Bad Math About Iraq

Dana tries to argue that the Lancet study’s numbers are too high because they compare unfavorably with Western strategic bombing in World War Two:

What those who accept the Lancet numbers unquestioningly are saying is that, if we account for population differences, Iraq has suffered between 3½ and 7 times as many civilian casualties as did the Third Reich, per capita, and that the vast majority of these were caused by what are essentially small weapons: car bombs and bullets. Thus far, we haven’t firebombed an Iraqi city the way Dresden was burned to the ground.

This is a good example of bad math. The Lancet’s numbers indicate the total number of excess deaths in Iraq. This includes Iraqis killed by the US military, Iraqis killed by insurgents, and Iraqis killed due to high crime rates. Only the number of Iraqis killed by the coalition is comparable to the number of Germans killed by strategic bombing; the other figures would correspond to, for example, the number of Germans killed by the Russians, or by a hypothetical civil war, for instance if Hitler’s plan for an insurgency campaign had materialized.

So right off, we’re down to 200,000, the number of Iraqis killed by the US and its allies. Compared with the figure of 600,000 dead German civilians, it’s not that much, even when scaling by population (for the purposes of the survey, Iraq’s population was taken to be 26 million; Germany’s was 70 million). And scaling by population is usually inappropriate, with this case being borderline.

Next, consider the fact of the occupation. Most Iraqis who died because of the occupation didn’t die in the war. In fact, the Lancet study shows that the death rate in Iraq is steadily increasing, as opposed to having a spike in March and April of 2003 and then a plummet to levels still higher than the prewar death rate.

In July of 2003 I saw a group claiming to have done a detailed study and gotten a figure of 37,000 civilian casualties, excluding Kurdistan. A study done in early 2004 of returning troops estimated that 41,000 Army men and Marines believed they killed at least one civilian in 2003. Another study done in October estimates the total number of Iraqi dead, civilian and military, in the 21,000-50,000 range.

Finally, we can estimate the intensity of bombing, which was naturally higher in Germany, as reflected in the higher number of dead. I can’t find any data on the bomb tonnage dropped on Iraq in 2003, but I can come with a very, very rough guestimate of the number of sorties. Desert Storm had a thousand per day, so guessing 50,000 sorties in Shock and Awe wouldn’t be out of line. This compares with 1.4 million bomber sorties during the strategic bombing, a figure almost 30 times higher, corresponding to almost 30 times as many dead civilians.

Although the munitions used in World War Two were less smart, and hence likelier to kill civilians, they were also likelier to miss any populated area. Today, smart bombs tend to hit their intended targets, and if they don’t, they’re guaranteed to kill civilians living nearby. In the 1940s, hitting anywhere within city limits was considered a success.

World War Two’s Allied bombing was actually not that intensive; the official estimate of dead civilians in Vietnam, 4 million over 14 years out of an average population of 40 million (it increased from 32 in 1961 to 48 in 1975), yields a death rate 3.5 times as high as this of Western strategic bombing in World War Two.

This is what happens when you try to give the argument from incredulity an air of respectability: you end up using bad math. I’m not saying any criticism of the Lancet study’s numbers is a priori wrong, but if you want to argue they’re implausible, the least you could do is not make specious historical comparisons.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: