2

Country Dates Death Toll Description WarMinor Armed

Conflict

Albania 1948-1952 5,200 Execution of political opponents by Communist regime N N

Afghanistan I 1978-1992500,000-1,000,000

Adherents of old regime targeted by revolutionary government Y -

Afghanistan II 1979-1985 20,000-25,000 Afghan civilians targeted by Soviets (mainly air force but other types of killing too) Y -

Afghanistan III 1998 5,000-8,000 Regime opponents targeted by Taliban Y -

Algeria I 1962 30,000-150,000 OAS supporters (Harki) victims of retributive politicideN—(retributionimmediatelyafter war)

N

Algeria II 1992-2002 60,000-100,000Civilians targeted by both sides during Muslim fundamentalist insurrection againstFIS and GIA

Y -

Angola I 1961-1975 30,000-50,000 Civilians targeted during war of independence N Y

Angola II 1975-1994300,000 -500,000

Government vs. UNITA war Y -

Angola III 1998-2002 70,000-100,000 Government vs. UNITA war Y -

Argentina 1976-1980 9,000-20,000 Right wing crackdown on socialists—disappearances, etc.Y (1976-1977)N (1978-1980)

-

Armenia-Azerbaijan

1992-1994 7,500 Civilians killed in war Y -

Bosnia 1992-1995150,000-250,000

Civil war—main perpetrators Bosnian Serbs Y -

Bulgaria 1945-1960 20,000-100,000 Opponents of Communist regime killed or killed in forced labour camps N N

Myanmar I 1948-1951 8,000 Ethnic violence against Karen NN (1948) Y (1949-1951)

Myanmar II 1978-1980 5,000 Crackdown against communists N Y

Appendix 1: Mass Atrocities and Armed Conflict 1945-2010

3


Conflict

Myanmar III 1981-1988 5,000-8,500 Ethnic violence against Karen N Y

Burundi I 1969 10,000-50,000 Hutu killed by Tutsi N N

Burundi II 1972-1973 80,000-210,000 Hutu killed by Tutsi N N

Burundi III 1988 5,000-20,000 N N

Burundi IV 1993- 50,000-250,000 Post-election killing N N

Cambodia I 1969-1973400,000-700,000

Civil war and US bombing Y -

Cambodia II 1975-19791,000,000-3,500,000

Khmer Rouge genocide Y -

Chad 1982-1990 40,000 Habre regime killing of political opponents Y -

Chile I 1973 5,000-20,000 Pinochet coup NY (battles associated with coup)

Chile II 1973-1976 5,000-10,000 Pinochet crackdown on socialists N N

China I 1946-19491,400,000-4,900,000

Civilians killed in civil war Y -

China II 1959 65,000 Tibet—civilians killed in crackdown Y -

China III 1966-1975400,000-850,000

Cultural revolution N N

Colombia I 1946-1958110,000-200,000

La Violencia—Political violence N N

Colombia II 1975- 50,000-200,000 Government and Communist FARC rebels—both sides have targeted civilians Y -

Congo-Brazzaville 1997-1999 5,000-11,000 Coup and civil war Y -

DRC (Zaire) I 1964-1965 10,000-50,000 Post-independence political violence Y -

DRC (Zaire) II 1977-1979 5,000-8,000 Angolan rebel attacks on Katanga Y (1977-78) -

DRC (Zaire) III 1992-1994 7,000 Ethnic clashes in Masisi (Hutu vs. Nyanga) N N

DRC IV 1997-20031,500,000-2,500,000

Civilians killed in war by all sides Y (1996-2001) -

East Germany 1949-1989 40,000-100,000Killing of former POWs from WWII and subsequent killing of political opponents inprison

N N

4


Conflict

Equatorial Guinea 1969-1979 50,000-100,000 Francisco Masias Nguema regime—killing of opponents, etc.N (violent revoltin 1979)

N

El Salvador 1980-1989 30,000-60,000 Civil war and targeting of civilians - killings principally by government Y -

Ethiopia I 1974-1991 30,000-200,000 Mengistu repression NN (except1976-1983)

Ethiopia II 1972-1978 30,000-38,000 2nd Ogaden War N (except 1977)N (except 1976and 1978)

Ethiopia III 1983-1985 1,000,000 Regime-induced famine mainly in Tigray region Y -

Ethiopia-Eritrea War

1965-1993 50,000-500,000 Civilians targeted as part of war of succession Y -

Guatemala 1978-1990 60,000-200,000 Civil war developed into genocide against some indigenous populations Y -

Guinea 1970-1984 6,000-35,000 Victims of a purge by the Sekou Toure regime N N

Haiti 1957-1986 10,000-60,000 Political repression by Papa Doc and Duvalier regimes N N

India 1947200,000-1,000,000

Rioting and ethnic violence associated with partition N N

Indonesia I 1965-1966300,000-600,000

Suspected communists killed in aftermath of attempted coup N N

Indonesia II (East Timor)

1975-1992100,000-200,000

Invasion and suppression of E. Timor by IndonesiaY (1975-1978)N (1979-1992)

Y (1979-1992)

Iran I 1953-1979 6,000-16,000 Civilian opponents killed by Shah regimeN (ends withwar 1979)

N

Iran II 1981-1992 10,000-50,000 Islamic revolution and crackdown on opponents and KurdsY (1979-1982and 1986-2001)

-

Iran-Iraq 1980-1988 12,420 Killing of Iranian civilians by Iraq Y -

Iraq I 1963-1975 30,000-60,000 Government suppression of Kurdish separatistsY (1961-3,1965-66, 1969,1974-5)

Y (other years)

Iraq II 1979-2003 60,000 General suppression (not including Anfal and Shiite campaigns) N Y

Iraq III 1988-1991 180,000 Campaign against Kurds in AnfalY (1988 and1991)

Y (1989-1990)

5


Conflict

Iraq IV 1991 40,000 Campaign against Shiites in the south Y -

Iraq V 2003- 22,000-25,000 Civilian deaths caused by anti-US insurgents and terrorists Y -

Korea 1950-1953535,000-1,129,000

Killing of civilians by N. Korea and US/UN in Korean war Y -

Laos I 1963-1965 18,000-20,000 Genocide of Meo tribesmen Y -

Laos II 1966-1972 35,000-200,000 US bombing and civil war Y -

Lebanon 1982 10,000 Killing of civilians by Israel Y -

Liberia I 1989-1997100,000-200,000

Charles Taylor and civil war—government v. National Patriotic Front Y (to 1995) -

Liberia II 2002-2003 50,000-150,000 Resumption of civil war and LURD insurgency Y -

Malawi 1964-1994 6,000-17,000 Opponents of Banda regime N N

Mozambique 1975-1992100,000-200,000

Civilians targeted by RENAMO during civil war Y -

Nigeria I 1966 9,000-30,000 Ibos in the north killed after coup Y -

Nigeria II (Biafra) 1967-1970600,0000-2,000,000

Government crackdown on rebellion in Biafra Y -

Nigeria III 2001-2004 54,000Civilians killed in communal clashes between Christians and Muslims in centralplateau state

N (except 2004) N

North Korea I 1948-1,200,000-1,700,000

Communist government repression (excluding famine)

Y (except 19491953)—war accelerates repression

-

North Korea II 1995-19982,000,000-3,500,000

Regime induced N N

Peru 1980-2000 30,000-50,000 Shining Path insurrection and government crackdown Y (1981-1988) -

Poland 1946-1949 8,000-30,000 Civilians killed as new Communist regime asserts its authority N N

Pakistan I(East)/Bangladesh

19711,000,000-3,000,000

Attacks on non-Muslim population by Muslim majority Y -

6


Conflict

Pakistan II 1973-1977 5,000-10,000 Government vs. Baluchi/Pathan separatists Y (1974) Y (1975-77)

Philippines 1972-1976 10,000-50,000 Guerrilla war against communists Y -

Romania 1948-1989 60,000-150,000 Politicide and forced labour by Communist regime N N

Rwanda I 1963-1964 12,000-20,000 Government oppression of Tutsi N N

Rwanda II 1994500,000-1,000,000 (usu-ally 800,000)

Rwandan genocide—Tutsi killed by Hutu Y (1990-2002) -

Rwanda III 1994-1995 25,000 RPF revenge killings Y (1990-2002) -

Russia (Chechnya) I

1994-1996 20,000-40,000 Chechen civilians killed by indiscriminate Russian attacks Y -

Russia (Chechnya) II

1999- 5,000-50,000 Chechen civilians killed by indiscriminate Russian attacks Y (1999-2001) Y (2001-)

Somalia I 1969-1990 50,000-60,000 Isaaks killed by Said Barre regime N (except 1978) N

Somalia II 1991-1993 10,000-50,000 Collapse of state and anarchy—warlord attacks on civilians Y -

Somalia III 2005-ongoing 5,000-50,000 Civil war between Government/Ethiopia and UIC/al-Shabaab Y -

Sierra Leone 1991-2002 50,000-100,000 Civil war—civilians killed by all sides—RUF, government, Kamajors Y -

Sri Lanka 1989-1990 13,000-30,000 Civil war—government crackdown on JVP Y -

Sudan I 1956-1972400,000-600,000

Government repression followed by civil war N (1956-1962)Y (1963-1972)

-

Sudan II 1983-2002 2,000,000North-south civil war. Both sides attack civilians, but especially the government—policy of extermination

Y -

Sudan III (Darfur) 2003-ongoing 250,000 Darfur—crackdown on rebels by govt and Janjaweed Y -

Syria 1981-1982 10,000-30,000 Hama massacre of conservative Muslims by al-Assad regime Y (1979-1982) -

Taiwan 1947 10,000-40,000 Crackdown on Taiwanese insurgents Y -

Uganda I 1971-1979100,000-400,000

Idi Amin government wages war on opponents Y -

7


Conflict

Uganda II 1980-1986200,000-300,000

Civil war between Obote government and National Resistance Army Y (1981-1988) Y (1980)

Uganda III 1987- 30,000-100,000 Lord’s Resistance Army Y (1987-1989) Y (1990-)

South Vietnam I 1955-1963 30,000-50,000 Killing of opponents of Diem regime Y -

South Vietnam II 1965-1975400,000-500,000

Atrocities against regime opponents and suspected communist sympathizers Y -

Vietnam I 1946-1952125,000-250,000

French counterinsurgency war against Viet Minh nationalists/communists Y -

Vietnam II 1957-1972 37,000-51,000 Civilians assassinated by communists during Vietnam War Y -

Vietnam III 1957-1972 50,000-220,000Civilians killed by indiscriminate bombing, artillery fire, and other means by US andallies during Vietnam War

Y -

Vietnam IV 1975-1990230,000-430,000

Reprisals/retribution/ forced movement by the Communist regime in former S. Vietnam

N (except 1979)N (1975-1977)Y (1978-1988)

Yugoslavia I 1945-1946 60,000-500,000 Reprisals by communists after end of WWIIN (immediatelyafter war)

N

Yugoslavia II 1991 25,000-50,000 Ethnic cleansing of Croats by Serbs Y -

Yugoslavia III 1998-1999 10,000 Ethnic cleansing of Kosovar Albanians by Serbs Y -

Zanzibar 1964 5,000-17,000 Massacre of Arabs N N

This is not a definitive list. Qualifying conditions:• In excess of 5,000 civilian deaths (according to low estimates)• Evidence that significant part of those deaths were deliberate• This includes: deaths caused by induced famine (i.e. famines in places withoutoverall food shortage) and death caused by indiscriminate targeting

8

Appendix 2: Prevention AgendasConflict

PreventionConflict

Within R2PGenocide R2P

SYSTEMIC Regulate trade in natural resources *

Regulate trade in arms *

Regulate trade in illegal narcotics *

Address HIV/Aids pandemic *

Address and mitigate environmental degradation *

STRUCTURAL Economic Tackling deprivation/poverty * * *

Reduce iniquities (esp. horizontal) * * *

Promote economic growth * * *

Better terms of trade/trade openness * * *

Structural reform * * *

Technical assistance * * *

Support community development/ownership * * *

Resilience to economic shocks *

Governance Institutional capacity and social services (inc. health systems) * * * *

Democracy (inc. electoral assistance) * * * *

Diffusing/sharing power/dealing with factionalised elite * * * *

Strengthen independent judiciaries/legal reform * * *

Exclusionist ideology *

Eradication of corruption * * *

Indigenous conflict resolution capacity * *

Security Rule of law * * * *

Addressing impunity (inc. criminal prosecution, truth and rec-onciliation)

* * *

Security sector reform (inc. training, civilian control, budgetcontrol)

* * * *

9

Conflict Prevention

Conflict Within R2P

Genocide R2P

Arms control and disarmament (esp. SALW)/inc. controllingmilitia

* * * *

Human RightsProtection of fundamental human rights/capacity building(inc. establishing HRIs)

* * *

Protection for minorities * * * *

Extend ratification of ICC * * *

Social Intergroup confidence building (inc. interfaith dialogue) * * *

Strengthen civil society * * * *

Press freedom/public participation * * *

Tackle incitement/hate speech * * *

Education for tolerance * * * *

Empowerment/Respect women’s rights * *

Protection of children * *

DIRECTStrengthenearly warning

United Nations * * *

National *

Diplomatic Fact-finding missions * * * *

Friends groups * * *

Eminent persons groups/envoys * * *

Good offices/mediation/facilitation * * * *

Arbitration (inc. ICJ) * * *

Track 2/unofficial dialogue * *

Human rights monitoring and reporting * *

Peace commissions *

Support indigenous conflict resolution * *

10

Conflict Prevention

Conflict Within R2P

Genocide R2P

Sanctions Travel bans * * *

Trade and arms embargoes * * *

Diplomatic isolation * * *

Asset restrictions * * *

Economic * *

Inducements Economic/trade incentives * * * *

Political * * * *

Military * *

Military Preventive deployment * * *

Stand-off reconnaissance/surveillance * *

Capacity for rapidly deployable intervention/threat * * *

Direct prevention of incitement (jamming) * *

Legal ICC referral (threat of) * * *

Humanitarian Relief aid *

Refugee and IDP assistance * *

Support children’s needs * *

Risk Factor Preventive Measures

1. Intergroup relations, including record of discrimination andhuman rights violations committed against a group

• Protection of fundamental human rights and building national capacity

• Specific protection of minority rights

• Intergroup confidence building, including interfaith dialogue

• Strengthening and supporting civil society

• Establishing freedom of the press

• Preventing and punishing incitement and hate speech

• Education for tolerance

Appendix 3: Risk of Genocide and the Common Prevention Agenda

11

Risk Factor Preventive Measures

2. Absence of structures to protect the population, includinglegislative protection, judiciary, human rights institutions,presence of international actors, neutral security forces, in-dependent media

• Strengthening the independence of judiciaries

• Strengthen rule of law

• End impunity

• Security sector reform

• Protection of fundamental human rights and building national capacity

• Specific protection of minority rights

• Establishing freedom of the press


3. Presence of illegal arms and armed elements• Arms control, disarmament, and management with particular reference to small arms

• Full range of direct measures

4. Motivation of national leaders, acts which encourage divisions

• Supporting the diffusion or sharing of power


• Freedom of press

• Strengthened civil society

• Inducements and sanctions

• Referral to ICC

5. Circumstances that facilitate perpetration of genocide• Structural measures reduce likelihood of these circumstances arising

• Direct measures mitigate their effect

6. Commission of genocidal acts

• Security Sector Reform

• Reduce impunity

• Direct prevention

• Escalation prevention

7. Intent to destroy whole group


• Direct prevention of hate speech

• Referral to ICC

8. Triggering factors (elections, adverse regime change, armedconflict, natural disaster)

• Building institutional capacity and ensuring delivery of social services

• Strengthening and supporting democracy

• Supporting the diffusion or sharing of power


• Strengthening indigenous conflict resolution capacity

Policy Analysis Briefs are thought-provoking contributions to the public debate over peace and security issues. The views expressed in this brief arethose of the author and not necessarily those of the Stanley Foundation. The author’s affiliation is listed for identification purposes only. Thanksto Sara E. Davies, Rachel Gerber, Gillian Kitley, Edward C. Luck, Serena Sharma, Eli Stamnes, Paul D. Williams, and Lawrence Woocher for veryhelpful comments on this paper. All errors of fact and interpretation are, of course, my own.

Mass Atrocities and Armed Conflict: Links, Distinctions,and Implications for the Responsibility to Prevent

Feburary 2011

Alex J. BellamyAlex J. Bellamy is Professor of International Security at the Griffith Asia Institute/Centre for Governance and Public Policy,Griffith University, Australia. From 2007 to 2010, he was Executive Director of the Asia-Pacific Centre for the Responsibilityto Protect and is currently co-chair of the Council for Security Cooperation in the Asia-Pacific Study Group on theResponsibility to Protect. He is co-editor of the journal Global Responsibility to Protect and his most recent books includeResponsibility to Protect: The Global Effort to End Mass Atrocities (Polity, 2009) and (with Paul D. Williams and StuartGriffin) Understanding Peacekeeping (2nd edition, Polity, 2010).

Key Points• Genocide and mass atrocity prevention requires an

“atrocity prevention lens” to inform and, whereappropriate, direct policy development and deci-sion making across the full spectrum of prevention-related activities.

• The relationship between armed conflict and massatrocities is highly complex and not yet well-under-stood. The strong empirical correlation between thetwo phenomena implies a direct link. However, notall conflicts give rise to mass atrocities, and manyatrocities occur in the absence of armed struggle.

• While there can be no meaningful and effective agen-da for the prevention of genocide and mass atrocitiesthat does not incorporate the prevention of armedconflict, atrocity prevention requires tailored engage-ment that targets both peacetime atrocities and thosecommitted within a context of armed conflict.

• The existing common prevention agenda, whichencompasses structural and direct conflict preven-tion, outlines the measures and programs appropri-ate to the prevention of both armed conflict and massatrocity crimes.

• However, the common prevention agenda points onlyto the most common measures that might be used andthe preventive capacities that are required. It does notindicate the appropriate balance of measures in agiven context or how those measures should be used.

• While the tools used to prevent mass atrocities andarmed conflict might be the same, their objectives aredifferent. The key to a more targeted approach togenocide and mass atrocities lies in using the tools inan appropriate and context-sensitive fashion.

• When mobilized for atrocity prevention, common pre-vention measures must be used appropriately to targetatrocity risk and avert the pitfalls of a conflict preven-tion-dominant mindset, such as a blind culture of neu-trality that treats all parties as morally equivalent, thepursuit of negative peace at any price in the face of acredible threat of atrocities, and the tendency tobelieve that prevention ends when violence begins.

• While an atrocity prevention lens would help outlineits broad parameters, the key to narrowing the atroc-ity prevention agenda lies in identifying strategiesthat target specific risks and capacity gaps in particu-lar country and/or regional contexts.

It is widely accepted that when it comes to genocideand mass atrocities, prevention is better than cure.1

Preventing atrocities saves lives, is less expensive thanreaction and rebuilding, and raises fewer difficultquestions about state sovereignty and noninterfer-ence. Little wonder that in 2001 the InternationalCommission on Intervention and State Sovereignty(ICISS) declared prevention to be the single mostimportant dimension of the Responsibility to Protect(R2P), an assertion often repeated since, including bymember states.2

However, it has proven difficult to translate rhetor-ical support for the prevention of genocide andmass atrocities into a cohesive strategy.3 At theheart of the problem lay doubts about the compo-sition of R2P’s prevention agenda and its relation-ship with the United Nations’ existing work on theprevention of armed conflict and prevention ofgenocide. Should R2P give rise to a distinct preven-tion agenda, or should it be subsumed within theprevention of armed conflict (or vice versa)? Theproblem is further complicated by the UnitedNations 2004 Action Plan to Prevent Genocide,which gave rise to the establishment of the Officeof the Special Adviser on the Prevention ofGenocide (OSAPG). In 2010, the secretary-generalproposed merging this office with the SpecialAdviser for R2P, creating a new joint office(OSAPG/R2P).

