Wednesday, October 25, 2006

Deaths in Iraq

What should we make of the Lancet paper showing 655,000 excess deaths in Iraq? My immediate reaction is that it is poppycock. To say why it is nonsense requires statistical expertise beyond my own. Readers can make their own minds up by going to websites here or here or here or here or here or here.

Statistical evidence is beyond the understanding of most people. Some people believe that statistics don't lie, other believe that there are lies, damned lies and statistics. I rather take the view that things that are obviously true don't need statistics to prove them, but statistical evidence is easlily abused by those with a strong point of view such as pharmaceutical companies who want to sell their wares, politicians who want to mislead the public and scientists who want to get published for the sake of career advancement. A short while ago a paper published in the BMJ claimed that only 30% of scientific papers used statistics correctly.

So I like to subject statistical claims to a reality check. Here is a quote from the first of the links:

If you finished a cohort study which concluded that the forest, by extrapolation was denuded yet aerial photography did not show any logged over areas then you would be exactly in the position of accepting Burnham's proposition that 655,000 people died without observing the masses of widows, orphans, mass graves and bomb craters that one would expect to accompany such an enormous loss. My parents were unfortunate enough to go through a major urban battle. It only killed 100K people but what it did was produce a flood of refugees and displaced persons whose existence was palpable. One quarter of Burnham's sample is from Baghdad. Where are the refugees? It is not widely appreciated but Iraq is host to millions of Shi'ite pilgrims every year. There is not a single historical instance I can think of where tourists continued to visit a country beset on the scale claimed.

Here's another:
Now let's consider Mr. Burnham's claim that "Overall, 13% of deaths were attributed to airstrikes" because this gives us a window into his data. A sample of his sample, so to speak. That's 85,000 deaths from airstrikes. The problem with this is that airstrikes virtually ceased since 2004. Now maybe "airstrikes" is really understood to be all kinds of fire. Could this be true?

Where would these airstrikes have happened? Where Burnham sampled, of course because he claims these deaths are backed by death certificates. Now recall that of his 47 clusters, 12 were in Baghdad, 2 were in Basra, and 3 were in Anbar. But wait! Basra is in the British sector and there have been no air missions in Basra to speak of. As for Baghdad, if which has 1/4 of the clusters, at least 20,000 people would have died in airstrikes within full view of the media since 2004. Anbar is the place where most airstrikes would probably have occured. But it's a sparsely populated area and contains 3/47 clusters. Could it be that most of the airstrikes were there?

So we come to the question of what his snapshots really look like. And as I said, I am disappointed.


Finally, it is worth looking at where the information is coming from. The Lancet under Rchard Horton has assumed a stance to the left of the Independent. It has always had a reputation for playing fast and loose with the peerr reviewing process, preferring to publish what is 'newsy' rather than what is scientifically correct, witness the MMR scandal. Richard Horton, himself, goes on 'Stop the War' marches. Under his editorship it has sunk from being the permier medical journal to perhaps the 3rd or 4th.

So, for many reasons, even without statistical expertise I think the Lancet article is poppycock.

8 comments:

Steve Madden said...

This is the second "paper" published by the lancet on this issue. The first being Oct 2004.

The estimate then was 100,000 using the same methodology.

They described airstrikes as "helicopter gunships, rockets or other types of aerial weaponry"

Two studies using the same methodology two years apart showing a 600% increase in deaths.

I have no idea what the real numbers are but the trend is alarming.

Terry Hamblin said...

There is undoubtedly a high death rate in Iraq, most of it Iraqi killing Iraqi, but the confidence intervals on this figure are immense and the technique suspect. It is also heavily politically driven. I have both the Lancet papers and in my opinion and that of many statisticians they are both flawed.

Anonymous said...

Of course they are lies! The left has NEVER let the truth stand in the way of making a political point.

Sad and pathetic.

vance esler said...

Most people do not understand statistics, and even if they did, few bother to think critically about such matters. It takes effort. Maybe busy laymen can be excused. However, it is shameful for journalists to report such stuff as fact without examining the claims more closely.

