Copyright ©2008 NPR. For personal, noncommercial use only. See Terms of Use. For other uses, prior permission required.

NEAL CONAN, host:

This is TALK OF THE NATION. I'm Neal Conan in Washington.

While the death toll for American troops in Iraq is known precisely. Counting the number of Iraqis killed since the U.S.-led invasion is difficult, dangerous and politically charged. There is no consensus and estimates vary. The U.S. government says it doesn't keep accounts, so the task falls to outside agencies. And no matter which estimate you look at, questions arise about methodology and about political bias.

The latest survey issued just yesterday was conducted by the Iraq Health Ministry and the World Health Organization and it estimated the number of Iraqis killed in the first three years, following the invasion, at about 151,000. Other estimates are much higher, others much lower. So today, we'd like to go through these figures and try to make sense of the difficulty and the value of counting the dead in Iraq.

Later in the hour, Andy Kohut of the Pew Research Center joins us to talk about how and why pollsters completely miss the story on the Democratic side of the New Hampshire primary earlier this week.

But first, counting the Iraqi dead. How do you count the dead in a war zone? How can you tell who's an insurgent fighter or an innocent civilian? How do you tell who killed who? And how does the number of Iraqis killed color your view of the war?

Our number is 800-989-8255. E-mail talk@npr.org. You can also comment on our blog. That's at npr.org/blogofthenation.

We begin with Ties Burma(ph) - Boerma, excuse me. He is the director of Measurements and Health Information Systems for the World Health Organization and one of the authors of the new WHO survey. He joins us from the studios of the World Health Organization in Geneva, Switzerland.

And it's good to have you on the program today. Thank you for your time.

Dr. TIES BOERMA (Director, Measurements and Health Information Systems, World Health Organization): Thank you.

CONAN: And why is it important, do you think, that this survey and others like it be conducted?

Dr. BOERMA: I think in any conflict or emergency situation, it's important to have an idea of how many people are affected, how many people are dying. That's an important statistic to estimate the seriousness of the situation. It's an important estimate of how to respond in terms of health services and other issues. So while it's a very difficult issue to get the number right, it is an important figure.

CONAN: And how difficult is it? It's a country in the middle of a war in many places.

Dr. BOERMA: Well, indeed and, obviously, in any country, the best way to count the dead is through a registration system like we have in many countries and Iraq had the system in place, although it even was incomplete before the invasion but the system is now not functioning so we don't get a good count of the dead and what people died of from the death registry system, so other ways have to be found. One way that is also often used in developing countries is through household surveys. And in the household survey, one tries to identify death in, say, last three or five years in this case, and for obtaining estimate of how many people have died from the sample of households in the country.

CONAN: From a sample of households. Obviously, you can't go to every house in Iraq.

Dr. BOERMA: Indeed. And that is already a difficulty, particularly in conflict situations because violence tends to be very much located and one, therefore, needs a very large sample with many clusters or sampling areas to try to get the best possible picture. So in the survey that the results of which the results were released yesterday, the Iraq Family Health Survey, there were more than a thousand clusters selected all over Iraq. In each cluster only 10 people or 10 households were interviewed and those were selected prior to the field work.

CONAN: And who then went and did the survey? Was it military personnel or doctors or nurses? Who did this?

Dr. BOERMA: The survey was field request done by the ministry of health with the assistance of the statistical office in Iraq and was taking a close distance from us. The household visits were done by doctors and nurses.

CONAN: A lot of people in Iraq don't have much trust for the government.

Dr. BOERMA: In this case, we're talking about doctors and nurses that I think - I hope that the people still have trust in the people their own health workers.

CONAN: And your findings and we're saying 151,000 - that's somewhere in the middle of a high and low end from just over 100,000 to more than 200,000. So this estimate is just that. It's an estimate.

Dr. BOERMA: Absolutely. We know that a survey is an imperfect method to estimate death even in peace time and, especially also in conflict situations. So we have to be very careful when we interpret the results of a survey and this range of 104,000 to 223,000 comes from several factors that we've taken into account. One is that some of the clusters, about 10 percent, could not be visited. Those were - because of high levels of insecurity. So those clusters have high mortality than the cluster that could be visited, so we actually made an adjustment for that. For instance, in Baghdad, for the clusters that could not be visited, we assumed the mortality to be four times higher than the clusters that we visited and we used the Iraq body count, the media based counting of civilian deaths as a way to come up with that number.

