Reprinted from the British Army Review Number 149, Summer 2010
There is growing interest in civilian casualties from war zones. Each week in the news, it seems, stories emerge from Afghanistan of civilians killed in the fighting; each report seems to be clouded by claim and counter-claim as to how many were killed and who was responsible. On 1 April this year, the American Civil Liberties Union made public over 13,000 pages of US Government data on reports of civilians killed or injured in Iraq and Afghanistan. Towards the back of the Summer 2009 edition of British Army Review1 there was an article specifically describing innovative methods for tracking civilian casualties in combat zones.
Traditionally armies have been reluctant to own up to civilian casualties. There are a number of reasons for this: it may be concern they might be unfairly blamed; or reports of civilian casualties may undermine militaries’ own reporting of improving stabilization; or it may be they are simply ‘none of our business.’ This article argues that far from it not being our business, or not in our interest, Britain and the British Army has powerful reasons for both collating and releasing information on civilian casualties in war zones.
The Dead in War
In war, as in many areas of modern life, huge amounts of data are generated, used, and then discarded forever. Treating information as temporary and disposable is hardly new, and may well be appropriate in many situations. Sensible operations managers, however, retain as much data as possible that might helpfully inform future operations; indeed, the US military is considered, since at least WWII, to be the exemplary retainer of every single piece of information it is possible to store. With data storage becoming increasingly easy and less expensive, modern militaries are able to record and store even the most apparently mundane information.
Wars produce one particular stream of information which, since the inscription of the first gravestone for a war casualty, survivors have always thought worth retaining. Knowledge of dead combatants – of the enemy or one’s own – has also long been of high value in military actions, indeed was once the most important datum. But for most survivors of wars, this is not knowledge that is principally retained for functional uses but is a means of retaining the memory of lost loved ones, an honouring and recognition of lives ended prematurely due to war.
Nothing speaks more clearly to this than the memorials created to the war dead. While civilians have been less well served than soldiers, individual gravestones for people whose lives were unnaturally ended by war are evidence that the same urge applies no less to their families as those of the military dead.
If wars are now considered to be about information, then the one item of information that has the deepest resonance and meaning to those directly affected by war must be treated with the same, and perhaps even more, seriousness accorded to functional data. Who can doubt that the bureaucratic machinery so capable of recording every event in war would be more than capable of recording its relatively far fewer deaths?
Although the “Holy Grail” of complete and absolutely final casualty data is yet to be attained, there are now sufficient case examples of high quality casualty recording in conflict to have stopped the argument being about whether it can be done. It can. But, unless states are helped to see more clearly that it is in their interest to do it, and unless international frameworks are developed that articulate the state’s responsibility in this matter, inertia will probably mean that they won’t.
The Current Landscape
At present it is generally not the military but various civil actors who collect, collate and disseminate data on civilian casualties. These may include civilian medics and police, often in some combination with official registration systems responsible for issuing death certificates. However, the collection and dissemination of casualty data that is the most visible, and therefore useful, to the public at large is that which is performed by the media. While it is not the business of news reporters or their employers to act as the primary investigators let alone the ultimate authority on casualties, that they serve this function at all (often at risk to reporters’ lives) tells of a public hunger for such knowledge, making it integral to all “frontline” reporting.
But just as vital registration systems even in relatively developed countries are often unable to sustain comprehensive monitoring during conflict, so the media – who are themselves heavily reliant on local informants, official or otherwise, for their primary data – are not generally in possession of first-hand data. Moreover, the middle task, of collating such information, is one for which commercial media are particularly ill-suited, given that each outlet is only fully aware of the content and provenance of its own reports.
In an effort to fill these gaps, NGOs have sprung up, making the best use of existing data through painstaking collation, or attempting to acquire new, on-the-ground information (of necessity, often post-conflict), or a combination of both.2 There have also been attempts to acquire broad estimates for deaths based on sampling methods, but these are inherently unable to connect such mathematically-generated numbers to particular individuals, and tend to stand outside the other methods which do.
