R2P and Prevention: The International Community and Its Role in the Determinants of Mass Atrocity

There has been increased focus on atrocity prevention and the preventative elements associated with Pillar ii of the Responsibility to Protect. Policymakers and academics have offered a range of short-term preventative measures available so that the international community can better fulfil its Pillar ii responsibilities. This article challenges this current R2P thinking by arguing that its short-termism insufficiently focuses on de-escalation of risk within already present cycles of violence while dealing superficially with long-term causes and the ways in which the international community is a contributing factor in underwriting systemic and structural determinants of violence which erode state resilience against mass atrocity. As an alternative, this article examines a number of ways in which key actors of the international community contribute to determinants of mass violence and further offer recommendations for how they could better discharge their long-term preventative responsibilities by first reforming their own practices.

'I prioritize prevention, as I believe all of us should do' -sg antonio guterres, 2017 Prevention of atrocities is a crucial part of the Responsibility to Protect (R2P), evidenced by discussions within and outside academia, including the UN Security Council.1 This commitment to preventing mass atrocities includes -at least rhetorically -both long-term 'root cause' prevention and more short-term 'operational' prevention. Prevention concerns both the state that has the primary responsibility to protect its own population (Pillar i) as well as the international community in its secondary assistance role (Pillar ii and iii). Despite commitments to prevention expressed by several UN Secretaries-General, there has been limited analysis in R2P literature of the international community's secondary assistance role in systemic, root cause prevention under Pillar ii commitments.2 This article seeks to fill this gap in the atrocity prevention literature associated with R2P by critiquing the understanding of the role of the international community in atrocity prevention under Pillar ii, particularly its role in systemic and structural root cause prevention. It argues that R2P documents, and the doctrine's supporters, fail to fully appreciate the ways in which the international community is a contributing factor in underwriting determinants of violence (which include or increase risks of atrocities) across the globe in a way that undermines and contradicts what is needed for root cause prevention under Pillar ii. R2P thus demonstrates an impoverished understanding of how the international community can actually help undertake its responsibilities in the R2P framework.
The article is presented in four sections. Section 1 demonstrates that R2P's prevention characteristics under Pillar ii have a short-term operational focus rather than a long-term root cause focus, both within R2P documents and within wider commentary. In doing so, this section challenges current thinking on the R2P and existing paradigms on prevention. Section 2 relates this timescale problem to R2P's view of troubled states as responsible for root causes versus the international community as responsible for short-term operational prevention, illustrating why this understanding of differentiated responsibility underplays the role of the international community in risk escalation. Section 3 demonstrates some ways that the relationship between internal state responsibility (Pillar i) and international responsibility (Pillar ii and iii) can hinder rather than help state resilience against mass atrocity under a state's Pillar i responsibilities. In other words, we argue that the international community can have direct and indirect effect on escalating risk factors which underwrite the four atrocity crimes. From this, Section 4 considers, beyond R2P's current assumptions, what some of the international community's prevention measures might be. By doing so, this article reimagines prevention in a way that emancipates the international community from its current stalled role and thinking regarding preventative measures so as to demonstrate a more effective concern for individuals at risk of mass atrocities.
However, before beginning it is necessary to set the conceptual parameters. First, by reference to the international community we denote its common use in International Relations as often presented in R2P literature, namely, a group of powerful states and their associating allies who are able to directly shape geopolitics, R2P norm compliance, and influence Bretton Woods instruments in attempts to advance a collective position. In the literature this community has largely been represented by Western countries, but it can also include permanent members of the UN Security Council when mutual interests align or when collective action decisions are needed in relation to peace and security. Second, we define systemic determinants as global system level determinants perpetuated and reinforced by key members of the international community, such as resource distribution, political alliances, trade relations, economic structures, and reinforced historical legacies, which can amplify the risk of increased conflict and mass violence at state and regional levels.3 Although there is often a distinction made between systemic prevention and structural prevention within the conflict prevention literature,4 we understand these as having overlapping elements and therefore that they amalgamate key structural considerations, such as notions of structural violence, as interwoven and complimentary to long-term systemic approaches. In this way, unlike many traditional discussions about structural prevention,5 we do not understand it as limited to state structures and risk, maintaining that many local structures 3 United Nations General Assembly, are in fact greatly influenced by, and interlinked with, external socio-economic systemic practices as part of a nexus of prevention determinants. Third, as will be suggested below, there are a number of potential overlaps between research on R2P mass atrocity prevention and those found in the conflict and civil war prevention literature. Furthermore, there is rich prevention research associated with specific crimes, such as genocide. What is surprising, however, is that these literatures do not always speak across and reinforce each other, nor are they often synthesised within R2P prevention research, which we believe undermines heuristic opportunities. To be clear, our aim in this article is not to develop these links. Our effort here is merely to highlight this lacuna while more directly challenging why some R2P scholars have tended to view mass atrocity prevention research as something uniquely different to the prevention of mass violence writ large.6 In making this point, the aim is to argue that current R2P debates fail to fully appreciate the ways in which the international community is a contributing factor in underwriting systemic determinants of violence (which include or increase risks of atrocities) across the globe in a way that undermines and contradicts what is needed for sufficient prevention under Pillar ii.

Prevention within R2P -Short-term Thinking
Prevention, according to Alex Bellamy and Edward Luck, 'is the single most important element of the R2P because it is morally, politically, financially, and prudentially better to prevent atrocity crimes than to react to stop them once underway' .7 However, despite such a recognition, the R2P (as well as advocates like Bellamy) undertake only a limited and cursory investigation of long-term prevention. Furthermore, where long-term root causes are identified, the preventive recommendations for action remain short-term and malnourished, primarily focused on reforming the national/domestic structure in which a population is at risk. Consequently, there is limited consideration of the extent to which global systemic factors influence long-term root causes of mass violence in general, and those instances of mass violence which meet the legal threshold for atrocity crimes in particular. There is also limited acknowledgement of the role of the international community in global systemic factors that underwrite violence. Below we identify four prevention weaknesses within the R2P lexicon.