The second thing is that, of course, as you said earlier, it is a sample so there's always some uncertainty around a sample, a sampling error. We took that into account. Thirdly, as also many people are aware, there's a lot of migration in Iraq and up to one or two million people may have left the country, so we took that range into account.

And then lastly, and this is a survey problem that is typical when one asks about death whether it's the last year or the last five years, is that households tend to move more after a death. And that means that those deaths don't get reported so one gets underreporting of deaths as well. So we also took arrange of 20 to 50 percent in our modeling to come up with the best estimates from this data and that's where the 104,000 to 223,000 comes from.

CONAN: We're talking with Ties Boerma of the World Health Organization about the numbers of Iraqis killed in the first three years after the U.S. invasion. 800-989-8255, if you'd like to join us. Talk@npr.org is the e-mail address.

And let's start with Brin(ph). Brin's calling us from Philadelphia.

BRIN (Caller): Yes. I'm concerned just by the introductory comments that were made, for instance, not counting as a war fatality if the family said it was a car accident because I think that kind - a number of situations in which a car accident might have occurred or any number of other bus fatalities, et cetera, that wouldn't have occurred had there not been a war. And I wonder if what kind of statistical - why that was not followed through with a follow up question when a person said they have a family member who died in the car accident, for instance.

CONAN: Okay. Ties Boerma.

Dr. BOERMA: The interviews - if a death was identified in the household, ask for the age and the sex of the respondents and then also they have a list of 23 courses including heart disease and infectious diseases and then for injuries. So for injuries, there were four categories: There was intentional injury, armed conflict, road accident and unintentional. And so we kept it quite simple because in that case if one asks about the cause of death from household members, it's often not a very reliable method to go into detail. Although with injuries, one could have perhaps added other questions, I think, as the other surveys done. But in this case, it wasn't done.

Now, you're right. Some of the road accidents could have been misclassified and could have been due to armed conflicts.

BRIN: High-speed and because for fear and things like that?

Dr. BOERMA: Yeah, that's right.

BRIN: I understand…

CONAN: Let him - Brin, let him talk, okay?

Dr. BOERMA: So if we look at the overall number of injuries of the all injuries together, post-invasion, about 7 percent were intentional injuries, 54 percent were armed conflict, 16 percent were road accidents and then 24 percent were unintentional injuries. So, if say, about a quarter of the road accidents could have been armed conflict it would add something to the number but it would have been relatively small proportion.

BRIN: Correct.

CONAN: Okay, Brin, thanks very much. We appreciate the call.

BRIN: Thank you.

CONAN: And Ties Boerma, we thank you for your time today we know it's late in Geneva, I appreciate it.

Dr. BOERMA: Thank you very much.

CONAN: Ties Boerma, the director of Measurements and Health Information Systems for the World Health Organization, with us from the studios of the WHO in Geneva.

Joining us now is Les Roberts. He's an associate clinical professor in the Program on Forced Migration and Health at Columbia University. One of the authors or a 2006 survey conducted jointly by Iraqi and Johns Hopkins University researchers. The study was published in the British medical journal The Lancet and he joins us now by phone from his home in Upstate New York.

And nice to have you on the program today.

Doctor LES ROBERTS (Associate Clinical Professor, Program on Forced Migration and Health, Columbia University): Thanks, Neal. Nice to be with you.

CONAN: And your estimate came in considerably higher than the WHO estimate over 600,000. Why the discrepancy do you think?

Dr. ROBERTS: Well, this is fascinating. I've had about a dozen interviews in the last day and I can't ever remember two studies having such similar results and having it being painted as so controversial. You know we did a study back on June and July of 2006 in which we tried to compare the death rate before the invasion with the death rate after. And our motive was exactly the motive of your first caller.