The broad range of approaches being adopted by civil society actors suggests that there is no single prescription for data collection that is suitable for all conflicts and regions.3 What it also suggests is that the only way to lift this particular fog of war is to integrate information from as many sources as can provide it. This naturally includes the military, who due to their closeness to the events in question often see and know things that others do not.
However, not every kind of information is helpful to integrative approaches – when, for instance, it is massively aggregated and impossible to connect to identifiable events. This has particular pertinence in the Afghanistan theatre, where highly divergent aggregate totals have been published by different interested parties, including NATO, leaving the ordinary citizen with absolutely no means of assessing whether these accounts are reconcilable or not.4
What Is Required For an Effectual Integrative Approach to Recording Casualties Besides Our Own?
First, integrative recording has to deal in concrete data and specifics, and must not remain at the level of numbers and timeframes alone. For instance, if one credible data source says 25 and another that 45 civilians were killed on a given week, it is necessary to know if one source refers only to bombings and the other also to shootings; if they refer to both, then it needs to be known which bombings and which shootings, where and at which times, by what kinds of bombs and guns, involving which actors. Only by cross-referencing data from different sources in this way is it possible to tell if the number of deaths for that week should be 45 or 70 or some other number. Even when incomplete and imperfect, such information can remove sufficient ambiguity that the data on casualties becomes distinct enough to be certain.
The gold standard for verifiability here is another datum not so far mentioned: the names of the dead. It is individuals with names who die, and it is by listing the dead by name that finality may be brought to the task of recording a conflict’s casualties. That this is also the best means by which a human identity rather than a cold number can be attached to the dead, and that the bereaved can be convinced that their loss has been recognised, simply adds a humanitarian imperative to the functional rationale for recording casualties in this way. Whenever possible, this standard is the one that should be striven for, with due allowance made for sensitivities that may apply, such as informing the next of kin first.
Based on the experiences of researchers attempting to integrate casualty information from different sources, we could propose a simple minimum for useful casualty data. For each incident in which people are killed, time, date, location, weapons used or incident type, perpetrators (if known), the number, demographics, and whenever possible, identity of those killed. If this work is to be universally useful, these indicators need to be accepted as internationally applicable data standards for every armed conflict.
In particular, openness about victim identity has a number of specific benefits, including that it helps to sharpen debates over whether victims were combatants (as often claimed by those who kill them) or civilians (as often claimed by those related to them). While this does not remove all argument (as exemplified by the Goldstone report on the Gaza incursion of January 2009), it provides the basis from which the issue can be investigated.
Second, an integrative approach must be transparent as well as specific, because its authority derives not just from the free flow of information necessary for cross-referencing, but from its openness to scrutiny. The more specific data is, the more easily it can be verified: stating that a death happened on a particular day is almost impossible to verify if nothing else is recorded; stating that the body was recovered at particular road junction at a particular time makes verification and cross-referencing far easier, especially in the absence of identifying information. Furthermore, exposing data from one source to scrutiny in light of data coming from elsewhere provides a valuable validity check for all data streams.
Who performs such scrutiny is of lesser significance than that the data is accessible to all interested parties, whoever they might be, and meeting the minimum standards specified above. Indeed, making the data available to all interested parties and allowing independent analyses offers the best prospect for convergence on an agreed casualty record. The independent data integrators, the scrutineers and analysts able to make sense of multiple data streams and collate it convincingly, need not be sought out or even anticipated: the availability of the data will itself ensure that those interested and capable of producing good work emerge, be they tenured academics, their students, NGOs, or individual investigators.
Whereas in the past there were always practical limitations to publishing even simple lists and tables, such data are now a few mouse-clicks (and perhaps a translation or two) from world-wide dissemination via the Web. (The importance and special needs of the next of kin, who may not speak a global language or have access to the Web, means that data pertinent to them needs to be delivered specially, but such is anyway already good and effective practice.)
Data providers such as the military may or may not want to take part in analytic integration or in interpretation of the data. However if they do so, their credibility, as it would for any participants contributing to the integrative process, depends upon on the extent to which they release data according to the standards suggested above, and are equally transparent about the analytic methods they employ. Those drawing conclusions from data are entitled to keep their interpretations to themselves. However, those holding important data may not have the same latitude to decide whether or not to disclose it. Material witnesses have inescapable obligations. Those who have access to relevant data in this poorly-documented field have special responsibility. We would argue that in this respect the obligations of the military, who often have unique access to information on casualties, are particularly clear-cut.
Third, an integrative approach by its very nature aims for comprehensiveness and completeness. This requires that all casualties are recorded and compared across sources, including on issues such as their combatant or non-combatant status. The priority is to establish that someone died; only after that is there a need to accumulate details about them. Only by ensuring that everyone has been accounted for can there be some confidence that more detailed conclusions, such as the civilian / combatant ratio among the dead, are valid.
Why the Army Has an Interest in Recording Casualties
We identify three major reasons why the military stands to gain from the recording of all casualties in the conflicts they are involved in:
Benefit 1: Casualty recording is a means of supporting reconciliation processes and emphasising the shared humanity between intervening forces and the populations among which they operate. The army that records and is open about casualties, including those it has caused, indicates that it respects locals enough to deal with them honestly. Conversely, failing to do so is easily interpretable as callousness, and can be seen as a disregard for civilian deaths by the military, an impression which may be far from the truth.
Benefit 2: Openness about casualties provides a credible foundation to military claims – frequently made by most military spokespersons – that they do all they can to minimise civilian casualties. An example is the NATO Secretary General’s claim that NATO “take extra measures to avoid killing or injuring civilians. This approach has already shown results, civilian casualties are significantly down [in Afghanistan].”5 This is a statement without a convincing evidentiary base to back it up – but that base could be easily provided. Furthermore, good data provides genuine counter-evidence to baseless propaganda.
Benefit 3: There are clear operational benefits to effective recording efforts. Modern operations recognise protection of civilians from insurgent coercion or intimidation as critical to success; hence one of the most important measures of effectiveness (MOE) of military operations is the level of civilian casualties. Identification and recording of individual civilian victims provides the data for this MOE. For example, after Operation Charge of the Knights in Basra in 2008, it was the reduction in violence against civilians that was the truest indicator of operational progress, and which enabled both Coalition and Iraqi commanders to progress from the clear to the hold phase of the counter-insurgency.
Anti-Recording Arguments Refuted
Despite these unambiguous benefits, there are three common objections. We outline and respond to each of these in turn.
Objection 1: Casualty recording involves too much time, money and risk: it is not feasible or practical for the military to undertake this. Response: Every undertaking has a cost, but so does every neglect. This is a neglect that we argue has the greater costs. It is possible that openness will reveal terrible blunders, not just picture-perfect operations – but these blunders are likely to become known one way or another anyway (certainly to survivors or their communities). If such events are covered up this could implicate the military and commanders in additional incriminating acts.
Objection 2: Recording casualties may make the governments of intervening forces liable for greater compensation. Response: This is all the more reason to ensure consensus through transparent, locally verifiable data. Compensation as determined by courts is a requirement of law-abiding nations as well as citizens – nations cannot behave like hit-and-run-drivers, and are unlikely to win the people’s consent if they do. Any actual harm needs to be compensated.
Objection 3: Recording will reveal blunders and other actions that caused civilian harm, leading to intensified local armed resistance. Response: Ultimately, high quality recording acts as a reality check on good intentions and statements. Only when the reality matches these, can local armed resistance genuinely be expected to be reduced or averted.
How the Military Can Move Forward With Recording
The British military should aim to do better in recording and releasing civilian casualty data. As outlined above, the Army should aim to collect the following basic information for every death:
- Date Time Group of incident or when bodies found;
- The number of casualties and anything known about them – preferably their names, but if not then at least their approximate ages, gender, and any other demographic information possible;
- Type of death and weapons used (if known);
- Perpetrators (if known).