Exiled in mainstream said...

Yes of course it's deeply dubious and probably wrong but...

my problem is that the links are themselves heavily politically driven, which to my mind points to one of the paradoxes of the bloggosphere, and particularly the bloggosphere of the right. While lazy, sloppy and mendacious writing by the mainstream journals is rightly being called, the people doing the calling are themselves not objective searchers after truth.

This seems to me particularly true on the right, and particularly in America where there is an echo chamber between right wing blogs, Fox News, the American Spectator and the Wall Street Journal where speculative, exaggerated or simply invented stories on the net are picked up by the more traditional right wing media without any critical appraisal at all, and then will be discussed on elsewhere as if they were undisputed facts e.g. “so Sean Hannity, what is the implication of Hilary Clinton being a baby-killing devil worshipper?”. Of course the Daily Mail operates in this way, entirely by itself, in the UK.

The traditional media (what the right erroneously imagines to be a liberal conspiracy) is slow to respond to this because they still have some journalistic standards. Don't believe me? The New York Times sacked Jayson Blair for just making things up; Ann Coulter gets plaudits and shedloads of cash for doing much the same. They also help to propagate this stuff because they have concern about trying to be balanced – witness the completely nonsensical stuff around Whitewater, and the extent to which the Today programme has followed Daily Mail front pages for its agenda over the last five years.

I hold absolutely no brief for Horton at all, and I think his behaviour over MMR was absolutely outrageous and unethical. If he had personal honour he would have resigned over it. The interesting question is why wasn't he sacked. The answer lies in the real bias of publishers: for stories that are sensational, to be first with the news, to appear to be fearless speakers of “truth to power” when they kowtow to establishments, and above all to increase revenue. Horton's editorial decisions all fit well with these biases I would suggest.

All this is a bad thing, and I'm glad that they are being attacked for it, but I'm worried that in doing so we open the door for something much nastier, with even less concern for the truth, and with an even more blatant political agenda.

Steve Madden said...

No matter how much you downgrade the number, even if it’s cut in half, it still amounts to more deaths in three years, as a result of the American intervention, than all the killings of the Saddam Hussein regime in a quarter century - and all the killings at Hiroshima, Hamburg, Tokyo and Dresden combined. It’s still more than all the American dead of World War I and II combined. It’s still more than all the American dead, northerner or southerner, of the Civil War, and by far more than all the American dead, soldier and civilian, in all wars involving Americans since World War II, combined.

Terry Hamblin said...

Steve,

That's where the reality check comes in. Is it really feasible that there have been that number of excess deaths?

And in what sense are the Iraqi on Iraqi the result of the American intervention? They could equally be said to be due to 25 years of oppression from Saddam. Before the war we were being told of hundreds of thousands of children dying because of sanctions. Was that just propaganda?

Untangling the facts from the biased reporting is very difficult. I find it hard to make a judgement, especially as the same people who were opposed to the invasion of Iraq are calling for armed intervention in Darfur.

In the former Yugoslavia the break-up of a forced federation only needed the death of Tito to precipitate it. Western intervention there was to stop genocide. Genocide had been taking place in Iraq before the invasion and continues today. Arguably there has been too little intervention, rather than too much.

Steve Madden said...

Terry

That is the whole point, what was the intention of invading iraq?

It has shifted over time from WMD, to removing an evil dictator to installing democracy.

Iraq will never be a democracy, the civil war will escalate, shias and sunnis will never live in harmony.

George Bush talks about "winning the war" this is not possible.

Eventually the shia majority will set up sharia law and join with Iran to create a shia state.

This was predicited in advance of the war.

The Kurds will then want their own state, destabilising Turkey.

The only way to change this outcome is the impositon of another ruthless dictator.

The can of worms caused by the invasion of Iraq will resonate for decades after we decide it is no longer palatable to remain and the real target of the war on terrorism will be stronger not weaker.

Saddam was an evil bastard but at least he was controlable unfortunately his succesor will not be.