Visit the Museum





Academic Research

Remember Survivors and Victims

Genocide Prevention

Antisemitism and Holocaust Denial

Outreach Programs

Other Museum Websites

Preventing Genocide: A Blueprint for U.S. Policymakers

Early Warning: Assessing Risks and Triggering Action

A destruction that only man can provoke, only man can prevent.

-Elie Wiesel

The first major element of a comprehensive system to prevent genocide and mass atrocities is a reliable process for assessing risks and generating early warning of potential atrocities. While some preventive strategies can be employed without respect to when and where risks are greatest (for example, advancing global norms and institutions, discussed in Chapter 6), most will need to be targeted to specific situations at specific points in time. At its most basic level, early warning means getting critical information to policymakers in time for them to take effective preventive action. Effective early warning does not guarantee successful prevention, but if warning is absent, slow, inaccurate, or indistinguishable from the "noise" of regular reporting, failure is virtually guaranteed.

In its popular conception, early warning is often equated with an alarm bell sounded just before disaster strikes. This notion is much too limited. If signs of genocide and mass atrocities are only detected once violence has begun to escalate, decision makers are left with only costly and risky options. In contrast, if underlying risks and evolving dynamics can be recognized and described accurately in advance of or at the early stages of a crisis, the full panoply of policy options will be viable. A range of instruments suitable for reducing risks in pre-crisis situations and halting escalation of emerging crises will be discussed in Chapters 3 and 4, respectively.

Beyond warning as such, accurate assessment of the relative risks of genocide and mass atrocities will permit efficient allocation of limited resources and policy attention. Furthermore, fine-grained analysis of specific contexts, actors, and dynamics of high-risk situations is a prerequisite for development of successful preventive strategies. There are few, if any, one-size-fits-all solutions. Effective strategies must be tailored carefully, based on a deep understanding of case-specific characteristics.

Chronologically, risk assessment and early warning can be considered the first phase of preventing genocide. The kind of information collection and analysis required for effective early warning, however, is equally necessary for supporting, evaluating, and fine-tuning ongoing strategies. For instance, assessing the impact of a regional mediation effort or financial sanctions against militia leaders requires analysis of the evolving capabilities, attitudes, and operational activities of key actors. Thus, improving early warning capacities should benefit U.S. government efforts to prevent and respond to genocide and mass atrocities across all phases of prevention and response.

Many people who have studied genocidal crises have asserted that "early warning is not the problem." Weighed against the challenges of generating political support for vigorous early action, getting bureaucracies to respond nimbly, and wielding preventive strategies that will halt escalation and relieve underlying pressures, early warning could be seen as a less severe problem. The Genocide Prevention Task Force, however, chose to frame the question differently: Instead of asking whether early warning is the problem or the biggest problem, we asked whether the U.S. government could improve its early warning systems in ways that would increase the chances of preventing future acts of genocide and mass atrocities. We are convinced that the answer is yes.

Core Objectives

Effective early warning should begin with global scanning and assessment of short-and long-term risks, move to detailed monitoring and analysis of high-risk situations, and end with reliable mechanisms for communicating results to policymakers in a way that will promote sound preventive action.

The impetus for global risk assessment is largely pragmatic. Given resource limitations, it is infeasible to engage in detailed monitoring of the entire globe. A global risk assessment will generate a "watch list" of states or situations at highest risk, based on structural or long-term characteristics known to be associated with genocide and mass atrocities (for example, history of mass violence; see page 25 for other risk factors). In addition, structural risk assessments provide a context for interpreting ambiguous information; for example, when there are significant uncertainties about events developing in a state that exhibits many long-term risk factors, an analyst should be more concerned, whereas the same pattern of events in a place with few or no known structural risk factors should raise less concern.

A watch list, as its name suggests, should be just the start of more detailed monitoring and analysis. For situations identified to be at elevated risk, early warning requires a thorough understanding of the fault lines along which genocide and mass atrocities might develop, the ideologies and organizational capacities-especially money and material resources-of potential perpetrators, the role of third parties, the extent of "negative support" in the population for genocide and mass atrocities, and so forth. As noted above, mass atrocities have unfolded differently in each case-for example, forced exile of Armenians into unlivable conditions, slave labor and starvation in Cambodia's "killing fields," and attacks by paramilitary death squads in Guatemala. It is crucial for early warning analysts to try to anticipate the range of plausible scenarios through which perpetrators could effect large-scale and systematic attacks on civilians. These scenarios enable analysts to identify case-specific indicators that genocidal dynamics are emerging or escalating-or that windows of opportunity for constructive action are opening-providing focus for ongoing monitoring.

We emphasize that early warning is about providing information and analysis that promotes effective preventive action, not predicting exactly where and when genocide or mass atrocities will occur. Forecasts need not be perfect to make early warning useful. Indeed, prediction implies both a degree of confidence in our knowledge of future events that is unrealistic and a passivity regarding future outcomes that is counterproductive. Nevertheless, past efforts at risk assessment and early warning have suffered because resulting lists of high-risk countries outstripped policymakers' capacity to take preventive action. Increasing the accuracy and precision of risk assessment is therefore critical.

Analyses about conditions or trends that might lead to genocide or mass atrocities, indicators of active movement in that direction, and forecasts of plausible future scenarios must be effectively communicated to decision makers. This component of early warning is sometimes overlooked, but is fundamental. Early warning is only as good as its weakest link, and this is frequently the communication of warning analysis to decision makers and a mechanism-or lack thereof-for using this analysis to support appropriate policy action. No matter how good information and analysis is, if it gets stuck in the bureaucracy or is presented to policymakers disconnected from any ideas about practical preventive measures, it will have little impact.

Major Challenges

Challenges to effective early warning fall into two broad categories: (1) generating timely and accurate warning analysis and (2) getting warnings to be heard by policymakers and taken into consideration in their policy planning.

Producing accurate warning analyses first requires thorough reporting on relevant actors and events. While it is the responsibility of U.S. embassies and missions to know what is happening in their host country, the tendency has been to report on developments in the capital rather than more remote rural areas, if only because of resource constraints. This was reportedly the case with the U.S. Embassy in Kigali, Rwanda in 1994, during the civil war in the Democratic Republic of the Congo in the late 1990s, and with diplomatic reporting from Nairobi prior to the violence in Kenya in 2007-08. The State Department's transformational diplomacy initiative, still in its early stages, aims to relieve some of these problems by shifting U.S. diplomats to developing countries and encouraging them to travel beyond the capital city.

The availability of news reporting on even remote parts of the world has tempered the information problem significantly. Counterintuitively, how ever, the bounty of information-which can only be expected to grow in the future-does not necessarily ease the analytic challenge. First, the amount of material can be overwhelming, and second, it is hard to judge the accuracy of the reporting. For example, a crucial and difficult task for analysts is to distinguish systematic killing of civilians from more generalized background violence, as most if not all mass atrocities occur in the context of a larger conflict or a campaign of state repression. When our diplomatic and intelligence reporting from the post is inadequate, analysts in Washington are left to make judgments from ambiguous and frequently conflicting information and assessments.

The accuracy of analysts' warnings will also depend on the extent to which they can identify warning signs or indicators of genocide and mass atrocities. While scholars have had some success in identifying long-term risk factors, it has proven much more difficult to find generalizable near-term indicators, "accelerators," or triggers. For example, pervasive hate speech is often cited as a warning sign of potential genocide, the Nazi propaganda machine and hate radio in Rwanda being just two examples. But we have observed many cases of pervasive hate speech that have not led to genocide or mass atrocities.

More research into the general dynamics of escalation to genocide and mass atrocities is warranted. In the meantime, analysts must consider how genocide or mass atrocities might manifest themselves in a particular context, generating a set of case-specific indicators. A related challenge stems from the fact that earlier warning, which is obviously preferable, typically means lower confidence that the apparent trends toward mass atrocities are real and significant. Earlier warnings are, therefore, easier to dismiss as being alarmist.

The recent post-electoral violence in Kenya illustrates the challenge of inaccurate warning. Most analysts anticipated some violence surrounding the election. But none imagined the scale, rapidity, and ferocity of violence and forced migration. Policymakers, as a result, were left to scramble in crisis response mode without having done advance planning-or set aside resources-for such a contingency. The collective response to the crisis in Kenya was impressive, but was far from assured given the lack of accurate warning.

The second major challenge is getting warnings to be heard by appropriate policymakers and taken into consideration in their policy planning. Many of us have had the experience in government of receiving warnings only at late stages, when violence had already gained momentum. Sometimes this reflects a reluctance to report bad news or accept that current policies are failing. Other times, cautious or risk-averse officials err on the side of transmitting too much or too generalized information, resulting in it being either dismissed or discredited. Given the information overload that senior officials experience, simply adding to their already overflowing inboxes more information on potential threats is not the answer. Meanwhile, policymakers are likely to be grappling with many other pressing issues and priorities-the more so at successive levels of seniority. Thus, for early warning to be effective, the response to warning must be built on an institutional mechanism that links analysis of plausible future scenarios with assessment of possible policy options. As one former official described it, a warning product is of little value unless it is linked to "an empowered process" of policy review and action.

Beyond these challenges, it is too frequently the case that parts of the U.S. government actively resist learning about grave risks of genocide or mass atrocities. Incentives for blocking efforts to illuminate imminent or ongoing atrocities can stem from a desire to avoid becoming entangled in complex situations or simply a natural desire to turn away from hard choices. Furthermore, many government officials perceive an interest in shielding their principals from accountability for possible failures if no obvious benefit is seen to balance the risk: it is easier to deflect charges that one should have known but was never informed than that one knew and yet chose not to act.

Few if any of these challenges are unique to the U.S. government, and many are common to the "warning-response problem" across a range of possible events. Warnings always entail a degree of uncertainty, and human beings naturally resist paying certain costs today, even if small, to protect against uncertain future costs; this is true of bureaucracies all the more so. Add to this the incentives for political leaders to focus on short-term costs and benefits, and the tendency for bureaucracies to resist risky action, and it should not surprise us that it is difficult to generate support for preventive action. Also, people with designated responsibility to provide warning have incentive to "overwarn" to shield themselves from criticism for failing to foresee a significant event, which in turn makes it easier for recipients of warnings to discount them. The task force believes these factors can be counteracted, if not completely overcome, through a combination of deft political leadership and innovative institutional design.

Readiness to Meet the Challenge

Principal U.S. Government Actors

U.S. embassies, USAID missions, and U.S. armed forces deployed around the globe represent the front lines of U.S. foreign policy, generating large amounts of information that could be relevant to preventing genocide and mass atrocities. The State Department and USAID employ roughly 6,600 and 1,000 foreign service officers (FSOs), respectively, and the U.S. military has about 500,000 troops forward-deployed worldwide. Although early warning of genocide and mass atrocities is mainly the province of the intelligence community and diplomats, all of these U.S. personnel can be valuable sources of information.

Task force consultations indicate that the intelligence community currently dedicates several to genocide, war crimes, and related issues. This includes a very small War Crimes and Atrocities Analysis Division within the State Department's Bureau of Intelligence and Research (INR). At the National Intelligence Council (NIC), the national intelligence officer for warning leads a team of analysts to oversee and coordinate all strategic warning, including that related to genocide and mass atrocities. In addition, the State Department's Office of the Coordinator for Reconstruction and Stabilization (S/CRS) has a monitoring function, though it remains underdeveloped and its ambit does not explicitly include genocide and mass atrocities. USAID's Office of Conflict Management and Mitigation (CMM) provides analytical and operational tools related to conflict and development, including some assessment activities. There may also be other isolated efforts within the U.S. government related to early warning of genocide and mass atrocities or associated events. For example, the Defense Advanced Research Projects Agency sponsors a project to develop an integrated crisis early warning system for forecasting and decision support vis-à-vis a wide array of crises, including genocide and mass atrocities. Lacking high-level focus or an effective coordination hub, however, the dispersed nature of these various efforts has limited their overall impact.

