This topic has come up in P-A’s UNFPA thread: rather than derail that discussion further, I’ve put up a new post, and transfered comments about the Lancet study from that post to this one.
It’s been almost a year since the British medical journal The Lancet published a study (pdf version here) showing that deaths among ordinary Iraqis have increased as a result of the American invasion and occupation.
This study is important because the invasion of Iraq, after the collapse of the WMD rationalization, is currently rationalized post hoc as justified by the good our invasion has done the Iraqi people. While it’s certainly true that getting rid of Saddam is wonderful – I mean that seriously – things on the other side of the scale – such as a massive increase in needless, violent deaths – need to be considered as well, if we’re going to take the “invade to help the people” argument seriously. Unfortunately, neither the American media nor the pro-Iraq-invasion folks on either side of the partisan divide have seriously dealt with the Lancet study or its implications; instead, they’ve argued that the study is bad science. They’re wrong. From the abstract of the study: Background In March, 2003, military forces, mainly from the USA and the UK, invaded Iraq. We did a survey to compare mortality during the period of 14.6 months before the invasion with the 17.8 months after it. Methods A cluster sample survey was undertaken throughout Iraq during September, 2004. 33 clusters of 30 households each were interviewed about household composition, births, and deaths since January, 2002. In those households reporting deaths, the date, cause, and circumstances of violent deaths were recorded. We assessed the relative risk of death associated with the 2003 invasion and occupation by comparing mortality in the 17.8 months after the invasion with the 14.6-month period preceding it. Findings The risk of death was estimated to be 2.5-fold (95% CI 1.6–4.2) higher after the invasion when compared with the preinvasion period. Two-thirds of all violent deaths were reported in one cluster in the city of Falluja. If we exclude the Falluja data, the risk of death is 1.5-fold (1.1–2.3) higher after the invasion. We estimate that 98 000 more deaths than expected (8000–194 000) happened after the invasion outside of Falluja and far more if the outlier Falluja cluster is included. The major causes of death before the invasion were myocardial infarction, cerebrovascular accidents, and other chronic disorders whereas after the invasion violence was the primary cause of death. Violent deaths were widespread, reported in 15 of 33 clusters, and were mainly attributed to coalition forces. Most individuals reportedly killed by coalition forces were women and children. The risk of death from violence in the period after the invasion was 58 times higher (95% CI 8.1–419) than in the period before the war. Interpretation Making conservative assumptions, we think that about 100 000 excess deaths, or more have happened since the 2003 invasion of Iraq. Violence accounted for most of the excess deaths and air strikes from coalition forces accounted for most violent deaths. We have shown that collection of public-health information is possible even during periods of extreme violence. Our results need further verification and should lead to changes to reduce noncombatant deaths from air strikes.
Lancet used to be a reputable medical journal. But after they published that “study” that estimated that over 100,000 civilians had died in the Iraq war, only to find in the fine print that the estimated error was about 90% of that, their work is a little suspect.
[Note: In Ron’s original comment, there he said more than just the above, but the bit I’ve transfered is the bit that’s germane to this thread. -Amp]
The Lancet authors give the upper and lower bounds of their estimate (8000-194000, with 95% confidence) on page one of their article, four words after the estimate of 98,000 deaths appears, in the same print size.
And the Lancet hasn’t been discredited among anyone but right-wing extremists – the same people who frequently argue that global warming is a myth and evolution is bad science. Not a very credible bunch when it comes to discussing science.
However, when RonF says that the fine print mentioned that the margin of error was 90% of the 100,000 additional deaths, I’m pretty sure he means the text of the article, not some obscure footnote. The 100,000 number was used in press releases to draw attention to the study, but the article itself was quite clear on the methodology. Crookedtimber.org has done a very nice series of posts over the last year debunking the Lancet trashers (main essay, list).
Amp, I haven’t read a copy of Lancet in years, so I’m not going to say that it’s definitely bad. However, there were a few years in my life where I spent a lot of time reading a lot of journal articles, so I’m familiar overall with the scientific publishing process. I never saw an article published where the 95% confidence level was almost equal to the value reported. Usually it was about 5% or 10% of the value reported.
It just seems to me that any journal that would publish such a study had, at least in that instance, compromised commonly accepted editorial standards. I do wonder if in this case politics overrode science.
I meant the text of the press articles on the study, not the study itself. In fact, some press articles didn’t even mention the statistical confidence level of the study.
There were also methodological flaws in the study as well; the samples were not normally distributed among the entire population, and there was no attempt (and no way, probably) to verify how accurate the reports of the causes of death were.
The press in general does an absolutely abysmal job in reporting the results and implications of scientific studies. They generally try to turn a scientific article into breaking news, seizing on an implication and blowing it up into a headline. It’ s common for them to stretch the study to conclusions that the authors themselves (and any scientists that read the study) would never support.
RonF Writes: I meant the text of the press articles on the study, not the study itself.
That’s very, very far from being what you said.
What you said was: “Lancet used to be a reputable medical journal. But after they published that “study” that estimated that over 100,000 civilians had died in the Iraq war, only to find in the fine print that the estimated error was about 90% of that, their work is a little suspect.”
Better to acknowledge that you erred than to backtrack and claim you meant something completely different from what you said.
RonF wrote:
BTW, when Robert had that link to the Lancet article about single parents and their kids, it was easy to find several dozen pages of rightwingers who ate the results up with glee. So I guess their distrust of the Lancet is conditional upon whether or not they like the results of its studies or not.
Well I could have told Chimpy and co invading Iraq would be a disaster back in 2003. Its no surprise there are more dead Iraqis than there was before.
Send Chimpy to the Hague he deserves it the most.
Ron wrote:
I’m not sure what you mean by “normally distributed.” Could you expand on this critique?
In a subsample, interviewers were able to view death certificates 80% of the time they asked; the other 20%, there were plausible reasons the death certificates weren’t at hand to show them.
Also, I think this line of argument – which boils down to, “obviously, a huge portion of Iraqis were lying” – is pretty lame. Do you have any evidence to support your implication that Iraqis, or perhaps people reporting deaths in general, are so likely to lie that the results of surveys are thrown off.
There are literally tens of thousands of peer-reviewed studies which use survey data – and which assume that the majority of respondants aren’t lying outragiously when they talk to surveyers. Are you seriously claiming that this perfectly standard, normal technique is a “methodological flaw”?
You’re not making any sense. Why do you think a wide confidence interval proves that a study is bad science?
Three points: First of all, as Daniel Davies has poined out repeatedly, the most important fact here is that even the wide confidence interval doesn’t include the value of “zero.” That is, this study shows that there is virtually no chance that, measured in death rates, things have improved for Iraqis since we invaded.
Why would you think that’s not an important or valid finding?
Second, a wide confidence interval doesn’t mean that all the numbers within the confidence interval are equally likely. It’s more-or-less a normal distribution curve; the numbers at the tail ends of the distribution are the least likely to be correct. The most likely number is the number in the middle: 98,000.
Third, it’s important to remember that the study, because it used a very conservative design, most likely underestimated how many Iraqis have died.
Those who want to argue about the Lancet study are advised to pick said fight with Tim Lambert, or at least read his lancet/iraq category pretty thoroughly. Ron, I think you’ll find your criticisms have been addressed there.
There are two perfectly valid criticisms of the Lancet study:
(1) The study was not conducted from a representative sample. People in areas of low population density were oversampled, people in areas of high population density were undersampled. Given that the sample was unrepresentative of Iraq, it’s hard to claim that an extrapolation from it can provide an accurate picture of what happened in Iraq.