The ICISS and former UN Secretary-General KofiAnnan argued that the prevention of armed con-flict should be incorporated wholesale into theR2P agenda.4 They maintained that because geno-cide and mass atrocities usually occurred within acontext of armed conflict, preventing armed con-flict would naturally reduce the incidence of geno-cide and mass atrocities. By contrast, current UNSecretary-General Ban Ki-moon and the Interna-tional Peace Institute’s 2009 “Blue Paper” on con-flict prevention and R2P have asserted that theprevention of armed conflict and of genocide andmass atrocities ought not to be conflated. Armedconflict and atrocity violence, they suggest, aredistinct problems, confirmed by the fact that massatrocities occur both within and outside the con-text of armed struggle.5

Resolving this question is impor-tant for moving the responsibilityto prevent from rhetorical commit-ment to political practice. Thisbrief argues that there can be nomeaningful and effective agendafor the prevention of genocide andmass atrocities that does not incor-porate the prevention of armedconflict and the measures com-monly associated with it. The pre-vention of genocide and massatrocities, however, requires morethan the prevention of armed con-flict—it requires tailored engage-ment targeting both peacetimeatrocities and those committed

2 within a context of armed conflict. In short, whatis required is an “atrocity prevention lens” whichinforms and, where appropriate, leads policydevelopment and decision making across the fullspectrum of prevention-related activities.

This brief outlines a number of avenues for accom-plishing these goals, including strengthening theOSAPG/R2P, ensuring an atrocity prevention per-spective in crisis decision making, developing amethodology to assess risks and needs as a founda-tion for strengthening the structural prevention ofmass atrocities, strengthening partnerships withregional arrangements, and appointing nationalfocal points to ensure that the atrocity preventionlens is applied in national decision making.

Armed Conflict, Genocide, and Mass AtrocitiesTo what extent are genocide and mass atrocitiescommitted within a context of armed conflict? Of103 episodes of mass killing (defined as a minimumof 5,000 civilians killed intentionally) observedsince 1945 (see Appendix 1), 69 cases (67%)occurred within, and 34 cases (33%) occurred out-side, a context of armed conflict. All except five ofthe peacetime cases commenced prior to 1980 andsince then only 15% of new episodes occurred out-side of armed conflict (see Figure 1). Of these, fourwere in countries that had recent experience ofarmed conflict in which mass atrocities were com-mitted (Burundi [twice], Democratic Republic ofthe Congo [DRC], and Myanmar).

Figure 1: Peacetime and Wartime Episodes of Mass Killing byDecade of Commencement: 1945-2010

25

20

15

10

5

01940s 1950s 1960s 1970s 1980s 1990s 2000s

Wartime cases

Peacetime cases

From this assessment, there is clearly a strongcorrelation between mass atrocities and armedconflict, one that has increased since the late ColdWar. Armed conflict provides an enabling contextfor most mass atrocities. This lends support tothe view that preventing armed conflict strength-ens efforts to prevent mass atrocities.6

It is important to recognize that not all armedconflicts give rise to mass atrocities.7 In fact, mostarmed groups do not massacre civilians—evenwhen they have the means and opportunity to doso.8 But this should not lead us to think that theincidence of mass atrocities can be reduced with-out action to prevent armed conflict.

International actors cannot realistically expect tolearn to distinguish, in advance, potential armedconflicts likely to generate mass atrocities fromthose that are not with sufficient confidence tosupport such a strategy.9 Studies show that gov-ernments of all types might resort to atrocitiesduring armed conflict if they consider the stakeshigh enough and fail to win at a reasonable costthrough conventional means. The likelihood ofatrocities by nonstate actors appears to be influ-enced by their relationship to the local communi-ty and intentions to signal resolve in order to geta seat at the table, factors very difficult to identi-fy in advance of armed conflict.10 A flexibleprocess of ongoing assessment alongside deter-mined efforts to prevent armed conflict is there-fore necessary. What is more, atrocity preventionshould not end when armed conflict begins.

While reaffirming a frequent correlation betweenatrocities and armed conflict, the evidence pre-sented here also supports the basic proposition ofthe alternative argument: that mass killing some-times occurs outside the context of armed con-flict. There appear to be three main forms of such“peacetime” atrocities:

1. State-directed suppression. Atrocities committedby nondemocratic regimes against opponents ormarginalized ethnic groups. The most frequenttype of “peacetime” mass atrocity, episodes usu-ally begin soon after an adverse regime change(e.g. Pinochet’s Chile) or attempted change (e.g.Zanzibar, Indonesia [1965-66]).11 However,some important episodes have occurred outsidethe context of regime contestation (e.g. China’sCultural Revolution).

2. Communal violence. Atrocities committed bygroups not organized by national governmentsor well-established nonstate armed groups.Violence is not entirely spontaneous and is usu-ally incited or orchestrated by local or nationalpolitical figures, often either state officials,politicians, or local leaders of different varieties(religious, ethnic, clan, etc.). Attacks are oftenreligious or ethnic in nature and can be trig-gered by a variety of national and local events.

3. Post-war retribution. Atrocities committed bystates and nonstate actors in the immediateaftermath of armed conflict as retributionagainst former enemy groups. Sometimes mas-sacres are intended to avenge specific atrocitiescommitted during the armed conflict (e.g. east-ern DRC), but are also perpetrated for a com-bination of political reasons not directlyconnected to the commission of prior atrocities(e.g. preemptively eliminating threats/perceivedfuture risk).

Many of the recent, smaller-scale cases of masskilling discussed in the context of R2P, such aspost-election violence in Kenya, the massacre ofcivilians by government troops in Guinea, andkilling of Uzbeks in Kyrgyzstan, occurred in oneof these settings. To address these cases, R2Pdemands strategies tailored to preventing atroci-ties outside a context of armed conflict.

What is required, as Lawrence Woocher recentlyargued, is an approach to preventing genocideand mass atrocities that reduces the risk of armedconflict (thereby reducing the primary enablingcontext), addresses the risk of peacetime atroci-ties, and includes steps to prevent atrocities with-in armed conflict. Utilizing the distinctionbetween structural/root cause prevention anddirect/operational prevention that is common tomost prevention frameworks gives us an atrocityprevention agenda with three main components(Table 1, page 4).

Because atrocities stem from numerous multilay-ered factors and incentives, our approach to pre-vention should be similarly multilayered. Wemight think of this in terms of Swiss cheese.12

Individual slices of Swiss cheese represent layersof societal resilience and preventive action. Holesin the Swiss cheese represent failings in the layersof (local, national, international) resistance toatrocities. Atrocities occur when the holes

3

ic mandate. Not only are the conflict preventionand R2P agendas similar in size, they espousealmost identical sets of measures. (Table 2 listsmeasures common to at least three of the agendas,including R2P.) Both contain structural and directmeasures to tackle underlying risks and imminentcrises respectively; structural measures commonto both include economic, governance, security,human rights, and social dimensions; they espousea similar range of diplomatic, economic, military,and legal measures to prevent imminent violence;and they endorse strengthening the UnitedNations’ capacity for early warning.

Why has it not been possible to identify a nar-rower set of measures that targets the preventionof mass atrocities?

First, the structural conditions that give rise toarmed conflict are similar to those that give riseto genocide and mass atrocities. Both requireintergroup competition and conflict, the presenceof political, economic, and/or social grievances, apreparedness and capacity (of at least one party)to use violence and violate human rights, and aperceived absence of legitimate pathways fornonviolent conflict resolution.

Second, these structural conditions are intercon-nected and cannot be effectively isolated. Forexample, in a context where group rivalry is ani-mated by perceived horizontal inequalities, theproblem is unlikely to be solved without someimprovement in the real economic, social, andpolitical opportunities open to group members.14

If it is necessary to reduce horizontal inequalitiesamong groups, then the full range of economicand governance factors come into play.

momentarily align, allowing the hazard to passthrough the defenses.

Given that, like Swiss cheese, human systemsalways have holes, the most effective way ofreducing risk is to introduce additional layers ofprotection. By including the structural and directprevention of armed conflict, the prevention ofpeacetime atrocities, and prevention of atrocitieswithin armed conflict within our system of atroc-ity prevention, we add layers of cheese and reducethe likelihood of the holes momentarily aligning.

A Common Prevention Agenda To understand where atrocity prevention, genocideprevention, and conflict prevention overlap anddiverge, we need to understand more about theircontent. This section analyses the principal meas-ures called for by four prevention agendas: (1) theprevention of armed conflict (as presented by UNSecretary-General Kofi Annan and Carnegie Com-mission on Preventing Deadly Conflict); (2) theprevention of armed conflict as part of R2P (aspresented by ICISS and Gareth Evans); (3) the pre-vention of genocide (as presented by the Office ofthe Special Adviser on the Prevention of Genocide,the Genocide Prevention Task Force, and leadingexpert Barbara Harff); and (4) the prevention ofR2P crimes and violations (as presented by UNSecretary-General Ban Ki-moon and the Asia-Pacific Centre for the Responsibility to Protect(APR2P). (See Appendix 2.)

There is a strong degree of overlap among the fouragendas. The prevention of armed conflict andR2P agendas are very similar in size and content,while the genocide prevention agenda is some-what narrower, perhaps reflecting its more specif-

4

CONTEXTS AGENDAS

• Evidence of general risk of violent conflict Structural Prevention

Measures to reduce risk of armed conflict; Mitigation of risk of peacetime atrocities

• Imminent armed conflict

• Serious political/social instability

Direct Prevention Prevent imminent armed conflict

or peacetime atrocities

• Armed conflictEscalation Prevention

Prevent armed groups from committing atrocities

Table 1: Atrocity Prevention

5Table 2: The Common Prevention Agenda13

STRUCTURAL PREVENTION DIRECT PREVENTION

Economic Measures • Reducing deprivation and poverty.• Reducing inequalities, especially horizontal.• Promoting economic growth.• Supporting structural reform.• Providing technical assistance.• Improving the terms of trade and trade openness.• Supporting community development and local

ownership.