Normally, when we talk about death in the American Civil War, deaths in the holocaust, we're not just interested in deaths from bullets. We're interested in the deaths that occurred because of that disruption. And the war in Biafra, the war in Congo, 90 percent of all deaths were from the disruption not from the violence itself. So we found a death rate after the invasion 2.4 times higher, that is, mortality a little more than double. And this new survey found the death rate before and the death rate twice as high after. They also - and Ties was - Ties Boerma was very, I think, honest about this in the paper and was quite clear about it a moment ago.

They knew they had underreporting. For example, before the invasion they had a death rate of about three deaths per a thousand people per year, that's about half the death rate in the United States and they knew it was too low. Most of the neighboring countries are more around five. In our survey we had found 5.5 and so they adjusted it up. So we actually agree on what the death rate was before the invasion. We agree on what the death rate was after the huge contrast between these two studies is that we think virtually all of that increase was from violence and they believe that only a third of the increase in mortality was from violence so…

CONAN: We're going to have to take a short break. Can you stay with us?

Dr. ROBERTS: Gladly.

CONAN: Les Roberts is an associate clinical professor at Columbia University. More with him and more on this controversy in just a moment. Stay with us.

I'm Neal Conan. It's the TALK OF THE NATION from NPR News.

(Soundbite of music)

CONAN: This is TALK OF THE NATION. I'm Neal Conan in Washington.

We're talking about surveys of the numbers of Iraqis killed since the invasion almost five years ago now. And the latest was conducted by the Iraqi Health Ministry and the World Health Organization and it estimates that 151,000 Iraqis died in the three years following the U.S. led invasion. Numbers are difficult to come by. This is one is just the latest in a series of Iraqi mortality estimates. You can get a comparison of these surveys and track the events on the Iraq at our Web site at npr.org/talk.

We're talking right now with Les Roberts one of the authors of a 2006 survey conducted by researchers in Iraq and at Johns Hopkins University published in The Lancet, which estimated that more than 600,000 Iraqis had died as a result of the invasion.

Probably what you're saying is your difference in World Health Organization's number is, if for example somebody died of a heart attack because of stress the stress you suggest may have been brought on by the conflict.

Dr. ROBERTS: Well, actually I'm not saying something quite that innocent. What I'm saying is that we interviewed households and we found a much higher, more than doubling of death rate after the invasion, and when we asked people how did your loved one die, the vast majority of the new excess deaths, in fact, the majority of all deaths were from violence. When this WHO and statistical bureau of Iraq survey was done and these government employees asked people how did your loved die, only about one in six deaths for them was said to be violence after the invasion. So for one reason or another we have pretty dramatic difference in the fraction of the excess deaths that were from violence.

CONAN: Hey, is there anyway to resolve this, do you think?

Dr. ROBERTS: You know the great thing is yes. This is a very easily resolved by National Public Radio because if we're right in our Lancet study of 2006, if National Public Radio were to go out to a handful of cemeteries across Iraq and every cemetery has a ledger were they keep track of who dies, the name, the date and they look at the death certificate and most ledgers they write down the cause of death. And if we're right over that first three years of occupation most deaths nationwide should have been from violence. And if this new WHO survey is correct only about one in six deaths will be from violence. And I would think, without doing a fancy scientific survey, people could go out and pretty quickly get a pulse on the causes of deaths in the last few years and decide which of these surveys is roughly correct.

CONAN: Les Roberts, thanks for your time today.

Dr. ROBERTS: My pleasure.

CONAN: Les Roberts, associate clinical professor in the Program on Forced Migration and Health at Columbia University; one of the authors of a 2006 survey, which was published in The Lancet that estimated more than 600,000 Iraqis died violent deaths in the first three years after the war. He joined us from his home in Upstate New York.

And joining us now is Neil Munro. He is a science reporter for the National Journal and author of this month's feature story, "The Data Bomb." He's critical of some of the previous surveys. And he joins us here in Studio 3A.

Thanks very much for being with us.

Mr. NEIL MUNRO (Science Reporter, National Journal; Author, "The Data Bomb"): Thank you very much.

CONAN: And in addition to these two survey results, there are other methods of trying to estimate the numbers of dead in Iraq.

