The Responsibility to Protect Turns Ten

Abstract

Ten years since its adoption by the UN General Assembly, the Responsibility to Protect (RtoP) has become an established international norm associated with positive changes to the way that international society responds to genocide and mass atrocities. In its first decade, RtoP has moved from being a controversial and indeterminate concept seldom utilized by international society to a norm utilized almost habitually. This is an assessment that stands in contrast to the widespread view that RtoP is associated with “growing controversy,” but is one that rests on evidence of state practice.

Ten years since its adoption by the UN General Assembly, the Responsibility to Protect (RtoP) has become an established international norm associated with positive changes to the way that international society responds to genocide and mass atrocities. ¹ In its first decade, RtoP has moved from being a controversial and indeterminate concept seldom utilized by international society to a norm utilized almost habitually. This is an assessment that stands in contrast to the widespread view that RtoP is associated with “growing controversy,” but is one that rests on evidence of state practice. ² Strong agreement on the meaning and scope of the principle has emerged in the annual informal dialogues on RtoP held by the General Assembly; the principle has been unanimously reaffirmed in its entirety no fewer than four times by the UN Security Council and has informed more than twenty-five other Security Council resolutions; and RtoP is now being utilized by the wider community of UN member states. With only a few exceptions, states accept that they have committed to RtoP and agree on the principle's core elements.

As UN Secretary-General Ban Ki-moon observed in 2012, international debate about RtoP has moved from a focus on the merits of the principle itself to matters of implementation. ³ Organizations and states once considered hostile to RtoP have themselves begun to utilize the norm. ⁴ At the September 2014 General Assembly Sixth Informal Interactive Dialogue on the Responsibility to Protect: Fulfilling our Collective Responsibility, China described RtoP as a “prudential norm,” argued that “states should establish relevant policies and mechanisms” for implementing it, and noted that it was appropriate for international society to adopt measures to support RtoP, including the use of force “as a last resort”; India noted that RtoP “was agreed [upon] by all” as early as 2005; Indonesia offered emphatic support, saying it “fully subscribes to the finest purposes and objectives of the concept of RtoP”; Nigeria declared that “RtoP is apt, based on humanitarian and human rights law, representing a global conceptual and policy shift in the notion of sovereignty and security”; Iran noted that “we cannot agree more with the Secretary-General” and his approach to RtoP; the Philippines observed that “we subscribe to our shared responsibility” in relation to RtoP; and Argentina declared that “since the beginning, Argentina supports the concept of RtoP.” These statements—all from the global South and including governments considered quite hostile to RtoP—provide strong support for the idea that RtoP, once considered deeply controversial, is now an established international norm.

The tenth anniversary of the adoption of RtoP by world leaders provides a useful opportunity to evaluate the progress made and the challenges that lie ahead. With that in mind, this article proceeds in four parts. First, it sets out the case for thinking of RtoP as an international norm. Second, the article defends this proposition against two prominent critiques: (1) that there is little observable change in international behavior since the genesis of the norm, and (2) that RtoP considerations have not influenced behavior. The third section briefly examines some of the principal reasons for the apparent shift in international society's attitude toward RtoP. In conclusion, the fourth section highlights some of the limits and enduring challenges to RtoP.

Is RtoP a Norm?

Two issues complicate the question of whether RtoP is a norm. First, governments and international relations experts mean different things when they use the word “norm.” Governments tend to view norms as binding legal principles: China, for example, questioned whether RtoP is a norm on the grounds that it is not legally binding. ⁵ International relations scholars understand norms as broader social phenomena: shared expectations of appropriate behavior for actors with a given identity. ⁶ From this perspective, norms need not be legally binding or viewed as synonymous with law. It is this understanding of norms that is utilized in this article.

The second complicating factor is that RtoP refers not to a single norm but to two quite distinct sets of norms: those relating to how states treat their own populations and those relating to a state's responsibilities to contribute to the protection of populations in other countries. It is now commonly understood that the first cluster (“pillar one” of RtoP) relates to well-established principles of international law. ⁷ No state has demurred from this view in the General Assembly.

Focusing on the second cluster of responsibilities, those associated with the protection of populations in other countries, this section will argue that RtoP, which was described as an “emerging norm” by the International Commission on Intervention and State Sovereignty (ICISS) in 2001, has now become an established international norm. ⁸ That is because there are now “shared expectations” within international society that (1) governments and international organizations do, in fact, exercise this responsibility; (2) they recognize both a limited duty and a right to do so; and (3) failure to fulfill this duty should attract criticism. Taken together, the combination of practice (in which measures have been adopted in the hope of fulfilling RtoP) and the ongoing consideration of the norm (in the General Assembly and Security Council) have been central to its emergence. Both have helped establish precedents and shared expectations that limit the decisions that can be legitimately taken in response to genocide and mass atrocities, making it more difficult (but by no means impossible) for those institutions charged with protection responsibilities, most notably the Security Council, to avoid acting upon them altogether. ⁹ The remainder of this section will examine the international dimensions of RtoP in light of the three “tests” for an international norm mentioned earlier.

To understand changes in behavior associated with the emergence of RtoP, we first need to understand earlier practice. It is certainly true that international society became more actively engaged in peacekeeping and humanitarian response after the end of the cold war. However, as Table 1 demonstrates, there were important limits to what international society was prepared to do. Between 1990 and the 2005 World Summit, the Security Council engaged formally in only a little over half of all episodes involving genocide and mass atrocities. Even where it did engage, however, the protection of populations from these crimes was seldom a priority, with the Security Council making specific provision for such protection on only four occasions. In the great majority of cases, therefore, to the extent that it was a consideration at all, the protection of people from genocide and mass atrocities was a peripheral concern. Given this fact, it is not surprising that UN peacekeepers in Angola, Bosnia, and Rwanda stood aside in the face of mass killing, as they were neither mandated nor equipped to do anything else. What is more, the Security Council frequently heard debate about whether civil wars and atrocities committed within states were matters of “international peace and security,” and therefore falling under its purview. ¹⁰ Although developments in the 1990s certainly informed the emergence of RtoP, the ICISS report and the international debate that followed it were self-conscious responses to the failures and controversies of that decade.

Table 1 International Responses to Mass Killing: 1990–2005 ¹¹

RtoP did not emerge as a fully formed norm from the 2005 World Summit. The Security Council was initially cautious about RtoP, reflecting more generalized attitudes among sections of the UN membership that were concerned RtoP might constitute a potential challenge to sovereignty and license for external intervention. Within that context, the inclusion of RtoP language and considerations in substantive resolutions proved difficult and controversial. In 2007 a reference to RtoP in the draft preamble of Resolution 1769 on the situation in Sudan had to be removed before the resolution was adopted; and in 2008 the Security Council decided not to refer to RtoP in Resolution 1814 on Somalia despite its inclusion in a report by the Secretary-General's Special Representative. ¹² After these unpromising early debates, it was several years before the Council referred once again to RtoP in relation to another country situation. The apparent success of preventive diplomacy in Kenya (2008), which explicitly utilized an “RtoP lens,” was set against a number of other crises where states were more reticent about accepting and acting upon their protection responsibilities. ¹³ During that time, there were major protection crises in Darfur, the Democratic Republic of Congo (DRC), Guinea, Kyrgyzstan, Somalia, Sri Lanka, and Sudan/South Sudan. The 2009 protection crisis in Sri Lanka, which saw up to 40,000 Tamil civilians killed, was not even placed on the formal Council agenda. As Ekkehard Strauss observed, “the lack of unity in the Council or, in fact, the opposition of some Member States prevented the Council so far from applying the responsibility to protect on a specific country situation.” ¹⁴

Since early 2011, however, and the adoption of a series of landmark Security Council resolutions on Libya (Resolutions 1970 and 1973) and Côte d'Ivoire (Resolution 1975), the situation has changed in at least two important ways. First, protection crises characterized by genocide and mass atrocities have tended to generate multifaceted international responses, usually marshaled through or supported by the Security Council, to a much greater extent than previously. Of the thirteen identified crises since early 2011, the Council acted in response to eleven, with one (Nigeria) under active consideration. The single outlying case was the situation of the Rohingya in Myanmar, which had not made it on to the Council's agenda but which had elicited responses through the Association of Southeast Asian Nations, the UN Secretary-General's Special Representative, and a range of bilateral initiatives. What is more, in almost all of these cases, international society marshaled a multifaceted response comprising the deployment of significant international resources, including military force, with the explicit intention of supporting the protection of populations. It is possible to see in all this an emerging pattern of response that is establishing precedents through practice and that is strongly suggestive of the emergence of shared expectations about international society's responsibility to protect. Simply put, in assuming an almost habitual quality, collective international action to support the protection of populations from genocide and mass atrocities has become “the norm.” Not only has complete failure to act become rare, so too have limited responses that stop short of addressing protection concerns.

These underlying changes of behavior are associated with the emergence of RtoP. As Table 2 shows, in nine of the ten crises on which the Security Council acted, it referred explicitly to RtoP and made provisions for the protection of populations. The two outlying cases were Gaza and the DRC. In both these cases, the Council pointed to the centrality of protecting populations from war crimes and crimes against humanity—the central purpose of RtoP—without directly mentioning the norm itself, though in the case of the DRC it has come very close to a direct reference. ¹⁵ Therefore, just as it has become almost routine for the Council to respond to genocide and mass atrocities, so has the adoption of RtoP language in its resolutions become commonplace.

Table 2 RtoP and the Response to New Crises: 2011–2015

* Individual states have referred to RtoP but not the Security Council.

Behind this trend is a pronounced shift in the level of controversy associated with RtoP. In sharp contrast to the controversy surrounding the inclusion of RtoP in Resolution 1706 (2006) on Darfur, in none of the resolutions adopted since early 2011 was the inclusion of RtoP difficult to negotiate. Several UN officials and diplomats from member states intimately engaged in the Security Council have expressed this view privately to this author, and public evidence also points in this direction. With the exception of Resolution 1973 on Libya, each of the resolutions mentioning RtoP since early 2011 was adopted unanimously. Further, even including the case of Resolution 1973, no Council member expressed concern about the inclusion of RtoP in these resolutions in their formal statements to the Council. Nor is there evidence suggesting that the inclusion of RtoP language in a draft resolution delayed its adoption.

There is therefore good reason to think that shared expectations have emerged within international society that governments and international organizations, especially the United Nations, have a responsibility to protect populations from genocide and mass atrocities. International protection practices have become routinized to such an extent that not only is a Security Council response to genocide and mass atrocities much more likely than it was in the past but that response will almost certainly include protection, once considered a peripheral concern. Moreover, the general notion that international society has a responsibility to adopt measures aimed at protecting populations from genocide and mass atrocities is now largely uncontroversial among states—though, of course, the question of how that responsibility should be exercised in actual cases is often hotly contested, and deep sensitivities remain about the use of coercive measures without state consent.

To what extent do states recognize a “right” or a “duty” to act? That, as I noted earlier, is the second test for an international norm. The question of “rights” is the more straightforward issue. In contrast to earlier times, when the Security Council's authority over “internal” matters was questioned, no state now disputes the right of the Council to adopt measures in respect to RtoP or the right of individual states and regional organizations to provide encouragement and assistance to states on a consensual basis. These points were clearly expressed by member states in the General Assembly's informal dialogues on RtoP. ¹⁶ Even the more cautious Council members, such as China and Russia, do not now openly question the appropriateness of including RtoP matters on the Council's agenda. When justifying their first Security Council vetoes on Syria in October 2011, Russia argued that it undertook “intensive, constructive efforts to develop an effective response on the part of the Council” (including an earlier Presidential Statement ¹⁷ ), while China called for the Council to do more to encourage domestic reform and indicated its support for an alternative draft resolution focused on political dialogue. ¹⁸ Neither questioned whether it was legitimate for the Security Council to involve itself in Syria, despite the fact that the situation was then still a largely domestic affair.

Beyond the Security Council, important questions remain about the extent to which institutions such as the UN's human rights and peacebuilding architectures, its Secretariat, regional and subregional arrangements, as well as individual states can take it upon themselves to advance protection efforts without the express consent of the states concerned. By and large, states hold the view that consent is a prerequisite for action not authorized by the Security Council, but there are indications that some other global institutions are beginning to identify their own roles in this area. Notably, some states have suggested that the Human Rights Council's Universal Periodic Review (UPR) process could be utilized to support RtoP goals. India, for example, has argued that “we need to activate an advance warning system of potential dangers to civilian populations by the UN Human Rights Council when the country concerned is being reviewed in the UPR system.” ¹⁹ Similarly, in December 2013 the UN Secretary-General unveiled a new “Human Rights Up Front” action plan, which called on UN missions and country teams to prioritize human rights protection. ²⁰

The associated question of whether there exists a “duty” to protect other populations from genocide and mass atrocities is altogether more difficult to untangle. It is also an issue that has animated moral philosophers. ²¹ While the 2005 agreement on RtoP did not extend international society's legal rights with respect to intervention and interference in the domestic affairs of states, it did award special responsibilities to the Security Council. These have made it more difficult for the Council to justify complete inaction in the face of genocide and mass atrocities. ²² In paragraph 139 of the World Summit Outcome Document, member states acknowledged an international responsibility to “use appropriate diplomatic, humanitarian and other peaceful means, in accordance with Chapters VI and VIII of the Charter of the United Nations, to help protect populations from genocide, war crimes, ethnic cleansing and crimes against humanity.” This responsibility should be exercised “through the United Nations” and specifically through Chapters VI (peaceful measures), VII (enforcement measures), and VIII (regional arrangements) of the UN Charter, awarding a special role to the Security Council.

One of the principal critiques of the claim that RtoP is an international norm rests on the observation that the Security Council has stopped short of recognizing international society's responsibility to protect. While it is true that the Council has not recognized an international responsibility to protect in its resolutions on individual country situations, no doubt owing to concerns that such a strong precedent could limit its future freedom of action, it is not the case that the Council has never affirmed the notion of an international responsibility to protect. On six separate occasions the Council has expressly reaffirmed the whole RtoP principle agreed to in 2005, including the special responsibilities that such agreement bestows upon the Council itself. ²³

Individual member states have also begun to acknowledge these responsibilities in their statements to the Council. On the occasion of the twentieth anniversary of the Rwandan genocide, Lithuania told the Security Council that “the international community has a duty and moral responsibility to make sure that genocide and crimes against humanity have no place in the twenty-first century.” ²⁴ Chile made a similar call: “We urge the Security Council, in particular its permanent members, to shoulder that responsibility.” ²⁵ Chad, too, argued that “the Security Council should react with urgency in the event of mass crimes based on its responsibility to protect.” ²⁶ There are also signs that the Council's permanent members are beginning to recognize this responsibility, though they are unsurprisingly more cautious in their approach. The United States, for example, has argued that “We have affirmed—all of us—the duty of each Government to protect its citizens from mass atrocities. We have stated our preparedness under the Charter of the United Nations to respond when States require help in fulfilling that duty.” ²⁷ Though stopping short of explicitly acknowledging a duty to protect, even China has recognized that international society ought to take steps to prevent genocide and mass atrocities, arguing that “the international community should use dialogue, good offices and mediation, among other tools, to promote the settlement of disputes and differences and contain the escalation of conflict and halt genocide and other crimes against humanity at the source” (emphasis added). ²⁸ Read in conjunction with the emerging practices of protection described earlier, these statements point to emerging recognition of positive international duties associated with RtoP.

