285-332 violent killings recorded at the Baghdad city morgue

The x460a-x460h series of entries in the IBC database is based on records of the Baghdad city morgue covering the period May-December 2004. As such it represents a continuation of previous Baghdad morgue-based IBC entries covering April 2003-April 2004. Then as now, the chief cause of fatalities is the post-invasion breakdown in civil security and unchecked growth in mostly criminal violence. About 60% of autopsies performed by the morgue are of violent killings ("homicides", principally from gunfire), and it is only these deaths which are added to IBC's database. And as before, IBC's entries do not include all of these deaths as recorded by the morgue but only those over and above the ordinary pre-war level of such deaths in the city, as recorded by the same institution (the Baghdad city morgue, also known as the Medico-Legal Institute).

One significant change since our earlier entries is the steep rise in the level of insurgent activity, which from our own data we estimate to have increased more than four-fold in Baghdad from May 2004 onwards. Our earlier morgue-based entries contained subtractions for the reportedly "rare" presence of insurgents among the dead; we have increased these more than four-fold in the x460-series to account for this on-the-ground change. Another change, this time in the reporting of the types of deaths covered by the morgue's "homicide" data, is that this no longer includes deaths from bombings; however this remains somewhat uncertain, as is the claim that only gunfire deaths are included among the violent killings reported by the morgue. These issues are of relevance to overlaps with existing IBC reporting and were dealt with as outlined below, where we list the methods employed to obtain earlier Baghdad morgue entries, with changes for the x460 series given in bold.

1. To allow for the sentence in the reports which reads "Also, the bodies of killed fighters from groups like the al-Mahdi Army are rarely taken to morgues," an estimate of "between 1 in 50 to 1 in 25" was used to represent the "rare" presence of fighters in the morgue statistics. AP's Baghdad figure was accordingly reduced by 4 per cent and 2 percent to produce "fighter-free" minimum and maximum estimates. The numbers produced by these estimates were increased 4.25-fold in the x460-series to account for a corresponding increase in insurgent activity documented after April 2004.
2. These numbers were further reduced by morgue data for the equivalent 2002-2003 periods (and locations) to provide an adjustment for normal "background" death rates unattributable to the war and its aftermath. It is only the difference between the pre- and post-invasion rates which are recorded here.
3. Deaths which may already have been recorded by IBC from other reporting for the locations and periods concerned were subtracted from the totals obtained after step 2 to avoid overlaps and double-counting (See the "Details..." note to x073 for an example of the Methodology used in such instances).
4. We also allowed for the statement from the director of statistics at the Baghdad morgue that "The figure [reported by the morgue] does not include most people killed in big terrorist bombings": Where there were existing, potentially-overlapping records of this nature identified in step 3, above, those overlaps were reduced by "half+1" for overlaps smaller than 10 and by 60 per cent for larger ones — "60 per cent" being our interpretation of the term "most" as used here. In the current crop of reports it is stated by the morgue's deputy director that deaths from bombings (not just "big terrorist bombings") are not dealt with by the morgue because the "cause of death is already known by the police". However, we judged that victims of certain types of "minor" explosions - including roadside bombs and mortars - and of US-coalition air strikes may continue to receive at least preliminary investigation or cause of death confirmation at the morgue, and that therefore victims of such incidents already in the IBC database may constitute overlaps with the new morgue data. To allow for this possiblity, those incidents were removed from the Min column of our x460 estimates, but not from the Max column (to allow for the counter-possibility).

If details later emerge that are of relevance to any of the calculations above, IBC will adjust its figures accordingly.