Johns Hopkins’ Iraq numbers

It would be nothing short of horrific to contemplate 600,000 excess Iraqi deaths as a result of the war. That would be just as true today as it was in 2006, when claims of such a figure first emerged, implying a new Rwanda of nearly genocidal slaughter. This number, produced by a Johns Hopkins research team shortly before the 2006 midterm elections and published in the British journal the Lancet, changed the terms of debate that year, dwarfing existing estimates by a factor of 10. It continues to be cited in public debate today — even though the number has never been independently corroborated, and even though reasons to doubt its origins and veracity continue to pile up.

Last week, the National Journal released what is by far the most extensive effort to get at the truth of the origins of the Hopkins numbers. Despite evident months of research, authors Neil Munro and Carl Cannon turned up few hard answers. The Hopkins authors continue to decline to release the full data, refusing Messrs. Munro and Cannon, citing security considerations. Instead, what the National Journal did turn up are further reasons to question the authors.

“They failed to do any of the [routine] things to prevent fabrication,” Fritz Scheuren, vice president for statistics at the National Opinion Research Center, told the National Journal. Messrs. Munro and Cannon report evidence of “data-heaping” — the statistician’s term for evidently faked numbers — which appears in one of the few disks of data the researchers released to outside investigators. Unusually low rates of unoccupied or uncooperative subjects raise questions of “curb-stoning,” or falsified data forms. Researchers also failed to ask respondents for fraud-preventing demographic data, such as ages and birth dates. Photographic evidence of death certificates purportedly examined amid data compilation has not been demonstrated, either. Then, the man who physically collected the data, Riyadh Lafta, a onetime child-health official for the Saddam Hussein regime, declined the National Journal’s request to be interviewed. As the only person fully capable of dispelling the questions, Mr. Lafta’s silence is suggestive.

For a number so often cited in the press and by politicians and so shocking, it ill serves the war debate that this number has never been corroborated. For instance, this past weekend, in The Washington Post, former Sen. George McGovern cited the number, calling the study “careful,” as he argued for President Bush’s impeachment.

Surely, the toll of death and destruction in Iraq is horrific by any account. Lest it seem a statistical quibble to try to corroborate this figure, though, consider that the most-often cited authority, the antiwar group Iraq Body Count, today finds 80,331-87,742 documented civilian war deaths in Iraq. This is the difference between epochal human tragedy and genocidal madness. It matters.