The Responsibility to

ProtectFive Years On
Alex J. Bellamy*

he Responsibility to Protect (RtoP) has become a prominent feature in

international debates about preventing genocide and mass atrocities and
about protecting potential victims. Adopted unanimously by heads of
state and government at the 2005 UN World Summit and reaffirmed twice since
by the UN Security Council, the principle of RtoP rests on three equally weighted
and nonsequential pillars1 : (1) the primary responsibility of states to protect their
own populations from the four crimes of genocide, war crimes, ethnic cleansing,
and crimes against humanity, as well as from their incitement; (2) the international communitys responsibility to assist a state to fulfill its RtoP; and (3) the
international communitys responsibility to take timely and decisive action, in
accordance with the UN Charter, in cases where the state has manifestly failed
to protect its population from one or more of the four crimes. The principle
differed from the older concept of humanitarian intervention by placing emphasis
on the primary responsibility of the state to protect its own population, introducing the novel idea that the international community should assist states in
this endeavor, and situating armed intervention within a broader continuum of
measures that the international community might take to respond to genocide
and mass atrocities. As agreed to by states, the principle also differed from the
proposals brought forward by the International Commission on Intervention and
State Sovereignty by (among other things) emphasizing international assistance
to states (pillar two), downplaying the role of armed intervention, and rejecting criteria to guide decision-making on the use of force and the prospect of
intervention not authorized by the UN Security Council. Five years on from

* Thanks to Sara E. Davies, Sarah Teitt, Nicholas J. Wheeler, Paul D. Williams, the editors of this journal, and the
anonymous reviewers for their helpful comments on earlier drafts of this article. Also thanks to Joanna Wenschler
at Security Council Report for assistance with my research on the councils activities.
Ethics & International Affairs, 24, no. 2 (2010), pp. 143169.
2010 Carnegie Council for Ethics in International Affairs

143

its adoption, RtoP boasts a Global Centre and a network of regional affiliates
dedicated to advocacy and research, an international coalition of nongovernmental organizations (NGOs), a journal and book series, and a research fund
sponsored by the Australian government. More important, RtoP has made its
way onto the international diplomatic agenda. In 2008, UN Secretary-General Ban
Ki-moon challenged the UN membership to translate its 2005 commitment from
words to deeds.2 This challenge was taken up by the General Assembly in 2009,
when it agreed to give further consideration to the secretary-generals proposals.3
RtoP has also become part of the diplomatic language of humanitarian emergencies, used by governments, international organizations, NGOs, and independent
commissions to justify behavior, cajole compliance, and demand international
action.
For all this apparent progress, however, disagreement abounds. Much as they
did before 2005, critics typically argue either that the RtoP is a dangerous and
imperialist doctrine that threatens to undermine the national sovereignty and
political autonomy of the weak or, quite the reverse, that it is little more than
rhetorical posturing that promises little protection to vulnerable populations. To
further complicate matters, profound disagreements persist about the function,
meaning, and proper use of RtoP, and the principle has been inconsistently
applied. For example, France (in relation to Myanmar) and Russia (in relation
to Georgia) used RtoP to justify the actual or potential use of coercive force in
contexts where there was no apparent manifest failure to protect populations
from genocide and mass atrocities. Conversely, the principle has not been used
by governments and diplomats in the context of Somalia, Afghanistan, and
Iraq despite the commission of many atrocities against the populations of these
countries.
This article takes stock of the past five years and uses this experience to address
three basic questions: (1) What is RtoPs function?; (2) Is it a norm, and, if so, what
sort?; and (3) What contribution has RtoP made to the prevention of genocide
and mass atrocities and the protection of vulnerable populations? It proceeds in
three parts. The first two parts assess the effort to implement RtoP within the
UN system and the use of RtoP in humanitarian crises since 2005, respectively.
The third section explores what these stories can tell us in relation to the three
questions posed above. In concluding, I argue that RtoP should be seen less as
a normative vocabulary that can catalyze action, and more as a policy agenda in
need of implementation.4
144

Alex J. Bellamy

Implementation in the UN
Debates about implementing RtoP through the UN got off to an inauspicious
start. Thanks largely to lingering concerns about RtoPs potential to legitimize
interference in the domestic affairs of states and other fears about abuse, several
states displayed what Gareth Evans labeled buyers remorse and launched a
revolt to prevent the principles development.5 Despite the commitment to RtoP at
the World Summit, it took six months of intense debate for the Security Council to
unanimously adopt Resolution 1674, reaffirming the World Summits provisions
regarding the responsibility to protect.6 Russia, China, and three non-permanent
Security Council members (Algeria, the Philippines, and Brazil) initially argued
that the World Summit had only committed the General Assembly to deliberation
on RtoP and that it was premature for the Security Council to take up the
matter.7 Changes in the councils non-permanent membership and the softening
of the language endorsing RtoP helped forge agreement, but it was a hard-won
consensus.8 This experience persuaded some of the councils RtoP advocates to
refrain from pushing the body to make greater use of the principle, for fear of
creating opportunities for skeptics to challenge the 2005 agreement.9
Since the passage of Resolution 1674, the council has reaffirmed its position
(Resolution 1894, 2009) but has referred to RtoP only once in relation to a specific
crisisin Resolution 1706 (2006) on the situation in Darfur. Several council
members expressed concern (China abstained on Resolution 1706) about the
diplomatic pressure brought to bear to secure this reaffirmation, and subsequent
resolutions on Darfur have shied away from reiterating it. Thus, a paragraph
indirectly referring to RtoP was deleted from a draft of Resolution 1769 (2007) on
Darfur; and Resolution 1814 (2008) on Somalia pointedly referred to the protection
of civilians and Resolution 1674 without referring to RtoP.10 This combination of
actions points toward a clear trend. Initially, the council displayed a willingness
to use RtoP in its consideration of ongoing crises, albeit reluctantly. It has shifted,
however, to referring to RtoP only in thematic resolutions, perhaps recognizing
that it is not appropriate for the council to use the principle ahead of further
consideration by the General Assembly. Resistance to implementing RtoP was also
evident in other UN bodies. For example, when the UN Human Rights Councils
High-Level Mission to Darfur reported in 2007 that the government of Sudan was
failing in its responsibility to protect Darfuris, the Arab Group, the Asia Group, and
the responsibility to protectfive years on

145

the Organisation of the Islamic Conference all questioned the reports legitimacy
and tried to prevent deliberation on its findings.11
Promising signs began to emerge with the election of the South Korean foreign
minister Ban Ki-moon as UN secretary-general in October 2006. Although Ban
has been criticized by some RtoP advocates for lacking charisma and leadership
skills and for negotiating with governments that have poor human rights records,
he has proven to be an effective norm entrepreneur and has succeeded in forging a
wider and deeper consensus on the principle.12 Campaigning under the slogan of
promise less and deliver more, Ban argued that the UN needed to close the gap
between its lofty rhetoric and its often less-than-lofty performance.13 In August
2007 the secretary-general indicated his intention to appoint UN expert Edward
Luck as his special adviser on RtoP, though Luck was not confirmed in his new
position till December of that year.
Lucks appointment, which was not without its controversies, represented a
turning point in the fortunes of post-2005 RtoP at the UN. Basing his approach
on a detailed dissection of the 2005 agreement, the special adviser engaged in
extensive consultation with member states. Vital to this approach was Lucks sharp
distinction between what states had actually agreed to in relation to RtoP and a
variety of alternative formulationssuch as recommendations of the International
Commission on Intervention and State Sovereignty (ICISS).14 These consultations
encouraged the secretary-general to identify a narrow but deep approach (one
that strictly limited RtoP to what was agreed to in 2005, but utilized the whole
prevention and protection tool kit available to the United Nations system, regional
arrangements, states, and civil society groups) and to develop the idea of RtoP
comprising three equally weighted and nonsequential pillars, as described above.15
In early 2009 the secretary-general released his report, Implementing the
Responsibility to Protect.16 The report clarified the nature of the 2005 agreement
and outlined a wide range of measures that individual states, regional organizations,
and the UN system might consider in order to implement RtoPs three pillars.
Ban saw the ensuing General Assembly debate as an opportunity for it to
consider whether and how to conduct periodic review of what member states
have done to implement RtoP and to determine how best to exercise oversight
over the secretariats efforts to implement the principle. As a first step toward
implementation, the secretary-general revived his plan for a joint office for
RtoP and the prevention of genocide and repackaged the proposal as a means
146

