On April 20, 2007, the United States Institute of Peace (USIP) and the Johns Hopkins Bloomberg School of Public Health (JHSPH) Task Force on Public Health and Conflict wrapped up its 2006-2007 activities with a public event featuring Dr. Christopher Murray of Harvard University School of Public Health. This USIPeace Briefing summarizes Dr. Murray’s presentation and the discussion that followed on armed conflict as a public health problem.

In his introduction to Dr. Murray, David Bishai, associate professor at the JHSPH, noted that we tend to spare ourselves the sight of human suffering but that the work of Dr. Murray quantifies such suffering in ways that make it hard to ignore. Thus medicine, economics, and public health work together to provide information with relevance for policymakers as well as for health care providers.

Two Approaches To Measuring Mortality

Dr. Murray identified two general approaches to the study of deaths from conflict that reflect two different realities: the social science approach, and the public health approach. How can we resolve the different perceptions of conflict that arise from these two different approaches and their respective bodies of literature? The social science literature on peace and security has suggested a decline in the number of conflict-related deaths since World War II, and has argued that conflicts have become less frequent in the post-World War II period. This literature, according to Dr. Murray, vastly underestimates the number of war-related fatalities, and has used these incorrect figures to imply that wars might even have a net benefit to public health. It suggests, for example, that more people would have died in the long run if the defeated party had remained in power.

Conflict studies in the field of epidemiology have taken a different approach to the effect of conflict on public health, with significantly different results. The most compelling of these studies have used household surveys to measure non-combat mortality in zones of conflict. Researchers go from household to household, asking a series of questions about which members of that household have died, and how, over a period of time. A study using this "verbal autopsies" approach attempted to document battle deaths and indirect deaths during five years of conflict in the Democratic Republic of Congo. This study, published in The Lancet in early 2006, estimated that the high-end approximation of conflict-related deaths in the DRC was 4.4 million in five years. This number is an order of magnitude higher than prior mortality figures for DRC that have been reported using the Uppsala Conflict Data Program dataset favored by the social science literature. In October 2006, another study used the household interview approach to research mortality in Iraq since the U.S. invasion of 2003. This study stimulated a heated controversy with its findings that as of July 2006, there have been 655,000 Iraqi deaths as a consequence of the war. Previous accounts had estimated war-related Iraqi deaths since the invasion at no more than 50,000. Differing methodologies mean that estimates of conflict-related deaths in the epidemiology literature tend to be far higher than in social science literature.

Methodology of Measuring Mortality

Dr. Christopher Murray speaks at the podium as panelists look on. (From left to right: Dr. Scott Barrett, Johns Hopkins School of Advanced International Studies; Dr. David Bishai, Johns Hopkins Bloomberg School of Public Health; and Dr. Linda Bishai, U.S. Institute of Peace.)

The World Health Organization standardized the methodology of the household survey approach in 2003 during a project in which researchers gathered sibling survival histories in 71 countries, including Kazakhstan, Vietnam, Sri Lanka, and Ethiopia. Dr. Murray noted that people are surprisingly able to remember details of war deaths over decades and decades, as demonstrated by the number of people who recalled details of World War II deaths in Kazakhstan. This Global Health Survey estimated that there have been 57 million conflict-related deaths worldwide, and that 5.2 million of these are injury-related deaths while the rest are caused by the conflict indirectly. In the social science security literature, war allegedly accounts for only 170,000 deaths. If this new epidemiological approach were to prevail, war would rank among the top-ten causes of death worldwide. But today, war is not even included on the WHO’s official list of risk factors for diminished longevity. Despite the staggering numbers, institutions continue to neglect war as a serious public health issue.

Dr. Murray noted that there are serious flaws with war mortality measurements that rely on reports from the media and individual combatants to determine mortality rates; there is no evidence that these reports are a valid measuring tool. But the epidemiological studies are not perfect either. Dr Murray mentioned three of their flaws. First, though the standard margin of error is larger when a study relies on household-sampling, a more significant problem is that when you go to a house and ask about deaths, people actually tend to under-report mortality. The demographic literature has tried to develop methods to remedy this. Secondly, there is the problem of "telescoping." This is the phenomenon by which people transfer deaths from the distant past into the more recent past. Telescoping varies by age and culture, but Dr. Murray noted that in China, interviewees reported more deaths having occurred in the past six months than the researches would predict based on the total number of deaths the same interviewees reported for the past 18 months. A third problem with the epidemiological approach to measuring mortality is "salience." People are more likely to remember, and to tell you about, violent or sensational deaths. This combination of problems makes it difficult to assess whether household mortality surveys have under- or over-reported conflict deaths.