Mr. MUNRO: Yes. And it is a very difficult task because after all, people are trying to kill each other in all directions. One basic method is simply to count up the news reports of the dead.

This is a very useful method because it gives you a minimum. After that, the other basic method is to do a survey. Surveys are complex. And it can lead to errors, but can find more accurate numbers and can also find how many people are not dying because of a change in government, for example.

CONAN: And the number - the survey you're talking about, the media survey it's called sometimes, the Iraq Body Count, I guess is the best known of that.

Mr. MUNRO: Yeah.

CONAN: Their estimate through the first three years of the war, the same period we're talking about here, was 45,000 to 48,000. And again, that's a sizeable difference between them and the survey numbers, either survey.

Mr. MUNRO: Correct.

CONAN: And that's going to be, by definition, you suggest more modest, more cautious.

Mr. MUNRO: One of the virtues of the IBC people is that they say this upfront. They say, we just count the report. We do not claim to be counting the full thing. And anyway, that's quite admirable. It serves also as a usual floor on casualty estimates, death estimates, and a useful guide to trend. Again, both incomplete, both imperfect, but useful. And very cheap.

CONAN: Could either the U.S. or the Iraqi governments or both have been more helpful?

Mr. MUNRO: Why would they want to be?

CONAN: That's a good question, but could they be?

Mr. MUNRO: Well, you ought to remember what the war is about. Now, people argue about what the war is about, fine. But from one faction's point of view, the al-Qaida guys, they want to show that the change in Iraq is bad, disastrous, a catastrophe, a mistake so they want to pump up the dead. And they do this by exploding bombs in the middle of market places.

The other factions, if you like, the Iraqi government, the elected Iraqi government now, the U.S. government want to show the war as a success. And therefore, they want to minimize the cost and maximize the visible benefits. The - neither side has no - doesn't intent to tell the truth here.

CONAN: Let's see if we can get another caller on the line. And this is Jeff(ph). Jeff calling us from San Antonio, Texas.

JEFF (Caller): Yes. I had a question regarding - it seems like recently, I had heard a quote from maybe "The Diane Rehm Show" that was upwards of like, 1.2 million dead Iraqis. So, I was wondering how that compared, if anyone else has heard that.

CONAN: Mm-hmm.

JEFF: And what the ratio may be was from citizen to military personnel (unintelligible).

Mr. MUNRO: Hey, I'm your man on that one. The - okay, there was a report put out as Petraeus was about to speak in Congress, in which a British polling firm announced a death toll of 1.2 million. Well, you know, the media kind of published this. But if you scratch at it as I did, I called up the Iraqi guy who did this.

It was a - the British poll had delegated - the British firm had delegated the poll to an Iraqi who spent his own money to the poll. And he tell me, I do these polls to drive the Americans out of Iraq. When the Americans invaded, I was an officer. I was teaching at the ministry of defense at a national defense university in downtown, Iraq.

And I wanted to fight the Americans and so I thought I'd use polling to get the Americans out of Iraq. And I timed this survey to coincide with Petraeus. And no, I'm not giving you the days. And no, I can't get a (unintelligible) reviewed yet and this kind of stuff. Now, some people believe that poll, others don't.

CONAN: Thanks, Jeff.

JEFF: All right. Thank you.

CONAN: Bye-bye.

In fact, timing of the release of these surveys has been significant in the past.

Mr. MUNRO: Yes. Gosh, there's a lot of suspicion in this business. And some people say, including the - one of the - some of the polls are timed for political events. And the two poll published in the - the two surveys published in The Lancet and done by John Hopkins were, they made every effort to release them before the 2004 poll, the 2004 election and the 2006 election.

It was said by the John Hopkins guys, they had to release them before the election, otherwise, our - their survey teams might have been killed, perhaps, but that's what they said.

CONAN: Well, joining us now is Sarah Sewall, she's director of the Carr Center for Human Rights Policy at Harvard University where she directs their program on National Security and Human rights. She was first deputy assistant secretary for peacekeeping and humanitarian assistance during the Clinton administration. And she joins us now from member station WBUR in Boston.

Nice to have you on the program today.

Ms. SARAH SEWALL (Director, Carr Center for Human Rights Policy, Harvard University): Thanks for having me, Neal.

CONAN: and I think you heard Neil Munro say why would the U.S. government or the Iraqi government release these numbers? Isn't it the responsibility of the U.S. government to keep track of the wreckage in Iraq?

Ms. SEWALL: Well, let's consider that we're responsibility, legal responsibility, moral responsibility, professional responsibility. I think you can debate what kind of response ability the U.S. military has. I like to think of their responsibility as one of minimizing civilian harm, which is what they claim to do. So, I'm less concerned about what they're counting than that they're counting the things that they need to count in order to learn and make good on the vows to minimize civilian harm.

And there's been some interesting developments on the ground in Iraq just in the last couple of years on that front, where I think the U.S. military is finally recognizing the value of using data so that they can learn and reduce swelling casualties.

CONAN: Well, it's not just the Defense Department. It's the United States government. I can understand that the Defense Department is busy with maybe other things. Is there no one else who takes this - or has this job?

Ms. SEWALL: Well, if you look at the - Iraq for example, there's really no one else in the field. So if this job is to be carried out in situations with a high degree of violence, it probably will be carried out by the military. And that begs the question of whether or not - it really raises a more fundamental question, which is why do you want to know the numbers of civilian dead?

And depending on why you want to know the numbers of civilian dead, you're going to have different requirements for your accuracy and you're going to have different corollary questions. So if you want to know the overall costs of war, then perhaps, surveys that include all causes of death and show you the change in the overall death rates or what you need them, and maybe to the person accuracy is not vital for that question.

If you want to know trend lines, you've got a different - we've got a different sort of approach whether your strategies weren't working and that's particularly important in counterinsurgency. Whether you want to be responding to specific problems, you may want to focus in on one particular type or cause or area of death. And if you're trying to do operational learning from a military perspective, it may be most important for you to identify the areas in which the largest number of civilians are being killed - that you really focus in on correcting the situation.

CONAN: Well, there's also a public relations and political aspect to this. I think, we all have to admit. And if the Defense Department or the U.S. government doesn't release numbers it's at that mercy of people who are showing estimates anywhere from 45,000 to 1.2 million.

Ms. SEWELL: Absolutely. And that's an argument that I've made for a long number of years. One of the interesting things about the Iraq Body Count, which is now considered sort of the - as we just heard, described the floor in the civilian casualty counting business.

I remember very well a couple of different conferences with military officials where everyone was questioning the method and the motive of the IBC's approach. And it wasn't until the first Lancet survey came out that everyone said, oh, well, goodness. The Iraq Body Count is so much more reliable. So the - our targets keep shifting. And I think the fundamental point is that if we really wanted to know, we could do all source data compilation and we could combine all the different methods that we know; each have their individual flaws.

And we could get fairly, good gross approximations of the kind of humanitarian impact from war. And there's a different question, and I think an important separate question about whether operational learning based on an understanding of the causality of civilian harm from the perspective of the U.S. military. And that I think the U.S. actually has a success story there in terms of looking at what they call escalation of force incidents. And when they started to - you know, the saying in the military is you inspect what you expect.

CONAN: Hmm.

Ms SEWELL: And when they started recording the numbers of civilian deaths at checkpoints and block points principally, they were able to really focus in on the tactics, techniques, and procedures they were using and reduced the number of civilians killed from 10 a week in July '05 to six in a week - I'm sorry - to one per week by June '06. That's a significant change. And it really points, I think, to the value of measurement in trying to achieve objectives.

CONAN: And one final question, we - I guess haven't addressed, and that's moral responsibility. The United States started this conflict. It initiated this.

Ms SEWELL: Are you asking me that question?

CONAN: Yeah, but…

Ms SEWELL: I think the U.S. ought to - I don't think if the U.S. issues casualty figures, it will have the credibility than it needs, but I think, that the U.S. ought to capture what it can and share that information with others who would be able to similarly share information and everyone would be able to reach their own understanding and assessment.

You know, one of the arguments I hear all the time from the Defense Department is, you know, not only - is it not really our job to count, but nobody will believe us. Well, just share what you have and everything is going to have its own limitations. They're going to be always questions of both bias and methodology.

And I think the more we know, the more we'll be able to likely triangulate our way to the answers that will help us accomplish humanitarian objectives in the face of what everyone hopefully would agree is war that we'd like to avoid.

CONAN: Sarah Sewall, thanks for your time today.

Ms. SEWELL: Thanks for having me.

CONAN: Sarah Sewell, director of the Carr Center for Human Rights Policy at Harvard. The - she was the first deputy assistant secretary for peacekeeping and humanitarian assistance during the Clinton administration with us today from member station WBUR.

And you're listening to TALK OF THE NATION from NPR News.

And still with us is Neil Munro, science reporter for The National Journal, author of the feature story, "The Data Bomb."

And let's see if we can get another caller on the line. This is Carl(ph). Carl with us from Buffalo, New York.

CARL (Caller): Good afternoon.

CONAN: Good afternoon.

CARL: The question of the U.S. military reporting is rather disingenuous one since clearly the factors of the war have taken an enormous toll. I think characterizing it as wreckage as opposed to carnage suggests a bit of subterfuge.

Additionally, the question of, are they hard attacks or have the Iraqi people suffered as they did in Fallujah by the possible use of chemical warfare, the use of depleted uranium poses a continual problem, and the infrastructural damage, obviously, has an enormous impact on health care.

Furthermore - and this may sound like a diatribe - what would I like this to be considered as a rebuttal. Furthermore, the Iraqi people suffer from the loss of heads of households and you have an enormous number of children who are at the risk of injury, death, et cetera. And that can't be - the total impact can't be considered without looking at a long-term generational dysfunction.

CONAN: Now, Neil Munro, does Carl have a point there?

Mr. MUNRO: I write about science. If Carl wants to say the results are good or bad, fine. But I write about science. And the Iraqis get to vote on their government. Whether they want to go through this or not is up to them.

CARL: Again, your response is disingenuous. Please, address what I have said. It is not electoral by any means. It is the U.S. imposing its will and the fallout of that.

CONAN: Okay. Carl, I think, we just going to have to leave it there.

As we look at this controversy, do you agree with Les Roberts that there are easy ways to resolve these conflicts, Neil Munro?

Mr. MUNRO: Easy is too strong a word because technically, there are some ways to resolve a variety of existing doubts. But as Carl made clear, this is a political issue. People want to believe certain things. They want to assign certain values to certain numbers of death and define them up and down and such like. The science is easier to decide. The harder thing is to decide the cause and benefits of the war.

CONAN: Is it easy to decide - to determine, on one hand, who caused casualties to who? Obviously, the conflict was initiated by the United States, but within that conflict?

Mr. MUNRO: Well, the first conflict was initiated by United States, knocking down Saddam. That was actually fairly low-destruction war as the wars go. The second conflict was initiated by the Sunnis when they were trying to regain power. Al-Qaida joined with them. Hmm, okay. Even Americans get to decide who is responsible for that. Whatever the situation is, Americans also get to decide what the moral value of life is and whether it should be counted and whether they want to continue going or not. But in terms of science, scientists can provide a lot of answers.

CONAN: Neil Munro, thanks for your time today. Appreciate it.

Neil Munro is science reporter for The National Journal, author of this month's feature story, "The Data Bomb." He joined us here in Studio 3A.

When we come back, we're going to be talking about polling and New Hampshire and why the pollsters got it wrong. Andy Kohut, head of The Pew Research Center joins us.

800-989-8255, if you'd like to join that conversation. E-mail, talk@npr.org.

This is NPR News.

Copyright © 2008 NPR. All rights reserved. No quotes from the materials contained herein may be used in any media without attribution to NPR. This transcript is provided for personal, noncommercial use only, pursuant to our Terms of Use. Any other use requires NPR's prior permission. Visit our permissions page for further information.

NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR's programming is the audio.

Comments

 

Please keep your community civil. All comments must follow the NPR.org Community rules and terms of use, and will be moderated prior to posting. NPR reserves the right to use the comments we receive, in whole or in part, and to use the commenter's name and location, in any medium. See also the Terms of Use, Privacy Policy and Community FAQ.

Support comes from: