Wednesday, October 11, 2006

This is Worse than Bad Statistics.

This could be considered election tampering. The Lancet et al say that 655,000 extra Iraqis have died since the start of the war. That’s around 550 dead per day. Well, that seemed a tad high to me, so I did a comparison. I compared the Iraqi losses with those of Imperial Japan during World War II, as a percentage of total population. Since the Lancet didn’t distinguish between combatants and non-combatants, I included both Japanese military and civilian deaths.

In 1931 Japan had a population of 73 million. During the entirety of WWII they lost 1.9 million military and 393,000 civilians for a loss rate of 3.1% of total population.

At the beginning of the Iraq War there were about 25 million Iraqis, and, according to the Lancet et al, 655,000 have died for a loss rate of 2.6% of total population.

This isn’t absurd, it’s profoundly absurd. There’s no way numbers can be that far off on their own. But we all make mistakes, and we learn.

Apparently not. It seems these same academics unveiled a similarly depressing study right before the 2004 elections.

Honestly, I have no idea if this is election tampering, but I’ve gotta wonder.


Anonymous David Gillies said...

It's even more dopy than that. Japan's loss of 3.6% of population occurred over eight years, between 1937 and 1945. So far we've been in Iraq a little over three.

Baseline death rate for Iraq, according to the CIA World Factbook is 5.37 per 1,000 p.a.. 665,000 over 3.5 years with a population of 27 million is 7 per 1000 p.a.. It took the Lancet to find this 130% increase in the death rate?

To summarise: we are being asked to believe two things. 1) Iraq's death rate is 230% of the pre-invasion figure 2) until now, no-one has remarked upon or even noticed this statistic, and it has taken a British medical journal to bring it to light.

1:57 PM  

Post a Comment

<< Home