Have 600,000 Iraqis died violently since 2003?

I'm sure most people have seen this figure, and it cropped up in another Iraq related thread. I thought it would be an interesting thread to talk about the study methodology and its accuracy.The report can be read here - registration is required but it is free.The method used is known as 'cluster sampling', I'll quote from the paper: They found that the pre-invasion mortality rate was 5.5 per 1,000 people per year. Post invasion the mortality rate was 13.3 per 1,000 per year. This translates to 655,000 more deaths than normal since the invasion. About 600,000 of those were violent deaths. The most common cause of death being gunfire. Some of the other 55,000 deaths could be indirectly related to war - things such as medical supply problems, fresh water supply issues and so on.Many people have criticized this figure, though those that agree that its findings might be accurate say that the method has been used successfully in other areas. Since there isn't a physical body count, we can only be so confident of the numbers, but the Lancet is confident that the confidence level is high. After all: There is an interesting test of the cluster sampling method (in general) here which finds that

quote:

---

: Immediately following a cluster survey to assess mortality retrospectively in a town in North Darfur, Sudan in 2005, we conducted a systematic survey on the same population and again measured mortality retrospectively. This was only possible because the geographical layout of the town, and the availability of a good previous estimate of the population size and distribution, were conducive to the systematic survey design.
RESULTS: Both the cluster and the systematic survey methods gave similar results below the emergency threshold for crude mortality

---

I have found no reasonable criticism of the report, only incredulity at the high number and how it compares with other figures.Comments?

The report discusses this very issue. It seems to state that if anything, it would cause final figures to be an underestimate, not an overestimate.There were 3 clusters that were discounted due to accounting problems, and some problems with violence in certain areas had to be dealt with. It's certainly not a walk in the park, but the Lancet was confident that it was handled the situation in a comparable way to other unsettled conditions (eg Darfur).As always, it probably had an effect - but passive methods are even more susceptable to errors than this (so they say anywho!)

That's unfortunately not the case. Right there in the OP is a link to the study in question. Why not go and read it? What part of the decrepit and war torn adjectives gives you the impression that the country is actually able to handle this influx? The Soviet Union had 11,500,000 civillian deaths in WWII which is over ten times the death rate in Iraq.

War/Country                    Civillian Deaths Per Day (rough est)
​
Vietnam                                     700
WWII Soviet Union                          5200
Darfur                                      270
Enduring Freedom/Iraq (Lancet)              460
Enduring Freedom/Iraq (Bush's est)           27
WWII/China                                 3200
WWI/Russia                                 1100
WWI/Germany                                 422
Second Congo War                           1944

It sounds like incredulity not a mathematical argument as your title seems to suggest.Edited by Modulous, : No reason given.Edited by Modulous, : added comparison tableEdited by Modulous, : added two different version of Iraq death estimates for comparisonEdited by Modulous, : More figures!

You need to read either the OP or the paper itself. They found the death rates before the war, and the death rates afterwards. The causes of death for the final figure were not considered.

quote:

---

. Pre-invasion mortality rates were 5·5 per 1000 people per year (95% CI 4·3-7·1), compared with 13·3 per 1000 people per year (10·9-16·1) in the 40 months post-invasion.

---

quote:

---

Aside from violence, insufficient water supplies, non-functional sewerage, and restricted electricity supply also create health hazards.9,10 A deteriorating health service with insecure access, and the flight of health professionals adds further risks. People displaced by the on-going sectarian violence add to the number of vulnerable individuals. In many conflicts, these indirect causes have accounted for most civilian deaths

---

quote:

---

The number of people dying in Iraq has continued to escalate. The proportion of deaths ascribed to coalition forces has diminished in 2006, although the actual numbers have increased every year. Gunfire remains the most common cause of death, although deaths from car bombing have increased.

---

Read it and you will gain this information. They calculated mortality rates before and after and then calculated how many extra deaths in total there had been since the war began. The entire point is that in countries of conflict, official 'body counts' have almost invariably been over 5 times too low (according to the Lancet, the OP discusses this and provides an appropriate quote). The point being, that journalists and soldiers cannot know what is happening to every single civillian since there are millions of them. Soldiers on the other hand, report more of their injuries which are then recorded. The OP asked for reasons as to what is fallacious about the methodology used and/or the conclusions reached. Declaration of them being a lie because you don't believe them doesn't count. I don't believe them, but I refuse to be swayed by my beliefs in the matter and will only accept reason.Edited by Modulous, : No reason given.

Other than basic incredulity, there is some reason to doubt the numbers.For a start, the death certificates. If 92% of deaths have a death certificate produced then an additional 603,000 death certificates should have been issued by officials in Iraq. Surely then, this number would be noticed? Perhaps the record keeping is particularly poor right now - as the Lancet reports, morgue watching is lucky to catch 20% of deaths. This lends us to question under what criteria these death certificates are being issued. The Iraqi Ministry of Health only reports having issued about 50,000 death certificates.Most of the deaths were by gunshot, but many were by explosion. Explosions lead to casualties who would be taken to hospital - often three times more are injured than die. Surely the hospitals would be reporting more injuries than they actually are? The Lancet report estimates 150 of car bomb/IED deaths occur daily, yet hospitals are only reporting a handful. Another peculiar thing is that hospitals have only reported 60,000 wounded for all causes. This puts a question on the concept of three people being injured for every one that dies. I suppose this is potential evidence that the deaths/injury ratio is not standard *or* that official reporting figures are skewed.The state of affairs must be very bad for the official channels to have the discrepencies that it does — if war is known for anything it is ensuring that he state of affairs is very bad.

---

Source of criticism figures: Iraq Body Count

If you look at page 8 of the other must read paper it gives you a map of the relative cost of war for the governates. Yes - the various areas of trouble and strife were all part of it...in a sense. The highest population areas were asked more frequently than the lowest one. It was cluster sampling so it was meant to get a picture of Iraq as a whole, not concentrate on 'bad' areas over 'good'.

Province     Mid-year 2004 population    Number of clusters
Baghdad                 6,500,000                    12
Ninewa                  2,521,300                    5
Basrah                  1,981,900                    3
Sulamaniyah             1,605,600                    3
Thi-Qar                 1,538,900                    3
Babylon                 1,408,700                    3
Erbil                   1,334,200                    3
Diyala                  1,271,300                    3
Anbar                   1,271,000                    3
Salah Al-Din              976,100                    2
Najaf                     950,200                    2
Wassit                    938,700                    1
Qadissiya                 915,600                    1
Tameem                    881,500                    1
Missan                    848,300                    1
Dahuk                     817,400                    0
Kerbala                   741,700                    1
Muthanna                  569,900                    0

Total                  27,072,200                   47

Edited by Modulous, : No reason given.Edited by Modulous, : adding tableEdited by Modulous, : page nuber

thanks - I was just reading through that when I came back here - I thought it would be odd trying to rebut myself (again), so I'm glad someone else did it

Why not? This kind of sampling has been used before succesfully. Indeed, it was used to determine mortality rates in Iraq pre-war which agreed with other methods. While there maybe inaccuracies, the contention is that the other methods so far used are fraught with more inaccuracies which are documentable. If you read the papers you will see that it stresses most of the deaths are from insurgency, a brewing civil war and possibly lawless murder. It states that only a small number are from coalition forces, and they only counted those cases where the people involved where very sure it was coalition troops and even then the study makes cautions about the figure.

quote:

---

The proportion of deaths ascribed to coalition forces has diminished in 2006,
though the actual numbers have increased each year. Gunfire remains the
most common reason for death, though deaths from car bombing have
increased from 2005.

​

...

​

Households were asked what party was responsible for the killing of their household member.
In many cases it was not clear. There was great difficulty in identifying which were criminal events.
Only when the household was certain that the death was as a consequence of coalition actions was
this recorded as such.

---

And look at table 2 where it says that only 31% of deaths are coalition attributed and 14% are air strikes. Of course, for those deaths the coalition are directly responsible for they should be held accountable for that. Likewise for those deaths for which they are indirectly responsible. It is not self-reputed. The figures are derived from basic statistical sampling mathematics. The same mathematics that gives percentages of religious groups, percentages of people who are anti-evolution in the USA, percentage of people who voted Republican at the polls etc.The reason it states the deaths are war related is straightforward reasoning. There have been 655,000 more deaths than would normally be expected in Iraq since the war began. Not all of them are directly war related, but the vast majority are. Unless you can think of some other factor in the deaths of civilians in Iraq that coincides with the timing of the war? Then please, sumarise and bring forward the argument to this thread. That is why I posted it, after all!Edited by Modulous, : No reason given.

That political bodies that have a political interest in rejecting the figures reject the figures is not surprising. Especially when two of those political bodies have mislead the public about the issue before. a) This is not relevant to the veracity of the figures.b) Your response is in stark contrast to the original statement you have made in Message 14 where you said "No, the purpose of the inquiry is to established the number of "civilian casualties" at the hands of US troops. " Agreed - which is why I don't lend much credit to those particular figures, and why the report goes to lengths to express that the figures are difficult to corroborate and states that they are only based on witness testimony. It was discussed in the paper and in the OP. The methodologies that come to the lower count demonstratably underestimate the true casualty rate. In the recent past the most accurate those methods have been is to capture 20% of deaths. In some cases they only capture 5% of deaths. It was a different report, from a much earlier stage in the war, and is not under discussion here. The report we are exmamining has different numbers, it discusses the range within which the 95% covers. The lower bound in this report is about 450,000. It states that the chances of it being much different from the 655,000 drops off dramatically as you move only a short distance away from this figure. Either way, we don't argue by link here, you should be able to sumarize the argument presented there and bring it here.Edited by Modulous, : No reason given.

You say "I'm saying that the Lancet report is being misused as some sort of tool of vengence to equate to "civilian deaths at the hands of coalition forces," to further stigmatize the war effort. If you type that in to a search engine, blog after blog is misrepresenting the issue." And provide links, without supporting quotes. This link simply rehashes the news story. It doesn't discuss hands of coalition forces, stigmatization or any such thing. Why do you think it supports your previous statements?Indeed - it merely states what is said in the Lancet:

quote:

---

The Lancet study, funded largely by the Massachusetts Institute of Technology's Center for International Studies, said that while the percentage of deaths attributed to the U.S.-led coalition has decreased over the past year, coalition forces were involved in 31 percent of all violent deaths since March 2003. Most of the deaths in Iraq, particularly in the past two years, have been caused by insurgent, terrorist and sectarian violence.

---

This one once againt hashes out the facts of the Lancet. It even offers up some counter-points.The other article seems to be the exact same article as this one. We need some details as to why the methodology is in question.
How is the 600,000 figure being distorted as being soley violent death, that's what the methodology has concluded. Since this is additional to the methodology issues, what distortions are you referring to? Who is saying anything about a careless coalition force? All that is being said is that 31% of deaths were attributed to the coalition.Why is the report improbable? I have addressed your "but that's lots of people" concern. The "How come nobody has noticed" concern is discussed. Address the rebuttals or provide another reason you find the figures improbable. Is it just incredulity? Indeed, a valid concern. Perhaps we should consider what percentage of Iraqi civilians are part of an insurgency and see how that might skew the results. This sounds like a more constructive avenue of debate.Edited by Modulous, : No reason given.Edited by Modulous, : No reason given.

To be the first active study of the toll of the war in Iraq. The reason being that the passive methods so far used rarely capture a fifth of deaths accurately, and more often it can be worse than that.Point of contention 1: This is the only active research method so far carried out. This kind of investigation is usually gives a more accurate figure than other kinds of investigation. If that were the case, the figure would be 30% not 87%. There were 629 deaths reported, and of those 545 were asked for documentation. They would have asked for more - but it was occasionally judged to be dangerous to question the word of a man in his own home. This is the only active study. Other methods are renowned for underestimating death tolls by a fifth at best. News reports are actually one of the methods used to tally these death tolls, so of course they are going to agree!There is evidence of over 400 people dying daily. Nearly 13,000 Iraqis were asked about deaths in their household. Of this group, their mortality rate was 13.3 per 1,000 per year.No major incident relies on physical body counts for a death toll. It's just not practical. That is why the American government invests millions of dollars into training people to use methods such as cluster sampling to gain an estimate of death tolls. A non-war example would be the Boxing Day Tsunami. Other examples include Darfur. Pure body counts as a method almost always underestimates the toll by a large degree in major incidents.Even before the invasion, some methods of body counting (counting death certificates issued) would only give us 30% of actual deaths.Point of contention 2: Sampling is used in similar situations to get an idea of deaths, and it doesn't get questioned. Even in the peacefulness of west Europe, decades after the World Wars, not all of the bodies of non-combatants have been found. Why would find them days/weeks/months later? It's not like there is an organized hunt for bodies going on. If they get buried, they won't get found unless an entire of a country is scoured. Can you imagine digging up the entirety of Idaho twice in the middle of a modern near-civil war with a barely effective government and a foreign power in place with foreign fanatical insurgents taking pot shots at everyone? That's the scale of the task you are asking here - just so you know. And how do we know that 10% of Americans are left handed unless everybody in America is put through dexterity tests every year? I don't know anyone who has done such a test. How do we know that belief in creation is growing in popularity, without testing the faith of every single person in intense psychological experiments (to make sure they don't lie)?We didn't see the 5.5 bodies per 1,000 people per year that were dying before the invasion. But the media are mostly concentrated in Baghdad, they cannot possibly report on every single death that happens in near-anarchy near-civil war conditions with Arabic foreigners coming in adding to the confusion.Point of contention 3: The media simply cannot report most deaths as they happen. There are many murders and gun battles going on that the media are ignorant of.

---

If you can answer the points of contention, I think we can move this debate forward. About half of your thread is allowing yourself to get drawn off topic into a general discussion on Iraq. Please don't let this happen, the thread is very specifically about this single report and its methodology. Thanks.Hopefully there are better arguments that can be levelled at this study than your incredulity. I started this thread with a 50-50 opinion of the study. Given the best arguments that have been raised so far - I'm leaning closer to accepting the figures might be accurate.Let's try and get this debate moving forwards, not sideways!Edited by Modulous, : No reason given.

There are about 27,000,000 people living in Iraq. Before the war the mortality rate was 5.5 per 1,000 per year. So every year 148,000 people died. This is approximately 406 people per day.When you say 1,000 Iraqis die per day, do you mean in total (an additional 600 per day) or do you mean in addition to this?If you mean in addition, how long have 1,000 Iraqis been dying every day? If that is a war-long average that means you believe over a million Iraqis have died since the war began, a very high figure indeed. Perhaps this is just a figure related to recent problems?Far from keeping me wondering, your figure has left me wondering all the more!

Do you think that the numbers outside of Ramadan are significantly different than during (say by a factor of three)? The way I'm looking at it, it seems you are agreeing that 655,000 is an entirely feasable amount of deaths in Iraq.

Thanks for asking, I think I wandered around the internet looking for figures - I'm assuming most of them came from wiki but it was a year ago now. Wiki lists a high end figure of 4million for Vietnam which is the kind of figure needed to give the death rate, so I might have made an error in choosing my figures - then again the death rate in the Lancet study is hardly conservative...

Thanks for this, this is a much more comprehensive criticism than I have seen and it looks to be pretty fatal, as the pundits observe.