International Commission on Intervention and State Sovereignty: Limited Determinants
The International Commission on Intervention and State Sovereignty (iciss) recognised in 2001 the existence of short-term conflict triggers and more deeply-rooted structural causes. These root causes include poverty, political repression, and uneven distribution of resources8 and are problems within the structure of a society, which predispose a society to violence if they remain unaddressed. The report urged states to become more serious at addressing root causes of the problems that put populations at risk,9 because ignoring root causes is to address the symptoms rather than the causes of conflicts.10 iciss endorsed the Secretary-General's prior report on conflict prevention11 and the importance of long-term strategies to address structural causal factors. This conflict prevention report refers to the higher costs of military interventions compared to preventive strategies, and emphasises the importance of structural causes of conflict such as development and equality problems.12 Problems at a more global systemic level are also loosely identified within the iciss report, such as Cold War debts and uneven terms of trade, and poverty was suggested to be tied to global interactions between states.13 Having set out some of the root causes of violence, the iciss then considered what prevention tools could address these issues. It suggested democratic capacity building, fact finding (political), development assistance, better terms of trade and market access, as well as International Monetary Fund (imf) rewards and sanctions (economic), rule of law promotion (legal), and security sector reform and peacekeeping operations (military).14 Some of these tools do clearly relate to root causes, such as poverty and inequality. Nevertheless, in many ways, recognition of these conflict causes represents 'business as usual' in their basic assumptions about how to deal with atrocity crimes. This is because none of these measures fully address the complex barriers to development and equality identified by iciss as a recognised determinant of violence. By and large, these measures also assume the international community is a benign actor that can provide prevention assistance, rather than looking at the international community's active systemic role in perpetuating poverty, repression, and uneven resource distribution.15 The iciss report did make a direct, though brief, link between international community actions and root determinants of violence (including Cold War debts, unequal terms of trade and market access) which increase domestic inequality and resource scarcity, and discourage the socio-economic development of many people within poor states.16 But this was a very brief paragraph embedded deep within a 108-page report. To pick one example from iciss's list, we challenge the idea that the imf's role in mass atrocity crimes is as a neutral actor that can provide rewards or sanctions for states who protect or put at risk their populations.17 Instead, as we discuss in Section 3, the imf's equal ability to increase instability, unrest, and disenfranchisement should be understood to better situate its potential relationship to violence in general and atrocity crimes in particular. While the Secretary-General's yearly reports show a strong rhetorical commitment to prevention aligned to Pillar ii commitments,18 there are two particular shortcomings in these reports which we wish to highlight. First, the time frames envisaged in the reports demonstrate that long-term structural prevention measures are significantly restrained compared to short-term measures. Second, the reports underplay the relationship between state resilience and international community assistance and the complex global-local relationship that plagues international community efforts under the international community assistance pillar of R2P. This article develops this last point later, for now the aim is to better highlight this lacuna in R2P thinking. that in Rwanda, atrocities 'started in hate speech, discrimination and marginalisation not in the massacres themselves' .19 Although the report is correctly trying to point to a more long-term view of atrocities than the moment that the perpetrators started killing, its example that the Rwandan atrocities started at the point of hate speech rather than the point where participants took up arms is still too short-term in the context of understanding the causes of the genocide. As will be explored further below, there is considerable evidence to suggest that the causes of the genocide had far longer historical roots and that many of those causes were influenced by key actors often associated with 'rescue' responsibilities in the international community.20 Moreover, it is important to note that this undervaluation of long-term drivers to mass atrocity is not limited to analyses of Rwanda alone.21

Secretary-General's Reports: Impoverished Time Frames
In addition, many of the reports refer to both short-term and long-term causes and preventive measures, whether using these exact terms or using similar concepts, such as 'structural' versus 'operational prevention' , or 'root causes' versus 'triggers' .22 Even the 2012 report on response mentioned the need for structural prevention to happen earlier rather than later.23 But, like iciss, the identification of long-term causes is often matched to prevention measures that are actually quite short-term. For example, several reports mention the root causes of uneven growth and uneven distribution of resources, poverty, inequality -and the need to implement welfare safety-net measures.24 Chronic under-development does not cause atrocities directly, but the reports do at times make the link between limited resources, civil strife, and increasing strains between communities. via free access the problematic global system that benefits wealthy countries to the disadvantage of poorer ones and that contributes to poverty, inequality, and weak state structures. Instead, international community assistance is recommended to be delivered in the form of capacity building in weak states,26 putting pressure on actors not to commit atrocities by signing up to relevant international treaties, and to provide education and training in areas such as good governance.27 Some of these measures are very short-term compared to the causes they are expected to address, such as putting pressure on potential perpetrators. Other measures might be longer term, such as capacity building. Nevertheless, none of these measures really address root causes such as poverty and inequality, even compared to iciss's few lines addressed to the international community's insistence on repayments of Cold War debts and unfair terms of trade imposed upon under-developed countries.28 As a result, this picture of causes and solutions is mismatched in terms of the times that the causal factors start versus the time that preventive measures are encouraged to be deployed. It is also mismatched in how blame and responsibility are situated.

Secretary-General's Reports: Undervaluing the Systemic-Structural Relationship
Many reports identify weak state structures as a significant cause of atrocities, in addition to armed conflicts or instability, a record of human rights or international humanitarian law violations and economic deprivation, and the resulting inability to manage diversity constructively.29 According to these reports, weak state structures mean that the state is not resilient, and is unable to ensure respect for international legal treaties on human rights. Alternatively, a resilient state has diversity within its judiciary, military, and police, and the international community's role is said to be in helping build state resilience, or helping with early warning mechanisms.30 The international community assistance, as envisaged by the 2014 report and repeated in others, is to fix problem structures within the target state,31 via educating national authorities, by disseminating standards of good governance, and the promotion of human rights and international humanitarian law.32 But, as mentioned above, this picture of the international community's role does not address what international community actors are already doing to seriously damage state structures, nor how the global system is structured in a way that can be damaging to resilience within state structures. Any acknowledgement of the relevance of long-term root cause prevention is not followed with discussion of how the international community could fulfil its own long-term commitments beyond assisting national structural resilience (for example, its influence on risk determinants of violence like uneven growth and resource competition). Even when taking a long-term view of root causes, the reports reflect the assumption that the state is the site of the problem, and that the international community is the site of solutions. Yet structural problems such as an armed conflict/instability/record of violations do not come from nowhere, emanating from a blank canvas of international relations and domestic politics with only the problem state to blame. These root cause risk factors themselves have causes which are enmeshed in that country's relationship with international community actors. While the 2016 report recognises the ability of external actors to influence atrocities, such as through arms sales, and recommends combined local, national, and global action, there is no vision of what this might entail, particularly at the international level.33 One particular example of this problematic conception of the system-structure relationship can be seen in relation to overseas development assistance (oda). Recognising that oda is not perfect, the 2009 Implementation report does suggest that increasing oda to the bottom billion, in a manner which does not worsen tensions, could help to reduce crimes and violence through capacity building, because often the poorest individuals lack the capacity to resist oppression and violence.34 This may seem promising, in that development assistance is an acknowledgement of a more long-term root cause of violence that results from poverty and inequality. But this treatment of oda repeats the iciss tendency to view internal problems in the weak state as the cause of atrocities, and to view international community activities as having the potential to help prevent these cases. What is lacking is consideration of the international community's role in creating/sustaining problems within the weak state. There is thus no consideration of international responsibility to cease such problematic activities as part of its prevention role.
The 2017 Implementation report also considers the interplay between national institutions and intergovernmental mechanisms, but again does not examine global systemic issues in a way that acknowledges potential responsibility of the international community for the very problems it then seeks to prevent. Examples of accountability for prevention given in the report include international human rights law requirements on states not to abuse their own people as well as legal obligations 'extending to root causes of atrocities, including persistent patterns of identity based discrimination, economic deprivation and state weakness' , although the report does not explain the source of this legal obligation on states.35 The international community's obligations are set out in relation to their obligation to prevent and punish genocide and respect and ensure respect for the laws of armed conflict, as well as obeying Security Council resolutions which refer to the R2P.36 This attempt at accountability is welcome but is couched only in legal terms. As a result, the report fails to address key behaviours by the international community that are not unlawful, yet are clearly unhelpful, in addressing atrocity determinants. As will be illustrated later, these behaviours often include the approving of questionable arms sales, the support or ignoring of 'friendly' authoritarians, and/or the pursuit of economic interests despite predictable and potentially adverse effects on domestic stability. Section 3 will explain the missing links between global systemic and state structural problems evident in these reports, and how the international community cannot effectively 'help' fix state structural problems without also limiting contributing determinants that derive from the international community's role in global structures. What is important to note here, however, as a major failing of the R2P discourse, is that where structural factors for prevention are mentioned, they are mentioned in terms of reforming a state's internal structures organisationally, versus how the international community can help relieve external pressures upon internal structures by outside forces. Indeed, 35 UN, Implementing the Responsibility Stephen McLoughlin notes that it is problematic to leave the definition of root causes, and their appropriate responses, to the 'rescuers' rather than giving voice to the potential victims.37

Secretary-General's Reports: Developments on Risk Factors
Following the 2015 report on implementation, the Secretary-General endorsed a Framework of Analysis developed by his Special Advisors for R2P, dealing specifically with risks specific to mass atrocities, rather than risks related to other forms of violence.38 The Framework of Analysis echoes those risks mentioned in the State Responsibility and International Community Assistance reports,39 referring to risk factors including armed conflict or instability, a record of past violations, weak state structures, motive and capacity of instigators, lack of local resilience, and particular triggers.40 Other research suggests mass atrocity risk factors of assassinations, battlefield changes, and territory takeovers -strategic issues that lead elites to consider atrocities as part of a strategic response to help retain power and defeat enemies.41 Others have identified risk factors of illegitimate regimes, economic inequality and decline.42 Mirroring many of the issues raised in the 2013 and 2014 Secretary-General reports, the Oxford Institute of Ethics, Law and Armed Conflict (elac) notes that root cause prevention is often overlooked compared to 'altering the capacity of perpetrators [of mass atrocities] to operate' .43 Despite this acknowledgment of long-term structural issues as a major determinant, elac's list of preventive tools is then divided only into short-term (crises) and very short-term (imminent threats) 37 Stephen tools. There are no suggestions for operationalising prevention of the longerterm risk factors that elac identifies.44 Thus, root causes are identified by many, but operationalisation focuses on short-term measures, and the international community's role in these measures is one of helping lower the risk of mass atrocity crimes through education, persuasion, or improvement of the actors within state settings. These approaches to risk factors again consider both long (inequality, past violations) and short (capacity of instigators, triggers) term risks, but focus only on short-term actions that might be taken in relation to risks. They also focus on risks at a local or national level rather than a global level.45 Of course, none of these risk factors are unique to mass atrocities (as the 2018 report notes46), nor do they necessarily always require unique measures different from more general violence or conflict prevention -for example, the presence of conflict is a risk factor for most atrocities, and war crimes (one of the R2P crimes) can only be committed in a situation of armed conflict. We return to this point in Section 2 below -for now, the point is that work on risk factors that are supposedly unique to atrocities, and not just any violence, pays the same lip service to long-term root cause prevention as does the R2P more generally.
The appreciation shown throughout recent reports of the importance of prevention, and acknowledgement of the need for long-term prevention at the early stage of a risk environment, is significant. It heralds a stronger commitment to a cosmopolitan idea that an ounce of prevention is worth a pound of cure.47 But, as described above, these long-term risk factors are met with calls for what are more short-term focused solutions, with far less discussion of operationalising long-term prevention in the R2P literature. Discussion of operationalising long-term prevention, such as oda, fails to explore the problematic nature of the international community activity in relation to longterm issues. As a result, most R2P approaches are focused on the de-escalation of immediate mass atrocity risks versus genuine 'preventative' commitments that address root causes as well as externally driven systemic determinants. The next section will consider why the importance of (and financial commitment to) long-term prevention measures has been understated, and to further present our argument in favour of a commitment to long-term systemic prevention. 44 Sharma and Welsh, 'Operationalizing the Responsibility to Prevent' , p. 9. 45 See, for example, UN, From Early Warning to Early Action, p. 7. 46 UN, From Early Warning to Early Action, p. 5. 47 Brown and Bohm, 'Jus Ante Bellum' .

Why Is Long-term Global Systemic Prevention Undervalued within R2P
There is a false assumption underwriting the fundamental relationship between R2P's focus on domestic/national problems and global international community solutions. Namely, long-term global systemic prevention has been predominately ignored in R2P because the problem of mass atrocity crime is assumed as occurring internally at the local level, within the problem state, so that the international community is viewed as an external actor helping to change domestic structures.48 Alex Bellamy and Edward Luck typify this position, arguing that 'the international community has only a limited capacity to prevent atrocities' and that 'international action to prevent atrocity crimes is therefore best understood as an exercise in supporting certain actors, institutions and capacities and influencing behaviours' .49 Delineating this relationship further, Bellamy and Luck claim 'the actors primarily responsible for determining whether or not a country will experience the horrors of atrocity crimes are those within the country itself' .50 Although Bellamy and Luck are of course technically correct in regard to identifying the primary agents and location of atrocity crime, we wish to challenge the assumption that the international community is a benign agent with only 'limited' involvement or capacity. Alternatively, we want to argue that recognising problems in global systems, and the international community's longterm relationship with problem states, helps to illustrate potential international community complicity in violence, including the four atrocity crimes. As a result, if mass atrocity prevention is indeed multifaceted, as is often suggested,51 then a true commitment to prevention by the international community would require looking beyond training locals on human rights and good governance (what is traditionally understood as structural prevention), so as to also include significant long-term commitments to reforming global systems that play their part in preventing problem states from developing resilience to atrocities.

The Prevention-Reaction Distinction and the Problem of Blurred Lines
Part of the short-term focus of R2P's prevention solutions can be explained by R2P's provenance. Although the R2P is a significant development from prior 48 Dunford and Neu, ' Already Existing Intervention' . 49 Bellamy humanitarian intervention debates, particularly in terms of breadth (including prevention and rebuilding responsibilities and not just jus ad bellum questions), early R2P debates retained much of its focus on reaction.52 In the 2012 Timely and Decisive Response report, Ban notes that an international community measure can be both a reaction to something and a preventive measure, meaning that it is not necessary to define under which R2P pillar an activity is being undertaken. For example, preventive diplomacy, or decreasing arms sales to problem areas, are a response to a pattern of circumstances and seek to prevent a situation from worsening.53 Ban gives the examples of Kenya and Kyrgyzstan, where peaceful settlement measures (commissions of inquiry, negotiation, and mediation) were used in response to escalating tensions and acted to prevent outbreaks of violence.54 The 2017 Secretary-General's report also played down the traditional 'just war' criterion (adhered to by iciss) that military intervention must be a last resort, arguing instead that 'there is no need for a rigidly sequenced strategy or for tightly defined "triggers" for action' .55 Yet, although it is correct to say that an international community act can be a reaction to an event and have the aim of preventing further violence (short-term), it is also correct to recognise that long-term prevention and reaction maintain a level of conceptual distinction and thus should not be blurred together.
In this light, long-term preventive measures addressing systemic and structural injustices (e.g. addressing poverty and under-development as a result of global policies) can be distinguished from reactive measures (e.g. sanctions or military intervention), because addressing issues of global injustice can/should occur well before there is any atrocity crime.56 The need for maintaining this distinction is seemingly echoed by Bellamy  that 'upstream or structural prevention can be differentiated from response because it is not triggered by particular actions or developments in the society in question' . As they continue, 'its generic quality sets it apart' .57 Thus, conceptually, there can be a clear line between long-term systemic and structural prevention and short-term prevention, even if there is a blurred line between short-term prevention and reaction. One result of blurring short-term prevention with reaction is the risk of elevating short-term preventive measures over long-term commitments, since they represent potentially easy options for the international community. Why long-term preventive thinking is not done more effectively in R2P debate might again be because mass atrocities are also usually perceived as unique acts, with different risk factors from those associated with conflict and violence more generally.58 Thus, long-term conflict and violence prevention strategies are seemingly not always seen as applicable to R2P situations, with targeted focus on the four crimes, leaving other long-term and conflict prevention related measures under-theorised within R2P. Furthermore, some fear that expanding issues of prevention too far from the immediate threat of mass atrocity will undermine the utility and distinctiveness of R2P. For example, Michael Lund warns 'with so much varied activity now being lumped under conflict prevention there is also a risk it will lose its distinctive meaning' .59 This is echoed by the iciss, when it stated 'taking such a broad approach may divert attention away from the behavioural origins of violent conflict that are ultimately political. Too heavy an emphasis on structural causes of conflict is also empirically inaccurate -social inequalities and resource scarcity do not in fact always lead to deadly conflict' .60 Nonetheless, even those who point out that the nexus between root causes and acts of atrocity crime is unclear do not suggest that we ignore economic risk factors.61 Where structural risk determinants are recognised, as done recently by Bellamy and Luck, the discussion quickly collapses into short-term preventative 57 Bellamy strategies applied to operational conditions that are already ripe within existing cycles of violence verging on one or more of the mass atrocity crimes.62 To be clear, we are not arguing that short-to-medium-term strategies are unnecessary or that having a sufficient focus on the triggers of political behaviour is problematic. This would be untenable, since it is likely that there will always be despots and genocidal fanatics to contend with despite coordinated best efforts. What we are suggesting, however, is that these shorter-term strategies remain necessary, but insufficient, unless they are meaningfully coupled to proactive long-term commitments to address key determinants associated with mass atrocity. We will continue this line of criticism with additional suggestions for better operationalising long-term prevention later in the article.