We note also that the State Department Bureau of Democracy, Human Rights, and Labor (DRL) prepares annual reports on human rights in nearly every country in the world. The reports are required by statute to include, "wherever applicable, consolidated information regarding the commission of war crimes, crimes against humanity, and evidence of acts that may constitute genocide." But unlike other subjects that are mandated to be discussed (for example, religious freedom), genocide and mass atrocities do not receive mention in every report, even if only to note that there is no evidence of problems.

Tools and Capacities

Substantial progress has been made in and outside of the U.S. government since the end of the Cold War in methods for identifying risks of genocide and mass atrocities. The task force's research suggests that, while it is impossible to anticipate exactly when and where the next genocide may occur, it is relatively easy to identify one to two dozen countries at highest risk. The most systematic effort in this direction has taken place through the U.S. government-sponsored Political Instability Task Force (PITF), a consortium of academic experts working since 1994 to assess and explain the vulnerability of states around the world to political instability and state failure.

Empirical analysis by the PITF and others indicates that the strongest and most reliable genocide risk factor is the existence of an armed conflict or a change in regime character. Virtually all instances of genocide or mass atrocities since World War II occurred coincident with or closely following a major internal conflict or the taking of power by more radical or more harshly authoritarian leaders. Examples include Cambodia, Guatemala, Algeria, the former Yugoslavia, and Sudan. Other conditions associated with elevated risk of genocide and mass atrocities include history of genocide, autocracy, state-led discrimination, and high infant mortality (see PDF version for sidebar). It is worth underscoring that there is little support for the conventional wisdom suggesting that religious or ethnic diversity in itself poses risks for genocide or mass atrocities.

The Atrocities Watchlist (AWL), issued quarterly by the NIC's Warning Staff since 1999, is the major regular product on these issues, drawing on systematic analysis of known risk factors and qualitative judgments by regional experts. The AWL is a classified product, reportedly distributed across the government to various offices in State, Defense, USAID, and the NSC. A short document, it identifies countries and situations at different levels of concern and describes briefly the current dynamics and potential for future changes. The NIC issues a separate Instability Watchlist biannually focusing on broader risks of political crises and conflict, including humanitarian emergencies.

Consultations with current and former officials lead the task force to conclude that the AWL is not as well known or useful as might be expected. Several current and former U.S. government officials said they did not recall ever having seen or even heard of the AWL. Respondents acquainted with the AWL gave it mixed reviews. Many judged it to be fairly thorough and accurate. Most noted, however, that it rarely points to situations not already known to be at great risk by policymakers-or, as some suggested, anyone following international news. The task force's sampling of opinions about the AWL, including those of us who have been on the receiving end, suggests that it adds only marginal value. The problem, however, is not necessarily with the AWL. Watch lists are helpful to the extent that they are linked to more detailed analysis of evolving dynamics and, most important, with a process for generating action. These linkages in the current system appear to be fragile at best.


As with other aspects of preventing genocide, the United States cannot satisfy the need for early warning by itself. The United States may have unmatched global capacity to collect intelligence, but other actors may have better access to specific situations of interest. Sometimes a local church group can provide more timely and accurate information than the world's largest intelligence service can. In addition, there is tremendous analytic expertise outside of the U.S. government-in academia, international NGOs, think tanks, grassroots civil society groups, other governments, and intergovernmental organizations. Yet cooperation between the U.S. government and other governments, intergovernmental organizations, and NGOs with respect to early warning remains relatively underdeveloped. Acknowledging the value of outside expertise, the DNI issued an intelligence community directive in July 2008 on analytic outreach prescribing a series of steps to encourage analysts to leverage this expertise more effectively.

NGOs such as Amnesty International, Human Rights First, Human Rights Watch, International Crisis Group, and Physicians for Human Rights, plus several organizations that have emerged more recently with a specific mandate related to genocide and mass atrocities-Enough, Genocide Intervention Network, Genocide Watch, and Save Darfur Coalition, for example-are just a few sources of relevant information and analysis. The U.S. Holocaust Memorial Museum's Committee on Conscience is another source for information on threats of genocide. Operational NGOs providing humanitarian assistance have long been crucial partners for U.S. government responses to human-made and natural disasters. Aid workers on the ground often have the best understanding of local populations and evolving conflict dynamics; they can be valuable partners for early warning as long as their neutrality and security are appropriately safeguarded. Likewise, U.S. policymakers should not overlook the role of indigenous NGOs and civil society groups in high-risk countries, notably including religious communities. A final component of the civil society sector is the growing community of academic experts and other scholars, both in the comparative study of genocide and mass atrocities and on specific situations of interest.

