Asymmetrical Information

The thing that you seem to be missing is that there are three parties to the discussion: people sitting around after the fact thinking that it seems impossible that the interviews could have been completed that fast; people sitting around after the fact thinking that it seems perfectly reasonable that the interviews were completed that fast; and the original interviewers and researchers who worked with them who say that the interviews were completed that fast. The first two parties may be honestly right or wrong -- the third party is either lying or truthful. Usually, if you're going to accuse the people who performed a piece of academic research of falsifying their results, you try to base that accusation on some higher standard of evidence than a belief that it would be really hard to do what they say they did.

The position being argued seems to be that "I doubt that the interviews could have been completed by the number of interviewers available in the stated time; I therefore believe that the results were falsified to some extent and the study is unreliable." And I haven't seen anything like the evidence necessary to support that claim.

Further, how is Davies' reasonable claim that one would expect someone living in a household to have no difficulty remembering who had lived there over the prior four years falsified by the statement that Arab households are larger and more fluid than middle-class British households? While the latter seems likely, they'd have to get awfully big and fluid before it became a non-trivial task to remember who had lived in the household at any given time.

Yes, LizardBreath, let me make it explicit: people who think that that the times are wrong are implying that someone, probably the interviewers, made shit up. This is not exactly unheard of among field workers who are NOT working in a dangerous war zone and do NOT have any political interest in the number produced, which is why it's common to audit a pretty high percentage of the interviewers. Daniel Davies is over at Deltoid handwaving . . . "What, so now all the dark implications are down to 'did they get informed consent'?" When of course the point is, that if they didn't get informed consent, which is a mandatory first step to conducting an interview and usually teh most scripted part of the whole thing, then what other procedures didn't they follow?

I don't think you can shift the burden of proof quite so neatly. In your formulation, the study is innocent until proven guilty, while any questions about it must be proven beyond a reasonable doubt. That's just silly. No one has to accept the figure just because someone wrote it down somewhere, particuarly if the implications (like how do they do the interviews so fast with a 98% response rate? Not a single householder was in the middle of dinner or on their way to the bathroom?)are odd. The more Les Roberts talks, the less confident I get in his results; he doesn't seem to have the faintest clue how the interviews were conducted. He keeps changing the number of interviewers, the number of houses (which has suddenly and without explanation dropped to 30 in some of the arguments) or who asked what. And somehow now all the interviews were done in 3 hours? At 20 households per team, that's 6 houses an hour, ten minutes a house including walking, peeing, and informed consent.

Take me: in the last three years, I've lived in four places. When did I move into my current digs? January? Later? Was my mom living with me when she was spending more time than not sleeping on my floor? What's the date? What's your cousin's birthday? If you have an average household size of seven people, instead of an average household size of 2.[mumble], then the law of averages dictates there will be a lot more movements. Then add the tendency for extended families to have people randomly stayng with them--did grandma stay with us in the winter of 2002 or the winter of 2003? It's harder to remember than you think; I just had to count backwards to figure out what year I broke up with my ex boyfriend in, and believe me, I remember that event all too vividly.

As someone who's lived in a "fluid household", I'd have to say it depends on the questions. If you want me to "name all of the people that lived in your house over the past four years", that's easy. If you want to know "When did X move in" and "When did Y move out", that's a lot harder. I could figure it out, but I'd have to think about it, and probably talk to other people in the house.

Right. The thing is, though, you don't, and the other people attacking the survey don't, know jack about the circumstances in Iraq that would affect the timing of interviews. You don't, for example, know the typical spacing of houses -- probably different in most Iraqi neighborhoods than in most American neighborhoods.

As a critique, this comes down to "Seems really unlikely to me from where I sit." Which is, you know, interesting, and there's nothing much that could be done to convince you to change that position, but there's not all that much of a basis for it.

The more Les Roberts talks, the less confident I get in his results; he doesn't seem to have the faintest clue how the interviews were conducted. He keeps changing the number of interviewers, the number of houses (which has suddenly and without explanation dropped to 30 in some of the arguments) or who asked what.

This, I'm really not seeing. I see a lot of accusations of inconsistency, but I've been following the discussions on this with reasonable attention, and the accusations don't seem to be supported by pairs of inconsistent quotes. For example, the '30 houses' bit you bring up, if I've got it sourced correctly, comes from a discussion of the 2004 study in which the clusters did consist of 30 households.

If you want to know "When did X move in" and "When did Y move out", that's a lot harder.

Probably gets easier if you're working in terms of "Did they move in before or after the war started," and "Were they living here when they died", don't you think? Tying it to notable events rather than dates is going to clarify the situation a lot.

(Drat -- in my comment at 2:23, the next to last paragraph should be italicized to indicate quotation. Should have previewed.)

Yes, there are three groups. Two have no direct knowledge of how the information was gathered and one does. Let's try to take the politics out of this and look at a different, but not entirely dissimilar, situation: Assume that Feelings Mutual has just published its annual financial statements. Two groups of outsiders debate whether or not it's a good idea to invest in Feelings Mutual. Among the questions these two groups debate is how reliable are the published financial statements. "How is it possible for ROE to be 120%? That's more than 3 times the highest ROE for any other company in the industry," wonders one group. "What? Are you accusing Feelings Mutual of fraud? You can't do that without proof," says the second. "You can trust us. We we wouldn't do anything wrong," says the third group -- those working for Feelings Mutual.