A New Global Institution?

There is no global institutional archive on deaths in conflict. No one is collecting international data in a comprehensive, systematic way. The World Bank and International Monetary Fund collect and monitor data on global economies much more closely than any institution does on global mortality. Dr. Murray proposed a new international institution to close this gap and to effectively monitor the impact of conflict on global health. Such an institution would need to embody three attributes: excellence in complex methods of measurement; relevance, so that the fields of the social sciences and public health are compelled by the findings; and independence, because politics cannot be allowed to undermine the integrity of such an institution. Finally, this institution would need a total commitment to data transparency, because a study will carry little weight if it cannot be replicated.

Dr. Murray concluded his remarks with an appeal to bring these two very different communities—security and public health—closer together. Health is at the center of the Millennium Development Challenge agenda, and conflict and security should figure prominently therein. Likewise, the security community should devote greater attention to public health concerns—for example the impact that a modern flu pandemic might have on global security.

The Role of the International Community in Public Health

Scott Barrett, the Director of the International Policy Program and of the Global Health and Foreign Policy Initiative at the Johns Hopkins School of Advanced International Studies, elaborated on the link between public health and security. He offered the example of a flu pandemic to illustrate the connections: if a deadly strain of flu breaks out, will people become more hostile towards each other? Wwhat will happen as some people have access to protections and immunizations while others do not? A flu pandemic will likely exacerbate the global inequalities that already exist; Dr. Barrett asked that we consider how we share the risk of a global health problem.

Dr. Barrett noted that possessing reliable data on public health and mortality is critical as data serves as a focal point for a discussion of what is happening. Conflict must be acknowledged as relevant for public health since conflict may be the "weakest link" that determines the success or failure of disease eradication. For example, smallpox was finally eradicated from the world in Somalia in 1977. What is the likelihood that we could successfully eradicate any particular disease from Somalia today? Polio was successfully eradicated from Botswana, but it has re-established itself today in Somalia. A comparison of disease eradication to the failed states index demonstrates that the "more failed" a state is, the more "spread of disease" exists in that country.

Dr. Barrett concluded with some thoughts on why the international community has not been good at preventing conflict. He suggested that we face a collective action problem, both globally and domestically. Domestic leaders in western democracies must consider future votes when they contemplate any action to intervene in a potential conflict. Globally, Dr. Barrett listed some of the problems with the administration of the United Nations: how can we convince members of the Security Council to show self-restraint? Could they pledge to refrain from use of the veto on any resolution to prevent or counter genocide? The problem is that countries have short-term incentives, and combating a long-term crisis is often unattractive. While measuring and collecting data on public health and conflict is critically important, it is not enough. We have to improve our institutions so that they are able and willing to respond to the crises in a timely fashion.

During the question-and-answer session, the audience drew out some nuances in the relationship between public heath and conflict. One participant asked what potential there is to use health as a tool for conflict mitigation or management. Dr. Barrett answered that the trick is first to be invited in. You have to get a wedge into the conflict—perhaps through health, perhaps through another avenue—and then you can use that wedge as a level to open up the conflict to mitigating influences. Dr. Murray added that we should be asking whether health issues have an impact on conflict, since we have already established that the reverse is true. He wondered if there are any efforts to measure the impact of health education and services on conflict management.

This USIPeace Briefing was written by Sarah Dye, a research assistant, and Linda Bishai, a senior program officer, both in the Education program at USIP. It does not represent the views of the Institute, which does not take policy positions.

The United States Institute of Peace is an independent, nonpartisan institution established and funded by Congress. Its goals are to help prevent and resolve violent international conflicts, promote post-conflict stability and development, and increase conflict management capacity, tools, and intellectual capital worldwide. The Institute does this by empowering others with knowledge, skills, and resources, as well as by directly engaging in peacebuilding efforts around the globe.