The third test for a norm is the extent to which clear compliance failures—in this case, failures to exercise the international responsibility to protect—attract criticism. Because norms set expectations of appropriate behavior, it is somewhat easier to see them at work when they are violated than when they are complied with. If clear failures to comply with a norm do not elicit criticism from within the relevant society, this is an obvious sign that the shared expectations related to it are nonexistent, weak, or else easily trumped by competing norms. ²⁹ Conversely, criticism is a clear indication that a society judges a course of action to be inappropriate because it does not comply with its norms.

The most obvious failure to comply with international expectations associated with RtoP has been in Syria, where the Security Council has been too weak and divided to lead a timely and decisive response. The General Assembly responded to this ineffectualness by signaling its strong disapproval of the Council's actions and endorsing a more resolute approach. Large majorities in the Assembly voted to recommend many of the measures vetoed in the Security Council. On February 26, 2012, the Assembly voted by 137 to 12 (with 17 abstentions) to condemn “widespread and systematic human rights violations by the Syrian authorities”; to call on all armed groups to end the violence; and to support an Arab League peace initiative. ³⁰ Among the supporters of that resolution were Brazil, India, Pakistan, and South Africa. In addition, the General Assembly expressly criticized the Security Council's handling of the crisis in Syria. On August 3, 2012, the Assembly voted 132 to 12 for a resolution that “deplored” the Council's failure to adopt effective measures to protect civilians. ³¹ Brazil and South Africa were among the countries that supported this resolution. More recently, the UN Human Rights Council's Commission of Inquiry on Syria, led by Paulo Pinheiro, argued that the Security Council “bears responsibility” for allowing the continuation of war crimes and crimes against humanity by all sides. ³² The willingness of the General Assembly and UN Secretariat to criticize the Council offers clear evidence of shared expectations within international society about the Security Council's responsibility to do what it can to protect populations from genocide and mass atrocities.

Taking the three tests together (practice, recognition of rights and duties, criticism of clear violations), there are strong grounds for concluding that over the past few years RtoP has emerged as an international norm. This is most obvious in the underlying changes to practice, whereby international responses to genocide and mass atrocities have become more common and more focused on protecting populations. It can, however, also be seen in the critical responses to failures to protect and in the emerging tendency of states, including permanent members of the Security Council, to acknowledge positive duties in this regard. Precisely what the norm requires in specific cases is difficult to determine, partly because the best course of action is seldom clear and always determined by context. But, as Jennifer Welsh argues, RtoP is primarily a responsibility to consider taking action to protect populations from genocide and mass atrocities—a “responsibility to try,” as Edward Luck puts it. ³³

Consistent Inconsistency?

There are two principal critiques of this view. The first disputes the claim that there has been behavioral change in the way that international society responds to genocide and mass atrocities, suggesting that RtoP is merely “sound and fury, signifying nothing.” ³⁴ According to this view, failures to protect in Darfur (a crisis that predated RtoP) and Syria are evidence of chronic inconsistency and selectivity, driven by the narrow self-interest of the great powers. Accordingly, RtoP “has contributed little of substance or prescriptive merit” and, as such, “it is difficult to understand why state behavior will change as a result.” That is because RtoP lacks “substance” and is little more than a “slogan employed for differing purposes shorn of any real meaning or utility.” ³⁵ From this perspective, RtoP “could well be the latest in a long line of grandiose declarations made by states that have had little influence on actual international relations.” ³⁶

This assessment rests on the empirical claim that there have been no substantive changes in Security Council practice; thus, intervention in Libya, for example, was merely an example of consistent inconsistency. ³⁷ That argument, in turn, rests on the twin claims that the Council remains generally antipathetic about responding to protection crises and that its response to Libya did not break new ground. Critics claim that the Council had acted similarly in the past—for example, calling on the British to use force to suppress the racist Ian Smith regime in Southern Rhodesia back in the 1960s; and, more recently, authorizing the use of force to secure the delivery of humanitarian aid in Somalia, to restore democracy in Haiti, and to prevent genocide in Rwanda in the 1990s. ³⁸

There are at least two significant problems with this line of argument. First, the claim that the Security Council has passed resolutions “comparable” to Resolution 1973 on Libya is flawed. ³⁹ In none of the other cases cited (Haiti, Rwanda, Somalia, and Southern Rhodesia) did the Council authorize the use of force for protection purposes without the consent of the de jure authorities concerned, as it did in Resolution 1973. Southern Rhodesia was never recognized as a state. When the Council authorized intervention in Somalia in Resolution 794 (December 3, 1992) it noted that it was responding to “urgent calls from Somalia for the international community to take measures to ensure the delivery of humanitarian assistance.” ⁴⁰ Haiti's ousted president, Jean-Bertrand Aristide, was still recognized by the Security Council as the legitimate leader of his country, and his letter requesting foreign intervention was annexed to Resolution 940 (July 31, 1994). ⁴¹ Likewise, French intervention in Rwanda in 1994 enjoyed the consent of the Rwandan Interim Authorities, who were still recognized by the Security Council as the country's government at the time. ⁴² Never in its history, prior to Resolution 1973, had the Council authorized the use of force to protect populations without the consent of the de jure authorities. Although such an authorization is not likely to be repeated often, this resolution marked a significant advance in signaling that the Council was no longer unwilling as a matter of principle to take such action, should it be judged necessary.

The second problem is that the critique focuses only on a small subset of cases—mainly Darfur, Libya, and Syria—and does not consider the totality of the Security Council's work. As Tables 1 and 2 show, there are clear underlying differences in international responses to genocide and mass atrocities, with strong trends toward more common engagement and the prioritization of protection. Over the past few years, nonengagement and the view that protection is a peripheral consideration have become clear exceptions. Even the Council's weak response to the situation in Syria, “deplored” by the General Assembly, belies the critics' claim. Despite the exceptionally complex situation and difficult geopolitics, the Security Council has referred to RtoP in its resolutions and statements on the crisis, has condemned the killing of civilians, supported negotiations, mandated investigations, authorized a monitoring mission (UNSMIS), mandated and supervised the removal of chemical weapons, and (for the first time in its history) authorized the delivery of humanitarian assistance without the host government's consent. ⁴³ In the past, such politically difficult cases yielding multiple vetoes have rarely—if ever—elicited this level of engagement from the Council.

Beyond Syria, the broad scope of the Council's engagement with crises characterized by genocide and mass atrocities stands in marked contrast to its much patchier and less focused engagement in the post–cold war period (1990–2005). At the time of writing, around 75,000 uniformed personnel were deployed to seven UN missions mandated under Chapter VII of the UN Charter to use “all necessary means,” including force, to protect civilians. ⁴⁴ This, too, is unprecedented.

A second critique accepts the novelty of the Security Council's response to the conflict in Libya but disputes RtoP's role as an influence on state behavior. As with the first critique, this too is built on an assessment of two cases (Libya and Syria) rather than on the Council's broader program of work identified in Table 2, and so overlooks the deeper transformations described earlier. According to Justin Morris, RtoP “was not cited as a justification for action [in Libya] because either it was not active in policy-makers' minds or, if it was, it was outweighed by other considerations.” ⁴⁵ While it is true that few states referred specifically to RtoP in the Security Council's formal deliberations on Libya, it is difficult to sustain the claim that RtoP was not a consideration. Most obviously, Resolution 1970 (which was adopted unanimously) and Resolution 1973 contained direct references to the norm, as did two subsequent resolutions on Libya. ⁴⁶ It would be difficult to explain why the Council would have adopted such specific references to the norm were RtoP not a factor in its deliberations. Moreover, states that led the initiative to secure Resolution 1973 and then spearheaded the intervention itself also utilized RtoP in their justifications for their actions. President Obama repeatedly employed RtoP language in his March 28, 2011, address to the nation, telling Americans that the United States has “a responsibility to act” and that NATO had “decided to take on [the] additional responsibility of protecting Libyan civilians.” ⁴⁷ At the time, some of the president's critics interpreted this as a clear invocation of RtoP. ⁴⁸ Likewise, British Prime Minister David Cameron referred specifically to RtoP, and the Security Council's reference to it, in his March 18 statement to Parliament on the intervention. ⁴⁹ Looking more broadly at international society's response to the crisis in Libya, it is hard to see how the response was anything other than consistent with the demands of RtoP and difficult to believe that this consistency was produced by chance.

It is well known that U.S. diplomats understand that, owing to lingering concerns about U.S. tendencies toward unilateralism on the use of force, RtoP would be better served by the leadership of others. Consequently, the United States has been quite reserved, especially in the Security Council, about its support for RtoP, but this should not lead us to think that the Obama administration is anything other than fully supportive. ⁵⁰ During the Security Council debates on Libya, most members acknowledged the imperative of protecting populations—RtoP's objective. ⁵¹ Even those objecting to military action recognized the imperative of protection that flows from RtoP. Russia argued that “our position regarding the clear unacceptability of the use of force against the civilian population of Libya remains unchanged. Any attacks against civilians and other violations of international humanitarian law and human rights must immediately and unconditionally cease.” China noted that it was “gravely concerned by the continuing deterioration of the situation in Libya. We support the Security Council's adoption of appropriate and necessary action to . . . halt acts of violence against civilians.” ⁵²

In most cases it is difficult to identify precisely why a government decided to support an armed intervention until many years after, when memoirs and official documents reveal intimate details about the decision-making process. In the case of Libya, however, we have remarkable clarity about the U.S. administration's decision-making and the unusually direct role played by RtoP-related considerations. Before a series of key meetings on March 15, 2011, the administration was skeptical, if not outright hostile, to British and French proposals for a no-fly zone over Libya. But faced with the clear threat of a massacre in Benghazi, the president decided to change course and support the use of force. Critical to this decision was the normative advocacy of foreign policy adviser Samantha Power and U.S. Permanent Representative to the United Nations Susan Rice, both of whom are longstanding RtoP advocates, as well as Secretary of State Hillary Clinton. These normative arguments, which drew on RtoP, were pitted directly against the national interest arguments offered by the Pentagon and others, and succeeded in altering the administration's approach. ⁵³ Libya is an especially bad example to use in order to cast doubt on the influence of RtoP on decision-making because, in this case, the role of normative argumentation based on RtoP was unusually direct.

To summarize, the principal critiques of the claim that RtoP's international dimension has become an established norm are not substantiated by empirical evidence. The critiques rest on flawed empirical propositions: the historical record suggests that Libya was not “business as usual” for the Council and that RtoP did influence decision-making, in an unusually direct fashion. Critics focus only on a small subset of cases rather than the whole range of relevant situations before the Security Council. Prominent cases like Libya and Syria should not be treated in isolation of the broader context, in which the Council engaged in protecting populations in the Central African Republic, Cote d'Ivoire, the DRC, Darfur, Liberia, Mali, South Sudan, and elsewhere. Looking at the Council's record more broadly points to the emergence of “habits of protection” both in terms of positive practice by the Security Council and negative responses from the wider international society when the Council is judged to have failed to fulfill its responsibilities.

Accounting for Change

How do we account for the emergence of RtoP as an established international norm? Although international society exhibited a degree of “buyer's remorse” ⁵⁴ immediately after its adoption in 2005, this should not obscure the fact that RtoP was unanimously endorsed at the World Summit and that several key states, including China, had made a conscious decision to support the formulation offered in 2005 as the best way of navigating thorny questions about how to protect populations while maintaining deference to sovereignty and noninterference. ⁵⁵ Agreement on RtoP in 2005 was itself largely a reflection of the fact that international society had already started to recognize human protection norms, including the protection of civilians in armed conflict, the protection and empowerment of women and girls, the protection of displaced populations, and international criminal accountability for grave crimes. ⁵⁶ In this context, RtoP needs to be understood as one component of an emerging “international regime” focused on human protection. It is related to, but distinct from, these other norms and practices. Viewed this way, international society's initial wariness toward RtoP owed more to uncertainty about its implementation—specifically, the concern that the concept could be hijacked to legitimize self-interested intervention—than to deep-seated antipathy toward what was agreed to in 2005. Given this, it is perhaps not surprising that practice has helped clarify what the norm should look like and has provided reassurance about its consistency with existing international law, especially relating to sovereignty, noninterference, and the preeminence of the UN Charter. These tendencies were reinforced with each new action related to RtoP.

Although it was the crises that beset the world from early 2011 that provided the initial breakthrough for RtoP, three practices helped clarify what had been agreed upon in 2005 and laid the foundations for the emergence of the new consensus.

First, the self-conscious use of an “RtoP lens” to frame Kofi Annan's diplomatic mediation in Kenya in 2008, which helped end intercommunal violence there, provided tangible evidence of the concept's capacity to support atrocity prevention through peaceful means. It also provided a powerful practical challenge to the still widely perceived association between RtoP and humanitarian intervention, which had hitherto stymied its development. ⁵⁷

Second, one of the principal sources of concern about RtoP was its perceived potential to legitimize self-interested or unilateral uses of force. However, practice since 2005 has shown that RtoP has limited capacity to legitimize such actions. In 2008, two permanent members of the Security Council used RtoP to justify the use of force in situations outside what was agreed upon in 2005. Significantly, both failed to legitimize their positions. In May 2005, French Foreign Minister Bernard Kouchner called for RtoP to be invoked in relation to Myanmar; and in August of that year Russia attempted to justify its invasion of Georgia by pointing to RtoP. Both claims failed to secure international support, even from allies. Ironically, these attempts to “abuse” RtoP indirectly helped the norm by clarifying its limits as a legitimizer for the use of force. This helped persuade cautious member states that RtoP did not provide an effective new means of legitimizing unilateral intervention. ⁵⁸ More recent Russian attempts to justify intervention in Ukraine on RtoP-related grounds have proven similarly unsuccessful, with none of its partners among the emerging powers, including China, lending their support to Russian actions.

Third, the establishment of a joint Office for Genocide Prevention and RtoP within the United Nations helped establish institutional foundations for a deepening global consensus by clarifying the concept (in terms of the “three pillars”) and engaging member states in detailed dialogue about its implementation. ⁵⁹ The Secretary-General's approach was welcomed by member states and provided them with reassurance about RtoP's aspirations and its focus on supporting states in protecting their own populations. ⁶⁰ This helped institutionalize RtoP, effectively routinizing or “mainstreaming” the norm within the UN system. His first report on implementing RtoP prompted the General Assembly to continue to consider the norm and to engage with its various facets. Since 2009 the Secretary-General has issued a report annually, which has guided the General Assembly's own deliberations. Meanwhile, the Office on Genocide Prevention and RtoP has engaged with member states on a range of issues, from training and education to early warning and assessment. Through these and other processes, member states have become increasingly familiar with RtoP. They have reaffirmed their commitment to it and considered—in detail—various aspects of its implementation. Partly as a result of this engagement, they have become more comfortable using the norm in practice.

