470-558 additional violent deaths at Baghdad city morgue

This series covers 4,587 deaths recorded at the Baghdad city morgue during the first four months of 2006. As in previous entries based on morgue records, a series of adjustments was made to the initial “raw” figure before Iraq Body Count (IBC) added the x493 series to its database. Principal considerations were:

1. Pre and post-war mortality levels for the periods covered
2. Overlaps with existing IBC entries
3. Possible presence of combatants among the dead

1. Pre and post-war mortality levels for the periods covered

IBC only includes violent civilian deaths in its count that are directly attributable to the military intervention in Iraq. One of the results of the intervention has been a sudden and calamitous breakdown in civil security, particularly in Baghdad, which continues to worsen as insurgency, counter-insurgency, sectarian, and less-readily identifiable killings compound increased deaths from “ordinary” crime. The principal role of the morgue is to identify the cause of death for legal purposes, meaning that all “suspicious” deaths, including those from accidents, are referred there. Such deaths, and those from crime, occurred during the pre-war period as they would in any big city, and would have continued at some level whether or not Iraq was invaded.

This ordinary, “background” rate of autopsies was subtracted by IBC from the 2006 morgue figures. Where known, the exact number of a corresponding pre-war month in 2002-2003 was subtracted. Where this was unknown, we subtracted “200-250 per month,” based on interviews with the morgue’s long-standing director, a figure that also accorded well with the exact numbers, which all fell within this range (250 being subtracted from IBC’s more conservative Min column, and 200 from IBC’s Max column).

The total adjustment for the ordinary, background rate of pre-war autopsies to the 4-month x493 series was 973 (subtracted from the Min column) and 873 (subtracted from the Max column).

2. Overlaps with existing IBC entries

Deaths in Baghdad in January-April 2006 already recorded by IBC and caused by gunfire and non-explosive weaponry were subtracted from both the Min and Max columns for the x493 series, since such deaths are typically referred to the morgue.

In some reports from the morgue it is stated that deaths from bombings are not handled there. However, in other and more detailed interviews, it is stated that it is deaths from “major” explosions that do not require their involvement because the circumstances of the deaths are already well known. To identify those incidents which might not be considered “major” in post-invasion Baghdad we looked at the frequency of incidents involving bombings in the city that had relatively small numbers of casualties. We included injured in this since an incident that kills 3 but wounds 19 is unlikely to be considered minor.

Our analysis showed a sharp drop-off in frequency after 8 combined casualties (killed and wounded). All bombing events recorded by IBC below this number were considered to be potential, but not definite, overlaps with the morgue data, and accordingly were only removed from the Min column of the x493 series.

In all, some 230 overlapping and potentially overlapping IBC entries were identified in the database, leading to the removal of 807 from the Min and 724 from the Max.

3. Possible presence of combatants among the dead

IBC does not record Iraqi army, insurgents or other fighters in its count unless they are killed post-capture (at which point they assume prisoner-of-war status). The morgue does not handle deaths among the Iraqi defence forces, but other combatants may be present in their records. To adjust for the possible presence of such fighters in its count, it was important that IBC find alternative sources for their potential number.

Figures from the Iraqi Interior Ministry obtained by the Associated Press provided monthly numbers for 2005 of insurgents killed during clashes with the government forces (police, soldiers and US troops), including those killed during the raids who resisted arrest. For this year only April figures were available from these sources. As there is no obvious month-by-month correlation between civilian and insurgent figures, we have elected to use 2005 figures as the proxy for January, February and March 2006, pending the publication of definite figures.

However, these are national figures and not all insurgents killed will have died in Baghdad. Two sets of figures outside IBC’s work point to about one third of all Iraqi (civilian and non-civilian) violent, conflict-related (as opposed to crime-related) deaths occurring in Baghdad: one is the ILCS survey to the end of May 2004, and the other, Iraqi Ministry of Health statistics.

In the absence of Baghdad-specific insurgent-related data, IBC has used this figure to provide an estimate of the number of insurgents that may have been recorded at the morgue.

Therefore one third of the monthly, national insurgent figures discussed above were subtracted from the Min and Max, leading to a total subtraction of 204 morgue-recorded deaths as possible insurgents over the 4-month period covered by the x493 series.

As a result of the steps described in 1., 2., and 3., above, a total of 1984 and 1801 were subtracted from the 4,587 media-reported raw total of the 4-month x493 series, leading to a final (combined) addition of 2603 (min) - 2786 (max) to the IBC database.

There are still outstanding April entries for Baghdad in the IBC processing pipeline - when these are added to the DB, any deaths which are overlapping will be retrospectively subtracted from the April entry in the x493 series. Similarly, future IBC entries based on hitherto uncorroborated reports for Baghdad 2006 incidents may be added to the database with the morgue figures acting as corroboration. Such entries will also be subtracted from the existing monthly, non-specific x493-series entries, again leaving these entries correspondingly reduced. These measures will not affect the IBC total, since in effect these deaths will simply be moved from non-specific, month-long entries to specific, incident-based ones.