Alex J. Bellamy

of strengthening the UNs early-warning and assessment capacitysomething

specifically called for in the World Summit agreement.
While Luck privately expressed cautious optimism that a consensus could be
reached, many advocates of RtoP expressed doubts about the secretary-generals
approach, fearing that a General Assembly debate could provide the opportunity
for skeptical governments to renegotiate the norm.17 They also expressed concern
that by distinguishing RtoP so sharply from the ICISS recommendations and
humanitarian intervention, the secretary-general risked weakening the principle
by diverting attention away from what they saw as the main issuepersuading
governments to intervene forcefully in the event of future cases of genocide and
mass atrocities.
Despite the best attempts of the President of the General Assembly (PGA),
Father Miguel dEscoto Brockmann (Nicaragua), to persuade the General Assembly to adopt a critical stance and not commit to implementing the principle,
the 2009 General Assembly debate vindicated Lucks cautious optimism and
revealed a broad consensus around the secretary-generals approach.18 Ninetyfour speakers, representing some 180 governments (including the Non-Aligned
Movement) participated in the debate.19 Of those, only four (Cuba, Venezuela,
Sudan, and Nicaragua) called for a renegotiation of the 2005 agreement. The
General Assembly largely agreed with the secretary-generals interpretation of
the principles fundamental elements. In particular, most governments welcomed
the secretary-generals report, noted that the 2005 World Summit represented the
international consensus on RtoP, and agreed that there was no need to renegotiate
that text.20 The challenge, the General Assembly agreed, was to implement RtoP,
not renegotiate it. The overwhelming majority also indicated their support for the
secretary-generals identification of the three pillars and the narrow but deep
approach to implementing the principle. After heated debate among RtoPs group
of friends about the merits of pursuing an assembly resolution, in which some of
them argued that a resolution was unnecessary and dangerous because it presented
an opportunity for opponents to revise the 2005 agreement, Guatemala led an
effort to secure a General Assembly resolution supporting the secretary-generals
position on RtoP. Although a small group of skeptics (Venezuela, Cuba, Syria,
Sudan, Iran, Ecuador, and Nicaragua) forced the resolutions endorsement of the
secretary-generals report to be toned down, the end product acknowledged the
report, noted that the assembly had engaged in a productive debate, and decided
that the General Assembly would continue its consideration of the matter.21 In
the responsibility to protectfive years on

147

addition, the resolution lent tacit legitimacy to the secretary-generals proposal for
the establishment of a joint office for RtoP and the prevention of genocide, though
it remains to be seen whether this will be sufficient to overcome potential political
and bureaucratic roadblocks.
The debate helped identify a broader consensus than was thought possible
in the immediate aftermath of 2005, but it also exposed some concerns and a
small group of determined opponents. Concerns were expressed about four issues
in particular. The first of these is the modalities and need for an early-warning
capacity. Several member states worried that information gathering and assessment
by the UN violates sovereignty and can be easily politicized. The second concerns
the respective roles of the Security Council and General Assembly, including the
question of reform of the Security Council. Several of RtoPs most prominent critics
maintained that expanding the councils membership should be a prerequisite to
the principles implementation, and that the General Assembly should formally
oversee the councils work in this area. Third, the perennial question of the
potential for RtoP to legitimize coercive interference and the lack of clarity about
the triggers for armed intervention was raised. Fourth, concerns were voiced about
the potential for RtoP to draw resources away from other UN programs without
adding additional value. 22
These concerns notwithstanding, it is fair to conclude that the secretary-general
succeeded in stemming the tide of the post-2005 revolt against RtoP, in presenting
an account of RtoP that commands a high degree of consensus, and in articulating
a manifesto for implementing the principle, which, in line with his overall
philosophy, represents a modest but tangible first step. Bearing in mind the state of
play when Ban took office in 2007, the rate of progress has been rapid. Through the
efforts of the secretary-general and his special adviser, consensus has been built,
the General Assembly engaged, and a plan for implementation set out.

RtoP in Practice
Concurrent with the General Assemblys consideration of RtoP, the principle has
been referred to in relation to nine crises (Table 1). Ranging from postelection
violence in Kenya, where RtoP was employed by Kofi Annan as part of a diplomatic
strategy, to the Russian invasion of Georgia, where the principle was invoked to
justify unilateral armed intervention, RtoP has been applied inconsistently. Some
of this inconsistency, however, may actually help to clarify the principles scope.
148

Alex J. Bellamy

Table 1
RtoP and Humanitarian Crises Since 2005
Situation

RtoP invoked/used by

Response

Outcome

Sudan (Darfur;
2003ongoing)

UN Security Council

No notable dissension

Peace operation
deployed (UNAMID)
Protection of civilians
mandated
Matter referred to the
International Criminal
Court (ICC)

Kenya (200708)

Kofi Annan (African

Unionappointed
mediator); Francis Deng
(UN Special Adviser on
Prevention of Genocide);
tacit implication by UN
Security Council

Broadly supported
AU commissioner
questions gravity of
crisis

Internationally
brokered end to
violence

Georgia (2008)

Russia

Russian claims widely

rejected by
governments and
analysts
Consensus that no
evidence of RtoP
crimes

No support for
Russian claims
EU blames both
parties for violence

Myanmar/Cyclone
Nargis (2008)

France and some NGO

advocates

French claims rejected

by the Association of
Southeast Asian
Nations (ASEAN),
China, UK, UN officials
Consensus that RtoP
does not apply to
natural disasters

No support for French

claims
UN and ASEAN
conclude humanitarian access
agreement with
Myanmar

Gaza (2009)

Palestinian Authority,
Qatar, Iran, World
Council of Churches

Little public
consideration given to
the claim

Reports of war crimes

subject of heated
international debate

Sri Lanka (200809)

India, Norway, Global

Centre for RtoP

Sri Lanka rejects

applicability
No general debate

Evidence of atrocities,
but fears of systematic
abuse not realized

Democratic Republic of UN officials

Congo (ongoing)

No general debate
No likely points of
dissension

Peace operation
deployed (MONUC)
Civilian protection
mandated

North Korea (ongoing) Havel/Bondevik/Wiesel

commission

No general debate

None
continued overleaf

the responsibility to protectfive years on

149

Table 1
Continued
Situation
Myanmar/ethnic
minorities

RtoP invoked/used by
Myanmar government in
exile/Human Rights
Documentation Centre

Response

Outcome

Edward Luck concludes

that RtoP not suited to
dealing with this case
No general debate

None

There were four additional cases (Table 2) where one or more of the four RtoP
crimes (genocide, war crimes, ethnic cleansing, and crimes against humanity) were
committed or credibly threatened but where RtoP has not been invoked. This
section focuses on a handful of these cases in order to understand how RtoP is used.
First, I consider two cases that have tested the boundaries of the principles scope
and the types of situations to which it applies (Georgia and Myanmar/Cyclone
Nargis). Second, I examine two cases where the application of RtoP was relatively
uncontroversial (Darfur and Kenya). Finally, I consider a case (Somalia) where the
state has manifestly failed to protect its population from three of the four crimes but
where RtoP has nevertheless not been part of the surrounding diplomatic discourse.
Debating the Limits of RtoP: Georgia and Myanmar/Cyclone Nargis
In each of these cases, a permanent member of the Security Council (Russia
and France, respectively) invoked RtoP to legitimize the deployment or threat
of coercive force to stem what it alleged to be the commission or imminent
Table 2
Crises Involving Commission or Threat of Genocide, War Crimes, Ethnic Cleansing, and/or
Crimes Against Humanity and the Absence of RtoP
Situation

Evidence of RtoP Crimes

Outcome

Sudan (north-south;
2008ongoing)

200809: more civilians

killed than in Darfur
250,000 displaced

UN peace operation deployed

(UNMIS)

Somalia (2006ongoing)

Approx. 16,500 civilians killed

1.9 million displaced

Partial and ineffective AU mission

UN reluctant to deploy
peacekeepers

Iraq (2003ongoing)

Approx. 70,000 civilians killed

Risk of heightened violence
after U.S. withdrawal

UN withdrew after Baghdad

bombing
Phased U.S. withdrawal

Afghanistan (2001ongoing)

Approx 5,000 civilians killed

since 2005

UN authorized enforcement
operation

150

Alex J. Bellamy

commission of one or more of the four RtoP crimes. The debates provoked by the
two cases have helped elucidate the scope of the principle and the limits on its use.
In the case of Georgia, the debate focused on the factual veracity of Russias claims
about Georgias abuse of civilians in South Ossetia. The debate over Myanmar,
however, hinged on the applicability of RtoP to situations where a government
fails to provide or permit sufficient humanitarian relief in the wake of a natural
disaster. In both cases, the claims advanced were roundly rejected by international
society, effectively placing two limits on the use of RtoP for coercive purposes:
(1) a requirement that the use of coercion be preceded by compelling evidence of
genocide or mass atrocities; and (2) a relatively narrow interpretation of crimes
against humanity that excludes crimes not associated with the deliberate killing
and displacement of civilians.
In early August 2008, Georgias nationalist government launched a military
assault aimed at restoring constitutional order in the breakaway region of
South Ossetia. Russia responded by quickly routing and pushing the Georgian
army back into Georgia proper, taking the city of Gori. A cease-fire brokered
by Frances President Sarkozy provided for the cessation of hostilities and the
staged withdrawal of Russian forces from Georgia proper. Shortly afterward,
Russia unilaterally recognized the independence of South Ossetia and that of
Georgias other breakaway province, Abkhazia. The Russian leadership argued
that intervention was justified by the commission and imminent commission of
mass atrocities by Georgian troops, which President Medvedev and Prime Minister
Putin insisted amounted to genocide.23 Foreign Minister Sergei Lavrov referred
explicitly to RtoP in justifying the intervention.24 Russia again implied that its
actions were consistent with RtoP during a heated exchange with Georgia in the
General Assembly.25 These arguments won little support. Two of RtoPs Canadian
progenitors described it as the misappropriation of RtoP, and the Global Centre
argued that the protection of nationals in foreign countries goes beyond the scope
of the RtoP, that the scale of Russias intervention was disproportionate, and that
RtoP does not provide a justification for the use of force without UN Security
Council approval.26 In addition, a special commission set up by the EU found that
Russia had acted disproportionately, and there has been little support for Russias
recognition of South Ossetia.27
On May 3, 2008, Cyclone Nargis struck Myanmar, devastating the Irrawaddy
delta area and leaving much of the region under water. Approximately 138,000 people were left dead or missing and 1.5 million were displaced.28 Despite the massive
the responsibility to protectfive years on

151

scale of the humanitarian catastrophe and the governments obvious inability to

respond effectively, Myanmars military regime initially denied access to humanitarian agencies, inhibiting the delivery of urgently needed supplies and medical
assistance. While offers of assistance poured in, Myanmar was slow to issue visas
and insisted on distributing aid itselfraising fears that much of the materials
would be siphoned off for other uses, including by and for the military.29 Frustrated
by the slow progress, French Foreign Minister Bernard Kouchner proposed that
the UN Security Council invoke RtoP to authorize the delivery of aid without
Myanmars consent, arguing that the denial of humanitarian assistance constituted
a crime against humanity. The proposal was flatly rejected by China and ASEAN,
which argued that RtoP did not apply to natural disasters. ASEAN governments
also maintained that Myanmar must not be coerced into accepting humanitaria
assistance. This view was shared by senior UN officials, who recognized the danger
that Kouchners argument posed to both the relief effort (which required Asian
cooperation) and the emerging consensus on RtoP.30 The British government
rejected Kouchners argument as incendiary and agreed that RtoP did not apply
to natural disasters.31 In the end, ASEAN and the UN secretary-general used diplomacy to secure the regimes acquiescence to the delivery of international aid and
organized a joint UN-ASEAN relief effort. There is speculation that the threat
of RtoP encouraged the regime to relent, and anecdotal evidence points in this
direction.32 If this was indeed the case, it was more likely the regimes paranoid
fear of a unilateral Western invasion rather than concern about a multilateral
RtoP response that prompted this shift, given that there was little likelihood that
coercion would be authorized by the Security Council.33 Although painfully slow,
uncoordinated, and ad hoc, the diplomatic effort secured humanitarian access and
delivered relief that helped avoid a much-predicted second round of deaths due to
disease and malnutrition.
Had one or both of these countries been successful in using RtoP to legitimize
their defense of intervention, they would have further confirmed the view that
RtoP is a Trojan horse that legitimizes great power interference in the affairs of
the weaka view that was fueled by the use of RtoP-related arguments to justify
the invasion of Iraq in 2003. However, the failure of Russia and France to legitimize
their positions by using RtoP suggests that while great powers might be tempted
to pursue this avenue, RtoP does not confer automatic legitimacy on coercive
interference in the event of a political or humanitarian crisis. This would seem
to suggest that fears about RtoP being used as a Trojan horse are unwarranted.
152

Alex J. Bellamy

In addition, the debates generated by Russian and French claims helped to clarify
the burden of proof that needs to be satisfied in order to successfully invoke the
principle to justify coercive behavior.
RtoP in Action: Darfur and Kenya
While the applicability of RtoP to the situations in Georgia and Myanmar was
hotly disputed, there was broad agreement of its relevance to the crises in Darfur
and Kenya. Yet assessments of the contribution made by RtoP to these two cases
could not be further apart. In relation to Darfur, RtoP is typically rated an abject
failure in that it failed to galvanize international action or, worse, exacerbated the
situation by distracting the relevant actors. In contrast, RtoP is widely credited
with having helped diplomatic efforts to stave off the escalation of violence in
Kenya.
The crisis in Darfur, which erupted in 2003, has been viewed more than any other
crisis through the prism of RtoP. For many, Darfur represents RtoPs primary test
case, one that it is generally reckoned to have failed.34 Although critics are right to
point out that RtoP has not propelled powerful states to contribute peacekeepers to
undertake complex and dangerous tasks in strategically unimportant regions, this
criticism needs to be tempered somewhat. Indeed, much of the killing occurred in
200304, prior to the worlds commitment to RtoP. In addition, the Darfur crisis is
closely interrelated with the peace process that delivered the 2005 Comprehensive
Peace Agreement (CPA) between the government of Sudan and the Sudan Peoples
Liberation Movement/Army (SPLM/A), which ended a decades-old civil war that
had claimed some two million lives. Key Western governments feared that taking
a robust line on Darfur would jeopardize these peace negotiations and undermine
the CPA.35 The conflict is also connected to the situations in Chad and the Central
African Republic, meaning that responses were calibrated for their impact on these
crises as well. However, and perhaps more important, since 2004 the Security
Council has been actively seized of the matter and has responded with a raft of
measures, including targeted sanctions, referral of the situation to the ICC (leading
to the indictment of Sudans president), and the authorization of a large peace
operation (UNAMID) with a civilian protection mandate. The key problem is that
these measures were slowly implemented and have proven insufficient to protect
vulnerable populations.
This persistent gap between what is needed and what is delivered cannot be
primarily ascribed to the situations complexity, but instead reflects the limited
the responsibility to protectfive years on

153

extent to which RtoP has the capacity to generate compliance pull in international society. Currently, there is no consensus in the council about the hierarchy
of relevant norms (with China insisting on Sudanese consent as a prerequisite for
the deployment of peacekeepers) and about what ought to be done (explaining the
lack of punishment and follow-up). There is also insufficient willingness on the
part of member states to commit forces and resources to implement the councils
will (explaining the under-resourced UNAMID mission).
In sharp contrast to the treatment of Darfur, the diplomatic response to the ethnic
violence that erupted in the aftermath of the disputed December 2007 elections in
Kenya is widely trumpeted as the best example of RtoP in practice.36 While up to
1,500 people were killed and 300,000 displaced, a coordinated diplomatic effort by
a troika of eminent persons mandated by the African Union (AU), spearheaded by
Kofi Annan, and supported by the UN secretary-general persuaded the countrys
president, Mwai Kibaki, and his main opponent, Raila Odinga, to conclude a
power-sharing agreement and rein in the mobs. This prevented what many feared
could have been the beginning of a much worse campaign of mass atrocities.
Reflecting on his successful diplomatic mission, Annan later observed that he
saw the crisis in the R2P prism with a Kenyan government unable to contain the situation
or protect its people. I knew that if the international community did not intervene,
things would go hopelessly wrong. The problem is when we say intervention, people
think military, when in fact thats a last resort. Kenya is a successful example of R2P at
work.37

Ban Ki-moon was also quick to characterize the situation as relevant to RtoP and
to remind Kenyas leaders of their responsibilities.38 And the secretary-generals
special adviser for prevention of genocide, Francis Deng, also called upon Kenyas
leadership to exercise their responsibility to protect, reminding them that if they
failed to do so they would be held to account by the international community.39
These efforts were supported by the Security Council, which issued a Presidential
Statement reminding the leaders of their responsibility to engage fully in finding
a sustainable political solution and taking action to immediately end violence.40
However, there was little evidence of support among council members for more
robust measures should diplomacy fail, and the African members in particular
expressed the view that the AU should play a leadership role.41
While many contend that Kenya provides an illustration of what RtoP can deliver
in terms of preventive action, others argue that RtoP itself played a marginal role.42
154

Alex J. Bellamy

It is fair to say that it was the AUs emerging, but still frail, peace and security
architecturewhich predates 2005that provided the immediate catalyst for
international engagement, not RtoP per se. Another note of caution was sounded
by AU Commissioner Jean Ping, who questioned the appropriateness of RtoP
in this case on the grounds that the killing had been relatively small in scale.43
Moreover, while the Security Council was content to support AU-led mediation,
it is not at all clear that had the process failed and violence escalated, it would have
found the consensus and willingness to step up its engagement.
In sum, in cases where there has been little dispute about the applicability of
RtoP, the principle has established a patchy track record. Nevertheless, both cases
elicited a consensus that the international community has a legitimate role to play.
Debates on Darfur and Kenya have hinged not on whether international actors
should intervene in one way or another, but how. Less promising, though, are the
limits that were exposed by these two cases. Consensus on Kenya was possible
because engagement was limited to diplomacy and had host-state consent. There is
little evidence to suggest that the council would have been prepared to adopt a more
robust stance had one been required, with African members especially insisting that
the AU play the primary role. In relation to Darfur, there have been serious problems in relation to consensus about the relative weight that should be afforded to
different norms relating to the protection of civilians, the principle of noninterference and host-state consent, and about what ought to be done; and there has been a
discernible lack of willingness to implement the councils decisions. Although RtoP
was utilized in both cases, in neither did it provide a catalyst for decisive action.
The capacity (or incapacity) of RtoP to serve as a catalyst for action is brought into
sharper relief by considering episodes of actual or imminent mass killing that have
not inspired the use of RtoP. The most prominent among these is Somalia.
RtoP Missing in Action: Somalia
While the international response to the crisis in Somalia in the early 1990s is often
referred to as an important precursor to RtoP, the crisis that has befallen that
country since the December 2006 U.S.-backed Ethiopian intervention that deposed
the Union of Islamic Courts (UIC) and ushered in a new period of conflict has
not commonly been viewed through the prism of RtoP, despite the commission
of war crimes, crimes against humanity, and ethnic cleansing, as well as the very
real threat of further escalation.44 Instead, international attention has focused
more on some of the crisiss symptomsmaritime piracy and some elements
the responsibility to protectfive years on

155

of support for Islamismrather than the crisis itself. In the two years after
December 2006, approximately 16,500 civilians were killed; and in 2007 alone up to
1.9 million Somalis were newly displaced, a higher level of death and displacement
than in Darfur and Kenya combined.45 Despite widespread attacks on civilians,
international actors have remained reluctant to link the situation to RtoP. Indeed,
ICISS co-chair Gareth Evans suggested that Somalia was not a classic [RtoP]
situation but that the imminent threat of mass atrocities warranted its placing
on a watch list of countries of RtoP concern.46 The international response to the
crisis has been slow and hesitant. The international community has responded to
events as they have unfolded and has tended to prioritize the interests of external
actors over those of Somali civilians.
After months of diplomatic wrangling, in November 2006 the United States
introduced a draft resolution authorizing the deployment of peacekeepers under
the auspices of the Intergovernmental Authority on Development (IGAD), a
subregional organization in East Africa. The proposed mission was tasked with
supporting the Transitional Federal Government (TFG) based in Baidoa and
defeating the UIC, which had established a firm hold on Mogadishu and had
encircled Baidoa. Although the resolution won the support of China, Russia,
and African members, European governments expressed concern that such an
obviously biased mission could escalate the crisis.47 There were also concerns
about IGADs capacity to deploy and fund an effective peace operation, given
a gap of almost $300 million between its projected annual cost and Western
donations, the absence of a viable political process, disagreements within IGAD
itself (Djibouti and Eritrea opposed the deployment), and concerns about the
availability of peacekeepers.48 However, the TFGs imminent defeat at the hands
of the UIC prompted Ethiopia to intervene with U.S. backing and forced the
council to authorize the IGAD deployment, despite the misgivings of several
council members. When IGAD proved incapable of deploying peacekeepers, the
Security Council instead turned to the AU, which deployed a small and ineffectual
peacekeeping force (AMISOM) in 2007. The AUs Peace and Security Council
insisted that AMISOM would be a temporary deployment (initially six months)
that would be followed up by a UN peace operation, but neither the council nor the
secretariat ever endorsed the idea of a UN mission. AMISOMs obvious incapacity generated renewed callsfrom the United States (the Bush administration),
China, Italy, and the AU especiallyfor the UN to take over peacekeeping duties,
though none couched their pleas in RtoP terms, and the United States especially
156

Alex J. Bellamy

was motivated more by war on terror concerns than RtoP concerns. Given
the fragile security situation and in the absence of a viable peace process
or clear commitments of troops, equipment, and financial support from the
West (or elsewhere), the secretary-general was reluctant to endorse the deployment of UN peacekeepers. The United Kingdom, France, and Russia shared the
secretary-generals view that the conditions were not right for the deployment of
peacekeepers, that peacekeepers would face significant threats, that the UN would
not be able to generate a sufficiently robust force, and that in the absence of a viable
political process there was no clear end state.49 As such, the secretary-general recommended that UN peacekeepers only be deployed once the security situation had
improved, a viable political process had begun, and the state had proven capable of
performing basic functions. The Security Council accepted this recommendation
in Resolution 1814 (May 15, 2008), and there has been little movement since.
In the case of Somalia, therefore, the commission of war crimes, ethnic cleansing,
and crimes against humanity has not prompted outsiders to view the problem
through the prism of RtoP, to take timely and decisive action, or to prioritize
the protection of Somali civilians. Where governments have called for and
supported intervention, this has been motivated more by strategic concerns
than by humanitarian concerns; and it is fairly clear that in terms of civilian
protection, the Ethiopian intervention did more harm than good, that AMISOM
has been relatively ineffectual, and that the council has not been proactive in trying
to create conditions that are conducive to the deployment of UN peacekeepers. As
such, this case highlights RtoPs limited capacity to act as a catalyst for action by
helping to manufacture political will. In the absence of states choosing to use RtoP
either as a diplomatic tool or to legitimize coercive intervention, the norm has
simply not been part of the international political discourse about how to respond.

Function, Status, and Contribution

Over the past five years RtoP has become a key part of the language used to
debate and frame responses to humanitarian emergencies, however selectively and
imperfectly. The challenge now is to consider what these efforts to implement
RtoP through the United Nations, and to work with it in responding to crises,
have to tell us about the state of the principle itself. This section poses three
basic questions about RtoP: (1) What is its function?; (2) Is it a norm, and, if
the responsibility to protectfive years on

157

so, what sort of norm?; and (3) What contribution has RtoP made to the prevention of genocide and mass atrocities, and to the protection of vulnerable
populations?
RtoPs Function
RtoP is commonly conceptualized as fulfilling two functions, but the two are not
complementary. The first is to use RtoP to describe a political commitment to
prevent and halt genocide and mass atrocities accompanied by a policy agenda
in need of implementation. This is how RtoP is most commonly referred to by
the secretary-general and diplomats at UN headquarters. Based largely on a plain
text reading of paragraphs 13839 of the World Summit Outcome document,
this position rests on two general propositions. First, as agreed by member states,
RtoP is universal and enduringit applies to all states, all the time. From this
perspective, there is no question of whether RtoP applies to a given situation
because RtoP does not arise and evaporate with circumstances. As such, much
of the debate about RtoP in practice, focused as it is on the question of whether
RtoP applies to a given situation, poses the wrong question. The question should
not be whether it applies, but how it is best exercised. Second, it follows that as a
universal and enduring commitment, RtoP gives rise to a policy agenda that needs
to be identified, articulated, and implemented. The secretary-generals 2009 report
Implementing the Responsibility to Protect and the ensuing General Assembly
debate marks the first step toward defining that agenda, but, as the secretary-general
recognized, there is much more work to be done in this field. The report and debate
mark out a framework in terms of the principles three pillars, but implementation
has yet to begin in earnest. Precisely what should be included in this policy agenda
is still to be debated. The secretary-generals report sets out some of the parameters;
and while academics dispute the precise contours, there is broad agreement around
a dozen or so key factors that ought to be addressed as part of a comprehensive
strategy to prevent genocide and mass atrocities, and about the most effective
forms of protection that might inform the development of capacities to better
exercise the three pillars.50 The effect of this approach is to emphasize RtoP as a
broad-based policy agenda focused on the upstream prevention of genocide and
mass atrocities through capacity-building and international cooperation, and to
focus attention on developing the institutions and capacities for effective response
within the prevailing normative framework, while keeping in mind the need to
marshal timely and decisive responses to atrocities when needed.

158

Alex J. Bellamy

This interpretation of RtoPs function accurately reflects what was agreed to by

states in 2005, and it is therefore no surprise that this view has been most commonly
expressed by governments, especially the global South. Critics, however, complain
that it weakens the principles capacity to mobilize action. For example, Gareth
Evans has argued that the preventive component of RtoP should be limited to
situations where mass atrocities are imminent and should not extend to longerterm prevention efforts.51 To widen the focus of RtoP in this way, he argues,
is dangerous from the perspective of undermining R2Ps utility as a rallying cry.
If too much is bundled under the R2P banner, we run the risk of diluting its
capacity to mobilize in the cases where it is really needed.52 The problem with
this critique, in my view, is that most scholars who work on the prevention of
genocide and mass atrocities seem to agree that prevention requires action well
before a rallying cry has been made.53 Moreover, in most cases, by the time
mass atrocities are apprehended as imminent, the window for prevention has been
closed. The oft-touted case of Kenya, where late-in-the-day prevention succeeded,
is a rare exception.54
Evanss argument brings us to the second function of RtoP language. Drawing
on Copenhagen School theories of securitization, Eli Stamnes has shown that the
intention behind the RtoP principle is to generate a speech act (that is, words and
sentences that perform specific communicative functions, such as promises and
warnings), which has the effect of elevating certain issues above normal politics as
a catalyst for decisive international action.55 In other words, RtoP is a label that can
be attached to particular crises in order to generate the will and consensus necessary
to mobilize a decisive international response. According to Evans, the whole point
of embracing the new language of the responsibility to protect is that it is capable
of generating an effective, consensual response in extreme, conscience-shocking
cases, in a way that right to intervene language simply was not.56 This is clearly
how RtoP has predominantly been used in relation to humanitarian crises. For
example, in the case of Darfur, advocates used RtoP to generate the political will to
intervene; in Myanmar, Kouchner invoked the principle as a catalyst for forcible
aid delivery; in Kenya (and to a lesser extent Sri Lanka), it was used by mediators
and foreign governments to persuade the countrys political leaders to step back
from the brink and also to galvanize international attention; and in relation to
Gaza, the World Council of Churches referred to RtoP to generate international
attention to the commission of war crimes. It is this mode of thinking that frames
debates as being about the applicability of RtoP rather than about the most
the responsibility to protectfive years on

159

appropriate manner of realizing the principle in a given situation. Although many

commentators and some diplomats switch between the first mode (RtoP as policy
agenda informed by commitment to normative principles) and the second (RtoP as
speech act and catalyst for action), the two are in fact incompatible, causing much
of the confusion about the way RtoP is used. One cannot sustain a commitment
to the long-term prevention of genocide and mass atrocities as part of RtoP while
also conceptualizing RtoP as primarily a speech act that acts as a catalyst for action.
While the second account dominates discussions about RtoP and humanitarian
crises, and may certainly be consistent with the concerns that animated the ICISS,
it is hard to find support for it in paragraphs 13839 of the World Summit
Outcome document. On the one hand, the commitment made by states was
open-ended, not time- or context-sensitive; on the other, the agreement made
it plain that RtoP-related crises would be dealt with through the UNs normal
peace and security mechanisms, and did not require exceptional measures. I will
return later to the question of how effective RtoP is as a call to action. For now, it
is important to note that actors have a choice about the way in which they employ
RtoP language, that the two options cannot be used conjointly with coherence,
and that this choice affects how we understand the principles role and impact.
RtoP as a Norm
Is RtoP a norm, and, if so, what sort of norm is it and what does it require? Norms,
recall, are shared expectations of appropriate behavior for actors with a given
identity.57 There is general consensus that RtoP is a norm, but much less agreement
on what sort of norm it is.58 There are two elements to this particular problem.
First, RtoP is not a single norm but a collection of shared expectations that have
different qualities. On the one hand, RtoP involves expectations about how states
relate to populations under their care. These expectations predate RtoP and are
embedded in international humanitarian and human rights law.59 Although there
are arguments about their scope (especially concerning crimes against humanity)
and the extent to which they are embedded or habitual, the basic proposition
that states are legally and morally required not to intentionally kill civilians is well
established.60 RtoPs first pillar is therefore best understood as a reaffirmation and
codification of already existing norms. However, RtoP also places demands on
states as members of international society: to assist and encourage their peers in
the fulfillment of their RtoP (pillar two), and to take timely and decisive action in
cases where a state has manifestly failed in its RtoP (pillar three). It is much less
160

Alex J. Bellamy

certain that pillars two and three constitute norms, and, if they can be so called,
there is a serious question as to whether they exert sufficient compliance-pull.61
The test of whether pillars two and three are properly called norms is the
extent to which there is a shared expectation that 1) governments and international
organizations will exercise this responsibility, that 2) they recognize a duty and
right to do so, and that 3) failure to act will attract criticism from the society of
states. There is some evidence to support the view that such positive duties exist.
First, in Bosnia vs. Serbia, the International Court of Justice (ICJ) found that states
have a legal obligation to take all measures reasonably available to them to prevent
genocide.62 Second, common Article 1 of the Geneva Conventions (1949) requires
that states ensure respect for international humanitarian law as well as obey
it themselves. Third, the International Law Commissions (ILC) Draft Articles
on the Responsibility of States for Internationally Wrongful Acts give states a
duty to cooperate to bring an end to breaches of the law.63 Fourth, following an
independent inquiry, the secretary-general recognized the organizations failure to
prevent the Rwandan genocide as a failure of the whole UN systemimplicitly
recognizing a shared expectation that the organization should act to prevent and
halt genocide.64 Combined with international societys commitment to RtoP,
these legal developments have given rise to claims that a positive duty to prevent
genocide and mass atrocities is emerging.65 But this claim requires a liberal
interpretation of the relevant law, and has been disputed.66 Moreover, the state
practice documented in the previous section suggests that mutual recognition of a
positive duty to exercise pillars two and three is inconsistent at best. While there is
evidence of RtoP at work in creating expectations about the worlds responses to
the crises in Kenya and Myanmar/Nargis, states were not criticized by their peers
for failing to do more to protect civilians in Darfur and Somalia.
Even if pillars two and three can be described as norms, they are weakened
by the problem of indeterminacy. Whereas norms shape shared understandings
and limit the behaviors that can be justified by reference to them, such that
actors will be inhibited from acting in ways that cannot be plausibly justified,
what a norm prescribes in a certain situation is never fixed and absolute.67 The
more precisely a norm indicates the behavior it expects in a given situation,
the stronger its compliance-pull.68 RtoPs first pillar, for example, is a relatively
determinate norm: it expects that states refrain from perpetrating four specific
crimes. By comparison, the specific demands imposed by pillars two and three
are indeterminate. It is seldomif everclear what RtoP requires in a given
the responsibility to protectfive years on

161

situation. This is partly because of the indeterminacy of the norm itself and its
place relative to other norms, such as noninterference. Indeterminacy is produced
by a combination of uncertainty about what is expected, disagreements about
what ought to be expected, and an interest in preserving flexibility for the future.69
Uncertainties and disagreements were evident in the 2009 General Assembly
debate on RtoP discussed earlier. The problem is amplified by the inevitable
effects of intervening variables, such as competing judgments about the facts of
the case (for example, about the cause, existence, and scale of mass atrocities)
and different assessments about the most prudent response. As a result, it is
plausible to legitimize both sides of the debates about forceful intervention and
no-fly zones in Darfur; the indictment of Omar al-Bashir; the deployment of
IGAD, AU, or UN peacekeepers in Somalia; heaping criticism and sanctions upon
Sri Lanka; criticizing Israel and/or Hamas; the deployment of Western troops
to the DRC, Sudan, and Somalia; sanctions, engagement, or the status quo in
relation to Myanmar; and much else by reference to RtoP. The indeterminacy of
what RtoP requires of external actors in these cases weakens its compliance-pull,
and therefore the extent to which actors are satisfying shared expectations of
appropriate behavior. Of course, as Quentin Skinner argues, actors cannot hope
to stretch the application of the existing principles indefinitely; correspondingly,
[the actor] can only hope to legitimate a restricted range of actions.70 It is therefore
fair to say that active support for genocidaires and defense of complete inaction
in the face of mass atrocities would be obviously inconsistent with RtoP. But
once states agree that something ought to be done, RtoP grants the international
community a relatively free hand to determine what. This indeterminacy severely
restricts RtoPs compliance-pull, and hence its ability to encourage states to find
consensus and commit additional resources to the protection of civilians.
RtoPs Effectiveness
To date, assessments of RtoPs effectiveness have tended to fall into one of three
camps. The most extreme, put forth by Alan Kuperman, holds that RtoP has
actually caused genocidal violence that would not have otherwise occurred by
encouraging rebels into risky armed uprisings that provoke genocidal responses
from governments.71 A less radical variant on this theme suggests that RtoP can
encourage external actors to pay undue attention to military responses to mass
atrocities rather than more promising political solutions.72 An alternative critique
suggests simply that the indeterminacy of RtoPs second and third pillars limit the
162

Alex J. Bellamy

extent to which it can exert tangible effects on behavior.73 On the other hand, as
noted earlier, some point to Kenya as a good example of RtoP in practice, and
argue that in this case RtoP helped galvanize international attention and bring
diplomatic pressure to bear on that countrys political leaders.
Systematic analysis of RtoPs effectiveness should be premised on a clear
understanding of what it is that RtoP is meant to achieve. This takes us back to
the principles function. The RtoP as policy agenda approach lends itself to the
view that the primary aim of RtoP is to prevent genocide and mass atrocities from
occurring in the first place. For this approach, while it is important that the world
responds to mass killing in a timely and decisive manner, the key long-term test
is whether there are fewer cases of mass killing to respond to. Although it is too
early to offer a definitive answer to this question, indications are mainly positive:
the Uppsala Conflict Data Program recorded a sharp reduction in the number of
fatalities caused by one-sided violence between 2004 and 2008;74 a survey of episodes
of mass killing involving the intentionally caused deaths of more than 5,000 civilians
shows an equally sharp decline, with no new episodes after 2005.75 However, this
data needs to be tempered by acknowledging that the number of ongoing episodes
has remained relatively static, and the fact that more civilians than combatants tend
to be killed in modern conflict.76 At the same time, the downward trends in relation
to attacks on civilians are particularly important because they come within a context
of an increase in the number of armed conflicts overall between 2004 and 2008,
suggesting that armed belligerents are generally less likely to resort to atrocities than
they were previously.77 Taken together, this data suggests that the period since the
adoption of the RtoP has been associated with a general decline in mass atrocities.
But we cannot conclude that RtoP caused this effect, because each of the trends
was evident prior to 2005; it is impossible to know from global figures whether
international activism was prompted by RtoP; and it is unclear whether it was this
activism or local factors that encouraged actors to refrain from mass atrocities.
One way around this problem of uncertainty would be to analyze cases individually. Of the thirteen cases identified in tables 1 and 2, five (Kenya, Myanmar/Nargis,
Sri Lanka, Gaza, and Sudan/CPA) were/are crises characterized by credible fears of
mass atrocities. In three of these cases (Kenya, Myanmar/Nargis, and Sri Lanka),
diplomats reminded governments of their responsibility to protect and of the fact
that the world was paying attention to their behavior. Although compliance was
neither complete nor unhindered, the feared mass atrocities did not eventuate.
Of course, in the case of Myanmar, the invocation of RtoP by France was widely
the responsibility to protectfive years on

163

rejected and criticized, so it would be difficult to credit the principle with fulfilling
a useful function in that case. Although there were calls for the Security Council to
step into the fray in all three situations, in no case did the situation get so bad that
the costs associated with nonconsensual measures were outweighed by the gravity
of the situation. RtoP has, therefore, been associated with the use of diplomacy
to prevent crises sliding into mass atrocities, and those attempts have enjoyed
a reasonable degree of success. There is, however, little evidence to suggest that
international concerns about RtoP persuaded either Israel or Hamas to moderate
their behavior. Finally, it is too early to tell whether international engagement with
the CPA will help prevent the potential slide back into civil war in Sudan.
This approach captures only part of the contribution that RtoP might make
to preventing mass atrocities. The further upstream we go in terms of structural
prevention, the more difficult it is to demonstrate RtoPs impact; yet this is
where the principle might have its greatest effect if it succeeds in galvanizing
efforts to strengthen early warning, build preventive capacity, and minimize root
causes. Upstream, RtoP may prevent mass atrocities in three principal ways: by
encouraging the internalization of the principle of discrimination within armed
forces; by helping states and societies to build the capacities they need to resolve
differences without recourse to violence and atrocities; and by persuading political
leaders that they are likely to pay costs for the commission of mass atrocities,
thereby encouraging them to adopt alternative strategies. These three factors may
deter potential belligerents from committing atrocities in the first place. This is
an inevitably long-term agenda that involves changing cultures and identities,
and it is impossible to draw a direct causal connection between these processes
and the prevention of specific atrocities at a later date. Moreover, these upstream
effects are contingent on RtoPs implementation. For these reasons we should
not expect to see moderating effects so soon. That said, there are numerous
examples of armed conflicts in the past five years that have not produced
mass atrocitiesincluding, among others, Chad, Mali, Pakistan, Guinea-Bissau,
Cote dIvoire, and Timor-Leste. In these cases, a combination of international
observation and engagement, and rational calculations influenced by expectations
about international engagement prompted by RtoP, encouraged actors to refrain
from committing widespread atrocities in situations where doing so might have
secured short-term gains. That RtoP and associated norms are having some effect
here is supported by the fact, noted earlier, that the incidence of mass killing is
164

Alex J. Bellamy

declining while the number of armed conflicts increases, suggesting that active
belligerents are more often choosing not to commit atrocities.
This brings us to the second function of RtoPas a speech act that provides a
catalyst for timely and decisive international action in response to the commission,
or imminent commission, of one or more of the four RtoP crimes. The effects of
RtoP in this regard are easier to measure, but the results are less promising and
suggest that the principle has had little success in mobilizing timely and decisive
international action. Since 2005 the largest crises in terms of the numbers of civilians
intentionally killed are (in order of the approximate number of civilians killed in
this period) Iraq, Somalia, Sudan (Darfur and south), the Democratic Republic of
the Congo, Afghanistan, and Kenya. Of these, Iraq and Afghanistan have not been
widely viewed through the prism of RtoP. In relation to Somalia, there has been
little RtoP talk, the UN and AU have proven reluctant to act decisively, and the
West tends to focus more on the situations external symptoms (piracy and links to
Islamist terrorism) than its civilian protection dimension. Despite talk of operating
through the prism of RtoP, international engagement with Sudan (UNAMID and
UNMIS) and the DRC (MONUC) has not perceptibly changed as a result of
RtoP, where the most significant developmentthe ICCs indictment of Omar
al-Bashirneeds to be balanced against continued under-resourcing of the UNs
peace operations, such that MONUC/DRC forces have thus far failed to bring the
murderous Hutu FDLR to heel, UNAMID continues to struggle to protect itself
and civilians in its care, and UNMIS has proven unable to stem the tide of local
conflicts. Finally, although Kenya is a notable success for RtoP, it is not clear that
RtoP itself caused the intense international engagement. More general data seems
to support these assessments. In the past few years, the RtoPs Western champions
have proven less, not more, willing to contribute to UN peace operations, and have
almost uniformly tightened their domestic regimes governing asylum for people
displaced by violence.78 The most charitable conclusion one can draw from all this
is that it is too early to tell whether RtoP can operate as a speech act that provides
an effective catalyst for timely and decisive action in the face of mass atrocities. But
even this assessment may be too generous in the face of the principles consistent
inability to generate heightened international activism.
The best explanation as to why RtoPs third pillar has thus far failed to generate additional political will relates to the problem of indeterminacy, discussed
earlier. Given that it is difficult to determine with any degree of certainty what
RtoPs third pillar requires in any given situation, it is unlikely thatwithout
the responsibility to protectfive years on

165

requisite transformation of national interests and institutional reforms that

reduce the costs associated with enhanced engagement by improving efficiency
and effectivenessthe principle will pull actors toward larger and more effective commitments to protecting civilians in danger. It is important to recognize
that there are different reasons why actors want to legitimize their behavior by
reference to a shared norm. At one end of the spectrum, norm compliance is
caused by instrumental decisions: actors obey a norm not because they think it has
inherent value (that is, because it is internalized), but because they believe that
norm violation will impose prohibitive costs. Close to this is noninstrumental
conformism or mimicry, where compliance is generated not by overt calculations of costs in relation to particular actions, but by a more generalized concern
about social exclusion.79 When a norm is indeterminate, actors who comply for
these instrumental reasons are unlikely to be pulled into changing their behavior.
Instead, they are likely to adopt the norms language while persisting with established patterns of behavior. They are able to do so precisely because it is so difficult
for other members of the society to determine whether or not they are acting in a
manner consistent with the norm. In none of the RtoP-related cases whose origins
predated the principle have actorsbe they RtoP supporters or criticsadjusted
their behavior to take account of the new norm, lending support to the view that
RtoPs indeterminacy permits states to formally commit to the principle without
having to change their practice. Alarmingly, this suggests that even RtoPs most
ardent supporters have not internalized the norm to the extent that it affects the
way they actually behave.
Given that indeterminacy makes it unlikely that RtoP will act in the near future
as a catalyst for international action in response to genocide and mass atrocities, it
seems reasonable to argue that the most prudent path is to view the principle as a
policy agenda in need of implementation rather than as a red flag to galvanize
the world into action. This view would certainly be consistent with the evidence
thus far that RtoP is best employed as a diplomatic tool, or prism, to guide efforts
to stem the tide of mass atrocities, and that it has little utility in terms of generating
additional international political will in response to such episodes. Over the long
term, if national, regional, and global institutions support the behaviors and build
the capacities envisaged by the UN secretary-general, it will be easier and less costly
to prevent and respond effectively to genocide and mass atrocities, reducing the
tendency of states to view prevention and protection work as inimical to national
interests and increasing the likelihood that they will undertake prevention work as
166

Alex J. Bellamy

a matter of habit and make use of the response capacities they have created when
needed. Of course, the prevention work should also reduce the overall number
of cases to respond to. We need to recognize that, by itself, RtoP is unlikely to
stimulate greater international activism in response to immediate crises in Darfur,
Somalia, the DRC, or, indeed, the next Rwanda, but that if the agenda outlined
by the UN secretary-general is implemented, there is a good chance of reducing
the likelihood of future Rwandas and ensuring that future responses are indeed
timely and decisive. For this reason, it may be more important that the UN
Secretariat establish a joint office for RtoP and the prevention of genocidethus
embedding RtoP within the UN systemthan that the Security Council refers to
the principle in the next resolution it passes on the protection of civilians.
NOTES
1

2
3
4
5
6
7
8
9
10
11

12
13

14
15
16
17
18
19
20
21

UN General Assembly (UNGA), 2005 World Summit Outcome, A/60/L.1, September 15, 2005, paras.
13840; UN Security Council (UNSC), S/RES/1674 (2006), April 28, 2006; UNSC, S/RES/1894 (2009),
November 11, 2009. On the three pillars, see Ban Ki-moon, Implementing the Responsibility to Protect:
Report of the Secretary-General, A/63/677, January 12, 2009.
Ban Ki-moon, On Responsible Sovereignty: International Cooperation for a Changed World, Berlin,
SG/SM11701, July 15, 2008.
UNGA, The Responsibility to Protect, A/RES/63/308, October 7, 2009.
I am grateful to one of the anonymous reviewers for suggesting this formulation.
Gareth Evans, The Responsibility to Protect: An Idea Whose Time Has Come . . . and Gone?
International Relations 22, no.3 (2008).
Draft Security Council Resolution on the Protection of Civilians, third iteration, December 1, 2005.
See UNSC, S/PV.5319, December 9, 2005, pp. 10, 19; and UNSC, S/PV.5319 (Resumption 1), December 9,
2005, pp. 3, 6.
Security Council Report, Update Report No. 5, Protection of Civilians in Armed Conflict, June 21,
2006; Update Report No. 1, June 18, 2007; and Update Report No. 1, March 8, 2006.
Security Council Report, Update Report No. 1, Protection of Civilians in Armed Conflict, June 18, 2007.
Ekkehard Strauss, A Bird in the Hand is Worth Two in the BushOn the Assumed Legal Nature of
the Responsibility to Protect, Global Responsibility to Protect 1, no. 3 (2009), p. 307.
High-Level Mission of the UN Human Rights Council, Report of the High-Level Mission on
the Situation of Human Rights in Darfur Pursuant to Human Rights Council Decision S-4/101,
A/HRC/4/80, March 9, 2007; and A/HRC/5/6, June 8, 2007.
See, e.g., Ramesh Thakur, The Model of a Mediocre Secretary-General, Ottawa Citizen, September
25, 2009; available at www2.canada.com/northshorenews/news/taste/story.html?id=2041865.
Joachim Muller and Karl P. Sauvant, Overview: The United Nations Year 2006/2007: A New Beginning
in Difficult Times, Annual Review of United Nations Affairs 2006/2007 (New York: Oxford University
Press, 2007), p. xv.
Edward C. Luck, The Responsible Sovereign and the Responsibility to Protect, Annual Review of
United Nations Affairs 2006/2007 (New York: Oxford University Press, 2007), p. xxxv.
Ban, On Responsible Sovereignty: International Cooperation for a Changed World.
Ban, Implementing the Responsibility to Protect.
International Coalition for the Responsibility to Protect (ICRtoP), Report on the General Assembly
Plenary Debate on the Responsibility to Protect, September 15, 2009, p. 3.
Office of the President of the General Assembly, Concept Note on Responsibility to Protect Populations
from Genocide, War Crimes, Ethnic Cleansing and Crimes Against Humanity, undated (July 2009).
Global Centre for the Responsibility to Protect, Implementing the Responsibility to Protect: Responding
to the UN Secretary-Generals Report, June 2009, p. 1.
See, e.g., statement by Maged A. Abdelaziz, Permanent Representative of Egypt, on behalf of the
Non-Aligned Movement, on Agenda Item 44 and 107.
UNGA, The Responsibility to Protect, A/RES/63/308, October 7, 2009.

the responsibility to protectfive years on

167

22

23
24
25
26

27
28
29
30

31
32
33
34
35
36

37
38
39
40
41
42
43
44
45
46

47
48
49
50

51
52
53

This analysis is based on Global Centre, Implementing the Responsibility to Protect, pp. 67; and
Asia-Pacific Centre for the Responsibility to Protect, Implementing the Responsibility to Protect:
Asia-Pacific in the 2009 General Assembly Dialogue, October 2009, pp. 1217.
International Crisis Group, Russia vs. Georgia: The Fallout, Europe Report no. 195, August 22, 2008,
esp. pp. 23.
Minister of Foreign Affairs of the Russian Federation Sergey Lavrov, interview with the BBC, Moscow,
August 9, 2008.
UNGA, Delegates Weigh Legal Merits of Responsibility to Protect, GA/10850, July 28, 2009, p. 14.
Lloyd Axworthy and Allan Rock, R2P: A New and Unfinished Agenda, Global Responsibility to
Protect 1, no. 1 (2009), p. 59; Global Centre for the Responsibility to Protect, The Georgia-Russia Crisis
and the Responsibility to Protect: Background Note, August 19, 2008; and Gareth Evans, Russia,
Georgia, and the Responsibility to Protect, Amsterdam Law Forum 1, no. 2 (2009).
Report of the Independent International Fact-Finding Mission on the Conflict in Georgia, September
2009.
Jurgen Haacke, Myanmar, the Responsibility to Protect and the Need for Practical Assistance, Global
Responsibility to Protect 1, no. 2 (2009), p. 156.
World Fears for Plight of Myanmar Cyclone Victims, Reuters, May 13, 2008.
Edward Luck, Briefing on International Disaster Assistance: Policy Options, to the U.S. Senate
Committee on Foreign Relations, June 17, 2008; and Julian Borger and Ian MacKinnon, Bypass Juntas
Permission for Aid, US and France Urge, Guardian, May 9, 2008, p. 20.
Borger and MacKinnon, Bypass Juntas Permission.
Evans, The Responsibility to Protect; and Haacke, Myanmar.
Andrew Selth, Even Paranoids Have Enemies: Cyclone Nargis and Myanmars Fears of Invasion,
Contemporary Southeast Asia 30, no. 3 (2008), pp. 379402.
See, e.g., Cristina Badescu and Linnea Bergholm, Responsibility to Protect and the Conflict in Darfur:
The Big Let-Down, Security Dialogue 40, no. 3 (2009), pp. 287309.
Discussed in detail in Alex J. Bellamy and Paul D. Williams, The Responsibility to Protect and the
Crisis in Darfur, Security Dialogue 36, no. 1 (2005), pp. 2747.
See, e.g., Evans, The Responsibility to Protect, p. 106; Desmond Tutu, Taking the Responsibility to
Protect, New York Times, November 9, 2008, available at www.nytimes.com/2008/02/19/opinion/19ihtedtutu.1.10186157.html?scp=1&sq=Taking the responsibility to protect&st=cse; and Donald Steinberg,
Responsibility to Protect: Coming of Age? Global Responsibility to Protect 1, no. 4 (2009), pp. 43241.
Cited in Roger Cohen, How Kofi Annan Rescued Kenya, New York Review of Books 55, no. 13, August
14, 2008.
Statement attributable to the Spokesperson for the Secretary-General on the situation in Kenya, New
York, January 2, 2008.
Ban Ki-moon, Address to the Summit of the African Union, Addis Ababa, January 31, 2008.
UNSC, Statement by the President of the Security Council, S/PRST/2008/4, February 6, 2008.
Security Council Report, Kenya update report, March 2008.
Sean Harder, How They Stopped the Killing, Stanley Foundation, June 2009. www.stanleyfoundation.
org/articles.cfm?id=596, accessed April 15, 2010.
Keynote Address by Jean Ping, Chairperson of the AU Commission, at the Roundtable High-Level
Meeting of Experts on The Responsibility to Protect in Africa, Addis Ababa, October 23, 2008.
Both Amnesty International and Human Rights Watch have found that war crimes have been committed
against the civilian population, including by Ethiopia.
See Somalia Conflict Kills More than 2,100 This Year, Reuters, June 26, 2008.
Matthew Russell Lee, Re-Branding Responsibility to Protect, Gareth Evans Says Somalias Not Covered, Inner City Press, September 17, 2009; available at www.innercitypress.com/r2p1evans091708.html
(accessed February 19, 2010); and Evans, The Responsibility to Protect, p. 76.
Colum Lynch, U.S. Peacekeeping Plan for Somalia Criticized, Washington Post, November 29, 2006.
See Security Council Report, Update on Somalia, MarchNovember 2006.
Security Council Report, Somalia Update Report, November 2008.
See, e.g., Barbara Harff, No Lessons Learnt from the Holocaust? Assessing Risks of Genocide and
Political Mass Murder Since 1955, American Political Science Review 97, no. 1 (2003); and Victoria
K. Holt and Glyn Taylor, Protecting Civilians in the Context of UN Peacekeeping Operations: Successes,
Setbacks and Remaining Challenges (study commissioned by the UN DPKO, November 2009).
Gareth Evans, The Responsibility to Protect in International Affairs: Where to From Here? (keynote
lecture at the Australian Catholic University, Melbourne, November 27, 2009).
Gareth Evans, The Responsibility to Protect: An Idea Whose Time Has Come . . . and Gone?,
pp. 29495.
See, e.g., Harff, No Lessons Learnt?; and John Heidenrich, How to Prevent Genocide: An Assessment
with Policy Recommendations (London: Greenwood, 2001).

168

Alex J. Bellamy

54
55

56
57
58
59
60
61
62

63

64
65
66

67

68
69
70
71
72
73

74
75
76
77
78

79

See Marc Saxer, The Politics of Responsibility to Protect, FES Briefing Paper No. 2, Berlin, April 2008,
p. 4.
Eli Stamnes, Speaking R2P and the Prevention of Mass Atrocities, Global Responsibility to Protect 1,
no. 1 (2009), p. 77. See J. L. Austin, How To Do Things With Words, 2nd ed. (Harvard: Harvard University
Press, 1975).
Evans, The Responsibility to Protect, International Relations, p. 294.
Martha Finnemore and Kathryn Sikkink, International Norm Dynamics and Political Change,
International Organization 52, no. 4 (1998), p. 891.
See, e.g., Evans, The Responsibility to Protect, p. 55; and Teresa Chataway, Towards Normative
Consensus on Responsibility to Protect, Griffith Law Review 16, no. 1 (2007).
Ban, Implementing the Responsibility to Protect, para. 13.
Hugo Slim, Killing Civilians: Method, Madness and Morality in War (New York: Columbia University
Press, 2008).
Thomas Franck, The Power of Legitimacy Among Nations (New York: Oxford University Press, 1990),
p. 49.
International Court of Justice, The Application of the Convention on the Prevention and Punishment
of the Crime of Genocide (Bosnia and Herzegovina vs. Serbia and Montenegro), judgment February
26, 2007, paras. 42838.
ILC, Draft Articles on the Responsibility of States for Internationally Wrongful Acts, 2001; available at
untreaty.un.org/ilc/texts/instruments/english/commentaries/9 6 2001.pdf; and ILC, Third Report on
Responsibility of International Organizations by Mr. Giorgio Gaja, Special Rapporteur, A/CN.4/553,
para. 10.
Report of the Independent Inquiry into the Actions of the United Nations During the 1994 Genocide
in Rwanda, S/1999/1257, December 16, 1999.
See Louise Arbour, The Responsibility to Protect as a Duty of Care in International Law and Practice,
Review of International Studies 34, no. 3 (2008), pp. 44558.
See, e.g., Jose E. Alvarez, The Schizophrenias of R2P (Panel Presentation at the 2007 Hague Joint
Conference on Contemporary Issues of International Law: Criminal Jurisdiction 100 Years After the
1907 Hague Peace Conference, The Hague, The Netherlands, June 30, 2007), p. 12.
This idea, and way of formulating it, is borrowed from Nicholas J. Wheeler, Saving Strangers:
Humanitarian Intervention in World Politics (New York: Oxford University Press, 2000). It is based on
Quentin Skinner, Analysis of Political Thought and Action, in James Tully, ed., Meaning and Context:
Quentin Skinner and His Critics (Cambridge: Polity, 1988), p. 117.
Franck, Power of Legitimacy, p. 52.
Ibid., pp. 5254.
Skinner, Analysis of Political Thought and Action, p. 117.
Alan J. Kuperman, Review of Gareth Evans, Responsibility to Protect, Political Science Quarterly, vol. 124,
no. 3 (Fall 2009), p. 591.
Alex de Waal, Darfur and the Failure of the Responsibility to Protect, International Affairs 83, no. 6
(2007), pp. 103954.
See, e.g., David Mepham, Darfur: The Responsibility to Protect (London: Institute for Public Policy
Research, 2006); Badescu and Bergholm, Responsibility to Protect; and Lee Feinstein, Darfur and
Beyond (New York: Council on Foreign Relations, 2007).
See www.pcr.uu.se/research/UCDP/graphs/one-sided fatalities 2007.pdf.
Documented in Alex J. Bellamy, Massacres and Morality: Atrocities in an Age of Non-Combatant Immunity
(New York: Oxford University Press, forthcoming), app. 1.
Ekaterina Stepanova, Trends in Armed Conflicts: One-Sided Violence Against Civilians, in SIPRI
Yearbook 2009 (Oxford: Oxford University Press/SIPRI, 2009), pp. 3968.
See www.pcr.uu.se/research/UCDP/graphs/conflict region 2008.gif.
Alex J. Bellamy and Paul D. Williams, The West and Contemporary Peace Operations, Journal
of Peace Research 46, no. 1 (2009), pp. 3957; and Brian Barbour and Brian Gorlick, Embracing
the Responsibility to Protect: A Repertoire of Measures Including Asylum for Potential Victims,
International Journal of Refugee Law 20, no. 4 (2008), pp. 53366.
Jonathan Symons, International Legitimacy: Contesting the Domestic Analogy (Ph.D. thesis, University
of Melbourne, 2007), p. 121. The mimicry label is taken from Alistair Iain Johnston, Social States: China
in International Institutions, 19802000 (Princeton: Princeton University Press, 2008).

the responsibility to protectfive years on

169