One characteristic of the Security Council's use of RtoP since 2011 has been the sheer variety of circumstances in which it has been employed. This suggests that the Council's embrace of RtoP has gone hand in hand with greater recognition that RtoP is not—and cannot be turned into—an interveners' charter. For example, Resolution 1975 (2011) on Cote d'Ivoire reaffirmed the primary responsibility of each state to protect its civilians, and authorized a strengthening of the UN mission there (UNOCI) to include the use of “all necessary means” to protect civilians. Resolution 1996 (2011) on South Sudan established a UN mission (UNMISS) mandated to advise and assist the government “in its responsibility to protect civilians.” Acting under Chapter VII, Resolution 2085 (2012) authorized the deployment of an African-led International Support Mission in Mali (AFISMA) to “support the Malian authorities in their primary responsibility to protect the population.” This mission was transferred to the UN (MINUSMA) in 2013. RtoP also featured prominently in the Council's response to the crisis in the Central African Republic, which involved mandating a peace operation (MINUSCA) to use force to protect civilians in 2014. The Council has also found consensus on RtoP in the context of Syria, referring directly to the norm in demanding that all parties take steps to protect civilians, especially minority groups (Resolution 2139, February 22, 2014).

Through these practices, the Council has come to understand that it can pursue the goals of RtoP by using the full range of measures at its disposal. This body of practice has helped establish expectations about how the Council ought to respond to genocide and mass atrocities. The transformation of RtoP from controversial concept to international norm was therefore facilitated by: (1) underlying agreement about the norm; (2) processes of engagement that provided reassurance about its scope and limits; and (3) practices of protection utilizing the full range of options available to international society within the bounds of existing international law. These processes were intensified by practices of protection that helped create precedents, clarify concerns, and provide more tangible understanding of what “implementing RtoP” looks like in practice. Together, these have helped establish expectations about appropriate responses to genocide and mass atrocities, making complete inaction more difficult to justify plausibly, and therefore less likely (but not impossible) in practice. With this development, however, the limits of RtoP and new challenges have come to the fore—what Edward Luck described as the “risks of relevance.” ⁶¹

Risks of Relevance: Limits and Challenges

Thus far, I have argued that RtoP has become an international norm associated with positive changes to the way in which international society responds to genocide and mass atrocities. I have shown that states now agree on what RtoP is (its “meaning and scope”) and accept that they have committed themselves to this standard (making it more difficult to avoid international responsibility altogether). I have demonstrated positive changes in behavior, namely, an increased likelihood that international society will respond to genocide and mass atrocities, and a strong trend toward the prioritization of protection within responses. But although significant diplomatic and practical progress has been made, it is important to recognize the challenges ahead.

First, there are limits inherent to the capacity of an individual norm to influence behavior. International norms function primarily by influencing states' own judgments about appropriate and inappropriate behavior and the way in which other governments will respond to that behavior. Behavior compliance with shared norms is likely to be praised; noncompliance is likely to meet with criticism, punishment, or social exclusion. Just as some people can cope better with ostracism than others, so too can some states; and there are a range of factors, including other norms, social contexts, and the role of interests, that influence the extent to which norms can influence particular behaviors at particular times. An established international norm like RtoP is a shared understanding of appropriate behavior. It can influence behavior through the power of its arguments, social pressure, and risks of punishment, but it cannot determine or compel behavior. As such, a degree of variation from case to case is to be expected. What matters in the long term is the extent to which a norm alters underlying patterns of behavior. RtoP clearly is associated with such underlying change, but we should expect that in some cases it will prove more difficult to realize the norm's objectives than in others.

Second, global agreement on the norm of RtoP does not exhaust the challenges associated with it. It is one thing for states to agree on a norm of behavior and another thing entirely for them to agree on how best to fulfill that norm both in general and in the face of specific crises. The challenges of implementation are no less daunting than the broader normative questions that preceded them. How well these challenges are addressed will inevitably shape the normative consensus in the future. That consensus is not permanent and immutable; it will be influenced by the experience of implementation. There are at least three important “clusters” of challenges.

First, the use of coercive measures to fulfill the goals of RtoP remains deeply controversial. This, of course, is not unique to RtoP but reflects deep-seated concerns among states about the use of coercion without consent in any setting. A key challenge is to improve the legitimacy and effectiveness of the Security Council's performance in this area. On this question, RtoP finds itself wedged between two positions. One, arising from Libya, holds that the Security Council and states acting on its mandates need to be held more accountable for their actions. The implementation of Resolution 1973 by NATO and its partners drew sharp criticism from Brazil, China, India, Russia, and South Africa, among others, who complained that the mission overstepped its mandate by pursuing regime change, employing disproportionate force, ignoring or outright rejecting opportunities for political dialogue, and violating the arms embargo mandated by Resolution 1970. ⁶² Particularly troubling for them was the Council's inability to hold NATO accountable. Russia subsequently argued that this experience influenced its thinking on Syria and justified its multiple vetoes. ⁶³ Some argued that these controversies would make it more difficult for the Council to reach consensus on similar cases in the future. ⁶⁴ It is not surprising that as the Council becomes more proactive in its pursuit of RtoP, demands for political accountability are becoming more significant. As Jennifer Welsh notes, “perhaps the biggest drawback of allocating collective responsibility to the Security Council is the lack of clear mechanisms of accountability.” ⁶⁵ Future agreement about the appropriateness of coercive military force against states as a tool of RtoP—a necessary response to international society's changing expectations—will likely depend upon concomitant steps to address accountability questions such as those raised by the “Responsibility while Protecting” concept developed by Brazil and the notion of “Responsible Protection” articulated by Chinese scholar Ruan Zongze.

The other critical issue for the Security Council, arising from Syria, stems from calls for the Council to become more decisive in responding to genocide and mass atrocities and less vulnerable to the use of veto. Calls for veto restraint have gained some traction among states, some of which see the veto as the principal barrier to effective protection. ⁶⁶ France has led the way on this issue, and in 2013 proposed an informal “code of conduct” to limit the use of the veto in situations characterized by genocide and mass atrocities. ⁶⁷ However, at least three of the permanent five members (China, Russia, United States) remain skeptical, meaning that the proposal “is likely to remain confined to the realm of theory for the foreseeable future.” ⁶⁸ That is perhaps no bad thing in the short term, since pressure on the veto issue could weaken consensus on RtoP by reawakening concerns about the norm's capacity to extend the legitimization of force beyond the prevailing legal status quo.

Second, there are political and practical challenges connected to marshaling international society's institutions, capacities, and resources to the goal of protecting populations from genocide and mass atrocities. Politically, the principal challenge revolves around international society's tendency to privilege states. In practical terms, this makes it much more difficult to build consensus about steps to prevent states attacking sections of their own populations. An additional political challenge is that RtoP, especially its international dimension, remains a relatively low priority for most governments, competing with more pressing security, diplomatic, and international economic concerns. While this “normalization” of RtoP assisted consensus-building, it has also limited the resources most states are prepared to commit to the effort.

Among the more important practical challenges is overcoming the tendency to see RtoP as disconnected from associated programs of work in areas such as conflict prevention, peacebuilding, the protection of civilians, international criminal justice, and the protection and empowerment of women and girls. Thus far, practitioners and analysts have tended to treat these agendas as “solitudes” within the UN system because of their differences, rather than recognizing their overlapping issues and mutual interdependence. ⁶⁹ This has limited the capacity of international society to develop a comprehensive approach to dealing with genocide and mass atrocities. Although recent research has begun to break down these barriers, there remain powerful institutional preferences for separation. ⁷⁰ Particularly significant is the common claim that Protection of Civilians (PoC) norms should be kept separate from RtoP because they are less controversial. ⁷¹ However, as Paul D. Williams shows, although RtoP and PoC are distinct concepts and in some respects cover different ground (for example, PoC applies only in conditions of armed conflict; RtoP relates only to four crimes), they also overlap in several important ways. Nor is it clear that PoC is less controversial politically than RtoP: in fact, RtoP was reaffirmed by the Security Council in the context of its deliberations on PoC. (Indeed, the two concepts have been featured together in more than a dozen Security Council resolutions.) It is the coercive measures that are controversial, and not the normative framework they inhabit. ⁷²

A third challenge is that the United Nations, interested member states, and advocates of RtoP need to do a better job of managing expectations. On the one hand, as I have previously argued, norms can influence but not determine behavior. On the other hand, outsiders have relatively limited influence on the conflicts that give rise to genocide and mass atrocities. Although concerted international action can sometimes prevent mass atrocities and protect the vulnerable, the primary sources of resilience rest within states and societies themselves. ⁷³ Determined international efforts can facilitate, support, and improve protection, but they cannot by themselves provide it in any comprehensive fashion, except through the type of massive interventions that are rarely contemplated.

At least two important points flow from this observation. First, modesty about the capacity of international actors to protect populations from genocide and mass atrocities is required. “Ending mass atrocities once and for all” is a task that will be achieved primarily within states and societies themselves, albeit with international assistance, and not within a short space of time. Second, and as a result of this fact, international engagement should be carefully attuned to local conditions. To paraphrase Alex de Waal in a different context, prevention needs to operate through the local “political marketplace.” ⁷⁴

Conclusion

In its first ten years, RtoP has emerged as an international norm. With only a tiny handful of exceptions, states accept that they have made a commitment to RtoP and agree on its fundamental components. Thus, as far as almost all the world's governments are concerned, the key debates about RtoP are not ones about whether to accept the principle or about its meaning and scope, but are rather focused on its implementation in practice. This point has been underscored by dozens of states from the global South. I have also argued that the norm of RtoP is not just a “convenient vocabulary” employed when powerful states find it useful to do so. It is a norm associated with profound underlying changes to the way in which international society responds to the problem of genocide and mass atrocities. International society is now more likely to respond to these situations than it was prior to 2005 and, significantly, is much more likely to prioritize protection in its response.

If the first ten years of RtoP were primarily about establishing the norm, the next ten should be about its implementation. Repeated failures to fulfill RtoP or the norm's association with controversial practices could undermine the very consensus that underpins it. This will involve concerted action to address the accountability and effectiveness issues associated with the Security Council in order to reassure member states concerned about coercive interference and improve the Council's performance. This, in turn, will require tangible work aimed at fostering comprehensive responses to genocide and mass atrocities, as well as efforts to create realistic expectations about what can be achieved in the decade to come.