Published online 15 December 2008 | Nature | doi:10.1038/news.2008.1303


Comparing the horror of wars

An index of atrocities might help to guide humanitarian intervention in conflicts.

SoldiersThe Dirty War Index allows different conflicts to be compared.Punchstock

How do the horrors of the civil conflict in Colombia compare with those in Northern Ireland in the 1970s or in the war devastating the Democratic Republic of the Congo?

The shocking nature of such conflicts might make such questions seem callous but a reliable index of war atrocities can help to focus policies and priorities for intervention, deterrence and monitoring.

Madelyn Hsiao-Rei Hicks of King's College in London and Michael Spagat of Royal Holloway College in Egham, UK, have developed a measure they call the Dirty War Index (DWI). The index quantifies the number of cases of a particular atrocity — such as rape, civilian deaths or torture — as a proportion of the total number of incidents1.

For civilian mortality, for example, the DWI would be the number of civilian deaths divided by the overall number of mortalities in the conflict, both civilian and combatant, multiplied by 100.

It might sound obvious, but Hicks and Spagat point out that these violations of humanitarian law are often discussed in terms of absolute numbers that complicate comparative assessments.

Unreported violence

Egbert Sondorp of the London School of Hygiene & Tropical Medicine agrees that such quantitative tools can be much more telling than more generalized statistics or qualitative descriptions.

"DWIs will potentially provide human-rights advocates with additional tools to make a case against certain practices and [offenders]," he says. "The relative simplicity of the indices may work wonders to get messages across to a wider audience."

Nathan Taback of the University of Toronto in Canada agrees that quantitative measures are valuable for interpreting data. "In general, gathering data on the health effects of conflict can illuminate patterns of violence that may not be apparent from anecdotal evidence, and that have not been taken seriously until quantitative evidence was available," he says.

But he wonders whether such tools would give a false impression of certainty for statistics that are often hard to gather in conflict zones. "The simplicity of DWIs might belie the complexity of the data used to calculate them, especially if the caveats and limitations of the data set are not taken seriously," he says. However, he acknowledges that this applies to any statistical measure.

Rape, for example, is generally under-reported. Hicks and Spagat argue that their index still allows comparisons if the incidence of under-reporting is more or less constant everywhere. But Taback says that this might not always be the case. "It is difficult to think of an example where researchers would have an estimate of the magnitude of the under-reporting," he writes in a commentary2 accompanying the paper.

Objective measure

Hicks recognizes such problems. "The DWI is not meant to be a tool that you just throw at any war data set that comes your way", she says. "It should be applied only when the quality of the data meets acceptable scientific standards."


All the same, Hicks and Spagat think that their index can tell useful stories. For example, in the Colombian conflict they show that 99% of the people killed by paramilitaries were civilians, whereas the proportions for both guerrillas and government forces were about half that magnitude. This shows at a glance that the paramilitary tactics are the 'dirtiest'.

Hicks says that their DWIs have already revealed issues that might not have been obvious from the raw data. "In our analysis on casualties from the Israeli–Palestinian conflict, we found that the proportion of females killed was significantly higher when Palestinian forces targeted Israelis (40%) than when they targeted Palestinians (3%), or when Israeli forces targeted Palestinians (5%)", she says.

The researchers acknowledge that their index leaves many ethical questions unanswered, such as precisely what should be judged 'dirty' in a war.

But, Hicks says, "people's opinions about conflicts are largely based on anecdotes, impressions and opinions, and on assumptions about whether a war is right or wrong". But "DWIs provide a way for individuals to test the accuracy of their opinions or assumptions," she adds. 

  • References

    1. Hicks, M. H.-R. & Spagat, M. PLoS Med. 5, e243 (2008). | Article |
    2. Taback, N. PLoS Med. 5, e248 (2008). | Article |
Commenting is now closed.