Start of Main Content


The 1979 Report of the President’s Commission on the Holocaust, which laid out the vision for the United States Holocaust Memorial Museum and its Committee on Conscience that oversees the Museum’s genocide prevention efforts, stated, “Only a conscious, concerted attempt to learn from past errors can prevent recurrence to any racial, religious, ethnic, or national group.” As the Museum’s Founding Chairman Elie Wiesel said when addressing the importance of preventing genocide today, “A memorial unresponsive to the future would violate the memory of the past.” 

The “lessons learned” project of the Museum’s Simon-Skjodt Center for the Prevention of Genocide is one way the Museum seeks to carry out the charge to identify lessons from history  that can potentially contribute to saving lives by preventing future genocides and related crimes against humanity.

The main goal of the project is to understand better how policy makers, across all levels of government, can take effective action to prevent mass atrocity crimes and protect civilian populations in situations where they face serious threats of group-targeted, systematic violence. To identify these insights, we reviewed academic articles and think tank reports, and interviewed experts. We then distilled this body of policy-relevant knowledge into an accessible, practical resource.

Compared with other fields of study and policy, the prevention of genocide and mass atrocities is relatively young. As a result, knowledge about how to prevent these crimes is imperfect and incomplete. Given what is at stake, however, the Center sought to summarize and make accessible what is currently known about the effectiveness of various atrocity prevention tools—both to support current decision making and to guide investment in future research. The Simon-Skjodt Center is committed to supporting scholars and practitioners to continue to learn over time about what works and how and to reflecting advances in knowledge through periodic updates of this web resource. 

As the 2008 Genocide Prevention Task Force wrote in a report co-sponsored by the Museum, “The United States has many tools at its disposal, a wide range of options between the extremes of doing nothing and sending in the Marines.” Congress reinforced this theme in the Elie Wiesel Genocide and Atrocities Prevention Act, which was passed into law in 2019, calling for a government-wide strategy to identify, prevent, and respond to the risk of atrocities using a range of tools. Each recent presidential administration has echoed these themes. 

Yet, history has shown that having a number of tools in itself isn’t enough. Leaders must be willing to employ them, acting early enough to prevent the worst atrocities, not just respond after atrocities have begun. Leaders must also make choices about which of the many tools available would likely be most effective in any specific situation and see that they are carried out wisely.

The Simon-Skjodt Center’s Lessons Learned project seeks to amass information that would help support our leaders in navigating these weighty decisions. This is not a matter of telling decision makers “what works,” as if preventing genocide were a purely technical matter, free from uncertainty, the uniqueness of every situation, and political judgment. 

Rather, we seek to understand when certain policy tools are more likely to prevent atrocities, and how policy makers can design and implement atrocity prevention tools to maximize their chance of success. Even where the evidence on these questions is strong, it won’t by itself lead to “do this” and “don’t do that”-type policy prescriptions. But it can help inform these difficult decisions. And where our review indicates that the evidence on a particular tool is weak or mixed, it does not mean that the tool doesn’t work or shouldn’t be tried. It means we need more evidence.

Uncertainty suffuses all major policy decisions. But the stakes in potential genocide are too great to give in to radical skepticism. No one should hold atrocity prevention tools to a higher evidentiary standard than tools for other important objectives, or view a lack of evidence as an excuse for inaction. Policy makers must act as wisely as they can, while committing to learn from experience.

Naomi Kikoler

Director, Simon-Skjodt Center for the Prevention of Genocide

July 2022

Executive Summary

Last updated April 2024

The Lessons Learned project of the Museum’s Simon-Skjodt Center for the Prevention of Genocide is one way the Museum seeks to carry out the charge to identify lessons from history  that can potentially contribute to saving lives by preventing future genocides and related crimes against humanity.

The main goal of the project is to understand better how policy makers, across all levels of government, can take effective action to prevent mass atrocity crimes and protect civilian populations in situations where they face serious threats of group-targeted, systematic violence.

Tools for Atrocity Prevention: A New Educational Resource

The Museum’s Simon-Skjodt Center reviewed nearly 400 research reports about selected atrocity prevention tools and summarized the findings in the Tools for Atrocity Prevention web resource. We focused on identifying the characteristics (or “factors”) that research suggests are associated with more effective use of the tools in helping prevent mass atrocities.

Focusing on How to Prevent

It is no accident that Tools for Atrocity Prevention is meant to help people think through how to help prevent mass atrocities. Focusing squarely on the how would actually represent an important shift in mindset.

The Tools for Atrocity Prevention website brings together information on four key questions:

  1. What can be done to help prevent mass atrocities?

    It highlights nearly two dozen different tools (or types of action) that external actors can use to help prevent mass atrocities;

  2. What does research say about when and how to use atrocity prevention tools most effectively?

    It summarizes findings based on systematic reviews of research on atrocity prevention tools to help inform discussions about which tools are most likely to be effective in different situations and how they can be implemented to greatest effect;

  3. Where is more research on atrocity prevention tools most needed?

    It helps identify policy-relevant gaps in research, thereby encouraging scholars to focus new research on topics where it will make the greatest policy impact;

  4. What informational resources exist regarding US government use of these tools?

    It complements the research summaries with brief illustrations of cases in which the tools have been used to help prevent or respond to mass atrocities and a set of links to external resources, such as reports on relevant US laws, government websites, and examples of relevant UN Security Council resolutions.

It is no accident that Tools for Atrocity Prevention is meant to help people think through how to help prevent mass atrocities. Focusing squarely on the how would actually represent an important shift in mindset. Too many people continue to question whether mass atrocities matter to the United States and whether there is anything that can be done to prevent them. If Tools for Atrocity Prevention helps encourage greater attention to what is already known about how to help prevent mass atrocities and what about the “how” has yet to be learned, it will have made a valuable contribution.

Publicizing the results of research reviews in a field as young as atrocity prevention is not without risks. Some have suggested that acknowledging the mixed and limited nature of the research could unintentionally discourage policy makers from taking any meaningful action to help prevent atrocities. Others have suggested that building a user-friendly resource that highlights specific findings supported by research evidence might give people unwarranted confidence in conclusions drawn from highly imperfect research.

These are both legitimate concerns. Yet, we believe that progress in the atrocity prevention field requires being open and honest about the limits to existing research evidence, and understanding that policy makers need to make decisions without perfect knowledge, making it easier for them to use the best available evidence. Regardless of the challenges, policy makers bear the weighty responsibility to try to prevent mass atrocities.

Understanding the Results

Before discussing the results, we underscore four points about interpreting our project findings:

First, research about the effectiveness of atrocity prevention tools is inherently very difficult.

The difficulty of research in this area undoubtedly contributes to the limited number of studies and should generally temper one’s confidence in specific findings (positive or negative). A combination of features makes research on this topic especially challenging: There are numerous tools, many potentially relevant factors, and a relatively small number of historical cases of mass atrocities including genocide, which vary tremendously. Many important aspects of policy actions are difficult or impossible for outsiders to observe. It is also difficult to account for the fact that factors affecting which actions are chosen in the first place also bear on the likelihood of their success. Finally, the way these tools work may change over time as both perpetrators and atrocity prevention actors learn and adapt. 

Second, it is important not to confuse a lack of clear evidence about effectiveness with evidence of ineffectiveness.

Especially given the difficulty of conducting research in this area, it is critical to understand what particular types of findings imply—and what they don’t. Where our review finds limited evidence about which specific factors contribute to an atrocity prevention tool’s effectiveness, one should conclude that we need more evidence on the question; it does not imply anything about how effective the tool is at preventing atrocities. Where our review finds that existing research has not established a reliable relationship between a particular tool or factor and a particular outcome, one should ask first whether the research was well designed to observe such a relationship; it does not imply that the tool is definitively ineffective or the factor makes no difference.

Third, we caution against the direct application of summary results from the research literature to any particular current case.

The wide variation across cases in the nature of atrocities and the way in which atrocity prevention tools are used makes it more challenging to apply research results to any specific situation. For instance, Russian atrocities following its invasion of Ukraine in 2022 were committed during an inter-state conflict, whereas most recent mass atrocity cases studied occurred during civil wars. In addition, while the US policy response has employed sanctions and security assistance—two atrocity prevention tools that we reviewed—the sanctions were far more extensive and the security assistance considerably greater than virtually all historical uses of these tools. The more a specific atrocity situation and the response to it differ from the historical cases that have been studied, the less likely it is that summary research findings will provide direct insights about the effectiveness of a policy response.

Fourth, a tool’s effectiveness in preventing mass atrocities is one of multiple considerations that informs policy decisions.

Decisions about which actions to take in response to mass atrocity crises are complex. Our focus on the question of effectiveness in preventing mass atrocities is not meant to suggest that this is the only information relevant to policy decisions in these situations. Other salient considerations, for example, might relate to other objectives that policy makers are pursuing simultaneously, the interest in establishing general deterrence against atrocity crimes, the degree of public support for an action, and risks and costs.

