The Times runs a story on the latest developments querying Burnham et al.
I’m one of the people that supported the study, not as a flawless piece of work, but as an important contribution to the study of mortality in post-invastion Iraq. The methodology – “proven” in epidemiological studies – was a useful one to adopt in this case. And I still think it’s appropriate to discuss statistical interpretation. The world may have focused on the 650,000 figure, but even the lower bound is a very high number.
But, I’m a Bayesian, or try to be, and it appears that cracks are appearing in the study. I’m willing to concede that these very well informed and expert critics may be right, in the sense that the implementation of the study may have been flawed. Indeed, in the worst case, there may have been fraud. But what is it likely to change, other than the job prospects for Dr. Richard Lambert, the Editor of the Lancet? True, there may be greater scrutiny of such politically-charged studies, which must be welcomed, but what of the war policy itself?
In fact, as with the situation when the study first came out, any sensible debate as to whether 650,000 deaths is a bad number is somehow meaningless unless compared to the benefits the Iraqi’s receive now or in the future. And of this there has been little mention other than ‘freedom’ which is questionable in itself; the freedom to vote masks the hugely inhibitions people place on simple freedoms like movement, work and dress. At best, for many of the population, there has been no change here. Indeed, which of ‘Life, Liberty and the pursuit of Happiness’ do current supporters of the war think has increased for the Iraqis?
If we are to do the rounds on this story again, perhaps we could focus on the benefit side, at least so we can make that cold calculation to estimate whether it really is worth it.
