by John Tirman

ONCE AGAIN, a controversy has erupted over how many people are being killed in Iraq. It's an important debate, not only for beleaguered Iraqis, but for Americans seeking stability and a timely exit.

Mortality figures alone can tell a compelling story. Add to that other numbers that fill in our understanding even more - such as the scale of the flow of refugees or the women widowed by the war - and we have useful information.

So what are these statistics, and what do they tell us about this nearly five-year-old conflict?

Two kinds of accounts have emerged on the question of mortality. One is a literal count, body by body, from reports in the English language press. Because the media, mostly based in Baghdad, cannot grasp most of the violence, this is an undercount (now about 84,000) even by the reckoning of its authors, the UK-based Iraq Body Count.

The second method is to go out and ask the question in surveys of randomly selected households. This has been done five times under very dangerous conditions. Surveys of this kind during war are relatively new, and, as a result, it's not surprising that the numbers they've produced have varied. But there is significant congruence.

The surveys agree that mortality is much higher than is typically held in political discussions about Iraq. The highest figure, from Opinion Business Research, a private survey firm in London, is 1.2 million through August 2007. It is also the most recent.

About 15 months ago, a survey commissioned by my center at MIT and published in The Lancet found that 601,000 had died by violence through June 2006. This figure has created a firestorm of criticism, but the methods are sound and none of the many peer reviews found anything greatly amiss. (One recalculation brought the death-by-violence total down to 450,000.)

Then last week, Iraq's Ministry of Health released its large survey, also ending in June 2006, finding that 151,000 had died by violence. But their data tables show an enormous "excess death" total of nearly 400,000 caused by the war, and a peculiarly flat rate of violence throughout the war. Because the interviewers worked for the government, it's likely that many respondents attributed deaths to nonviolent causes, in order to protect themselves from unwanted attention.

What to make of all this? The first conclusion is that hundreds of thousands of people have died as a result of the war - this seems incontrovertible. It is buttressed by the large number of displaced - some 3 million to 3.5 million caused by the war - and a reported total of 500,000 war widows.

The second conclusion, which helps us understand the violence, is that such a human catastrophe accounts for the insurgency in ways that no other explanation does. Whatever one makes of these insurgents, they appear to be fighting to defend their towns and tribes (apart from Al Qaeda's foreign operation). Violence begets violence, especially when foreigners are involved.

The third conclusion is that Iraq's devastation runs deep and wide. A generation of young men is being wiped out. Many of the most educated have left. The poverty of widespread widowhood may become chronic. The healthcare system is in shambles. Neighborhoods and towns ethnically cleansed means long-lasting displacement for tens of thousands. The humanitarian aid challenge is vast, and will last for many years.

How this affects US strategy is complex, of course, but two things stand out. First is that strategies to reduce violence against civilians and to increase economic and physical security are paramount. US leaders seem to grasp this, but their actions (arming Sunni militias, for example) may prove foolhardy.

Second, Iraq's neighbors must be part of the solution, given the scale of misery. President Bush has never embraced this idea, but it seems more and more obvious as the war drags on. Yet on Bush's recent trip to the region, Iraq was nearly absent from his agenda.

The lessons from the killing fields and refugees and widows won't go away. The sooner we fully realize the scale of this catastrophe, the better we may be able to work on reconstructive remedies.

John Tirman is executive director and a principal research scientist at MIT's Center for International Studies.
© Copyright 2008 The Boston Globe