While U.S. officials increasingly recognize the potential value in cooperating with NGOs for early warning of genocide and mass atrocities, institutional culture, lack of established mechanisms for collaboration, and even legal restrictions on sharing information (for example, high-resolution satellite imagery that can indicate destroyed villages) all limit the current extent of cooperation. Most cooperation with partners in this domain appears to be ad hoc and driven by individuals rather than systems. For example, the ambassador-at-large for war crimes issues convenes monthly meetings in Washington with human rights NGOs, and at least some ambassadors at post convene similar meetings with civil society groups. These appear to be undertaken entirely on the initiative of individual officials, however. A rare systematic effort is the Global Futures Forum, initiated by the U.S. intelligence community in 2006 as "a multinational, multidisciplinary intelligence community that works at the unclassified level to identify and make sense of emerging transnational threats." The forum includes a "community of interest" on genocide prevention that, while inchoate, may hold promise as a venue for deepening cooperation between the U.S. government, scholars, and NGOs as well as other governments.

Enhancing cooperation on early warning with the United Nations and regional intergovernmental organizations has its own set of challenges and potential rewards. The United Nations has at various times attempted to develop more significant capabilities for early warning of political crises or violent conflict, to little effect. The UN secretary general appointed a special advisor on the prevention of genocide with an early warning mandate in 2004, but this office remains under-resourced and challenged bureaucratically. The adoption of the "responsibility to protect" at the 2005 World Summit, which explicitly calls on the international community to "support the United Nations in establishing an early warning capacity," may provide a new opportunity. The secretary general has expressed a commitment to institutionalizing the responsibility to protect in the UN system, appointed a separate special advisor to develop the concept and explore potential mechanisms, and is working with member states to move this agenda forward.

The African Union (AU) and Africa's regional economic communities-for example, the Economic Community of West African States (ECOWAS) and the Intergovernmental Authority on Development-are engaged in systematic early warning efforts to support their efforts to prevent violent conflict, if not genocide and mass atrocities specifically. The U.S. government has provided financial and technical support to these initiatives. Development of other regional early warning systems, especially in Asia and the Middle East, would be desirable, but there has been little progress in this direction outside of Africa and Europe.

Responding to the Challenge

Recommendation 2-1: The director of national intelligence should initiate the preparation of a national intelligence estimate (NIE) on worldwide risks of genocide and mass atrocities.

NIEs represent the most rigorous and thorough process of reaching an intelligence community-wide judgment, going through several iterations of drafting, briefing, comments, and revisions, and ultimately requiring approval by the National Intelligence Board, composed of the heads of all relevant intelligence agencies. An NIE on global risks of genocide and mass atrocities would, first, engage a wide array of senior analysts and policymakers in defining the policy-relevant questions, considering available evidence, and generating key judgments and dissenting views. This process would raise the profile of the issue and sensitize intelligence officers and analysts across the community. NIEs frequently include a summary of opinions of nongovernmental experts, so the process would engage outside experts as well. Second, NIEs are typically briefed to the president, members of Congress, and other senior officials, contributing to the effort to attract attention from high-level decision makers. Third, an NIE would highlight areas of poor knowledge or lack of consensus, pointing to actions that would improve analytic capacities in this area.

An NIE on global humanitarian emergencies, which included a short section on genocide and crimes against humanity, was released in 2001; it merits updating and more in-depth analysis. The national intelligence assessment (NIA) on the national security implications of global climate change to 2030, which was briefed to Congress in June 2008, could serve as a useful model. The study drew on substantial bases of knowledge outside of the U.S. intelligence community to generate key judgments on global and regional trends and consequences for the United States. In briefing the NIA to Congress, the chairman of the NIC concluded by highlighting several challenges to collection and analysis and outlining future research plans. As with most NIE/NIA processes, the value was more in the high-level attention and analytic rigor feeding into the policy debate than in precise forecasts.

Recommendation 2-2: The national security advisor and the director of national intelligence should establish genocide early warning as a formal priority for the intelligence community as a means to improve reporting and assessments on the potential for genocide and mass atrocities.

Sending a clear signal to the intelligence community that genocide and mass atrocities represent threats to U.S. national interests that demand attention would almost certainly increase the amount and improve the quality of relevant information collected. The National Intelligence Priorities Framework (NIPF) is the current mechanism for establishing U.S. intelligence priorities, by issue and by country. Risks of genocide and mass atrocities relate to a number of the twenty-plus issues currently in the NIPF-for example: human rights and war crimes, regional conflict and crisis, democratization and stability. We believe, however, that genocide and mass atrocities should be explicitly mentioned as a priority. Not all situations of human rights abuses or even regional crises need concern the president and his top advisors. All cases of genocide or mass atrocities should. Thus, genocide and mass atrocities deserve a more prominent place in the NIPF-or whatever process the incoming administration uses to establish intelligence priorities.

Recommendation 2-3: The State Department and the intelligence community should incorporate training on early warning of genocide and mass atrocities into programs for foreign service and intelligence officers and analysts.

No matter what policies, systems, and structures are adopted, effective warning depends on the knowledge and skills of individuals. Therefore, it is imperative that the front lines of America's foreign policy apparatus be equipped to understand genocide and mass atrocities, to recognize conditions that might lead to these crimes, and to employ analytic techniques specially suited to warning (for example, scenario gaming). A training initiative would help analysts distinguish critical warning signs from background noise. It would also promote a shift in the culture of FSOs and analysts writ large, to "get early warning and prevention of genocide into the DNA of regular analysts," in the words of one former official.

The ambassador-at-large for war crimes issues has engaged in some training activities, but these do not appear to be standard. DRL conducts some training at the Foreign Service Institute (FSI) more broadly on human rights and religious freedom, as directed by the International Religious Freedom Act of 1998. Training focused on genocide and mass atrocities could be paired with these and existing courses at FSI on conflict assessment. It should become required training at least for FSOs being deployed to high-risk countries. Training initiatives should aim to reach higher-level officials (for example, via the Ambassadorial Seminar and deputy chief of mission training) as well as junior officers.

Analyst training for early warning of genocide and mass atrocities could be a joint initiative of the national intelligence officer for warning, given his explicit mandate to promote analyst training in techniques that might contribute to improved warning, and the State Department's Office of War Crimes Issues or DRL, given their substantive expertise and reach to FSOs. Both the FSI and the Sherman Kent School for Intelligence Analysis would be important partners for a new training initiative.

Recommendation 2-4: The national security advisor should create a "mass atrocities alert channel" for reporting on acute warning of genocide or mass atrocities akin to the State Department's "dissent channel."

This mechanism would reduce the chance that concerns about impending atrocities might fail to reach high-level policymakers for bureaucratic or political reasons. Such a new mass atrocities channel would be designed to be a seldom used fail-safe mechanism. If other elements of the system worked well, there would be no need for a dedicated mass atrocities channel. Yet, the specter of genocide and mass atrocities is sufficiently grave-and history indicates warnings sometimes get stuck in the system for political or bureaucratic reasons-that the task force believes a special mechanism is warranted as a last resort.

This channel should be reserved for situations when a U.S. official in the field judges there is a significant risk of massive atrocities or genocide in the near term and finds the standard lines of reporting to be blocked. The channel would transmit a message directly to the co-chairs of the Atrocities Prevention Committee (APC)-the director for crisis prevention and response at the NSC (a proposed new position, as discussed in Chapter 1) and the assistant secretary of state for democracy, human rights, and labor-in addition to a senior official at the official's home agency (for example, the State Department director of policy planning or the under secretary of defense for policy). One or both of the recipients would be required to send a prompt response from Washington addressing the concerns. All communications sent through this channel would also trigger immediate discussion by the APC.

Recommendation 2-5: The national security advisor should make warning of genocide or mass atrocities an "automatic trigger" of policy review.

There is a balance to be struck between too sensitive a trigger that compels review too often, becomes burdensome, and gets ignored, and too high a bar that fails to trigger review in serious situations until it is too late or too difficult to take action. In addition, because each case is different in significant ways, it is hard to specify which actions should be triggered on warning. The main point, however, is not to prescribe specific steps, but to institute a mechanism that at least forces policymakers to consider the situation. They could always choose to discount the warning or judge that ongoing government actions are sufficient. The fundamental advancement over the current system is that this kind of mechanism would make it much harder for U.S. government officials to avoid making a decision vis-à-vis the situation.

We propose a new warning-response mechanism that is calibrated based on the severity and urgency of the warning, so as to avoid a single, all-or-nothing trigger. The most acute level of warning, based on the AWL and a request by any member of the APC, should trigger a discussion of policy options at an NSC Deputies Committee meeting. Less acute but still serious warnings (for example, high level of concern on the AWL) should trigger deeper analysis of the evolving dynamics and preparation of crisis response plans by the APC. Any time a country appears on the AWL for the first time, reappears after having fallen off, or remains at moderate concern, it should trigger actions such as enhanced information collection, consultation with independent experts, and preparation of crisis prevention plans.

Recommendation 2-6: The State Department and USAID should expand ongoing cooperation with other governments, the United Nations, regional organizations, NGOs, and other civil society actors on early warning of genocide and mass atrocities.

This expanded cooperation would at the same time improve the analysis available to U.S. policymakers and promote early preventive action by others in the international community. We offer the following ideas to enhance cooperation with the major categories of key partners. These recommendations relate closely to broader strategies for cooperation with partners, discussed in depth in Chapter 6.

The State Department should launch a major diplomatic initiative to create a permanent network of like-minded international actors to continuously exchange information on risks of genocide and mass atrocities. The group of states should extend beyond the circle of allies with which the United States regularly shares intelligence, as there is a broad community of interests against genocide and mass atrocities and little relevant information is derived from sensitive sources and methods. This network could also serve as a hub for cooperation with NGOs and regional organizations including the European Union. Recommendation 6-1 describes the network and its functions in greater detail.

At the United Nations, the United States should pursue information sharing in both directions. UN agencies with significant presence in the field-the Office for the Coordination of Humanitarian Affairs (OCHA), UN High Commissioner for Refugees, Department of Peacekeeping Operations (DPKO), World Food Programme, United Nations Development Programme (UNDP), and United Nations Children's Fund-are a rich potential source of information relevant to early warning. On the response side, the United States should offer information and analysis, within normal constraints, to the UN Department of Political Affairs, DPKO, the UN Office of the High Commissioner for Human Rights (OHCHR), the secretary general's special advisor on the prevention of genocide, and any forthcoming institutional mechanism related to the responsibility to protect. In addition, U.S. representatives at the UN Security Council should welcome briefings by the secretary general's special advisor on the prevention of genocide. It could be valuable for the special advisor to brief the Security Council regularly, as the emergency relief coordinator does, while still leaving space for ad hoc briefings to the council on emergent crises.

The United States should continue to support the development of regional early warning systems at the AU and African subregional organizations and push to incorporate specific attention to genocide and mass atrocities. U.S. diplomats should encourage the possibility of a regional early warning system for Asia supported by the Association of Southeast Asian Nations (ASEAN). ASEAN has been slow to engage on political issues, but its commitment to creating a regional human rights body is encouraging, and the ASEAN Regional Forum's "Concept and Principles of Preventive Diplomacy" provides a natural basis for discussing how early warning can be strengthened across the region. The United States should also coordinate with the European Union, which is engaged in its own early warning activities and has been a major supporter of capacity building by other regional organizations.

NGOs and civil society, broadly defined, are critical partners for information collection and analysis. The long-term, ongoing engagement exemplified by regular meetings between American diplomats and NGOs, in the field and in Washington, produces tangible rewards. It should not be left entirely to the discretion of individual officials, but should become standard practice. Local religious leaders in particular are often cognizant of social, political, and economic dynamics in rural areas that U.S. diplomatic missions cannot easily access or monitor. Given this reality, U.S. embassies should explore means by which to engage religious leaders and institutions in early warning efforts.