(2) Samples were not independently distributed, as clusters in pairs of provinces were reallocated to one province by chance. The methodology they used requires independence of samples as a necessary assumption for the generation of confidence intervals. Without it you can generate confidence intervals, but you can’t have any certainty that they accurately provide an assessment of how reliable your inference is.
So we have an estimate, but (1) we can’t be sure that it is representative of Iraq and (2) we can’t be sure how accurate it is. There are far better ways to criticise the War than to base your case upon this study.
In the realm of statistics, “normal” is a technical term, not a social descriptor. When a population or a sample is “normal”, that means that the distribution of the variable in question among the population follows the classic “bell curve”, and can be described by a particular set of mathematical formulae. The distribution of a variable (say, age of death) in a sample should match the distribution of a variable in the entire population the sample is taken from.
Actually, my use of the word “normal” was not precise; what I was trying to say was what Nilolai presented much better in his first point; the sample that the authors of that study used was not representative of the population of Iraq as a whole.
Jesurgislac said:
“Better to acknowledge that you erred than to backtrack and claim you meant something completely different from what you said.”
As I’ve said, I’m well aware of how a scientific article is written. The locations and structures of the descriptions of error rates, standard deviations, methodologies, etc., are fairly standardized in a scientific article and are highly standardized in a given journal; it’s not possible to hide them, and I had no concept that they were in the actual article as published in the Lancet.
BTW, when Robert had that link to the Lancet article about single parents and their kids, it was easy to find several dozen pages of rightwingers who ate the results up with glee. So I guess their distrust of the Lancet is conditional upon whether or not they like the results of its studies or not.
Alsis35, I’m not familiar with that study or any press reports on it. Maybe that report is accurate, and accurately being interpreted, and the right wing won one there. But I am under no illusions that the left has a monopoly on distorting or misinterpreting what a scientific study says into what they want to hear, so I make no judgement.
RonF, since you’ve apparently backed down from your “wide confidence intervals are always bad science” and “obviously the people surveyed were liars” claims, and folded your sole remaining claim into Nikolai’s post, I’ll just respond to Nikolai. But first…
Nikolai. before I respond, could you please provide sources for your claims? Thanks.
Amp;
The nature of the sampling procedure – which includes going to a random point on a map, and pairing provinces and redistributing clusters – is all in the paper. I’m just repeating it. The assumptions needed for statistical inference the contruction of confidence intervals and the nature of a representative sample are covered in most statistics texts. Beyond this I’m not really sure what you’re asking for. If you could provide more detail I’d be happy to clarify.
Amp, much of my answers to your points would basically boil down to what Nikolai posted, so I’ll leave that stand.
As far as the significance of such a wide confidence level goes, what it means is that there’s something very, very wrong with the sample; that there is very little mathematical basis for having confidence that the characteristics of the sample are a good match with the characteristics of the population it came from. That calls into question the validity of any conclusions drawn from the characteristics of that sample.
By disparaging the methodology, I’m not talking about the fundamental concept of a survey (which I agree is quite valid), but the method by which this particular survey was implemented.
One thing I haven’t seen comment on is the baseline for this study; how many deaths were reported for a similar time period prior to the war. What did they use to establish that? The civic records of the Saddam regime? Given how many of his subjects the Saddam regime killed and why, how reliable are those records? Did the civic authorities write up death certificates for all the people gassed in the Kurdish areas? What about the marsh Arabs that were starved to death? Were death certificates filed in a village hall in an empty village? We don’t know. And right now, we can’t.
I accept that a study like this can be valid. But I have little confidence that this study is valid.
One can also speculate (yeah, I know, one can always speculate) as to what political pressures the doctors who were writing the death certificates were under when they put down the cause of death. After all, the use of the causes of deaths during this war for political gain didn’t start with this study; it was going on even before the war during the sanctions. There were certainly people with guns running around who might have decided to make sure the records showed that the Americans were responsible for the slaughter of civilians. The study authors presumed the veracity of the records they examined. However, under the conditions of the conflicts of various kinds that have been going on in Iraq since the 1991 Gulf War and the Iran-Iraq war and the use of statistics for propaganda, I don’t think you can assume that the records used are clean data. This study cannot help but mix politics and war and science, and I don’t think the variables of war and politics were accounted for in it’s conduct.
Yes, the study didn’t include the value “zero”. But maybe it would if the sampling problems were overcome and the study was repeated.
But let’s say for the sake discussion that it’s true; more people died because of the war than would have if there had not been one. Then we get into the question of “what price freedom?”, and what constitutes “better off”. It seems that both the war’s opponents and it’s supporters find surveys and polls to support their views that the Iraqis think that they are or are not better off. What I don’t hear, outside of part of the Sunni areas, is “Bring Saddam back, we were better off with him.”
Now that last paragraph is off topic from the validity of this study. It’s on topic, though, about what the meanings and consequences of the number being non-zero are.
RonF says: As far as the significance of such a wide confidence level goes, what it means is that there’s something very, very wrong with the sample;
No, it doesn’t mean that.
“The effects of clustering on standard results of sampling theory are known, and there are standard pieces of software that can be used to adjust (widen) one’s confidence interval to take account of these design effects. The Lancet team used one of these procedures, which is why their confidence intervals are so wide (although, to repeat, not wide enough to include zero). I have not seen anybody making the clustering critique who as any argument at all from theory or data which might give a reason to believe that the normal procedures are wrong for use in this case.” link
RonF, I really suggest that before you try to rubbish the Lancet survey, you go read the Crooked Timber article I linked to, and see if you can come up with any criticisms of the Lancet study that haven’t already been rather thoroughly dealt with. (Hint: so far, you haven’t managed it.)
Well, aside from your assumption that the data can’t be right because Iraqis lie.
Completely off topic thread-hijacking post:
Iraq has a consitution approved by a majority that seems to be vote fraud proof. While there is a split in Iraq on this, consider the splits we’ve had in our own elections lately. I wonder if you’d get as strong a vote in the United States for our own Constitution. In, what, 2 or 3 months, there’s going to be a parliament elected on democratic principles that a large majority of the population will at least accept as being representative of the country as a whole, even if they don’t like the fact that their own ethnic group isn’t large enough to dominate it. And a security apparatus will be moving along in it’s development; it won’t be ready to secure the country and enforce it’s laws thoughout the entire country on it’s own, but it’ll be on it’s way towards that goal. Iraq will be the second Middle Eastern country, and the first Arab one, to have a democratically elected central government based on the rule of laws, not men. It seems to me that these, at least, are good things.
Some Sunnis will be worried that they will be shut out of being able to share in oil sales and any prosperity that develops in the country. Numerous Iraqis of all ethnic groups will be worried that religion will have excessive influence on the law, and that their civic rights will be curtailed – especially women. Sunnis and Kurds will be worried that Iran (especially) and other local regimes will work to destablize their country and turn it into some kind of theocracy or dictatorship. That will be in part on religious grounds and part because a stable and democratic Iraq, with the second largest proven oil reserves in the world and a well-educated populace can become the most powerful country in the Middle East and that constitutes a threat to these other regimes.
Some will be so worried about these things, or in favor of them, that they will faciliate the entry of foreign money, arms and fighters to opposed the new government. Others will continue to take up arms against it personally.
But it appears likely that many others will pursue democratic means to deal with these issues, by participating in the new parliament and trying to get just laws written, passed, and enforced. Will the new government address the concerns of it’s various minorities, like Lincoln tried to do with the South after the Civil War? Or will it adopt the kinds of Reconstruction policies that came out of the Federal Government after Lincoln’s assasination? That would have been a lot different if the citizens of the conquered South had had access to AK-47’s, 200-pound and 500-pound bombs, and C4; hopefully the new legislators will take that into account. Will the security forces gain enough strength and training to secure the country? Will the combination of those forces and the political machinery be able to extend the authority of the central government across the country? Will the religious parties become ascendent? Will there be Civil War, or will a democracy survive?
I have my guesses, and my hopes. But regardless of what anyone says, there’s only one thing that they or I know; we’re going to find out.
Let me ask you this: is there any possible outcome of this situation that any of you would consider worth the expenditure of the 2000 American lives and however many Iraqi lives that may have been lost? If there is, what are they?
I read through a good chunk of that link. I’ll back off some. I’ll buy that the math, based on the data, justifies concluding that the number of increased deaths were not zero or below. The extremely broad (and extreme is the right word, here) CI does mean that the identity of the right number is still pretty damn fuzzy. To say that 98,000 is the most likely number according to this study is valid. To say that 98,000 is the most likely number isn’t.
And I am not convinced that the underlying data are valid. To say that my position is “Iraqis lie” is overly simplistic. Would you dispute that Saddam’s Baathist regime maintained power though terror and slaughter? That it’s not well within that regime’s capacity to have manipulated death numbers up or down for whatever reason – say, to mislead Iran or the United States on their military capabilities? Or for someone to throw wads of death certificates for the people who died in various massacares on doctor’s desks and say, “these people all died of heart attacks; write it down. Or else.”? Neither you nor I can say, and neither can the authors of this studies.
I’m not saying that it’s proven that they did. I’m saying that it’s very possible that they did. Unless there were steps taken (or even possible) to control for that, the study is suspect.
This is about a regime that used poision gas on it’s own subjects. That’s pretty extreme. I think they were capable of doing stuff a lot less extreme, like pencil-whipping or losing death certificates.
To say that my position is “Iraqis lie” is overly simplistic.
It is, nevertheless, your position with regard to the data post-invasion.
Your argument about how we don’t know about the baseline data for the 18 months prior to the invasion is somewhat legit: you’re assuming it has to be wrong, on no particular evidence except that you’d prefer it to be wrong, but it’s true I don’t know how the study sampled the baseline data prior to invasion.
Your argument that there have got to be more deaths by violence in the data for deaths prior to invasion because Iraq was a violent country before the US invaded is really a shoddy piece of logic, though: you want to believe that the invasion and insurgency didn’t change things much, so you assume that the data showing that it did must be faulty data. That’s really bad scientific thinking.
This is what I was hoping for you to provide a source for. Where, precisely, does it say this in the study? I don’t see it.
I think you’re mistaken. As long as it’s the case that each household in Iraq had an equal chance of being interviewed, and as long as they properly accounted for the effect of their clustering methods (which they did – that’s why they have such a wide C.I.), their results are valid.
As far as I can tell – I’m just a lay reader, not an expert. (On the other hand, as I’m sure you know, plenty of actual experts seem to agree with me – see the Chronicle of Education article, for example.)
Ron F wrote:
Ron, your question is answered in the abstract of the study, which I quoted in my post. They established the baseline by surveying respondents about deaths in the household since January 2002. Your speculation that they used numbers gathered by Saddam’s government for their baseline figures is mistaken; all data used in this study were gathered by the study group.
However, I should back down a little from my previous position, and admit that if the survey instrument is returning honest answers is a legitimate concern. However, I don’t buy that it’s such a huge concern in this study that it delegitimizes it.
Furthermore, the interviewers did not ask to see death certificates until after the interview was concluded. It seems unlikely that 80% of study subjects were planning ahead for exactly this sort of lying to a random western interviewer, and had forged death certificates at hand just in case.
By the way, there are plenty of reasons to think that this study may have seriously undercounted Iraqi deaths.
1) The largest portion of deaths counted by this study took place among people living in Falluja. However, the study authors decided to exclude Falluja as an outlier, greatly reducing the number of deaths this study included.
2) If the Iraqi householders did lie, they probably did it in exactly the opposite direction than Ron F imagines. Because rations in Iraq are distributed by household size, Iraqis have an incentive to lie about the number of (living) people in their households, and some may routinely add a person or two to the total when asked “how many people live in this household?”
3) Surveys that use clusters are likely to undercount deaths caused by highly concentrated, high-casualty events – like bombings and air strikes.
4) Because the survey only counted deaths among people who were residing in the house for at least two months prior to their death, two groups of people who may have had higher than average death rates – soldiers and the homeless – were undercounted.
RonF, I assume that Amp knows what “normal” means, as he has presented statistics in previous posts. I think what he meant was that he didn’t understand what a “normal distribution” of samples would mean.
I’m not certain what a “normal distribution”would mean in this case, unless they did many, many (>20) independent stuides, and each came up with a different result, and then grouped the different results to see how well the variation in study results fit a bell curve. (e.g. 1 study says 8 000-39 000 deaths, 5 say 39 000-70 000, 10 say 70 000-101 000, 10 say 101 000 -132 000, 5 say 132 000 – 163 000, 1 says 163 000-194 000).
I also don’t recall saying any data were definitely wrong, but that there could be reasons for a regime such as Saddam’s to falsify data and that it *could* be wrong, and that the authors had no way to control for that.
Now; I’m now confused about the structure of the data. IIRC, the point was to compare pre-invasion death rates and causes with post-invasion death rates and causes to see what effect the invasion had had. Am I right so far? It’s late and I’m going to confess I’m not going to go back through the study, so how did they get the pre-invasion data? Were those not based on data gathered by the Saddam regime?
RonF,
To quote Amp from comment # 23:
Your speculation that they used numbers gathered by Saddam’s government for their baseline figures is mistaken; all data used in this study were gathered by the study group.
RonF,
You strongly implied that publishing the study was an extremely significant black mark against the Editors of the Lancet, and that publishing this study alone was sufficient to remove the Lancet not merely from the category of highly esteemed medical journals, but entirely from the category of journals that a reasonable person would trust. You used the fact that the Lancet had published this study as a reason that you could simply ignore the results of a completely different study. I find it bizarre enough that you did this without ever having read the study, but that you don’t feel any compunction about now weaseling that down to maybe suggesting that there were some uncontrolled factors in the study is pathetic.
It is particularly pathetic because it shows that you still don’t understand the basics of the methodology you feel qualified to critique. Not only have you still not read the study (“go back through the study” my eye), but you still haven’t read the abstract, even though Amp was kind enough to post it for you. Or, if you have read the abstract, even more damning, you have not understood it.
Here is the key passage of the abstract for answering your question concerning the methodology:
Now, probably you remember when the war started but, in case that has also slipped your mind, please remember that it was started in March of 2003. Therefore, as the last sentence of this section of the abstract states, the study asked about deaths within the household both before and after the start of the war. To state that again, the survey subjects were asked to list the deaths and the causes of those death for a little over a year before the war started, and for the year and a half between the start of the war and the survey date. No information was gained from the official records of the Hussein regime. Any concerns about the study being tainted by lies by the Hussein regime are simply proof that you don’t understand the methodology of the study, despite implying that you have in fact read the study, and despite the fact that the abstract of the study explains this portion of the methodology and is freely available at the top of the thread we are both posting in.
I do not feel that I am at all qualified to defend the study in any detail, as I don’t have sufficient training (and I’ve never bothered to read it). However, it is very clear that you are in no way qualified to criticise it even in the broadest terms. At the moment, it seems that your only argument as to why this study was so bad that a willingness to publish it places the Lancet beyond the pale of legitimate medical journals is that maybe a significant portion of nearly a thousand Iraqi households engaged in an elaborate concoction of lies about family deaths (no evidence of this, mind you, just maybe).
Sad.
Ampersand: If the Iraqi householders did lie, they probably did it in exactly the opposite direction than Ron F imagines.
Of course not! Iraqis love America, and love democracy – after all, it saved them from a terrible military occupation and a life of capricious violence. When you love something, you never want to hurt it. And what would hurt democracy more than to have it know about all the torturing and death-squads and things that went on before they were Liberated? You’d make Lady Liberty cry.
The honourable thing to do is to steadfastly not mention it. Sure, because 50% of Iraqis were regularly shot by Hussein personally in the middle of the night, this will make death rates before the invasion seem strangely skewed, but that’s a small price to pay… for freedom!
RonF: I also don’t recall saying any data were definitely wrong, but that there could be reasons for a regime such as Saddam’s to falsify data and that it *could* be wrong, and that the authors had no way to control for that.
Indeed. Because Saddam was Bad, all figures that may apply to before 2003 are inherently tainted. Because he was Bad. And before 2003. QED. Throw them all out.
Therefore, the entire population of Iraq can, statistically, only be considered to exist after the invasion. And we know that existing anywhere on God’s earth is better than floating in some kind of demographic limbo, right? Waaaay better. We’re talking milk-and-honey territory here. Pity they don’t realise it. Because they hate America.
(Weird love-hate thing they have going on for America over there, I guess. Must be the lack of education. It’d explain why so many of them forgot to vote yes for the constitution)
Amp;
The study doesn’t explicitly say this. It does say that they located each cluster by generating a random point on a map and going to the nearest (30, I think) households. This procedure will oversample people in areas of low population density and undersample people in areas of high population density by its very nature.
The procedure accounts for clusters of households as opposed to randomly distributing them thoughout Iraq (given some assumptions – such as clusters being hetrogenous) . It doesn’t account for making the locations of clusters dependent upon the locations of other clusters – in fact the maths they used assumes they are independent.
RonF said:
I also don’t recall saying any data were definitely wrong, but that there could be reasons for a regime such as Saddam’s to falsify data and that it *could* be wrong, and that the authors had no way to control for that.
AG said:
Indeed. Because Saddam was Bad, all figures that may apply to before 2003 are inherently tainted. Because he was Bad. And before 2003. QED. Throw them all out.
O.K., how did we get from A to B here. AG? Anyone?
It’d explain why so many of them forgot to vote yes for the constitution.
A hell of a lot more didn’t forget to vote yes. In our elections, 75 – 25 (or whatever it was, 70 – 30) is a damn landslide. Sometimes I think you’d have a hard time getting that level of approval out of the American electorate for our own Constitution. I bet the First Amendment would get hammered.
O.K., how did we get from A to B here.
You asserted that the data from before the invasion could be wrong because of Saddam Hussein’s regime. Given that the data was gathered by talking to the households surveyed, it’s hard to see what else your assertion could mean besides either “Anything anyone said about what happened in Iraq under Saddam Hussein’s regime is tainted!” or else “All Iraqis lie. You can’t trust what they tell you.”
nik,
Here is the critical passage for the aspect of the method that you are describing:
As you can see, the points were not chosen at random in the sense of picking random numbers within the range of lat and lon of each Governate, going to those points and seeing who was nearby (which would, as you say, inherently oversample rural areas). Instead, using the best available (but still very rough) list of where people lived, the researchers effectively randomly selected a person, and from that selected their community, then overlayed a fixed grid of points over the community, randomly selected one of those, and then surveyed the nearest 30 households to that point. So the locations picked were distributed in accordance with the distribution of the population.
Actually, a small amount of information from the previous regime was used in this study, as no one has managed to do a valid census since the war. However, it is unclear to me why we should expect the pre-war census to exert a huge influence here. It is likely (and certainly argued) that the Hussien regime distorted population numbers to make it seem that there were more Sunnis than there actually are, but I would be surprised if this distortion were sufficient to bias the results significantly. Certainly, this is a line along which I think one could legitimately attack the study, but it would take a stronger argument than simply pointing out the potential problem to produce one that was at all convincing.
Just as I was crafting my remarks on this thread, along comes This American Life in their audio report “What’s in a Number?” and beats me to it. The show interviews the study’s lead author, reviews the studies’ methodology (including a discussion of the author’s fanatical devotion to randomness in selecting the households to be studied), and interviews a guy who was involved in selecting targets for bombing during the Iraq war. Highlights:
– The study’s author is perhaps the world’s foremost authority on calculating deaths due to war. His work had been widely cited and praised, even by the State Department, prior to his study of Iraq. He uses the same methodology for all his studies.
– Typically his studies show that war results in greatly increase death due to simple chaos: loss of clean water and public health. The Iraq study revealed a different trend: the principle cause of increased death was violence. And the largest single source of violence leading to death was Coalition Forces.
– The report was released five days before the election. A human rights worker remarked that he found the results to be unbelievable. When interviewed, the worker acknowledged that 1) he had not actually reviewed the study when he gave his remark, 2) he had actually been part of the Pentagon’s targeting team during the war, and this may have influenced his perceptions, and 3) he found the report credible now.
– The study’s wide confidence range reflects the fact that 1) violent death is not as randomly distributed as, say, infant mortality, and 2) the authors had only one month and $24,000 to gather the data. Given more time and money, the confidence band could be narrowed. Nevertheless, it is wrong to suggest that it is just as likely that 8000 people died as 98,000; to the contrary, the study showed that there is a 90% likelihood that more than 44,000 people had died.
– To provide some perspective: Saddam is believed to have killed 230,000 people during his decades-long reign, not counting the half-million people that died as part of his pointless wars with Iran. Also, the US killed many more civilians at many points during WWII than we have killed as a result of the current war. And while expert opinion is valuable for calculating the cost of war, each of us must rely on our own judgments in evaluating the benefits.
The show also contains a segment on a GI who has the thankless job of appologizing to innocent Iraqis when the US blew up their family. Worth a listen.
Charles;
I don’t see how anything you’ve said contradicts my point. If you go to a random point on a map and pick the nearest X households you’re going to oversample households in areas of low density. This wasn’t all they did, but the other stuff they did won’t stop this from happening.
Nik,
As Charles already explained, they didn’t pick a random point on a map until after they had already picked the general location a cluster was assigned to, so your critique is mistaken.
When it came time to pick a particular location within a Governorate, the resarchers did this:
It is at this point that either a low-population or a dense-population area was chosen; and the choice was made by randomly selecting a person in the population, not by randomly selecting a place on a map. Using this method, they’d only be more likely to pick a low-density area to the extent a randomly chosen person in the Governorates would be more likely to live in a low-density area.
(By the way, even if your critique wasn’t mistaken, oversampling low-density areas would probably lead to undercounting deaths from bombings, since more people die when a dense area is bombed than when a sparsely populated area is bombed.)
Amp;
I think you are deaf to my argument. I’m fully aware that clusters were assigned to governorates and communities based on populations. But this makes no difference to the point I’m trying to make. All I need to establish is that each possible cluster did not have an equal chance of being selected for the sample not to be representative of Iraq. The sampling procedure was to:
(1) Assign clusters to Governorates (thorough SESS).
(2) Assign clusters to individual communities within Governorates (at random based upon community populations).
(3) Draw map, generate random XY, visit the nearest 30 households.
The map stage oversamples areas of low population density. That they didn’t do this until after they had already picked the general location of a cluster within Iraq (1), or within a Governorate (2), makes absolutely no difference. Low density areas of each “community” in Iraq are oversampled, with the consequence that low density areas of Iraq as a whole are oversampled. It’s true that intra-community, rather than intra-governorate or intra-Iraq density variations will determine how un-representative the sample is, but unless there are no variations in density within “communities” the sample will still not be representative of Iraq.
Nik, low density areas, as Amp points out, are less likely to have bomb casualties because people are more spread out, which means you’re giving an argument for saying that the Lancet study underestimates the number of people killed by explosions. Which gives another reason for thinking the 98,000 figure was on the low side.
Speaking for myself, I think that on the one hand the Iraq Body Count number undercounts the number of people killed by Americans, because the American and Iraqi officials are much more likely to provide information about civilians killed by insurgents than about those killed by Americans or Iraqi government forces (or death squads). But on the other hand I had trouble believing the undercount could be as great as the Lancet study implies (roughly 60,000 of the 98,000 would be violent deaths and the IBC number at the time was about 15,000). You’re giving a reason for believing the Lancet study undercounted the deaths.
To my mind, the most revealing aspect of the Lancet paper is what they said about Fallujah. Forget the overall death toll–they went into Fallujah in September 2004, before the final assault, and saw vast areas that looked as bad or worse than the neighborhood they ended up sampling. They visit 50 -something homes and find around 20 have been abandoned, with large numbers of deaths according to the neighbors. They find about 50 deaths in the 30 homes still occupied, roughly a quarter of the neighborhood. This suggests–doesn’t prove, but suggests, that Fallujah was a place where thousands and quite possibly tens of thousands of civilians died. (Certainly tens of thousands if this neighborhood was typical.) And so far as I can tell, the mainstream press did nothing, nothing at all, to try and determine
if this was the case. I haven’t seen stories where the NYT tried to interview former Fallujah residents to ask them about deaths they knew about. Maybe it’s too dangerous for reporters to do this, but I don’t buy it. If they really wanted to know they’d probably find Iraqi stringers who could do some of the interviewing for them. There were 200-300 thousand people in Fallujah and nobody bothers to try and find out if the Lancet neighborhood was typical of what people in Fallujah went through last year.
Dear Sir/Madam,
As of April 2006, how many people have died AVOIDABLY in post-invasion Occupied Iraq and Afghanistan? UK-US Mainstream Media will simply NOT report the post-invasion Iraqi and Afghan avoidable mortality (excess mortality) and under-5 infant mortality that now total 2.3 million and 1.8 million, respectively, as assessed from the latest UN data.
Below are the LATEST assessments deriving from Web-accessible, authoritative UN and UNICEF reports but which are NOT REPORTED by racist, lying, holocaust-denying Mainstream Media.
Whether a child is killed VIOLENTLY (by bombs or bullets) on NON-VIOLENTLY (through deprivation and malnourishment-exacerbated disease) the end result is the same and the culpability the same; further, the Ruler is responsible for the Ruled, notably in war-time as set out in the Geneva Conventions (this site ).
According to the LATEST, Web-accessible UN Population Division data (see: this site ) and UNICEF data (see: link ), the “under-5 infant deaths per 1,000 births” in oil-rich Iraq versus its impoverished neighbour Syria were 200 vs 170 (1953), 50 vs 44 (1990) and 125 vs 16 (SIXTEEN) (2004) i.e. infant mortality decreased enormously under the dictator Saddam Hussein but increased hugely after 1990 due to Western intervention.
According to the latest UNICEF report (2006), in 2004 the under-5 infant mortality was 122,000 in Occupied Iraq, 359,000 in Occupied Afghanistan and 1,000 in the occupying country Australia (noting that in 2004 the populations of these countries were 28.1 million, 28.6 million and 19.9 million, respectively) (link ).
About 1,300 under-5 year old infants die in Occupied Iraq and Afghanistan EVERY DAY and 0.5 million die ANNUALLY (about 90% AVOIDABLY) due to non-provision by the US-led Coalition of life-preserving requisites demanded by the Geneva Conventions (link).
Using the latest UN Population Division data (2004 revision) it is possible to calculate “avoidable mortality” (“excess mortality”), which is the difference between the ACTUAL deaths in a country in a given period and the deaths EXPECTED for a peaceful, decently-run country with the same demographics (see: link ).
The post-invasion avoidable mortality now (April 2006) totals about 0.5 million (Occupied Iraq) and 1.8 million (Occupied Afghanistan); the post-invasion under-5 infant mortality totals 0.4 million (Occupied Iraq) and 1.4 million (Occupied Afghanistan) (see MWC News:
link).
The post-1990 avoidable mortality and under-5 infant mortality in Iraq now total 2.2 million and 1.6 million, respectively, due to Coalition-imposed Sanctions, war and occupation (see MWC News:
link).
Post-invasion avoidable deaths due to the Coalition now total 2.7 million, comprising 0.5 million (Occupied Iraq), 1.8 million (Occupied Afghanistan) and 0.4 million (post-2001 opioid drug-related deaths due to Coalition restoration of the Taliban-destroyed Afghan opium industry) and the cost to the US is estimated at US$1-2 TRILLION (see: link ).
According to “Layperson’s guide to counting Iraq deaths” (see MWC News: link ), (a) “under-5 infant mortality” in “bad outcome” Third World countries is numerically about 0.7 of the “avoidable mortality” and (b) under-5 infant mortality data for Occupied Iraq and Afghanistan and other Third World countries is regularly up-dated and reported by UNICEF (see: link) – the awful truth is only a click away.
A detailed, formal complaint has been sent to the International Criminal Court charging the Coalition with war crimes in Occupied Iraq and Afghanistan. Bush, Blair, Cheney, Rumsfeld, Dr Rice (DR DEATH) and their Coalition confreres should be arraigned before the International Criminal Court (see Countercurrents, 21 December 2005: link ).
Those who IGNORE, DENY or SUPPORT man-made mass mortality of ANY PEOPLE (let alone UTTERLY INNOCENT INFANTS) are utterly beyond the Pale and “proto-Nazi”. Unlike Nazi Germany, the UK, the US and its Coalition partners such as Racist White Australia are all democracies – and their complicit, citizens variously IGNORE, DENY or SUPPORT Coalition racist violence that is killing 0.5 MILLION MUSLIM INFANTS EVERY YEAR.
Faced with the realities of the Jewish Holocaust, in 1945 ordinary Germans claimed that “We didn’t know”. By deliberately IGNORING the horrendous mass mortality in the Occupied Iraqi and Afghan Territories, Mainstream Media are providing the same excuse for Coalition citizens and are guilty of lying by omission, holocaust denial and complicity in the continuing crimes of UK-US state terrorism and UK-US “democratic imperialism” (democratic tyranny, democratic Nazism).
In relation to the continuing Coalition carnage, DECENT PEOPLE around the world are OBLIGED to (a) ESCHEW racist, holocaust-denying Mainstream Media, (b) INFORM everyone of the continuing catastrophe and (c) AVOID (where possible) any personal or business dealings with those complicit in Coalition war crimes through ethical exercise of “free market choice”. Would you buy soap made in Auschwitz?
Yours sincerely,
Dr Gideon Polya
Melbourne, Australia
website: link
Credentials: Dr Gideon Polya published some 130 works in a 4 decade scientific career, most recently a huge pharmacological reference text “Biochemical Targets of Plant Bioactive Compounds” (Taylor & Francis, New York & London, 2003), and is currently editing a completed book on global avoidable mortality (numerous articles on this matter can be found by a simple Google search for “Gideon Polya” and on his websites: link and link).
Dr. Polya –
I’d say that your contention that the media concentrate on the deaths of Americans and don’t comment much on deaths of Iraqis and Afghanis is accurate.
Regarding your charge that the Coalition is responsible for all deaths that fall under the classification of “avoidable” in these countries, you are of course aware that the actual source of the vast majority of voilent deaths of civilians there is at the hands of terrorists and insurgents who oppose the Coalition. It appears to me (correct me if I’m wrong) that it is the job of the Coalition to stop those deaths.
How would you propose that the Coalition stop those deaths? They can try to defeat their opponents in a measured fashion while training up local forces to take over (this seems to be the method currently in use). They could commit larger forces in the country and try to kill them faster with more force. However, this to me seems likely to kill more civilians.
How would you propose that the Coalition try to reduce or eliminate the death toll?
Do you propose that anyone else do anything to reduce the death toll?
Also, you seem to blame the U.N. approved sanctions on the Hussein regime for an increase in infant deaths in Iraq during the period it was in effect. If you apply the same Geneva Convention rule to pre-invasion Iraq as you do post-invasion Iraq, doesn’t that mean that Saddam Hussein was responsible for those deaths?
Additionally, consider that it was Saddam, not the U.N., who decided that vast palaces and military bases were more important than food and medicine for his subjects.
Even one dead kid is a horror. But I differ from you with who is responsible. Terrorists and insurgents are not automata who are not responsible for their actions. There is a popularly elected government in Iraq, elected by Iraqis for Iraqis. If the terrorists would accept their authority and stop trying to kill both Coalition troops and civilians who wish to accept and work with that government, the Coalition troops could leave tomorrow. Iraq could sell enough oil to start restoring their place in the world.
Do you want the Coalition troops to leave Iraq? Or do you want them to stay, but put the American administration on trial anyway? Or do you want them to leave Iraq and Afghanistan?
RonF wrote:
From the Lancet report’s abstract (quoted in the original post), emphasis added by me:
As for the Iraqi Sanctions policy, you seem to be saying that if Person A and Person B have a tug-of-war in which the rope is wrapped around the neck of Person C, strangling person C to death, person “A” should be able to get out of all responsibility for the consequences by saying “it was B’s fault! B could have acted better!”
The sanctions were a tug of war between the US and Saddam, with Iraqi civilians in the middle. Saddam is responsible for what happened, but so is the US.
IIRC, the Lancet report covers the invasion period, so regardless of whether the study is questionable or not it wouldn’t cover post invasion Iraq and Afghanistan up through April 2006, which is what Dr. Polya is talking about.
That does seem to be Dr. Polya’s position. He supports his positions by saying “the Ruler is responsible for the Ruled, notably in war-time as set out in the Geneva Conventions.” Saddam Hussein certainly ruled Iraq during the time of sanctions, and between the money he made in compliance with the sanctions and the money he made under the table in defiance of the sanctions, he could have taken care of his subjects. So it seems to me that Dr. Polya’s argument that the Coalition is solely responsible for deaths after Saddam was gone means that Saddam was responsible for them while he was in office. I’m curious to hear Dr. Polya’s comment on this. I wonder if he’ll be back?
I saw a short documentary on this study just after it was published. I believe the main things that was said was that in all propability it vastly undercounted the amonts of Iraqui dead.
The reasons were as Amprsand said in post 24 and as Nik said all along – they didn’t count in the heavily populated areas where most of the fighting took, or had taken, place.
RonF,
Dr. Polya’s piece (why, without having seen it appear anywhere else, do I feel that this is comment spam, posted unchanged in numerous blog comment sections? At least it was posted on a relevant -if very old- thread.) is not primarily concerned with violent deaths, it is concerned with avoidable deaths. The overwhelming majority of these are non-violent. Dr. Polya is faulting the occupiers for not restoring (in Iraq) or creating (in Afganistan) the necessary infrastructure to provide basic functional third world (say, Syrian) living conditions. In the case of Afganistan, this seems questionable (has Afganistan ever had the infant mortality rate of post-1950’s Syria? Certainly not in the past 2 decades). It is hard to imagine a reconstruction effort that would transform Afganistan into a functioning state in 4 years. On the other hand, it is obvious that what we are doing there isn’t anywhere close to the best we could be doing. In the case of Iraq, the failure of the post-war reconstruction has been overwhelming, and is certainly significantly the fault of the occupiers.
Indeed, it is this falure to carry out effective reconstruction that has been a major impetus for the insurgancy (which has, itself, also been responsible for the failure of the reconstruction). I remember hearing an interview (on This American Life, can’t find the specific program right now) with a group of tribal leaders in Iraq (this was in the summer immediately after the war, when the insurgancy was still in its very early stages). They had just come out of a Occupation run meeting to discuss the reconstruction that had gone badly, and they all agreed that they were giving the occupation a few more months to get its shit together and start a real reconstruction, and that if we didn’t they were joining the insurgency. We all know how things went from there.
Dr. Polya didn’t deny Saddam Hussein’s responsibility for deaths under sactions, he merely implicated all of us. It is you who are trying to argue that it was all Saddam Hussein’s fault. So Amp’s analogy to the deadly tug-of-war stands. It was certainly Saddam Hussein’s fault. It was also certainly your fault and mine.
Whether or not the Lancet story is true, B, it has no applicability to most of the time period that Dr. Polyna’s comments cover.
Dr. Polya didn’t deny Saddam Hussein’s responsibility for deaths under sactions, he merely implicated all of us.
I disagree. It seems to me that he is in fact placing total blame for the “avoidable mortality” deaths in Iraq and Afghanistan squarely and solely on the Coalition. The only mention of Saddam Hussein he makes to to favorably compare the death rate of his subjects prior to the sanctions to that after the sanctions. Otherwise all he talks about is the Coalition and how they are responsible for what’s going on, and quotes the Geneva Conventions when talking about the Coalition (but not Saddam).
In the case of Iraq, the failure of the post-war reconstruction has been overwhelming, and is certainly significantly the fault of the occupiers.
People with guns are trying to kill anyone who cooperates with the reconstruction efforts of the Coalition and the Iraqi government. It is the responsibility of the Coalition and the Iraqi government to try to stop them. But it is not the Coalition and the Iraqi government’s fault that disaffected elements of the previous government, local Islamic terrorists, and foreign terrorists are trying to take over the country. The superior moral position here does not belong to the people busying themselves purposely killing civilians to spread terror and destroying infrastructure to try to destabilize a democratically elected government.
So Amp’s analogy to the deadly tug-of-war stands.
I didn’t address Amp’s analogy. I pointed out that Dr. Polyna seems to have argued against it. I haven’t seen that addressed.
It is hard to imagine a reconstruction effort that would transform Afganistan into a functioning state in 4 years.
Afghanistan hasn’t been a functioning state for a very long time. Hell, if you want to get strict about it, no one in Kabul has told the warlords in the highlands what to do since Alexander the Great. The central government in Kabul won’t have the same kind of control over the highlands that Washington D.C. has over Anchorage and L.A. for a very long time, if ever. But they might have control equivalent to what it has had in the recent past in the next few years.
On the other hand, it is obvious that what we are doing there isn’t anywhere close to the best we could be doing.
Every milblog posting from Afghanistan and Iraq that I’ve read that has addressed the subject has said that the U.S. could get a lot more stability and cooperation in these two countries if they got a lot more serious about pushing the reconstruction efforts and pumping more money in. Of course, this is no secret to Iran, Syria, etc.; that’s exactly the reason why they are doing whatever they can to force the Coalition to take resources out of reconstruction and into security. So you’ll have no argument from me that we need to do more. Given the security situation, it’s an open question whether or not we can do more. I’d like to think we’re doing what we can, but this is, after all, a government project. So I’m sure there are screwups and massive inefficiencies. And Pres. Bush and Sec. DOD Rumsfield and the rest have to wear the jacket for that.
Tell me; what would you recommend should be done to solve this?
They had just come out of a Occupation run meeting to discuss the reconstruction that had gone badly, and they all agreed that they were giving the occupation a few more months to get its shit together and start a real reconstruction, and that if we didn’t they were joining the insurgency. We all know how things went from there.
Yeah. In fact, we do. They’ve had 3 elections with increasing participation from previously disaffected elements in each one. The government has formed and after some predictable delays (“Hey, we’re the biggest faction, we get to pick the PM. What? What do you mean ‘They get to have some say, too.’?”) the leadership is starting to get sorted out. Deaths due to combat and terrorism if anything decreasing; they’re certainly not trending up. More and more Iraqi Army units are getting equipped and trained. The most influential cleric in Iraq, Al-Sistani, has told everyone that private militias must be folded into the army, and that the central government must be in complete control of all men under arms, and you might notice that when he speaks, the Shiites tend to listen.
All is not well. People are still getting killed by terrorists. Oil sales are not what they should be, or need to be. Infrastructure projects are going more slowly than they should be. But overall, the trends are good. This is not going be something that’s over in 12 months, but I’m optimistic that the end result will not be a dictatorship or oligarchy that will slaughter it’s subjects with bullets, poison gas and environmental disasters, reward the families of dead foreign terrorists with tens of thousands of dollars for killing civilians, and watch its subjects die of disease and starvation while building 70+ palaces and military bases with gold-plated faucets and heroic statues of “The Leader”.
BTW, I have a problem with the usage by the mainstream media of the word “insurgent”. They apply it to anyone not under arms under the authority of the Coalition or the Iraqi government. I think that the word is being generally misapplied by the mainstream media and it masks who is acting in Iraq and what their moral standing is. In my opinion, someone who kills civilians as their primary target is a terrorist, not an insurgent; their object is to terrorize and destabilize civic order so that they can take over. On the other hand, if someone attacks armed men, then they are fairly described as insurgents. Some one who sets off a roadside bomb to try to kill Coalition troops in a humvee is an insurgent. Someone who sets off a car bomb near a market or somewhere else a lot of people gather, or who kidnaps or kills people trying to rebuild the country is a terrorist.
Amp, as far as your analogy goes, I won’t argue the situation you outline because I don’t accept it as a representation of the actual situation. My visualization is a picture of Saddam with both hands on both ends of the chain and his foot up against the neck of his subjects, choking more and more life out of them so that he could use it to build up his military and personal power. Meanwhile we stood around for years hitting him with pieces of paper trying to make him stop, until someone finally picked up a shovel and whacked him over the head with it.
Now, the victim lies on the ground, choked near the point of death, still trying to recover their breath and ability to defend themselves. Should we stand by while hoodlums from the neighborhood come by to steal his wallet, loot his house and kill him?
The idea of “causation” is difficult enough. The idea of “fault” seems hopelessly mired in issues of forecasting as well as cultural norms of autonomy and rights. (We’ve discussed this somewhat in the context of when a woman engages in “risky” behavior and gets raped.) After all, there is no distinction between mutual exchange, bribery and extortion “but that thinking [and law and cultural norms] makes it so.”
For example, the first Gulf War could have been avoided if only Saddam had simply not invaded Kuwait … or if people had refrained from resisting the invasion. If we equate fault with foreseeable “but for” causation, then we could assign fault on either side or both. If we equate fault with a violation of norms of rights and autonomy, we tend to find fault with Saddam for failing to respect the autonomy of Kuwait, a sovereign nation.
Arguably the first Gulf War ended when Saddam agreed to certain conditions designed to limit future aggression. Arguably Saddam reneged on those conditions. This left the Coalition Forces with the choice of 1) resuming the Gulf War, 2) abandoning the conditions or 3) finding some middle path such as sanctions. Whatever the Coalition chose, it was foreseeable that Saddam might a) reduce spending on his military and use his resources to promote public health or b) reduce spending on public health and use his resources for his military or c) reduce spending on both to use his resources elsewhere or d) maintain or increase spending on either or both and reduce spending elsewhere.
Given these dynamics, if you wanted to design a policy to maximize public welfare you would need to compare the consequences of 1) renewing war, 2) letting the world know that cease-fire agreements have no teeth, or 3) imposing economic sanctions. And you would need to do this without knowing what Saddam would do to his own population. It is far from obvious to me what the optimal course of action would be. Absent such knowledge, I don’t know how to assign fault for failure to adopt an optimal course of action.
In contrast, I can find fault with people who elect a course of action that is foreseeable less desirable than an alternative. Even if I could not fault your choice to impose sanctions, I might fault you for failing to design an optimal sanctions scheme. Even if I could not fault your choice to abandon the anti-aggression conditions, I might fault you for failing to seize any opportunity to mitigate the lessons to be learned by other tyrants. Even if I could not fault your choice to resume the war, I might fault you for conducting the war ineptly. I may not be able to know whether a well-executed sanctions regime would produce better results than a well-executed war, but I pretty well guess whether a well-executed war would produce better results than a poorly-executed war.
disaffected elements of the previous government
Hook.
local Islamic terrorists
Line.
foreign terrorists
Sinker.
I’ve got a lot more to say. Really. But it’s exhausting and I wonder if it’s really worth it.
(Not to say that RonF’s comment was worthless – it truly was not.)
RonF,
Deaths have certainly trended up since the summer after the invasion (the point at which I cited the description of tribal leaders taking a wait and see attitude). There have been many peaks and troughs, and March 2006 was a trough for US soldiers (although Jan, Feb, and Apr were all pretty typical 2+ deaths per day months), the past 4 months have seen a steady spiraling upwards of Iraqi deaths, both military and civilian (although not surpassing the peak of last August), and have seen more extreme and concerted campaigns of ethnic cleansing through terror and intimidation than have been seen at any other point in the war.
A large part of the reason we aren’t seeing many of the 4+ US deaths a day months like we were last year is that the US ground forces are being pulled back to bases (air strikes and their massive casualties are up recently), which I suppose you can view as Iraqi government troops taking over their own part in the war (along with interior ministry death squads taking up their part in the war), but I have also seen it described as US troops staying out of what is rapidly turning from an anti-occupation insurgency to an inter-ethnic civil war, as US troop involvement would only enflame things further (as when they recently destroyed a religious center while pursuing Mahdi Army forces).
I vacilate between whether we should stay or withdraw. If I thought we could do any good at all (and maybe we can) then I would favor us staying. While the people responsible for this horrible mess will not be the one’s dieing for their mistakes, we are all culpable. To the extent I think that we are only going to make things worse, I think we should go.
