« Your Voting Rights | Main | 'Twas The Night Before Elections... »
October 31, 2004
Lancet Study on Iraqi Deaths - Revisited
Facts are stubborn, but statistics are more pliable.
There are lies, damned lies and statistics.
- Mark Twain
In looking more closely at the Lancet report and in reviewing comments by other respected pundits (see Fred Kaplan at Slate, among others), it seems clear that the only thing we really know is that we really don't know much. The Lancet study tried to use a statistical technique known as clustering, in which random clusters of a given population are sampled and the results are then extrapolated across the entire population. While this technique works well in epidemiology, it can have problems when applied to other areas. And, of course, the first problem we run into is that of randomness.
Mathematicians know that true randomness is an extremely elusive property. And in an environment such as currently exists in Iraq, a truly random sample would be almost impossible to obtain, mainly due to the fact that not all areas in Iraq are equally accessible. We also know that Iraq consists of a widely divergent environment. That is, some areas, such as Falluja, are extremely violent, while other large areas of the countryside see little or no fighting.
We have written before concerning the underlying bias of polling. And this study is, in essence, nothing more than a poll. The difference is, instead of speaking in terms of "margin of error," the Lancet study speaks in terms of a "confidence index." And what the study actually says is "We estimate there were 98,000 extra deaths (95% CI 8000-194 000) during the post-war period." What that means in plain English is that the authors of the study are 95% confident that the number of deaths is somewhere between 8000 and 194,000.
Does this mean, as others have been quick to say, that the study is "junk"? I don't think so. It's important to view this study in light of information that we do know, and just as importantly information that we don't know.
According to the British Foreign Minister, Jack Straw, the British government estimates the toll at around 15,000. Iraq Body Count, a respected group of academics and peace activists who base their numbers on deaths reported in the news, puts the count at between 14,219 and 16,352.
Given that information, I think it can safely be said that the civilian toll is at least 15,000. So we are now left with an imprecise figure somewhere between 15,000 and 100,000. Why the imprecision? According to Wired News:
While no issue is as contentious in the Iraq war as civilian fatalities, no organization -- with the exception of Iraq Body Count -- appears to be keeping score. No one in the media, the U.S. military, the Iraqi government or humanitarian organizations like the Red Cross is estimating the conflict's running cost in Iraqi civilian lives.
Yes, that's right, no one is counting. According to General Tommy Franks, U.S. Central Command, "We don't do body counts." And since the U.S. Military refuses to survey, our government also refuses to issue any numbers. They only offer vague reassurances that the casualty levels are "acceptable". The Iraqi government, if it can be called a government at this point, also refuses to provide figures.
And what about the International humanitarian organizations that have been so crucial to operations in other wars throughout history? The Red Cross, the United Nations, Medicin Sans Frontieres, all have been forced to pull out due to the rapidly deteriorating stability in Iraq. These are operations that have decades of experience in dealing with the aftermath of war; organizations such as the Red Cross were on the front lines in World War II. To be somewhat crude, this isn't a bunch of pansies. And yet, each one has been unable to deal with the level of violence and terror that we are seeing in Iraq.
And so, while others sit by and argue over the validity of the numbers, like children arguing over the score of a sandlot baseball game, we say that the mere fact that the numbers are so open to argument puts this conflict in a class by itself. Our leaders are making decisions that directly impact my life and the lives of many others. We demand accountability.
And we demand the truth.
Posted by bcoffee at October 31, 2004 07:54 PM
Trackback Pings
TrackBack URL for this entry:
http://www.icantthink.com/cgi-bin/mt-tb.cgi/76
Comments
Excellent demands.
Any concrete ideas on how to actually obtain a truthful estimate though?
A general remark : there have been many criticisms of the way the Iraq war has been conducted: many valid, the majority not. But worse, the MSM has let valid criticisms go by without raising them, while concentrating on invalid ones.
I personally give a lot more time to people who say "X is wrong, we should have done Y, and it should have been foreseen because of Z" rather than saying "X is wrong, but no-one could have foreseen that, and there's no solution" , or, worse, just "X is wrong".
Over to you. FWIW IMHO Iraq Body Count, although tainted by it being a product of "Peace Activists" is at least using a reasonable methodology, and I consider it a reasonable upper-bound. I'll even concede the true figure may be a bit higher, though think it most unlikely.
If only I knew a good, non-biassed and non-partisan way of approaching this very important issue. It's made more difficult because I'm partisan, and am objectively unable to trust my own objectivity.
One more thing: in a previous post you asked the quite reasonable and relevant question :
"Think before you vote. Does America want to be the outcast of the world?"
I'll quote in return from something the Australian Foreign Minister said :
“In a nanosecond, I said we will not change our vote, we will vote against this, even if we’re the only country in the world that votes with Israel on this resolution, we’ll still do it because this resolution is wrong.”
But that's us. In 1999 we too went into East Timor "unilaterally", and only got UN backing afterwards. That was a country of 20 million opposing the largest Islamic country in the world, with 180 million people. Call it 8 Iraqs. Fortunately, the other guy blinked. No-one has accused us of being "cowboys", but we've been called "deputy sheriffs" by our detractors. Actually, we feel pretty good about that. We feel better about stopping the massacre of civilians though.
You must act as you see fit in your own country's best interests, of course.
One thing I think we can agree on : may the best man win on November 2nd, and may the result be clear-cut.
Posted by: Alan E Brain at November 1, 2004 07:58 AM