Governance Measures• Building institutional capacity and ensuring

delivery of social services.• Strengthening and supporting democracy.• Supporting the diffusion or sharing of power.• Strengthening the independence of judiciaries.• Eradicating corruption.• Strengthening local conflict resolution capacity.

Security Measures • Strengthening rule of law.• Ending/preventing impunity.• Reforming the security sector.• Encouraging disarmament and effective arms

control/management with particular reference tosmall arms.

Human Rights Measures • Protecting fundamental human rights and build-

ing national capacity, with specific protection ofminority, women, and children’s rights.

• Supporting the work of the International CriminalCourt.

Social Measures• Intergroup confidence building, including

interfaith dialogue.• Strengthening and supporting civil society.• Establishing freedom of the press.• Preventing and punishing incitement and hate

speech.• Educating on diversity and tolerance.

Early Warning• Establishing a UN early warning and assessment

capacity.

Diplomatic Measures• Fact-finding.• Forming “groups of friends” among UN member-

ship.• Deploying eminent persons/envoys.• Exercising the good offices of the secretary-

general.• Pursuing arbitration (including International

Court of Justice).• Supporting indigenous conflict resolution

processes.

Sanctions• Banning travel.• Embargoing trade and arms. • Freezing assets.• Imposing diplomatic sanctions.

Inducements• Promoting economic or trade incentives.• Offering political inducements.

Military Measures• Mobilizing preventive deployments.• Developing and/or threatening rapid deployment

capability.• Jamming and other means of preventing

incitement.

Legal• Referring matter to the International Criminal

Court.

ESCALATION PREVENTION

• This agenda has not yet been articulated.

Armed conflict can create new incentives andopportunities to target civilians. Once a conflicthas begun, for example, target populations asso-ciated with combatants to the conflict are oftenleft exposed, virtually unarmed or too weak toprotect themselves (e.g. Tibet [1959], East Timor[1976-77], and Guatemala). In such cases, effec-tive measures to prevent armed conflict wouldalso prevent the commission of mass atrocities.As soon as a conflict begins, however, new risksand vulnerabilities develop. In these situations,the distinction between prevention and reactionbecomes fuzzy.17

Moreover, escalation prevention raises difficultquestions about the appropriate relationshipbetween humanitarian action, the protection ofcivilians, and the prevention of genocide andmass atrocities. Awarding humanitarian action apreventive role is especially fraught with dangerand complexity given the centrality of concepts ofneutrality and impartiality to humanitarianwork. But if we take seriously the view that atroc-ity prevention does not end when armed conflictbegins, we need to tackle these and other thornyquestions. More work is needed to identify con-cepts and strategies for preventing atrocities with-in a context of armed conflict and to build aconsensus around them.18

Preventing Peacetime AtrocitiesWhile it has little to say about preventing theescalation of armed conflict, the common agendadoes provide a framework for preventing peace-time atrocities. Properly conceived, structuralprevention reduces the risk of both armed conflictand peacetime atrocities. To demonstrate this wecould ask whether the common agenda sufficient-ly addresses the risk factors associated with geno-cide that have been identified by the OSAPG,which establish risk irrespective of the presence ofarmed conflict (Appendix 3). Doing this, we findthat the common agenda articulates a range ofmitigating factors for each of the risk factorsidentified by the OSAPG.

Alternatively, we could revisit the three circum-stances in which peacetime atrocities occur andexamine the common agenda’s suitability:

Scenario A: State-directed suppression. If a regimeis committed to abusing and killing its population,no amount of structural prevention can dissuadeit. This assertion is largely true but not wholly so,

Here again, atrocity prevention and the more gen-eral conflict prevention and economic develop-ment agendas overlap. Redistributing slices of theeconomic pie so that disadvantaged groups havegreater equality of opportunity, thereby reducingintergroup tensions that can precede mass atroci-ties, is more palatable for privileged groups if theeconomic pie itself is growing—suggesting thatredistribution goes hand-in-hand with a more tra-ditional focus on overall growth. Furthermore, theperception of horizontal inequalities is sometimescreated and exploited by political figures to servetheir own purposes. Thus, a range of measuresaddressing political leadership, hate speech andincitement, and intercommunal trust and under-standing might also be needed.

Third, direct prevention of all varieties is aboutchanging the behavior of specific actors, and wehave a good idea about the repertoire of measuresinternational players can use to induce, persuade,or coerce actors they engage. That said, while thetools used to prevent mass atrocities and armedconflict might be the same, their objectives aredifferent. Atrocity prevention targets a specificactor and seeks to dissuade it from committingatrocities. Conflict prevention/resolution targetsseveral actors in the hope of arriving at a consen-sual agreement among them.15 This distinction isimportant, and carries significant implications forpolicy development.

Fourth, it makes no sense to consider reducingthe range of possible policy options. In mostcases, policymakers already have too fewoptions from which to choose in order to pre-vent mass atrocities. As the secretary-generalhas made clear, policymakers benefit most froma broad range of available options and the flex-ibility to apply them in ways appropriate to thespecific context.16 This not only requires thearticulation of a broad-based toolkit but also,and more importantly, the strengthening ofinternational capacity to give practical meaningto those options.

Preventing Escalation to AtrocitiesWe know comparatively little about preventingatrocities within the context of ongoing armed con-flict. Preventing armed and actively engaged com-batants from resorting to mass, civilian-targetedviolence is an area that is conceptually and practi-cally underdeveloped.

6

because structural prevention promises to reducethe incidence of abusive government in the firstplace by supporting democracy, the rule of law,and other barriers to it. Assuming this ship hassailed, however, there is little that structural pre-vention can do to avert atrocities. The best thatcan be done is to detect the risk through earlywarning and utilize direct measures to: (1) per-suade, deter, induce, or coerce the potential perpe-trators; or (2) to deny them the means toperpetrate atrocity crimes. The primary measuresto do so are already part of the common preven-tion agenda.

Scenario B: Communal violence. Several aspects ofthe common agenda speak directly to this type ofproblem, not least those relating to strengtheningthe capacity of local security forces to fulfillnational directives to protect populations throughsecurity sector reform, deterring violations by end-ing impunity and reinforcing the rule of law,dampening communal tensions by reducing eco-nomic inequities, creating economic opportunities,building confidence and trust among groups, edu-cating for tolerance, and supporting local path-ways for conflict resolution. If the risk persists,direct measures such as the use of envoys, induce-ments, and peacekeepers might be called upon.

Scenario C: Post-war retribution. The third sce-nario is similar to Scenario A and likely to requiresimilar measures because reprisal killings are usu-ally committed by the victors—and hence by gov-ernments themselves. Sometimes, however, thenational government itself is not directly involvedin retributive killing (e.g. Algeria, 1962), openingopportunities for prevention through immediatesecurity assistance, assistance to strengthen thesecurity sector and rule of law, and the use ofdiplomacy, sanctions, and inducements as neces-sary to encourage the state to step in to protectthe victims.

This exercise at least demonstrates that the pro-grammatic content of the common preventionagenda is as well suited to the prevention ofpeacetime atrocities as it is to the prevention ofarmed conflicts that can give rise to atrocities.The prevention of mass atrocities does not, there-fore, require an entirely new agenda set apartfrom existing prevention agendas, nor is there aneed to radically rethink the core components ofprevention for R2P purposes.

However, while preventing armed conflict is animportant component of preventing mass atroci-ties, and while there is a common repertoire ofmeasures that might be used for both purposes,there is a critical need to ensure that activities areguided by an atrocity prevention lens and arecarefully tailored to the unique objectives andcontext of mass atrocity scenarios.

This conclusion is grounded in the followingobservations:

• There is no one-size-fits-all approach to pre-vention. Prevention strategies must be careful-ly tailored to each country’s individual needs,defined by a contextual assessment of risksand vulnerabilities that accounts for interna-tional political will and the availability ofinternational capacity.

• The common prevention agenda points only tothe most common measures that might be usedand preventive capacities that are required. Itdoes not indicate the appropriate balance ofmeasures in a given context or how those meas-ures should be used.

• While direct prevention measures employ com-mon tools (such as diplomacy) for the preven-tion of genocide and mass atrocities, thesetools must be used appropriately to targetatrocity risk and avert the pitfalls of a conflictprevention-dominant mindset, such as a blindculture of neutrality that treats all parties asmorally equivalent, the pursuit of negativepeace at any price in the face of a crediblethreat of atrocities, and the tendency to believethat prevention ends when violence begins.

• Early-warning analysis must pay due attentionto the circumstances that give rise to peacetimeatrocities and not fixate on armed conflict.

• Strategies and assessments should ensure thatprogress in one area (such as fostering econom-ic growth or strengthening security capacity)does not come at the expense of progress or cre-ate problems elsewhere (such as increasing per-ceived horizontal inequalities or creating newsecurity capacities that can be readily mobilizedfor atrocities).

• The key to narrowing the atrocity preventionagenda lies in identifying strategies that target

7

ing these atrocities requires direct measuresdesigned to persuade, deter, and even coerce,elites and/or impede their ability to commit suchcrimes.21 Such direct measures are rarely appro-priate for more general conflict prevention.

Conflict prevention strategies are sometimesunsuited to the prevention of mass atrocities. InBosnia (1992-1995), for example, the internation-al community prioritized conflict resolution overatrocity prevention in the face of ongoing massatrocities committed primarily by one party.Internationally sponsored negotiations gave equalweight to the views of perpetrators and victimsand an arms embargo applied to both groups,inhibiting the victims’ capacity to protect them-selves. Likewise, the international community’spreoccupation with maintaining the peace settle-ment for Rwanda agreed in the Arusha Accordsblinded it to evidence that Hutu extremists wereplanning genocide. More recently, the internation-al community privileged the Comprehensive PeaceAgreement for Sudan over the prevention of massatrocities in Darfur by refusing to adopt measuresto deter, prevent, or punish the Sudanese govern-ment’s actions.

Conflict prevention activities may inadvertentlycreate incentives to commit atrocities. When con-flict prevention takes priority, rebel groups maybe encouraged to commit atrocities during transi-tional phases in order to “earn” a seat at thenegotiating table by signalling resolve.22 Similarincentives can be prompted by the deployment ofimpartial peacekeepers. It is worth rememberingthat more civilians were killed after peacekeeperswere deployed to Bosnia, Rwanda, and the DRCthan before.23

These points underscore that while the generictools for preventing imminent armed conflictappear similar to those used for preventing immi-nent atrocities, application of each tool should beadapted to its particular context. There is a clearneed for an atrocity prevention lens to be appliedat every stage of the conflict cycle, operatingalongside and informing more general conflictprevention/resolution activities and, whereappropriate, guiding those activities.

In addition to improving preventive strategies,the inclusion of an atrocity prevention lens prom-ises to make the best use of scarce resources.Although the different prevention agendas are

specific risks and capacity gaps in particularcountries or regions.

Although the metaphor is well worn, it is usefulto think of the common prevention agenda as atoolkit that can be used in multiple ways toachieve different outcomes depending on the cir-cumstances.19 Just as artisans put identical toolsto different uses, the institutional and bureaucrat-ic arrangements established for the prevention ofarmed conflict should be used in a targeted, tai-lored, and flexible fashion to achieve maximumeffect in specific contexts.

To do this, we need to know more about how andby whom the tools might be used. What is more,while we know what general tools we ought tosee in the kit, we do not know how full the kit is,or whether all the tools are in working order.Finally, we need a clear idea of the jobs that needdoing—not only the immediate work identifiedby early warning and assessment but also thelonger-term, pre-crisis structural work.

An Atrocity Prevention LensEffective prevention of mass atrocity crimesrequires the development of an atrocity preven-tion lens within existing institutional frameworksdedicated to the prevention of armed conflict.Such a lens would identify the risk of mass atroc-ities and advise policymakers and political leaderson the most appropriate courses of action. Thismay involve tailoring ongoing prevention workto specific risk factors in individual countries or,when risk is high and imminent, prioritizingatrocity prevention and directing the whole rangeof preventive activities toward this goal.

Although there is significant overlap betweenmass atrocities and armed conflict, as well asamong the general tools employed to preventarmed conflict and mass atrocities, the objectivesisolated in the prevention of each are different—and likely to require different strategies. Thereare three main reasons for this:

Conflict prevention and atrocity prevention havedifferent purposes. As Lawrence Woocher pointsout, while the prevention of genocide and massatrocities targets a specific actor and seeks to dis-suade it from committing atrocities, conflict pre-vention targets several actors and pursues aconsensual agreement among them.20 When animminent threat of atrocities is detected, prevent-

8

similar in scope and require analogous genericinternational capacities, it does not follow thatevery aspect of the common prevention agendaneeds to be given equal weight in every context,producing an impossibly comprehensive and bur-densome agenda unlikely to be implemented.Indeed, there may be tensions between the toolsthemselves—such as between curtailing hatespeech and supporting freedom of the press—thatcan only be resolved through dialogue amongpolicymakers in relation to specific contexts.

It is not possible to identify in the abstract a uni-versally applicable, discrete agenda for genocideand mass atrocity prevention because the relativemix of risks differs from case to case and overtime. There is no substitute for detailed examina-tion of individual countries to determine thenature and source of risks and the extent ofresilience and protective capacities (state-basedand otherwise) as a basis for determining theappropriate combination of programs, measures,and strategies that might be utilized to reduce therisk of genocide and mass atrocities.

The key to a more targeted approach, therefore,lies in using the tools in an appropriate and context-sensitive fashion. It is also important tolearn lessons from past cases and desktop exercis-es and to think about the general repertoire ofmeasures and tactics that might be employed inspecific kinds of situations in order to bridge thegap between totally generic and totally country-specific prevention strategies.

An atrocity prevention lens would identify atroc-ity risk, as well as provide insight into how bestto tailor preventive action to respond to suchrisks and avoid the pitfalls associated with prior-itizing elements of a preventive strategy ill-suitedto addressing mass atrocity dynamics. As the UNsecretary-general has recognized, an atrocity pre-vention lens should not be used to provide a one-off evaluation of a situation, but should rather beintegrated as a central element of ongoing assess-ment and dialogue with stakeholders about theemergence and treatment of risk factors.

This brings us to two questions about the appli-cation of an atrocity prevention lens. The firstrelates to how an atrocity lens might identify spe-cific risks so that policymakers can make judg-ments about which tools and strategies to adoptand, more generally, about whether to prioritize

9atrocity prevention or conflict prevention or giveequal weight to both. The second considers howinstitutional meaning might be given to the atroc-ity prevention lens.

The following sections explore these questions,first in relation to the “preconditions” for massatrocities requiring structural prevention, andthen in relation to a “path of escalation” requir-ing direct and escalation prevention.24

Mass Atrocity Preconditions andImplementing Structural PreventionThe preconditions of genocide and mass atroci-ties can be thought of as necessary but insuffi-cient risk factors that are best addressed throughthe structural measures identified in the commonprevention agenda. The presence of one or moreof these conditions is necessary for the futurecommission of genocide or mass atrocities, buttheir presence does not mean that genocide ormass atrocities are inevitable or that these crimeswill occur within a given period of time. Thepresence of several preconditions and heightenedseverity increases the likelihood of future geno-cide and mass atrocities but, theoretically at least,mass atrocities are possible with the presence ofonly one precondition and may not occur even ifall are present.

Based on the existing literature, Table 3 (see page10) identifies the most prominent risk factorsassociated with mass atrocities.

This framework is a relatively accurate indicatorof general risk. To test it, we used data availablein 1997 to conduct a risk analysis as if it were1998. We attached numerical scores to the datato represent the gravity of the problem. On a five-point scale ranging from extreme risk to negligi-ble risk, the analysis identified ten countries atextreme risk. Of those, nine did indeed succumbto atrocities over the following decade. A largernumber of countries were judged to be at “high”risk and, of these, a little over half succumbed toatrocities in the following decade (see Table 4,page 10).

The only case of mass atrocities between 1998and 2008 that did not rate “extreme” or “high”on the basis of this framework was Russia’s sec-ond war in Chechnya. This framework thereforeprovides a useful guide to the sorts of factors thatan atrocity prevention lens needs to investigate in

10 Table 3: Preconditions of Genocide and Mass Atrocities

Table 4: Risks in 1998 and Mass Arocities 1998-2008

Key Elements

Social Factors• politicization of religious or ethnic divisions.• social, economic, or political discrimination.• history of genocide and mass atrocity.

Regime Factors• human rights violations.• absence of rule of law.• absence of democracy.

Economic Factors• low GDP per capita.• low economic interdependence.• horizontal inequalities.

Armed Conflict

• presence of multiple armed groups/illicit armsflows.

• establishment of militia (government and nongovernment).

• group-based recruitment practices.• presence of armed conflict.

Threat LevelAtrocities Committed

Atrocities Not Committed

Percentage

Extreme

DRCLiberiaSudanAfghanistanIraqSierra LeoneSomaliaMyanmaRwanda

Burundi 90%

High

AngolaDPR Korea25

UgandaChadYugoslaviaIndonesiaAlgeriaNigeriaKenya Cote d’Ivoire

EthiopiaIranAzerbaijanBangladeshChinaEritreaLaosPakistanSyria

47%

order to identify a generalized risk of mass atroc-ities and the sorts of issues that ought to be takeninto account in developing a tailored program ofstructural prevention utilizing the tools identifiedin the common prevention agenda.

Using the Atrocity Lens to Tailor Structural PreventionStructural prevention programs to reduce the riskof genocide and mass atrocities before they mani-fest themselves in clear early-warning signals needto be carefully tailored to individual contexts.This should be a reflective and ongoing processthat addresses the root causes of mass atrocitiesand reduces the overall risk of armed conflict.26

Structural prevention encompasses work alreadyundertaken by different arms of the UN System aswell as by some regional arrangements, NGOs,and private businesses, and usually takes the formof economic development programs, capacitybuilding, human rights assistance, humanitarianrelief, democracy support, rule of law support,and/or security sector reform, to name a few areas.

To ensure that this ongoing work adds value to thecapacity of states and societies to reduce the risk ofgenocide and mass atrocities, existing programsmust incorporate an atrocity prevention lens andthe United Nations must “deliver as one.”

The first step to achieving this is to establish acommonly accepted methodology for assessingthe mass atrocity risks and protection needs inparticular countries and to test this methodologyin relation to actual cases. Needs assessment,which could be based on a framework similar tothe one identified above, would focus on identify-ing risks in need of mitigation and gaps in protec-tion capacity, and provide a basis for developingpreventive strategies. Among other things, needsassessments could:

• Identify the presence and nature of specific nation-al factors associated with the risk of atrocities.

• Assess sources of social resilience and stability.

• Assess national capacity and identify protectiongaps.

• Map existing preventive activities.

• Identify areas requiring additional support.

This would need to be a dynamic process ofassessment/reassessment because, in the words ofthe special adviser for R2P, “we need a movingpicture, not a snapshot…[a]ssessment entailsunderstanding the mosaic not the pieces, the pat-tern not a single act.”27

Significant progress has already been made indeveloping conflict assessment tools, and suchtools might be adapted for this purpose.28 The ini-tial research needed to develop a methodologycould be conducted by the OSAPG/R2P or by anindependent research body commissioned by theOSAPG/R2P.

From this, a methodology for needs assessmentand strategic planning could be developed by theOSAPG/R2P. This could be used by theOSAPG/R2P as an assessment and planning toolto: (1) guide inputs into other UN interagencyframeworks (such as those relating to peacebuild-ing/post-conflict reconstruction, humanitarianaffairs, economic development, human rights, orthe prevention of armed conflict) to ensure that aR2P/atrocity prevention lens is mainstreamedacross the UN System and to strengthen the orga-nization’s capacity to “deliver as one”; (2) informpartnership with regional arrangements andmember states about building capacity andstrengthening preventive action; and (3) strength-en early warning and assessment by providingadditional context and a deeper knowledge baseto inform policymaking.

The methodology might be shared with and usedby regional arrangements to guide their own pre-vention work. It might also be appropriate for theGeneral Assembly to consider how best to strength-en the structural prevention of genocide and massatrocities as part of its annual informal interactivedialogue on R2P.

Paths of Escalation andDirect/Escalation PreventionAnalyzing the preconditions can tell us whichcountries or regions are especially vulnerable, dis-tinguish situations in which the outbreak of armedconflict might be expected to produce mass atroc-ities, and help identify specific areas of vulnerabil-ity in a country that might be addressed throughtargeted structural prevention efforts. However,this is a notoriously imprecise science and suchanalysis cannot tell us whether, when, and how therisk of mass atrocities will be realized.

11

that analysts might use to identify the risk of atroc-ities and contribute atrocity-specific inputs tointernational policy development, especially exist-ing conflict prevention and mitigation efforts.

Step 1: Crisis. Most episodes of genocide or massatrocity are directly preceded by some type of cri-sis. Crises often provide actors with the reasons andopportunity to commit mass atrocities. Without acrisis, even actors predisposed toward mass atroci-ties would have little justification to commit atroc-ity crimes. Crises also tend to weaken domesticconstraints and institutional resilience, creating anenvironment more conducive to an atrocity-basedstrategy. An absence of such a crisis partly explainswhy some nondemocratic states are able to endurepersistently high levels of risk without succumbingto mass atrocities.

Four principal forms of political crisis provide thecatalyst for genocide and mass atrocities. Eachform has its own subsets and the list is notexhaustive:

1. Armed contests: civil war, external interven-tion, reneging on peace agreements.

2. Unconstitutional regime changes: coups andattempted coups, disputed elections, contestedsuccession, contested secession.

3. State incapacity: new states with low legitima-cy, failed/failing state.

4. Radical revolutionary government: ideologicalcommitment to radical transformation.

Other forms of crises such as economic collapseor natural disasters can also provide the catalystfor escalation, but usually they are the immediatecause of a political crisis, which ultimately precip-itates a process as outlined above.

Step 2: Mobilization. Although the intervalbetween the eruption of a crisis and the commis-sion of genocide and mass atrocities can be veryshort, a crisis in itself is still insufficient to sparkmass atrocities. Mass atrocities require somedegree of organization aimed at preparing andstrengthening a particular group and weaken-ing, excluding, or targeting victim groups.Typically, mobilization involves at least one ofthe following:

In order for risk to manifest as mass atrocities, atleast three additional factors need to be in place.First, there needs to be a reason to commit massatrocities. Typically, actors select mass atrocitiesas a rational strategy for pursuing their objec-tives, such as countering a serious existentialthreat, coercing compliance from or exploitingunfriendly civilians, securing a seat in interna-tional negotiations, and/or changing the ethnic orreligious composition of a particular territory. Inweighing their options, some actors make astrategic determination that viable alternativesare either nonexistent or too costly.29

Whether committed in the course of armed con-flict, to eliminate a challenge to the regime, or toeffect radical social transformation, atrocities aremeans to an end, not an end in themselves. Unlessthere is reason to think that their use might servesome purpose, even actors that are strongly pre-disposed toward committing atrocities will beunlikely to do so. This reason is usually providedby an acute crisis, but can also be generated by anelite’s ideology.

Second, potential perpetrators require the meansto commit mass atrocities. At the very least, theyrequire a sufficient number of people who areprepared to commit atrocity crimes. Third, theyrequire the opportunity to commit mass atroci-ties, whether enabled by a weakening of domesticinstitutional restraints or the support and/oracquiescence of external actors.

A useful way of thinking about these issues is toconceptualize them as processes of escalation inthree distinct stages: (1) the emergence of a crisiswhich produces social conflict; (2) the mobiliza-tion of an armed group or groups, and (3) the out-break of limited violence against civilians which, ifleft unchecked, escalates into full-blown atrocities.

Where the preconditions indicate high levels of risk,this pattern of crisis, mobilization, and violence isquite likely to produce atrocities unless stemmed byone of two principal factors: (1) the regime’s capac-ity to end the crisis before resorting to atrocities or(2) effective international engagement.

Of course, real cases are unlikely to proceed in pre-cisely this fashion and the temporal gaps may bequite long (i.e. years) or very short. The point ofthe framework is not to pinpoint when and whereatrocities will occur, but to highlight some red flags

12

• Marginalization of moderates within the elite.

• Organization of “hate groups” dedicated to thevilification of the target group.

• Purging the security forces of minority groupsand those thought disloyal, and expandingrecruitment among dominant and radicalizedgroups.

• Establishment, arming, and training of militias.

• Escalation of unpunished human rights abusesagainst targeted groups.

• Publication of hate propaganda.30

Sometimes, there are also moves to marginalizeminority groups through segregation, displace-ment, the establishment of camps or ghettos, andexclusion from schools and employment. Ofcourse, elements of mobilization are likely to beevident prior to the emergence of a crisis, espe-cially in countries suffering from high levels ofrisk. But when a crisis erupts, we might expect tosee the acceleration of mobilization if mass atroc-ities are imminent.

Step 3: Violence. It is common for organized, ifstill low-level, violence to erupt prior to the com-mission of mass atrocities. The nature of this vio-lence provides some clues as to the likelihood thatan imminent armed conflict will degenerate intogenocide or mass atrocities. Some rules of thumbcan be used here—though again this is an impre-cise science and assessment needs to be done onan ongoing and context-sensitive basis.

If one or more of the following factors are evidentin these early violent exchanges, then the poten-tial for mass atrocities ought to be considered agenuine and imminent risk:

1. Violence intentionally targeted against civilians.

2. Impunity for the perpetrators of these earlycrimes.

3. The existence of a significant threat to the sur-vival of the governing regime and failure of ini-tial attempts to counter that threat.

Early violence becomes especially indicative ofthe potential for a wider campaign of mass atroc-

ities when it takes the form of “trial massacres.”Typically, these target the victim group in rela-tively small numbers and are sometimes conduct-ed as a test for negative repercussions in the formof arrests or substantive international engage-ment. If the perpetrators enjoy impunity, this maybe seen as a “green light” to expand a genocidalor mass atrocity campaign.31 In other words, aculture of impunity will have been established.

These early massacres provide what ManusMidlarsky refers to as “continuity and validation”:(1) they give future perpetrators something to iden-tify with; and (2) they are evidence that the com-mission of further massacres are likely to gounpunished.32 This creates a cycle of impunity thatfacilitates the escalation of massacres (see Figure 2).

Figure 2: The Cycle of Impunity

Another rule of thumb that policymakers oughtto be aware of is that the greater the instability,the bigger are the opportunities for armedgroups to target large numbers of civilians andwhole groups—and to hide this beneath the gen-eralized chaos.33 For these reasons, understand-ing the nature of early violence is absolutelycrucial in terms of identifying the risk of immi-nent mass atrocities.

This brief framework identifies a series of factorsassociated with the risk of mass atrocities and anumber of red flags that might give warning ofthe potential for mass atrocities in the midst of

13

Commission of atrocities

Perpetrators praised,rewarded, or leftunpunished

Culture of impunity established

sion and use a prevention lens in their program-ming. Many regional arrangements, individualstates, and nongovernmental organizations havealso consciously developed conflict prevention pro-grams and others conduct prevention-relevantwork, though without labeling it as such.

While there is plenty of prevention-related activi-ty, there is relatively little coordination and noclear picture of global preventive capacities. Toknow where the capacity gaps are, we need acomprehensive audit of global preventive capaci-ty covering structural, direct, and escalation pre-vention and including assessment of (a) thecapacities of global institutions, regional arrange-ments, individual states, and global civil society;(b) how those capacities are currently used andhow they might be employed to prevent atroci-ties; (c) the level of commitment to employingthese capacities for preventive purposes; and (d)the extent of coordination and coherence.

In addition to identifying capacity gaps, mappingexisting and potential global preventive capacitywould itself go some way toward enabling coor-dination by informing prevention actors aboutthe activities of their peers.

At the United Nations. It is important that whencrises emerge, an atrocity-focused perspective isprovided to guide every stage of decision making.Within the United Nations, an atrocity-specificoffice such as the OSAPG/R2P is well placed toassess the risk of mass atrocities and recommendstrategies for addressing the risks within existingprograms or through new modes of engagement.The OSAPG/R2P would participate in UnitedNations’ decision making about conflict preven-tion and engage in dialogue with relevant memberstates and regional arrangements. Through thisdialogue, the OSAPG/R2P would provide specificadvice about the likely risks of mass atrocities andthe policies and strategies that ought to be adopt-ed. This would include advice about the potentialfor conflict prevention measures (such as blindimpartiality) to inadvertently encourage or rewardthe perpetration of mass atrocities.

Where the risk of mass atrocities is thought to behigh and its realization imminent, the OSAPG/R2Pmight argue for the prioritization of atrocity pre-vention. In high-risk situations where time is limit-ed, this would be the most effective way ofdeveloping coordinated atrocity-specific responses

major political crises. Of course, actual crises arecomplex and multifaceted. It is likely to be verydifficult to distinguish the warning signs of armedconflict from the signs of mass atrocities until arelatively late stage when the early patterns ofviolence provide a clear indication of what is tocome. Likewise, regimes that suffer a political cri-sis in a context rife with atrocity preconditionsmay not resort to an atrocity-based strategy.Frameworks such as this should be used only asgeneral guides and as aids to—not replacementsfor—detailed country assessments. Moreover,when put into practice, any framework of analy-sis should be subjected to regular review and revi-sion.

Implementing Prevention: Coordinating Structural, Direct, and Escalation-Focused EffortsPreventing genocide and mass atrocities does notrequire a new repertoire of measures, but ratherthe appropriately tailored, carefully targeted, andproperly coordinated application of existingcapacities. It is important, first of all, to identifythe risk of imminent genocide and mass atrocities,and the previous section has identified a numberof red flags that analysts might use to identifyheightened risk. The strengthening of the early-warning capacity within the OSAPG/R2P, sup-ported by a mechanism for convening discussionabout policy options within the UN Secretariat, isan important and appropriate first step.34 Ongoingneeds assessments, of the kind envisaged earlier,would play an important role in providing theknowledge base necessary to facilitate the rapiddevelopment of direct prevention policy options.

Auditing Global Prevention CapacityAlthough the United Nations has primary respon-sibility for international peace and security and isappropriately thought to be the principal vehiclethrough which to implement R2P and preventgenocide and mass atrocities, there are importantroles for regional arrangements and national gov-ernments. The first step in identifying these rolesand maximizing their contribution is to strength-en understanding of global preventive capacitiesand to pinpoint areas where there might beimportant gaps.

Since 2001, the United Nations has mainstreamedconflict prevention. Most departments and agenciesrecognize that their work has a preventive dimen-

14

that avoid the common pitfalls associated with themisapplication of tools identified earlier. In suchcircumstances, it might be appropriate for theOSAPG/R2P to assume responsibility for develop-ing a coordinated plan for a unified response, mak-ing use of all UN capacities in line with the“narrow but deep” ethos established by the secretary-general. This would occur at the requestof the secretary-general and on the basis of advicefrom the OSAPG/R2P and elsewhere. Of course,the office’s long-term credibility would depend ona cautious approach that invoked this mechanismonly when the risk was high and imminent.

In addition, direct prevention strategies should becalibrated with the longer-term structural strate-gies discussed earlier, where appropriate and fea-sible.35 In the long run, the problem of incoherencebetween short- and long-term measures might bereduced by making the same UN office responsi-ble for coordinating both sets of activities.

Matters are complicated somewhat when atroci-ties are threatened within the context of an armedconflict. It should be understood that the risk ofatrocities is heightened by the onset of armedconflict and that the path from threat to commis-sion might be very short indeed, sometimes tak-ing only hours. We need to know more about themeasures and strategies that might be employedto prevent armed conflicts degenerating intoatrocities. Greater clarity is also required regard-ing the relationships between prevention andreaction, humanitarian aid, and the civilian pro-tection agenda.

In addition, it is vital that (1) UN staff on theground are trained to detect the warning signs ofatrocities; (2) UN headquarters has the capacityto monitor information coming from the field inas close to real time as possible in order to assessthe risk of mass atrocities; and (3) there is aprocess for fast-tracking early warning, assess-ment, advice, and decision making. This mightrequire strengthening OSAPG/R2P with thecapacity to receive and assess real-time informa-tion from the field.

When atrocities are imminent, time is usuallyshort and the chances of success mixed because,to return to the Swiss cheese analogy, there arevery few layers of cheese left. Because the stakesare so high and it is so difficult to know how bestto utilize preventive measures, it is important to

learn lessons from past cases and desktop exercis-es. On the basis of this analysis, the OSAPG/R2Pcould develop and share best practice guidance.

Engaging Regional Arrangements. It is importantto strengthen collaboration between the UnitedNations and regional arrangements. Recent caseswhere prevention succeeded in stemming atroci-ties in Kenya and Guinea involved collaborationbetween the United Nations and relevant region-al arrangements (AU and AU/ECOWAS respec-tively), with the latter taking the lead. Thus, asthe UN secretary-general argued, “UnitedNations decision making concerning the respon-sibility to protect should be informed andenriched, wherever possible, by local knowledgeand perspectives, as well as by the input ofregional and subregional organizations.” He con-cluded, “This puts a premium on regularizing andfacilitating the two-way flow of information,ideas, and insights between the United Nationsand its regional and subregional partners.”36

The Secretary-General’s Special Adviser on R2P,Edward Luck, identified three areas where relationsbetween the region and the United Nations mightbe productively strengthened: (1) two-way provi-sion of R2P relevant information and assessment;(2) cooperation in responding to imminent emer-gencies; (3) facilitating cooperation between theregion and the United Nations in supporting oper-ations authorized by the UN Security Council.37

Regional arrangements should also be fullyincorporated into the development of strategicplans for structural prevention. This partnershipmight be strengthened through direct coopera-tion between OSAPG/R2P and regional lead-ers/officials, and by developing the role of theUnited Nations regional offices such as the UNOffice for West Africa (UNOWA) and UNEconomic and Social Commission for Asia andthe Pacific (ESCAP).

National Governments. National governments arepivotal to the implementation of R2P and preven-tion of genocide and mass atrocities. Besides ful-filling their own, internal responsibility to protect,ensuring that the United Nations and regionalarrangements have the political support andresources they need to implement their atrocityprevention plans, and making resources availableto preventive efforts when international action isneeded, national governments should also give

15

nate national responses, thereby underminingthe role and priority of the atrocity lens innational decision making.

Although a modest first step, the appointment ofnational focal points creates the potential for anatrocity prevention lens to inform national-leveldecision making. Over time, this would strength-en national and regional approaches to prevent-ing mass atrocities, as well as facilitate the UnitedNations activities in this area.

Conclusion and Recommendations

General 1. Significant overlap between the tools needed to

prevent armed conflict and those required toprevent genocide and mass atrocities suggestsno need for a fundamental redevelopment ofpreventive measures. Instead, the prevention ofmass atrocities requires application of an atroc-ity prevention lens to wider prevention efforts,including detailed analysis of specific situationsand the careful tailoring of commonlyacknowledged preventive tools to suit their par-ticular context and purpose.

2. Armed conflict dramatically increases the riskof mass atrocities. Preventing atrocities there-fore constitutes a call to deliver on past com-mitments in the field of conflict prevention.However, it also demands focus on the preven-tion of peacetime atrocities and atrocities with-in armed conflict. An atrocity prevention lenswould ensure that concerns and analysis spe-cific to the prevention of mass atrocities isincorporated throughout the internationalcommunity’s engagement with conflict preven-tion and management.

3. The principal purposes of an atrocity preven-tion lens are to (1) identify the risk of massatrocities; and (2) advise policymakers andpolitical leaders on the most appropriatecourses of action. This may involve tailoringongoing prevention work to address specificrisk factors in individual countries or, whenrisk is high and imminent, prioritizing atrocityprevention and directing international engage-ment entirely toward this goal.

4. Although existing prevention tools are wellsuited to the prevention of peacetime atrocities,we know much less about preventing genocide

effect to their international commitments bymainstreaming R2P into national policy.

A useful starting point is the Global Centre forR2P’s proposal, adopted by the Group of Friends ofR2P, for governments to appoint a national focalpoint for R2P/mass atrocity prevention. Amongother things, national R2P focal points would:

• Provide early analysis of emerging situations.

• Provide advice directly to the executive aboutmatters relating to the prevention of genocideand mass atrocities.

• Coordinate national responses to mass atroci-ties—thereby providing an atrocity lens tonational decision making.

• Spearhead cooperation with the OSAPG/R2Pand other relevant agencies and offices.

• Help foster international consensus on theresults of early analysis.

• Make it possible for governments to respond tomass atrocities in a timely and decisive fashion.

• Collaborate with other focal points.38

The US administration has already taken the leadin this area by appointing David Pressman asDirector for War Crimes, Atrocities, and CivilianProtection within the Office of the NationalSecurity Advisor. Denmark has nominated itsHuman Rights Ambassador, Arnold Skibsted, asR2P focal point.

It is imperative that focal points be located with-in the national government, have sufficient stand-ing to deliver inputs directly into executivedecision making—including early-warning advicethat the executive might not wish to hear—andhave the capacity to coordinate policy across anumber of relevant government departments andagencies. In some situations, the national focalpoint may be called upon to direct whole-of-gov-ernment responses to the threat of mass atrocitiesand the appointment should have the seniorityand capacity to fulfill this role. Assigning therole of focal point to a diplomat housed in acountry’s permanent mission to the UnitedNations would limit the position’s capacity toadvise executive decision making and coordi-

16

and mass atrocities within a context of preex-isting armed conflict. There are several complexquestions to resolve, not least relating to viablestrategies, the role of humanitarian actors, therelationship between prevention and the pro-tection of civilians, and the relationshipbetween atrocity prevention and conflict reso-lution. More research and dialogue are urgent-ly needed on this subject.

5. The path to strengthening global capacity forpreventing genocide and mass atrocities andmaking full use of available capacity beginswith a comprehensive assessment of existingpreventive capacities. This assessment shouldcover structural and direct prevention andinclude the capacities of international institu-tions, regional arrangements, member states,and civil society. A useful place to begin wouldbe an audit of direct prevention capacitieshoused within international institutions andregional arrangements.

For the United Nations6. As the body with primary responsibility for

international peace and security and the organ-ization specifically charged with implementingR2P, the United Nations ought to strengthen itscapacity to bring an atrocity prevention lens toits work. The OSAPG/R2P is the most appro-priate vehicle for achieving this goal. Its actual-ization requires:

• A strengthened capacity to provide early warn-ing of genocide and mass atrocities, as agreedby member states in 2005. Early warningshould not only alert the United Nations toimminent dangers of atrocities, but should alsohighlight areas of risk where longer-term pre-ventive strategies are urgently needed.

• The capacity to conduct detailed assessment ofindividual country risks and needs and, on thatbasis, provide advice about appropriate pro-gramming, priorities, and risks to UN agencies,funds, and programmes, as a basis for develop-ing coordinated programming targeted at reduc-ing the risk of mass atrocities. A methodologyfor assessing risks and needs should be devel-oped and used by the OSAPG/R2P and sharedwith regional and national partners to bring anR2P lens to existing programs and activitieswithin and outside the UN System.

• The capacity to contribute to UN conflict pre-vention policymaking, especially country-specificpolicy development, to identify emerging andongoing crises that contain the risk of massatrocities, and to ensure that decision makingacross the UN System is informed by the atroci-ty prevention lens.

• In the event of a crisis likely to produce massatrocities, the capacity to take the lead in coor-dinating the United Nations mass atrocity pre-vention activities, as recommended by thesecretary-general, and to bring an R2P lens tothe way that the UN System responds to crisesand armed conflicts.

• A growing capacity to assess lessons learnedand develop guidance on how direct measuresare best employed to prevent genocide andmass atrocities, particularly thorough past caseanalysis and desktop exercises.

For Regional Organizations7. More attention should be paid to the role of

regional arrangements in preventing atrocitiesand the strengthening of UN/regional partner-ships. It is therefore appropriate that theGeneral Assembly has agreed to address thisissue in its 2011 informal interactive dialogue.This dialogue should be used as a catalyst forstrengthening the partnership between theUnited Nations and regional arrangements,particularly in reference to information shar-ing, coordinating crisis response, and support-ing Security Council-authorized operations.

For National Governments8. Individually, member states have an important

role to play in translating the promise of R2Pinto practice. In addition to supporting the UNsecretary-general’s proposals for strengtheningthe organization, member states should appointnational focal points for R2P. The R2P focalpoint would bring an atrocity prevention per-spective to national policymaking and ensurethat foreign, defense, and aid policies are care-fully calibrated and coordinated to achieve themaximum preventive effect and minimize thepotential pitfalls. To achieve this goal, thenational focal point should have direct accessto executive decision making and the capacityand authority to coordinate—and sometimesdirect—relevant policy across a range of gov-ernment departments and agencies.

17

Early Warning, Assessment and the Responsibility toProtect, A/64/864, 14 July 2010, p. 4.

6 Report of the Secretary-General, Progress Report onthe Prevention of Armed Conflict, A/60/891, 18 July2006, para. 42.

7 As suggested by IPI’s “Blue Paper.”

8 Daniel Chirot and Clark McCauley, Why Not KillThem All? The Logic and Prevention of MassPolitical Murder (Princeton, NJ: PrincetonUniversity Press, 2006).

9 Lawrence Woocher, “Developing a Strategy,Methods and Tools for Genocide Early Warning,”report for the Office of the Special Adviser to theUN Secretary-General on the Prevention ofGenocide,” September 26, 2006.

10 See Alexander B. Downes, Targeting Civilians inWar (Ithaca: Cornell University Press, 2008),Jeremy M. Weinstein, Inside Rebellion: The Politicsof Insurgent Violence (Cambridge: CambridgeUniversity Press, 2007) and Lisa Hultman,Targeting the Unarmed: Strategic Rebel Violence inCivil War, Report, Uppsala University, Departmentof Peace and Conflict Research, 2008.

11 The Political Instability Task Force identified “tran-sitions” and “adverse regime changes” as periods ofelevated risk. See http://globalpolicy.gmu.edu/pitf.

12 James Reason, Human Error (Cambridge:Cambridge University Press, 1990).

13 Bracketed measures were identified by only twoagendas, but these were found in the UN secretary-general’s reports on the prevention of armed con-flict and R2P and therefore warrant inclusion.

14 See Frances Stewart (ed.), Horizontal Inequalitiesand Conflict: Understanding Group Violence inMultiethnic Societies (London: Palgrave Macmillan,2008).

15 Woocher, “Responsibility to Prevent,” p. 17.

16 Ban Ki-moon, Implementing the Responsibility toProtect: Report of the Secretary-General, A/63/677,January 12, 2009, p. 2.

17 Woocher, “The Responsibility to Prevent.”

18 See Paul D. Williams, “Enhancing CivilianProtection in Peace Operations: Insights fromAfrica,” Africa Centre for Strategic Studies ResearchPaper, September 2010.

Appendices

Appendix IThis statistical chart catalogues mass atrocity cam-paigns between 1945 and 2010, indicating whetherthey occurred in contexts of war or minor armedconflict. All included campaigns resulted in anexcess of 5,000 civilian deaths and demonstratedevidence of deliberate civilian-targeting.

Appendix II: This chart cross-compares the policy instrumentsassociated with systemic, structural, and directprevention with existing prevention agendas artic-ulated for conflict, the Responsibility to Protect(R2P), and genocide.

Appendix III: This chart relates elements of the common preven-tion agenda to the specific indicators of genociderisk identified by the United Nations Office of theSpecial Adviser for the Prevention of Genocide.

These appendices are found at www.stanleyfoundation.org/resources.

Endnotes

1 See Eli Stamnes, “Speaking R2P and the Preventionof Mass Atrocities,” Global Responsibility toProtect, 1 (1) 2009, p. 71.

2 International Commission on Intervention and StateSovereignty (ICISS), The Responsibility to Protect(Ottawa: IDRC, 2001).

3 Notable exceptions are Lawrence Woocher, “TheResponsibility to Prevent: Towards a Strategy” inAndy Knight and Frazer Egerton (eds.), RoutledgeHandbook on the Responsibility to Protect(London: Routledge, 2011) and Stamnes, “SpeakingR2P.” Woocher also notes this problem.

4 A point made by Woocher, “Responsibility toPrevent.” See ICISS, Responsibility to Protect;Gareth Evans, The Responsibility to Protect:Ending Mass Atrocity Crimes Once and For All(Washington, DC: The Brookings Institution, 2009),pp. 80-81 and Kofi Annan, In Larger Freedom:Towards Development, Security and Human Rightsfor All (New York: United Nations).

5 International Peace Institute, “Conflict Preventionand the Responsibility to Protect,” IPI Blue Papers,No. 7, 2009, and Report of the Secretary-General on

18

19 This is entirely consistent with the “narrow butdeep” approach to R2P articulated by the UNsecretary-general.

20 Woocher, “Responsibility to Prevent,” p. 17.

21 Woocher, “Responsibility to Prevent.”

22 See Lisa Hultman, Targeting the Unarmed:Strategic Rebel Violence in Civil War, Report,Uppsala University, Department of Peace andConflict Research, 2008.

23 Lisa Hultman, “Keeping the Peace or SpurringViolence? Unintended Effects of Peace Operationson Violence against Civilians,” Civil Wars, 12(1&2), 2010, pp. 29-46.

24 This discussion is based on research on early warn-ing funded by the Australian Research Council andconducted with Sara E. Davies and StephenMcLoughlin.

25 The government’s role in the 1990s famine is wide-ly considered a crime against humanity.

26 Edward Luck, for example, maintained that, “earlywarning should not be the beginning of the UnitedNations engagement in a situation of concern.” SeeStatement by Edward C. Luck, Special Adviser to theUnited Nations Secretary-General at the InformalInteractive Dialogue on Early Warning, Assessmentand the Responsibility to Protect, United NationsGeneral Assembly, August 9, 2010, p. 1.

27 Statement by Edward C. Luck, p. 3.

28 See Javier Fabra Mata and Sebastian Ziaja, User’sGuide on Measuring Fragility, GermanDevelopment Institute/UNDP, 2009 and JennaSlotin, Vanessa Wyeth and Paul Romita, “Power,Politics, and Change: How International ActorsAssess Local Context,” International PeaceInstitute, June 2010.

29 There is a relatively large literature which holdsthat mass atrocities appear rational from the pointof view of the perpetrator. See for example, Peterdu Preez, Genocide: The Psychology of MassMurder (London: Boyars/Bowerdean, 1994); FrankChalk and Kurt Jonassohn, The History andSociology of Genocide: Analyses and Case Studies,pp. 29-40; Barbara Harff and Ted Robert Gurr,“Toward Empirical Theory of Genocides andPoliticides: Identification and Measurement ofCases since 1945,” p. 363.

30 Gregory H. Stanton, “Could the RwandanGenocide have been Prevented?,” p. 215.

31 Stanton, “Could the Rwandan Genocide have beenPrevented?,” p. 216.

32 Manus Midlarsky, The Killing Trap: Genocide inthe Twentieth Century (Cambridge: CambridgeUniversity Press, 2005), pp. 43, 62.

33 Barbara Harff, “No Lessons Learned from theHolocaust? Assessing Risks of Genocide andPolitical Mass Murder since 1955,” p. 62.

34 Ban Ki-moon, Early Warning, Assessment and theResponsibility to Protect, report of the Secretary-General, A/64/864, 14 July 2010, especially para. 18.

35 The problem of incoherence between direct andstructural measures was identified by Serena K.Sharma, “Blueprint or Bandage? Assessing the Roleof Preventive Action in Kenya’s Post-electoralCrisis,” 2010.

36 Report of the Secretary-General, Early Warning,Assessment and the Responsibility to Protect,A/64/864, 14 July 2010, para. 11.

37 Council for Security Cooperation in the AsiaPacific Study Group on the Responsibility toProtect, report in its first meeting, Jakarta 2010.

38 Global Centre for the Responsibility to Protect,“Fulfilling the Responsibility to Protect: StrengtheningOur Capacities to Prevent and Halt Mass Atrocities,”meeting summary, September 24, 2010.

19

Stanley Foundation reports, publications, pro-grams, and a wealth of other information are avail-able on the Web at www.stanleyfoundation.org.

The Stanley Foundation encourages use of thisreport for educational purposes. Any part of thematerial may be duplicated with proper acknowl-edgment. Additional copies are available. Thisreport is available at www.stanleyfoundation.org/resources.

The Stanley Foundation209 Iowa AvenueMuscatine, IA 52761 USA563-264-1500563-264-0864 [email protected]

The Stanley FoundationThe Stanley Foundation is a nonpartisan, privateoperating foundation that seeks a secure peacewith freedom and justice, built on world citizen-ship and effective global governance. It bringsfresh voices and original ideas to debates on glob-al and regional problems. The foundation advo-cates principled multilateralism—an approachthat emphasizes working respectfully across dif-ferences to create fair, just, and lasting solutions.

The Stanley Foundation’s work recognizes theessential roles of the policy community, mediaprofessionals, and the involved public in buildingsustainable peace. Its work aims to connect peo-ple from different backgrounds, often producingclarifying insights and innovative solutions.

The foundation frequently collaborates withother organizations. It does not make grants.

20The Stanley Foundation209 Iowa AvenueMuscatine, IA 52761 USA

Address Service Requested

Nonprofit Org.US POSTAGE

PAIDCedar Rapids, IA

Permit 174

Printed on recycled paper

02/11 850

mass atrocities and armed conflict: links, distinctions...

Documents