An oft-quoted statistic draws attention to the terrible impact that certain wars have had on civilian populations, but as a generalisation about all wars since 1990 it is based on shaky foundations.

Throughout the post-Cold War period there has been a widespread view that war has changed radically since the early twentieth century to the point where some 80–90% of war victims are now civilians. This view was reflected in the European Union’s European Security Strategy, adopted by the European Council in Brussels in December 2003, which stated as fact that ‘since 1990, almost 4 million people have died in wars, 90% of them civilians’. Many other individuals and institutions have made similar statements.

This proposition rightly draws attention to the terrible impact that certain wars have had on civilian populations, but as a generalisation about all wars since 1990 it is based on shaky foundations. A range of sources provide evidence of a lower percentage of civilian casualties in certain recent wars. Moreover, generalising about war in this way is damaging, not least because of the capacity of bad statistics to drive out good and to be believed by international bodies, governments and publics alike. Major controversies such as that over the high figures for Iraqi war deaths published in The Lancet in October 2006 serve as a reminder that rigour is needed in the compilation of statistics in this field.

Problems of assessing civilian war casualties

Generating reliable assessments of casualties of war is a notoriously complex process. Civilian casualties present particular difficulties. One problem is that the attribution of the label ‘civilian’ is contested in some cases. On the surface, the definition of a civilian, at least in the context of international armed conflicts, is relatively simple: a civilian is any person who does not belong to the armed forces of a party to the conflict and is not among the categories entitled to prisoner-of-war status. In practice, however, there are debates about whether, say, civilian contractors working with the military, or terrorists, or certain part-time participants in a civil war, should be considered civilians. A more serious problem arises from the variety of ways in which civilians may become casualties. To make effective use of such statistics as there are about civilian casualties of war, it is necessary to be explicit about the criteria for inclusion. All too often, there is a lack of clarity about which of the following categories of civilian casualties are included in any given set of figures:

  1. Those killed as a direct effect of war;
  2. Those injured as a direct effect of war;
  3. Those dying, whether during or after a war, from indirect effects of war such as disease, malnutrition and lawlessness, and who would not have been expected to die at such rates from such causes in the absence of the war;
  4. Victims of one-sided violence, such as when states slaughter their own citizens in connection with a war;
  5. Victims of rape and other forms of sexual violence in connection with a war;
  6. Those uprooted in a war – that is, refugees and Internally Displaced Persons (IDPs);
  7. Those who, even after a war is over, die prematurely from injuries sustained in war.
Online Access & Digital Download £10.00
Product variations
Online Access & Digital Download £10.00

Adam Roberts is President of the British Academy and Senior Research Fellow at the Centre for International Studies in Oxford University’s Department of Politics and International Relations. He is also an Emeritus Fellow of Balliol College, Oxford. He is joint editor of The United Nations Security Council and War: The Evolution of Thought and Practice since 1945 (Oxford: Oxford University Press, paperback edition 2010).

Back to content list

How to access Survival: Global Politics and Strategy

The bi-monthly journal of international and strategic affairs.