What I definitely think is that we should never have been there in the first place. Saddam Hussein’s Iraq was largely defanged, no threat to the outside world, and less of a threat to its own people than it had been earlier (the last big massacres were at the end of the Gulf War, and the major internal enemies (and subjects of brutal repression), the Kurds, were well protected by the no-fly-zone and the peshmerga). Sanctions had been made less brutal with the Oil for Food program (which is credited with a huge decrease in infant mortality in the late nineties/early oughts). Other than being able to vote for who will run the official death squads (and therefore who will be forced to have illegal death squads instead), I can’t think of any way in which conditions have improved since before the war, three years ago.
But are we really doing anything other than repeating cant? If you want my opinion on the war, you can find most of it (along with Juan Cole’s own creepy obsession with the Isreal lobby) over on http://www.juancole.com. My summary of violence levels comes from http://icasualties.org/oif/. Anyway, it seems both of us have merely been proved right in our own minds by the war, and if 3 years of reporting of facts on the ground haven’t gotten our positions within any reasonable range of ewach other, it is doubtful another exchange here will do so.
I do agree that Dr. Polya’s drive-by Teal Deer (“Too long- didn’t read” in fan-shprek) is using distinctly inflated methods to come up with US responsibility death counts, particularly for Afganistan.
I’m willing to go along with “maybe we shouldn’t have started this in the first place.” I’m no Bushite. And I definitely think that, having started it, we haven’t pursued the optimal strategy. I will say that I think no one anticipated the very rapid advance into Baghdad and the minimal opposition presented by the Iraqi army. I think there was a lack of planning for “O.K., we have now occupied Iraq. Now what?” And everyone – Coalition troops and government troops and the Iraqi population – is now suffering for that, and will continue to suffer regardless of whether we stay or leave.
I figure we should stay. My reasoning is that there is right now the start of a representative government in Iraq, and without our continued support it will take it a much longer time to overcome internal and external resistance. In fact, it might even fail. Such a failure will probably make what’s going on now look like the Easter parade; if we leave, or stay and fail, there won’t be any further argument on whether or not a civil war is going on there. It really will be a civil war, with a greatly increased death toll on all sides. Iran will either invade, or else pour huge amounts of money and munitions into a radical Shiite faction that will be a lot worse than what’s in there now. Then Kurdistan will revolt, and Turkey will go on alert because they have a fractious Kurdish population and they don’t want to see an independent Kurdistan right next to their Kurdish population, etc., etc.
It might fail even with our continued presence, but it’s got a much better chance. I think we are morally obligated to finish what we’ve started.
Other than being able to vote for who will run the official death squads (and therefore who will be forced to have illegal death squads instead),
That’s a very pessimistic viewpoint of the value and functions of the Iraq government that I don’t share.
So you don’t believe that the SCIRI controlled interior ministry is carrying out illegal murders along ethnic and sectarian lines?
Certainly, the government is doing other things as well, and it is conceivable that the government might play a role in negotiating a peace with the various parties to the civil war (although, since at least one party to the civil war has been carrying out a fairly succesful campaign of murder against the families of everyone who participates in the government, it currently seems unlikely that they are ready to negotiate).
I will be very happy if I am proven wrong, but…
So you don’t believe that the SCIRI controlled interior ministry is carrying out illegal murders along ethnic and sectarian lines?
I’ll respond to that if you can quote where I said that, or even implied it.
Jake, you seem to have a problem with my terminology. Different people are involved in this conflict for different reasons, and are using different methods. To lump them all under “insurgents” when many of them are not Iraqi citizens and/or are deliberately targeting civilians and civilian infrastructure to gain their ends is in my opinion misleading.
I think the government is is playing a role in negotiating a peace with the various parties to the civil war. I also think that they are playing a role, and an increasingly effective one, in dealing with those who don’t want to negotiate. The interior ministry has reportedly got some people in it who are using their authority as a cover to accomplish their own ends by killing their opposition – the death squads. But that activity is being fought by other parts of the government and I don’t anticipate that this will be a long-term problem. It’s too reminiscent of what went on under Saddam.
No, RonF, it is not a problem with your terminology. I have a problem with the fact that you parroted the bushcoadmin line & left out a significant portion of the people involved in the insurgency. It is a lot more than just disaffected elements of the previous government, local Islamic terrorists and foreign terrorists. In fact, most credible people believe that the number of “foreign terrorists” is insignificantly small. I’ve never seen it suggested that “local Islamic terrorists” make up a large portion of the insurgency (unless you mean that all the Iraqis [save those who are “just disaffected elements of the previous government”] who are part of the insurgency are terrorists – which I don’t buy).
You minimize the signifigance of the insurgency and the civil war by reducing those opposed to the US & the current Iraqi government to those three groups.
Also, insurgencies routinely use terror tactics against civilian populations, so insisting that people who use terror tactics against civilians aren’t insurgents, they are terrorists makes little sense, except as a rhetorical tactic. It both allows people to claim that fighting the insurgency is part of the war on terror (see, we’re fighting terrorists!) and to minimize the significance of the insurgency (we, correctly, don’t associate terrorism with 800 deaths a month, month after month, for three years).
Also, I implied that the Iraqi government runs death squads, and that if a different party had been elected, they would have been running death squads too. You replied that that was a pessimistic view you didn’t share. Apparently the part about running death squads wasn’t what you disagreed with. I suppose it was merely my rhetorical overstatement you disagreed with.
But then, as you say, you believe it is just a few bad apples who run death squads, and not a coherent policy of the interior ministry.
As I keep saying, I hope you’re right, but I see no evidence for it.
Well, Charles, my point is that anyone who uses terror tactics against a civilian population isn’t an insurgency; it’s a terrorist organization. In my opinion, the term “insurgency” is misapplied to such a group in the first place.
I’d say that there a definite faction of the Interior Ministry that is running these death squads. I think it’s something that the Iraqi Government is fighting against, and that it’s lessening. But I haven’t see a lot of hard and fast information either way about how many people are involved in these as either perpetrators or as victims. I agree they’re there, but it’s not something I’ve seen good data on.
“local Islamic terrorists” would be people who, in the name of Allah (as opposed to Saddam, the Baathist party, or just anger at Americans) commit terrorist acts. As far as “foreign terrorists” go, there used to be a number of them, but the reports I’ve seen say that the locals are getting sick of them and are reporting/betraying them to the Coalition forces at increasing rates. The most current reports are that Al-Queda’s activities in Iraq have been greatly diminished and are becoming a non-factor there. The foreign issue now is becoming arming and financing of the locals by Iran and (to a much lesser degree) Syria.
In reviewing this, though, you make a good point. There are people that are motivated by nationalistic (or sub-national/tribal) anger towards the Coalition occupation (oh, yeah, it’s an occupation) who are not acting on the basis of the desire to put a theocracy in place. But it seems to me that most of the people doing the killing are aligned with either an Islamic group or an ex-Baathist group. I don’t think there’s too many folks out there that are something else. I think the tribal groups are starting to figure out that fighting on the same side as the more idealistic groups and the remnants of the old regime is a losing proposition.
As far as who’s a terrorist or not, I fall back to my definition. Anyone doing stuff like blowing a bomb up in the middle of a cafe or next to a bus is a terrorist. Their ultimate aim is immaterial, as is their party affiliation, their tribal membership or their national origin.
This blog is historically interesting. By now, there’s an updated report in Lancet which is consistent with the first report. The current critics are not noticing the degree to which these reinforce one another. The 655K is quite consistent with the greater period of time and the rapidly rising death rate. It’s still possible to note the wide error range at 95% confidence, but it is now much harder to challenge methodology.