translate into the individual agents' increased vulnerability to blame when duties are not discharged.⁷⁹

Accompanying Individual Responsibilities

This is not to say, however, that all moral responsibilities are somehow discharged when an ad hoc group of agents comes together and acts in concert to respond to a crisis in the absence of a viable alternative. Here it might be useful to return briefly to Held's example of the victim of the collapsing building. Even if individual actors have an immediate responsibility to form a temporary, informal association and cooperate in order to save a victim facing a profound threat, such emergency measures, and even their possible success, do not mitigate the prior and ongoing responsibilities both to create the conditions conducive to avoiding this sort of accident in the first place (by promoting the construction of safe buildings) and to ensure that a more suitable agent is in place to respond (by establishing robust and reliable emergency services). In the international realm, alongside the immediate—and transient—imperative that institutional agents act in concert when faced with a grave crisis that would otherwise be unattended, it is necessary to recognize the ongoing prospective moral responsibilities both to create an environment in which such crises are prevented and to build and bolster formal “agents of justice” with the will, resources, and procedures in place to act.

This final moral responsibility is crucial because the informal associations that can satisfy the immediate imperative are, in many ways, imperfect alternatives. Informal associations such as coalitions of the willing may be capable of joint purposive action, and may even boast a speed and flexibility in responding to crises that some formal organizations (such as the United Nations) lack when their rigid, formal decision-making structures are slow to arrive at a course of action or result in stalemate. However, these associations also have limits. Unlike (effectively functioning) structured institutions, they have neither the potentially sophisticated capacity for deliberation manifest in highly developed mechanisms for accessing and processing information nor the capacity for institutional learning whereby an organization is able to reflect on past experiences (and the consequences of previous acts and omissions) in a way that allows calculated revisions to policies, practice, codes of conduct, and organizational culture.⁸⁰ Nor is there the same potential within such informal associations to integrate coherently the roles of their constituents and thereby achieve a comparably complex level of coordinated action. Accordingly, both individual human and institutional moral agents also have an obligation to create, empower, or reform those formal organizations best able to respond to crisis so that such ad hoc arrangements do not exhaust our options in the future.

Conclusion

It is now possible to respond to the three questions posed at the start of this article. Coalitions of the willing cannot coherently be considered bearers of duties. Our calls to action and cries of condemnation must be directed instead toward the states, nonstate actors, and intergovernmental organizations that variously constitute (or could constitute) them. And even while such statements of moral responsibility are reasonably directed toward the members (or potential members) of these ad hoc associations, our attention should also remain focused on reforming and enabling the formal organizations for which coalitions of the willing act as either necessary supplements or, more controversially, unauthorized substitutes. Yet these are only some of the conclusions that follow from the argument above. The analysis that has led to these deceptively straightforward answers has also yielded a more nuanced account of moral responsibility in relation to informal associations in world politics than is currently on offer. This account, in turn, has far-reaching implications for our responses to practical problems, such as how to make sense of the widely accepted, and detrimentally vague, claim that “the international community” has a moral responsibility to respond to gross violations of human rights, and how to organize ourselves to realize this collective commitment.

Informal associations (which lack established organizational structures and decision-making procedures) do not qualify as institutional moral agents and cannot bear moral burdens at a corporate level. This judgment is important; however, the proposed model of institutional moral agency dismisses these types of collectivity too quickly with such a stark assessment and, therefore, fails to address important problems of moral responsibility in world politics. Fortunately, the same model allows us to understand certain bodies in world politics as themselves constituted by institutional moral agents. It thereby makes accessible valuable lessons from the work of philosophers such as Held and May, who focus on informal associations at the domestic level, populated by their individual human counterparts.

Consideration of the work of Held and May helps us to identify two categories of collectivity that lack formal organizational structures and decision-making procedures: “random collections” of agents that contingently share a time and place, and associations made up of agents that share a common purpose and have developed an informal organizational structure and deliberative capacity in order to pursue it. Both are distinct from those formal organizations (such as the United Nations and NATO) that are themselves made up of institutional moral agents and are at least contingently capable of deliberation and action at the corporate level. While the members of the former are only able to achieve (very limited) joint action that does not require prior deliberation and special coordination, the constituents of the latter can realize more sophisticated forms of cooperation, or what I have labeled joint purposive action.

The members of informal associations capable of joint purposive action can address injustices and respond to crises (and, indeed, produce harm) in ways that far surpass what they could each achieve if acting on their own. Our prescriptions and evaluations of their actions should therefore be different in significant ways to those that we would make if these agents were acting independently. In short, their enhanced capacities within these associations lead to magnified responsibilities on the part of each individual agent. Significantly, however, this argument does not support the simple conclusion that we should only have greater expectations of agents if and when they become members of such informal associations. (There could then be a perverse disincentive to form such associations in the face of crises.) When confronted with what is commonly held to be a duty-generating crisis, and in the absence of a viable alternative, our expectations should attach, in the short term, to those individual agents able to contribute to establishing the sort of informal associations needed to discharge the duty. In the wake of duty-generating crises that were neither prevented nor mitigated, blame should be apportioned not only to those individual and institutional moral agents who caused harm (by acting maliciously, recklessly, or negligently), and to those individual and institutional agents who could have responded but did not, but also to those who failed to collectivize to create a temporary, informal, and arguably imperfect association necessary for urgent remedial action.

With respect to the hard case addressed in this article, this argument suggests that in order to discharge the responsibility to protect vulnerable populations in situations of gross human rights abuses, and in the absence of a single institutional agent willing and able to respond, individual states, nonstate actors, and intergovernmental organizations each have a duty to contribute to establishing an informal association capable of joint purposive action and then to participate in an effective response. This conclusion will understandably raise a number of concerns, particularly when we are talking about vigilante coalitions of the willing. Primary among these concerns must be whether the imperative that agents contribute to such informal associations can be appropriated as a convenient cover for some to circumvent intergovernmental organizations (and international law) for their own advantage—in other words, to surreptitiously bypass the United Nations as the designated “custodian of collective legitimacy.”⁸¹

I do not deny that this can be—and has been—a motivation for establishing such associations. Nevertheless, if the value of such informal, collective action is brought out of the shadows cast by denial and wishful thinking (manifest in the mythical abilities of the so-called international community, or misplaced confidence in a frequently weak and constrained United Nations), and if the coalition of the willing is acknowledged as an alternative that is sometimes necessary when designated “agents of justice” are unwilling or unable to respond, then explicit guidelines can be set regarding the conditions under which such collective action is genuinely required. Important questions of what these guidelines should look like, which institutional agents have an obligation to come together and act in concert in a particular case, and how resulting costs are to be distributed, necessarily follow. By proposing that the nature of informal associations, such as coalitions of the willing, should profoundly affect our prescriptions and evaluations of the acts and omissions of their constituents (and potential constituents), this article has aimed both to offer a more demanding account of moral responsibility in relation to such entities in world politics and to set the stage for these crucial, further questions.