2.2
The Assumption of Mass Atrocity Exceptionalism R2P advocates justify its narrow focus on mass atrocities (and use this to assuage fears of the doctrine being so broad it can be used to justify any and all military interventions) by reference to the uniqueness of mass atrocity crimes, and this can be seen clearly in the work on risk factors, with R2P work developing tools to assess risks specific to the atrocity crimes/R2P crimes rather than conflict or violence risks more generally. This means that the links with conflict prevention research (or the prevention research associated with the individual four crimes) is not often well explored nor synthesised within R2P research, due to a stated perception that mass atrocity prevention is uniquely different from other forms of mass violence and conflict. This distinction when made in categorical terms is unfortunate, since, prima facie, R2P prevention work could learn a great deal from conflict and violence prevention research more generally. Bellamy makes this point in addressing the 'common prevention agenda' where he argues that the structural prevention of communal violence would decrease the likelihood of both armed conflict and peacetime atrocities.63 Our point is that the global system feeds into structural prevention, and the international community's role at the global level needs to be assessed and critiqued.
Despite claims of the uniqueness of atrocity crime risk factors, there is actually significant overlap between the risk factors associated with mass violence and conflict (and its preventative research) and the four mass atrocity crimes. In fact, comparing the list of usual risk factors between conflict prevention and mass atrocity prevention suggest more similarities in terms of risks than of uniqueness. The risk factors referred to in the Secretary-General's Framework 62 Bellamy and Luck, From Promise to Practice, Chapters 4, 5 and 6. 63 Bellamy, 'Mass Atrocities and Armed Conflict' .
of Analysis are similarly mostly very general, including armed conflict or instability, a record of past violations, motives, incentives and capacity, weak state structures, lack of resilience, and enabling circumstances and triggers.64 The risk factors specific to each different atrocity crime (rather than those common to all atrocities) in the Framework of Analysis (such as 'signs of intent to destroy a group' or 'signs of widespread attacks') are themselves short-term acts, and the point of this article is to challenge the short-term focus within R2P work and to address the relationship of the international community to long-term problems (and the behaviour changes required in international community actors to implement long-term prevention). The long-term risks do not seem so unique to atrocities. The academic literature also provides multifarious underwriting risks associated with acts of mass atrocity, but their distinctiveness from general violence and conflict risk is not always clear-cut nor certain. For example, in addition to those mentioned in Section 1 above, Welsh and Dieng65 identify eight risk factors common to all atrocities, noting that mass atrocity crimes occur in similar settings and share features despite their differences. This suggests a commonality between what are supposedly very different individual R2P crimes. As another example, Bellamy and Luck have recently identified a list of five 'pre-existing' risks associated with atrocity crime, including a history of communal violence, a history of election related violence, a weak and corrupt government, marked economic and social decline, and the presence of armed militia affiliated with opposed political groups. In all cases, argue Bellamy and Luck, these conditions were politicised in such a way that mass atrocity was 'uniquely' mobilised for the maintenance of power, for counter-insurgency, to promote radical social transformation, for rebellion, to terminate a protracted war, and/or in response to intercommunal strife.66 McLoughlin notes the 'remarkable range of factors at play' and the diversity of views over the nature and causes of genocide, which challenges attempts to pinpoint usefully distinct approaches to take when trying to prevent a specific atrocity crime.67 Nevertheless, these risk factors can be viewed as determinants for mass violence and conflict more generally, and not just as unique risk factors for atrocities. This is because although there is something unique about mass atrocity in terms of its ultimate mobilisation in scale and scope, many of the risk factors that precede mass atrocity are the same risk factors for general non-ideological violence and conflict. A history of instability or a record of international legal violations can lead to violence that does not meet the threshold for the legal definitions of a mass atrocity crime. But this violence is just as deserving of preventive efforts, whether or not it meets certain legal definitions. Indeed, as Schabas notes, prevention of crimes against humanity 'involves addressing violations of human rights that may not yet warrant qualification as "gross and systematic"' .68 This suggests that there is heuristic value in making stronger and more systematic interconnected links between conflict risk research and research involved in the four mass atrocity crimes.
For example, akin to more general conflict situations, Straus has referred to mass atrocities as a strategic response by elites to events, rather than being deeply rooted in inevitable ideological, identity differences.69 Similarly, Simon draws attention to the motivation of actors to commit atrocities, not just their ability (power) to carry out attacks, and these motivations are not always purely ideological.70 Harrison points to five key factors present in African atrocities since the 1990s, including the strategic nature of the violence, enabling access to resources, and contested trade routes and labour.71 The literature on the international community's role in contributing to violence and conflict can be enlightening when considering the international community's contribution to mass atrocities. If atrocities can be a strategic choice for actors, out of a range of options, rather than inevitable ideological conflicts, then decreasing the need to control limited resources/trade/labour is an important way to address root cause poverty and inequality within a state, and in how the global system operates.
Because there is complex interaction between the underlying root cause factors (such as deprivation and social exclusion or poor trade relations and external economic conditionalities) and short-term 'proximate drivers' (such as global price shocks, violent competition for resources), it is hard to demonstrate a clear and direct link between root cause risk factors and specific mass atrocity crimes.72 McLoughlin points this out when he argues that work on international community assistance often ignores important local resilience sites and situations when, despite the presence of risk factors, there are no mass atrocity crimes.73 Nevertheless, it should be noted that this does not mean it is right to dismiss root determinants that are related to violence more generally, rather than focus on the proximate drivers of atrocities specifically.
McLoughlin's critique of the indirect link between root causes and atrocity crimes is aimed at third party intervention and generalised approaches, rather than denying the importance of addressing root causes such as under-development.74 Moreover, although political motivations for atrocity crimes are crucial for understanding and preventing mass atrocity (as noted in the iciss report above), those political motivations are not hatched in socio-economic vacuums and significantly draw upon and utilise systemic conditions in such a way to mobilise mass atrocity. As a result, logically, diminishing the acuteness of systemic and related structural risk factors will also diminish the existing 'fuel cells' available to politically mobilise mass violence, including the four crimes associated with mass atrocity. If international community actors' behaviour is weakening local resilience, then a broader view of international community assistance preventive duties can only help atrocity prevention. Lastly, the very fact that 43 per cent of all conflicts resume within five years of their ending suggests that broader long-term systemic conditions matter for delivering peaceful outcomes. Although it may not be clear exactly how systemic and related structural conditions matter, it is nonetheless reasonable to assume that a focus on short-term solutions will remain insufficient as a general policy for diminishing risk factors that lead to violence and mass atrocity.75

Local Problem, Global Solution
In addition to a focus on short-term solutions, even when long-term problems are identified, what is particularly problematic in the R2P literature is that, having located the source of trouble ( report and 2015 Implementation report) to be poor state structures that are not resilient, the 2014 International Community Assistance report naturally looks at 'positive' assistance by the international community towards improving these structures (e.g. by dissemination of standards of international human rights and humanitarian law).76 In the 2013 State Responsibility report, the link between socio-economic justice and violence is noted, and a possible solution for a domestic welfare safety net to reduce socio-economic disparities in troubled states is mooted. In the 2014 International Community Assistance report, socio-economic justice and development are again stated to be important, but there is no consideration of the possible global solutions available -no equivalent of the 2013 report's welfare safety net.77 The benevolence (or benign nature) of the international community is again assumed in how the international community can help change state problems and domestic injustice, rather than also looking at global injustice and the international community's role in perpetuating unjust systemic conditions and the effect of these global conditions in problem states.78

Global Problems Causing Local Problems
As stated before, the 2013 State Responsibility report looks briefly at risks within the problem state that enable mass atrocity crimes. Yet the 2014 International Community Assistance report does not address fully the ways in which the international community can make these local problems worse. In other words, the report does not say what the international community needs to stop doing before it can actually start assisting vulnerable states and people. For example, if the imf loan conditionality prevents states from enacting social security measures,79 then the international community can be reasonably understood to be contributing directly to determinant factors associated with mass atrocities by preventing the suggested welfare safety net (while at the same time, perhaps hypocritically, offering human rights training as a form of positive assistance programs worsen local inequality and distribution of resources, heightening tensions between communities,80 then the international community's development programs could be exacerbating the underlying socio-economic risk factors associated with mass violence, rather than being a positive actor in prevention. For example, increasing research suggests a correlation between poor health systems and conflict, noting the relationship between poor health systems, poor population health, a lack of 'smart' development policy by the international community,81 and increased risk of violence,82 which has exacerbated the risk of mass atrocity, particularly ethnic cleansing and genocide.83 In order to give more detail beyond these small examples, and to challenge the idea of an inherently benevolent international community seeking to help problem states, we argue that there are a number of links between socio-economic injustice and civil and political strife (whether mass atrocity crimes or not); between global international community-dominated systems and national structural problems -specifically between international community behaviours and intra-state violence, including atrocities.
The importance of socio-economic justice in resilience to atrocities comes from the fact that socio-economic rights support the meaningful exercise of civil and political rights, not least in fragile states where local resistance to oppressive policies might be weak.84 The relationship of root causes to conflict is noted in Article 55 of the UN Charter,85 which provides: With a view to the creation of conditions of stability and well-being which are necessary for peaceful and friendly relations among nations based on respect for the principle of equal rights and self-determination of peoples, the United Nations shall promote: via free access a) higher standards of living, full employment, and conditions of economic and social progress and development; b) solutions of international economic, social, health, and related problems; and international cultural and educational cooperation … Thus, it makes sense to enable socio-economic development and support of those unable to exercise meaningful civil and political participation -more so than, for example, offering training to target states in human rights and good governance, or telling a state to sign the 1998 Rome Statute of the International Criminal Court (icc statute).86

Economic Drivers Undermining State Resilience
Beyond the general theoretical and empirical link between socio-economic inequality and violence,87 quantitative analysis of conflicts over many years makes the same point in relation to real conflicts.88 Poverty is a significant factor contributing to civil war,89 whereas economic development reduces the risk of civil war.90 This shows a general link between socio-economic conditions and decreased violence, including cases where a mass atrocity crime has occurred. We have already made the point that mass atrocities often occur during conflict or violence more generally, and many risk factors for the four atrocity crimes are similar to risk factors for conflict. In relation to mass atrocity, economic inequality (particularly horizontal inequality between groups) is widely accepted to increase the risk of atrocity crimes.91 In R2P terms, the international community could enable local resilience by supporting socio-economic rights through positive behaviour with target states,92 not just through training target states in civil and political rights and good governance.
In relation to the international community's role in promoting socio-economic injustice and violence, rather than helping support human rights in problem states, the global finance regime has been linked to local human rights problems, and it has been argued that the effects of the global financial crisis showed that we should not assume that free markets are the best way to maximise human welfare.93 The 2008 global financial crisis (and, more generally, any instability in financial markets) resulted in a loss of livelihoods, having a detrimental effect on human rights such as access to food, housing, education, healthcare, and non-discrimination.94 In R2P terms, deteriorating state finances means states are not able to fulfil their primary responsibility to protect their citizens' socio-economic rights (again, a good example would be the inability of a state to maintain welfare safety nets that are associated with a decreased risk of atrocities). Mary Dowell-Jones and David Kinley make the link between financial insecurity, rising unemployment, reduced welfare spending and increased social unrest (though not specifically atrocities).95 The structure of the global financial system might seem too remote from any particular state, or particular part of the international community institutions, to blame for atrocity violence. Yet scholars like Vijay Mehta charge Western states with actively fuelling unemployment and unrest (despite preaching good governance), allowing businesses to support the local politicians who enable them to do business and who siphon public money away from national investment in good governance structures and social welfare.96 As just one example, Shell (British and Dutch owned) continues to enjoy drilling rights in Nigeria that were established during colonial rule,97 from which numerous human rights violations continue to be documented including links to militias engaged in mass violence (and potential acts of ethnic cleansing).98 Beyond the general global financial system, there is a strong correlation between imf involvement in a state and state collapse. Although it is not easy to show causation (and these states were hardly healthy before imf intervention) the link is still troubling. In addition to colonial exploitation, incompetence, and 'chaotic decolonisation' , Liberia, Somalia, Zaire, and Sierra Leone all spent over 50 years in imf structural adjustment programs (a key policy of which is for the government to cut subsidies necessary for basic subsistence) before eventual state collapse in the 1990s.99 Ghana's colonial legacies include economic marginalisation and violence towards locals from settler farmers.100 In 1990-1991 the imf recommended 'structural adjustment' policies for Rwanda including removing economic protections afforded to local producerssomething that led to increased poverty and inequality after Rwanda's exposure to the market, resulting in rapid increases in poverty, horizontal inequality between groups, and social unrest.101 Prior to this, Belgium's colonial differential treatment of Tutsi and Hutu populations (and the stoking of violence and conflict in neighbouring states such as the Democratic Republic of the Congo, from colonial times onwards) has been linked to rising ethnic tensions that culminated in the 1990s.102 These tensions were not helped by the development enterprise, which continued the colonial policy of hiring elites -generally Tutsi -if any locals at all were hired, and failing to involve the national population in the development process.103 Ban is correct that the atrocities in Rwanda did not start in the massacres themselves, but his assertion that the atrocities started with hate speech and discrimination still hides a deeper history and role of the international community in building up conditions that helped enable the atrocities. It is troubling that R2P reports, including iciss, view the World Bank and imf exclusively as a source for solutions to atrocity risks, by rewarding 'good' states with extra funding and withholding funds from problem states.104 Yet this same behaviour has shown to also weaken and threaten state resilience, particularly in states where underlying tensions exist and help is needed most.105 Developing this correlation between 'international community' involvement and violence within a state, Anne Orford and Susan Woodward both point out the effects of the World Bank and imf structural adjustment programs in the former Yugoslavia before the mass atrocities in the early 1990s, with economic decline leading to social unrest, calls for autonomy, and increasing instability.106 This suggests that international community assistance should begin with fundamental redrawing of the relationship between the international community and troubled states, rather than merely offering human rights training, diplomatic missions, or the exertion of pressure.

3.2
Overseas Development Assistance and Resilience Building While R2P literature references to oda can be praised for recognising its potential long-term relationship to mass violence, we argue that it is wrong to view the international community role as a benign distributor of aid that helps prevent violence. There is a vast field of development literature critiquing oda and financing for development more generally. Development literature demonstrates either that the amount of aid given by the rich to the poor is largely falling107 or that the entire enterprise of aid is corrupt, self-sustaining, fails to achieve its aims, and potentially contributes to violence around the world, and is therefore in need of significant reform.108 For example, despite 104 See Section 1 of this article. 105 Bellamy and Luck, From Promise to Practice, p. 124. 106 Alexandra ; Ajay Mehta points out that of the limited amount of oda given by rich countries, much of this goes to states such as India which has its own nuclear programme and could therefore spend more wisely on its own citizens rather than needing aid. 108 Moyo, Dead Aid. This literature is not mere speculation, listing many examples of inappropriate oda which does not help those most in need. This is troubling in itself, but also because iciss and the Secretary-General have mentioned oda as a solution to the inequality associated with conflicts and mass atrocities. However, this raises the question about whether oda can be tied more specifically to mass atrocities via its role in perpetuating poverty and horizontal inequality, which then underwrites conditions that increase the risks of any of the four atrocity crimes. the development programs contributed directly to increased social unrest and tensions between Hutus and Tutsis. 115 Even if oda was put to good use, the money received in aid is dwarfed by protectionism in markets, something the iciss mentions as a root cause, but which has received little attention in R2P literature (although development scholars have suggested that poor countries ignore demands to open their markets fully to multinational corporations).116 For example, it is estimated that the impact of agricultural subsidies (US$250 billion a year) by developed countries on African states significantly undermines market access and competitiveness for what is often the main industry in African economies. One result of these subsidies for US and EU farmers is reduced market access, reduced incomes, and lower gdp productivity for African agricultural states, which in turn undermines and reduces available tax bases for social projects (not to mention population well-being). This is problematic, since if there is a connection between economic prosperity, economic hardship, inequality, and conflict, as we are suggesting, then these market restrictions will have longterm effects on developing states and their resilience, which increase the likelihood of violence, and, over time, increases the risk of mass atrocity (e.g. as documented in Nigeria).117

3.3
Rethinking the Socio-economic Link to Mass Atrocity Although development aid is increasingly understood as problematic, the global financial system and other socio-economic factors also have effect on human rights and structures of resilience. Dowell-Jones and Kinley recommend a more regulated global financial system to give states the ability to protect themselves from the worst externalities of the global market rather than demanding a liberalised financial system as well as liberalised markets. 118 Again, related to markets and finance, but a slightly separate point, is the activities of transnational corporations and other private actors. The ability of the private sector to increase inclusivity has been mentioned within R2P reports as well as recently championed by Bellamy and Luck.119 However, the private sector, as much as the development industry, can be a significant source of human rights abuse, such as the controversial use of private military and security contractors, or the extractive industries, both of which often leave victims struggling to gain redress. 120 Furthermore, a link has been made between resource control and mass atrocity, such as Sierra Leone's diamond mines and other extractive industries in eastern drc.121 What is less often mentioned is who is buying these resources -the role of Western companies, governments, and consumers in fuelling the trade in these minerals. If market demand helps to drive violence,122 which we believe plays a key role, then responsible action by demand side actors needs to be linked to R2P preventative obligations under Pillar ii. If not, then R2P will remain symptom based versus tackling root causes, diminishing the effectiveness and sincerity of mass atrocity commitments.
In response we suggest that there is a correlative, if not fully determinant, link between the international community's economic interventions into troubled states and the risks of violence which spring from the resulting socio-economic deprivations. As a result, this suggests a research agenda that requires far more attention than it has previously received, and the aim here is to highlight this gap. Moreover, as has been recently argued, the international community is already making significant 'interventions' within fragile states and these actions have significant moral implications for the legitimacy of prevention and intervention narratives within R2P.123 This link does not mean that those carrying out acts of violence should not bear any responsibility for their actions. Nor does it mean that those acts of violence will automatically escalate into any of the four mass atrocity crimes. As with most things, causes of violence are often complex and multifarious. What we argue is that the relationship between powerful agents in the international community and mass atrocity remains under-developed within R2P debates and that this merits further investigation. Moreover, even if this relationship is only at the moment correlatively true, then the assumptions made by scholars and R2P advocates like Bellamy and Luck pertaining to the 'limited' role the international community plays in atrocity prevention is suspect and requires better investigation of long-term systemic and related structural root causes. This is particularly required if the international community really is committed to assisting states at risk of atrocity violence.
As a means to start this reinvestigation, we have suggested that socio-economic injustice plays a key role in violence, and that the international community has a role in perpetuating injustice and violence through its design and preservation of global systems that produce injustice. iciss's root causes -of poverty, unequal resource distribution, poor market access and terms of trade -have not been helped or even addressed by the international community. Thus, instead of solely focusing on capacity building and education in human rights/international humanitarian law standards, the international community would do well to also address its behaviours that maintain corruption and poor governance, or that remove welfare safety nets or market access in vulnerable countries.

4
Long-term Root Cause Prevention Strategies

4.1
Address Global Development and Horizontal Inequality iciss mentions Cold War debts, poor terms of trade, and unequal access to markets as root causes of violence. Given these root causes, the root cause prevention envisaged by iciss includes development assistance, promotion of economic growth, and encouragement of economic reform.124 Ban's Secretary-General's reports to the General Assembly are somewhat more vague in their suggestions, referring to oda and the need for this assistance to be given in ways that do not worsen tensions of inequalities. Nevertheless, the Secretary-General has been understandably hesitant to set out in detail what sort of commitments would be expected of the international community to meet these 124 iciss, The Responsibility to Protect, pp. 21 and 23. conditions. Within academia a large amount of work has been undertaken on duties of global redistributive justice from the global North to the global South, as stand-alone duties unrelated to violent conflict or atrocities. Global justice arguments include that we have duties to help the world's poorest because to permit gross misery and suffering just by reason of the coincidence of someone's birthplace is unfair; and that we have 'thick cosmopolitan' duties because we participate in a global regime that is unjust to the world's poorest and therefore have duties to make reparations.125 These arguments are not specifically about prevention of violence, but go to strengthen the argument that the international community should be undertaking redistributive justice efforts as part of its responsibility to prevent. This is particularly the case if it should be doing so anyway as a matter of fairness and if under-development is linked to inequality and violence. Given the problems with oda, it is plausible that the international community is failing in its responsibility to prevent through mutually consistent development, which needs to be addressed. It could address this by not giving development assistance at all,126 or, by doing so in a way that empowers local resilience rather than contributing to weak structures at the state level. What is also necessary, if prevention via socio-economic resilience is desirable, is to create mutually consistent trade relations that allow for real market access and competition on an even playing field, without the sorts of subsidy blockages often inflicted on LMICs by the United States and European Union. For example, there is significant literature highlighting the adverse effects of agricultural subsidies and intellectual property restrictions on agriculturally dependent economies in Africa, in which uneven trade barriers 'lock out' African countries from participating in the global market as well as undercutting African markets with subsidised products from Western states. R2P envisages that international financial institutions can help with prevention through encouraging good governance,127 but we argue that many global institutions actually contribute to gross inequality (and indirectly, the resulting violence). If we agree that an ounce of prevention is worth a pound of cure, then, as well as overhauling the way we do aid, we will also need to overhaul the way we govern global finance and trade.

Raise Funds to Pay for Prevention
In the wake of recent revelations about how little tax some global corporations pay,128 there have been demands for a fair global tax system that would increase global revenues and ensure that more funds were available for all the various costly measures suggested in the prevention literature.129 However, the role of Western governments in allowing Western corporations to avoid taxes is only one side of a double-edged sword, since many of these same Western governments also encourage 'capital flight' from lower income countries via investment loopholes that allow wealthy individuals from developing countries to 'offshore' money away from local tax systems. As one example, it is often reported that half of Zambia's gdp is lost to capital flight, much of which is funnelled into offshore tax havens.130 These 'shadow investments' not only protect revenue from taxation in developing countries, but also, in many cases, allow ill-begotten revenues to avoid any form of public accountability. To be clear, this is not merely a consequence of local corruption, from which to assign responsibility. For example, the London property market has traditionally been seen as a convenient and easy place to hide wealth away from tax collection due to the UK's lax actions on prohibiting capital flight. The result of capital flight and its adverse effect on state resilience is well documented in the development studies literature. And the logic of this negative relationship is simple, smaller tax base due to capital flight, lower resources for a state to offer economic equity measures, welfare safety nets, self-development programmes and, ultimately, its ability for population protection under its Pillar i obligations. Yet, recognition of the potential relationship of capital flight on conflict risk is near absent within the R2P prevention literature (although present in the conflict prevention literature). As a result, questions about Western participation in allowing capital flight and its risk potential remains underexamined and ignored in favour of traditional narratives that understand the risks for atrocity crimes as resting squarely on problem states.