Key Conclusions

Significant gaps exist in research knowledge on when and how to use atrocity prevention tools most effectively.

While we found many studies of certain atrocity prevention tools (e.g., prosecutions, peace operations, mediation), we found a small number of studies on others, including some that are of great interest to policy makers (e.g., targeted sanctions, naming & shaming). We recorded findings on dozens of discrete factors as being related to the effectiveness of atrocity prevention tools, but the large majority of these factors were cited in just one or two studies.

Existing research doesn’t offer a clear answer to the question, “what works?”

For every atrocity prevention tool we reviewed, we found a mix of findings in the research literature as to whether the tool was effective. Our research review underscores that the effectiveness of atrocity prevention tools depends largely on factors related to the context in which the tool is used and the manner in which the tool is designed and implemented.

Existing research indicates that certain factors–strong commitment, international coordination, use of multiple tools–are associated with greater likelihood of success across several atrocity prevention tools.

Our review found relatively strong evidence that commitment on the part of the preventive actor, international support or coordination, use of multiple atrocity prevention tools concurrently, and use of prevention tools without bias toward any group are associated with greater effectiveness of multiple tools. We found evidence, albeit somewhat less strong, supporting another set of success factors across multiple tools: cooperation or support from the national government for the use of the tool, support for the tool’s use from the local population where atrocities were committed or threatened, use of the tool early in the course of the crisis or conflict, the tool addresses root causes of the conflict, and the preventive actor being well-informed and/or skilled at use of the tool.

Implications for Policy Makers

When developing responses to atrocity crises, consult the list of factors that are associated with greater likelihood of success across several tools; when considering the use of a particular tool, consult the list of factors associated with its success.

None of these factors should be read as a prerequisite for success, but the presumption should be that atrocity prevention tools are more likely to be effective when these factors are present—that is, when the preventive actor is highly committed, international support or coordination is high, multiple atrocity prevention tools are used concurrently, and prevention tools are used without bias toward any group. For some tools, evidence supports additional success factors—for example, peace operations are generally more effective when the number of troops is large.

Pay particular attention to findings that are supported by research and experienced practitioners.

Empirical research and opinions of experienced practitioners ought to be complementary sources of policy-relevant insights. Policy makers can place more confidence in conclusions that are supported by both. For example, our research review and practitioner interviews on targeted sanctions both found that commitment or credibility and international coordination are linked to greater sanctions effectiveness.

Look to research for questions that should be asked as much as for answers.

Even where relatively strong evidence exists, research on atrocity prevention tools mainly reminds policy makers that a particular factor is important, which should prompt them to ask further questions. For example, the finding across many studies that atrocity prevention tools are more effective when there is international support or coordination leads to multiple policy questions: How much international support exists in a particular case for action to help prevent mass atrocities? What can be done to increase the level of international support? What mechanisms can be used to increase international coordination? Likewise, findings that atrocity prevention tools sometimes lead to unintended consequences should encourage policy makers to routinely think through potential adverse outcomes in advance and take measures to mitigate their risk.

Implications for Researchers

Work to develop a 10-year collaborative research program to help fill critical gaps in policy-relevant knowledge.

Despite growth over time in the number of research reports on atrocity prevention tools, more of the same will not address policy needs. Greater collaboration among researchers and between researchers and policy makers could help ensure that research effort yields the greatest benefit for atrocity prevention practice. The Simon-Skjodt Center is committed to facilitating this type of collaboration and plans to convene researchers and policy makers to discuss potential avenues for further research as an early follow-on to this project.

To maximize policy relevance, focus research on (1) topics where the gap is largest between policy maker interest and existing research knowledge, and (2) identifying factors that are associated with greater likelihood of a tool’s success, more than estimating a tool’s average effects.

The current distribution of studies across different atrocity prevention tools appears to be driven much more by data availability and scholarly interest than policy salience. This is understandable, but problematic. In addition, policy makers rarely ask about the average effect of a prevention tool across many cases; their challenge is to decide how to act in specific cases. Therefore, findings on specific factors associated with a tool’s success are more useful.

To facilitate knowledge cumulation, use common definitions and increase data transparency.

Researchers’ use of different definitions for similar concepts, inconsistent reporting about factors that influence tool effectiveness, and confusing or ambiguous writing made it surprisingly challenging to summarize and aggregate research findings. Using common definitions and making standard data routinely available should make each new study more valuable by enabling simpler comparison and/or synthesis with existing research.