Clearly in combat it is sometimes not possible to stay and find out all the information we would like. But even partial information is useful; for example "one unknown woman seen dead in street at GR ... at ... hrs with gunshot wounds to head; perpetrator unknown" could be cross referenced to other civilian sources to help complete a casualty report.
Once reported, the information needs to be released openly, where other agencies can use it to compile impartial and authoritative civilian casualty lists. The obvious method would be on a freely available website. For our own casualties we delay publication of names until after next of kin have been informed; for civilian casualties the military will rarely, if ever, be responsible for informing next of kin so we cannot be exact in the timing of release of details. We suggest a standard delay of 24 hours between initial reporting of the casualties and public release of the information which would allow some time for next of kin to be informed.
What would it take to achieve this? Firstly the Army need to include the requirement to report civilian casualties in its Standard Operating Procedures and include in pre-deployment training. We do not envisage any need for additional staff branches or personnel below National Contingent Command (NCC) level; the information would simply be treated as standard operational information to be passed up through the chain of command, handled by G3/J3 Operations staffs as appropriate. At the NCC there would probably need to be a specific collation desk, the manning of which would be driven by the level of civilian casualties in theatre; perhaps one or two in total. This is probably the level of command which would post civilian casualties on the internet, although this could equally be done at the Permanent Joint Headquarters or the Ministry Of Defence.
A final remark is in order here. This article is a civilian-military collaboration of the kind that will be required to make such a system work. It is an indicator that such collaborations are possible. Civil society is already hard at work in this area, but hampered through lack of the information that is available only to the military. The onus now lies with the British Army.
Hamit Dardagan is co-founder and principal researcher at Iraq Body Count, which remains the only continuously updated source of event-based information about civilian casualties in the ongoing Iraq conflict, and is the Oxford Research Group’s Consultant on Civilian Casualties in War.
John Sloboda is the Director of Oxford Research Group's Recording Casualties in Armed Conflict programme and is co-founder of the Iraq Body Count project. He is Emeritus Professor of Psychology at the University of Keele, and an Honorary Professor in the School of Politics and International Relations at Royal Holloway, University of London.
Richard Iron is a British Army colonel, currently a Defence Fellow at the University of Oxford.
1 Cameron, E., Spagat, M., & Hsiao-Rei Hicks, M. (2009) Tracking Civilian Casualties in Combat Zones using Civilian Battle Damage Assessment Ratios. British Army Review, 147, 87-93
2 This is not the place for an exhaustive review of what has been achieved by NGOs. However, a growing list of organisations bound by common purpose in this respect is to be found at http://www.oxfordresearchgroup.org.uk/projects/recording_casualties_arme..., convened as an International Practitioner Network, with active projects for countries as diverse as Bosnia, Columbia, Guatemala, Iraq, Nepal, the Occupied Palestinian Territories, Somalia, and Sri Lanka. In addition, organisations such as the American Civil Liberties Union and the Campaign for the Innocent Victims of Conflict, have a particular mission to get casualty information collected by the US military released into the public domain to foster increased democratic accountability of the military to American voters and tax-payers, and to foster proper public debate about the nature of our responsibility to victims and their families.
3 See, for instance: Spagat, M., Mack, A., Cooper, T., & Kreutz, J. Estimating War Deaths: an Area of Contestation. Journal of Conflict Resolution, 2009, 934-950. http://www.hsrgroup.org/images/stories/Documents/JCR_EstimatingWarDeaths...
4 Sloboda. J. (2009) The Need to Acquire Accurate Casualty Records in NATO Operations. Oxford Research Group, May 2009. http://www.oxfordresearchgroup.org.uk/sites/default/files/casualties.pdf
5 Fogh Rasmussen, A. (October 26th 2009). Video address On civilian casualties in Afghanistan. Available at: http://andersfogh.info/2009/10/26/on-civilian-casualties-in-afghanistan/ (Last accessed on 24/02/10)