Our collective failure has been to take our political leaders at their word. This week, the BBC reported that the government's own scientists advised ministers that the Johns Hopkins study on Iraq civilian mortality was accurate and reliable. This paper was published in the Lancet last October. It estimated that 650,000 Iraqi civilians had died since the American- and British-led invasion in March 2003.

Immediately after publication, the prime minister's official spokesman said that The Lancet's study "was not one we believe to be anywhere near accurate". The foreign secretary, Margaret Beckett, said that the Lancet figures were "extrapolated" and a "leap". President Bush said: "I don't consider it a credible report".

Scientists at the UK's Department for International Development thought differently. They concluded that the study's methods were "tried and tested". Indeed, the Hopkins approach would likely lead to an "underestimation of mortality".

The Ministry of Defence's chief scientific advisor said the research was "robust", close to "best practice", and "balanced". He recommended "caution in publicly criticising the study".

[snip]

http://commentisfree.guardian.co.uk/richard_horton/2007/03/counting_the_cost.html