But a new survey by U.S. and Iraqi researchers suggests the civilian death toll as a direct or indirect result of the U.S.-led invasion may be close to 100,000 -- far higher than anyone previously thought.
The survey was published on the Internet today by "The Lancet," a prestigious British medical journal. It's conclusions -- and methodology -- look certain to spark controversy just four days before the U.S. presidential elections.
The administration of U.S. President George W. Bush publicly says as a matter of policy it has not estimated civilian war casualties in Iraq. But in recent remarks before the Senate Armed Forces Committee, Defense Secretary Donald Rumsfeld criticized the media and opposition Democrats for focusing too much on casualties in Iraq.
"Is [Iraq] dangerous? Yes. Are people being killed? Yes. Is it a violent country? You bet," he said. "Were there 200 and some odd people killed in Washington, D.C. last year? Yes. Were they on the front page of every newspaper? Were they on the television every night? No."
Previous estimates by independent scholars and organizations put the civilian death toll in Iraq at between 10,000 and 30,000. But the new study paints an even grimmer picture.
Last month, research teams led by scholars at Baltimore's Johns Hopkins University interviewed 1,000 Iraqi families in 33 locations. The families were asked about deaths and births in their homes in the 15 months before the March 2003 invasion and in the 18 months since.
The figure of 100,000 civilian deaths excludes findings from hard-hit Al-Fallujah -- a center of the antioccupation insurgency and the site of heavy, frequent combat. Including Al-Fallujah, the study says the death rate would be higher still.
Just how they died is also fairly clear. In 15 of the 33 communities visited, residents reported violent deaths in their families, most of them due to attacks by U.S.-led air strikes. Researchers concluded that the risk of violent death was 58 times higher than before the war.
And most of those killed were women and children, a point Al-Sadr City resident Ali Mutlaq made to reporters after a U.S.-led raid on the restive city last month. "As usual, the warplanes dropped bombs without any reason, there was no resistance," he said. "They used to say the Al-Mahdi Army attacked them, but there was nothing yesterday. What is the guilt of the children and the women and the houses that were destroyed?"
The Bush administration has not yet commented on the survey. Democratic Party challenger Senator John Kerry has made Bush's handling of the Iraq war a key criticism in his drive to replace him at the White House.
The study purports to be scientific and makes no political judgments. But its leader, Dr. Les Roberts, told AP that he opposed the Iraq war from the start and that the survey was intentionally released to have an impact on public opinion just before the elections on 2 November.
The researchers themselves say they sought to be extremely technical in the way they chose their interview subjects and sites. However, they acknowledge that locations were limited by the decision to cut down driving time to reduce the risk to researchers.
Some experts, such as Oxford University's Richard Peto, have said that because of those location limitations, the researchers may have focused on hot spots that are not representative of the death rate across Iraq.
Roberts, for his part, criticized the Bush administration's refusal to release estimates on Iraqi civilian deaths. He said his study proves that with moderate funds and researchers willing to risk their lives, "a useful measure of civilian deaths could be obtained."
Carl Conetta is a scholar with the Project on Defense Alternatives, Washington, D.C. think tank. Earlier this year, Conetta released a survey that said that up to 15,000 Iraqi combatants and noncombatants had been killed during the first two months of the U.S.-led invasion.
Conetta said it shouldn't be too hard to estimate an Iraqi death toll. He told RFE/RL the Iraq war has been the most scrutinized conflict in history -- beamed to people worldwide by satellite television and the Internet. "Never has this been the case before. So if we don't have any idea of how many people are killed here, then we don't have any idea of any war: we don't have any idea of how many people Saddam Hussein killed; we have no idea of how big the Holocaust is, because of we don't know this, we don't know any of those things, either," he said. "But of course, we have some sense [of the death toll]."
In London today, British Foreign Secretary Jack Straw told BBC radio that the estimate appeared "very high." But he vowed that his government, the United States' main ally in Iraq, would study the survey "in a very serious way."
[For the latest news on Iraq, see RFE/RL's webpage on "The New Iraq".]