4.3
Control the Arms Trade A commitment to reducing and controlling global arms sales would be a key long-term prevention measure (rather than temporary arms embargos as a short-term measure). The quantity of arms circulating in vulnerable states is particularly worrying given the role of several wealthy states within the international community in the arms trade. For example, between 1994 and 2004 the arms trade generated US$269 billon with US$185 billion worth of arms earmarked for developing countries. 131 In 2013 alone, the global arms trade was estimated to be above US$85 billion.132 Nevertheless, as argued by Perkins and Neumayer,133 despite some regulatory compliance, there is little evidence that Western countries have systematically discriminated against arm sales to states with poor human rights records and democratic freedoms as well as from buyers in areas where acute human security risks are present. This condition is exacerbated by arms deals associated with a lack of documentation on the recipients and increasing complaints of corruption in the arms industry. 134 The Arms Trade Treaty attempts to limit the sale of arms if they may be used to commit atrocities; with around 60 ratifications at present. Yet it is widely accepted that the treaty is far from achieving its aims, and it has been heavily criticised. In particular, the treaty does not include ammunition within its scope and excludes defence cooperation agreements, leaving wide scope in exporter assessment of the risk of export and increasing exporter responsibility for aiding or assisting international crimes.135 As one example, despite the systematic killings in Rodrigo Duterte's 'war on drugs' in the Philippines, in which evidence of crimes against humanity as understood under the 1998 Rome Statue exists, the UK completed an arms deal worth £1.7 million in 2016. This suggests that there is a weak or non-existent relationship between the UK's responsibility to prevent mass atrocity and the pursuit of self-interest in the arms trade.136 Moreover, there is increasing evidence of a relationship between the arms trade and increased incidence of political violence. For example, a multivariate binomial study of political violence between 1967 and 1997 in Africa showed that arms transfers were significant and positive predictors of an increased probability of political violence and war, including mass atrocity crimes.137 Consequently, there is a general assessment that there is a lack of attempt at actually reducing arms production or to sufficiently control trade in light of heightened mass atrocity risk. 138 Some authors have expressed concern that a focus on how the international community can carry out its responsibility to prevent will have an unintended consequence of downplaying, or even weakening, local resistance and resilience to atrocities -taking agency away from the potential victims, making problem states accountable to the international community, rather than to the local population, and focusing on international community rescue, which will prevent people from rescuing themselves.139 Moreover, some prevention activities by external actors, like training local actors on human rights, could well remove local agency and resilience. 140 In response, we agree that it is certainly problematic to view the international community as rescuer and saviour, given the ways in which the international community has contributed to the violence from which it wishes to save people. To be clear, our focus on the international community is not intended to take attention away from local resistance to atrocities. Instead, given the prominent role the international community has in assisting states with their R2P duties, we have simply aimed to show some ways in which the international community could better support local resilience through modified global systems. In other words, if the international community is already damaging local resilience by its ongoing global activities, such as arming belligerents or violent groups, then in the name of prevention we need to seriously reconsider the role that the international community plays in increasing risk, violence, and mass atrocity.
Lastly, an expected critique of our argument would be that it suffers from a number of feasibility constraints, which renders the types of reforms we have suggested unrealistic and unrealisable. For example, the UN has tried to institutionalise conflict prevention in the Secretariat since the 1980s, yet it has received little funding support or political will from member states. Furthermore, there have been numerous formal amendments in efforts to better regulate the arms trade, yet many signatories continue to exploit loopholes and actively sidestep compliance.
In response, we think it is important to make a distinction between illuminating a particular lacuna within the R2P lexicon and offering thoroughgoing reforms to address this gap. Although not mutually exclusive, they are different activities with separate objectives. Moreover, logically speaking, the identification of existing weaknesses in R2P would necessarily need to precede any recommendations for reform. In our case, the effort here has been to merely expose what we see as a major oversight within current R2P prevention discussions and to give some examples of where meaningful reforms could take place. Thus, we have not claimed that the aforementioned reforms are either feasible or likely, that is beyond the scope of this article. Our claim is simply that reforms in the aforementioned sectors will be necessary if long-term R2P prevention strategies are to meet any meaningful sufficiency standard. While we appreciate the practical difficulties that this involves, we believe it is still crucial to outline a normative compass from which practice should be guided.

Conclusion
Prevention is increasingly viewed as crucial in R2P (whether because it is cheaper or more morally acceptable than reaction), but that prevention is often understood short-term and focused on fixing problems at the local level. Long-term prevention is acknowledged as part of R2P, but largely bypassed in discussions on operationalisation. We have argued that this is a major shortcoming of the current prevention paradigm and thus represents a considerable failure in the international community's responsibility to protect under Pillar ii. This article has further argued that the international community's commitment to atrocity prevention should go beyond early warning mechanisms and capacity-building activities such as human rights training in problem states. This is not to say that these aspects of prevention are not necessarily useful, but early warning mechanisms can still be insufficient, or potentially narrow in what they address, and human rights training can be similarly narrow in its understanding of how to prevent atrocities. In response, we set out some evidence which suggests that global poverty and inequality are correlated with violent conflict (of the type that greatly increases the determinants of mass atrocity), and further argue that the international community has a responsibility to prevent atrocities by operationalising long-term root cause prevention, which address systemic issues such as poverty and inequality. This includes addressing the international community's own role in perpetuating systemic conditions of economic deprivation which can lead to violence and mass atrocity. Without doing this, R2P is only responding to symptoms of crises, not addressing the root causes. Finally, we offered some suggestions for how to 'operationalise' long-term prevention measures that address global, rather than local, root causes. While these measures may not all be linked clearly to specific atrocity acts, it has been argued that there is enough of a link to make these measures necessary. If the international community is unwilling to offer better terms of trade, to reduce arms sales, to stop demanding unjust financial conditions, to stop incentivising reduced social protection in problem states, and unwilling to force their transnational corporations to alter problematic behaviour, then the international community will be rightly treated with suspicion. Particularly when they send people to offer human rights training, or suggest a sanctions regime, or drop bombs -all in the name of protecting the vulnerable.