Our experience throughout human history is that people have an unfortunate tendency to be inaccurate in ways that tend to favor outcomes they prefer. Whether intentional or not, it's likely that Feelings Mutual's financial statements cannot be relied upon unless an independent auditor has reviewed them (and performed the necessary audit tests). If we are going to insist on such safe guards in the financial markets, why not require something similar in the social sciences? Oh, wait, we do! As Jane pointed out, a standard feature of a well designed study is an audit to ensure the interviews actually took place. That the survey was conducted in a war zone make it more critical, not less, that such an audit be performed. If it was not, why should we place any more faith in the study than we would in Feelings Mutual's unaudited financial statements?

The obvious - really obvious - problem with the Lancet study is right there in plain sight.

Death certificates, and their assumptions of how many get issued.

Ninety-two percent. More than nine out of ten households had one on hand.

The Lancet study then goes on to assume that about ninety percent of deaths in Iraq were never reported or recorded by the government..

This right here kills their whole conclusion. If the deaths were real, and reported, and recorded by the government, then the easy way to find out the number of civilian deaths in Iraq would be to see how many were recorded. When people do that, the get numbers that agree with other sources, and are a small fraction of the number in the Lancet study.

But... according to the Lancet authors, only about one in ten civilian deaths were ever officially recognized. So out of those households, only about one in ten should have had a real, honest, official death certificate on hand.

They can have one or the other. Not both at the same time.

But... on the other hand, if someone either fudged the list of households to interview (preselecting for deaths from official rolls), or arranged for faked death certificates, or lied about seeing the death certificates altogether, then they'd certainly get the results they were looking for.

The difference is that an ROE more than three times as high as as any other company in the same industry is fairly good evidence that chicanery is going on. The fact that someone sitting in the US with no experience of doing surveys under Iraqi conditions thinks that it's really unlikely that a two-person team of interviewers could have given a five-question survey to twenty households in a day is not similarly strong evidence.

The fact that someone sitting in the US with no experience of doing surveys under Iraqi conditions thinks that it's really unlikely that a two-person team of interviewers could have given a five-question survey to twenty households in a day is not similarly strong evidence.

Then that follows for those defending the survey as well. Given that all other evidence points to the survey being incorrect, and the unlikely possibility that the government has lost track of 90% of those who have died, the onus is on Roberts et al. Throw in that they can produce no evidence to corroborate their findings they must be viewed with skepticism. All tests (you know, falsification in science being important) of their results have raised alarm bells. Yes, if no evidence other than a persons word is what we are relying on then we do get to ask whether they are possibly making things up and ask them to give us some kind of corroborating proof.

We should also allow that while a survey in Iraq is unique, lessons from other settings do bear on how we evaluate the survey. In addition, many who have done surveys in a variety of settings have raised alarm bells. You are asking us to believe that no other setting is relevant, that Iraq was uniquely amenable to such a survey relative to almost every other survey conducted. I find that unbelievable, but if they are going to make the claim then they should document it or audit it in some fashion.

Finally, it is rich that somehow asking such questions with what they imply about the surveys veracity is improper without proof, when the survey itself makes some pretty extreme claims, is defended with rationales which imply some pretty extreme claims and for which Roberts et al, Tim Lambert, etc. provide no proof at all. The burden of proof lies as much with them. Recordkeeping is that far off? Recorded deaths are that far off? Etc. It is circular. It amounts to "Our survey is reasonable, despite all the contrary evidence because of the flaws in other sources of data, and we know that data is as flawed as we say it is because of our survey."

That isn't scientific and no amount of hand waving will change that whether the survey is bogus or not.

Right. The thing is, though, you don't, and the other people attacking the survey don't, know jack about the circumstances in Iraq that would affect the timing of interviews. You don't, for example, know the typical spacing of houses -- probably different in most Iraqi neighborhoods than in most American neighborhoods.

Pfffft. Here's a mental exercise for you: Contemplate going to an apartment complex in order to ask a five-question survey about car ownership and driving habits. First, of course, is the informed consent portion. Then, one of your questions is about tickets, and requests permission to inspect the ticket stub if possible. Another requests a viewing of any available drivers' licenses in the household, and the others relate to the dates (exact, if possible) that various records were filed or vehicles in the household were obtained/sold.

Unless you heavily cherry-pick the interview parameters, right down to the neighborhood and which households in the complex you are planning to visit, you're going to be hard pressed to fulfill, on average, six household interviews per hour.

If you actually succeed and obtain enough data to fulfill the survey parameters, try out your walking skills on a nearby lake.

I cannot agree with you more. I'm living in a developing country (Samoa) which is more or less peaceful (unless you go to the bars on Friday or Saturday night) and it can take forever to accomplish the most trivial of tasks. I'm a Peace Corps volunteer, and during training I tried to figure out the relationships of the different people in my host family. I spend 2 months thinking that a cousin was a brother. Something that may be unique to Samoa, but when someone has a child here they are raised communally, so the child may not actually know who its parents are.

No shit! I'm an RPCV Samoa, from 93-94; I taught at Vaipouli.

Having lived in the region for a couple of years, it strikes me as dubious that this type of information could be obtained in the time claimed. There are typically social traditions adhered to when interacting with strangers within one's home, or even at one's door, which would preclude such rapid interviews. Nope, a war wouldn't likely speed things up, either.

I didn't read the post yet. I have know idea what it's about. But if Daniel Davies is right, I'll kiss my own ass.

Well, I won't go so far as to promise self posterior-puckering, but I will note that using the person who sits next to you in an office in the U.K. as a surrogate for knocking on a stranger's door in Iraq is a very stupid thing to do.

Jane, when you misunderstand someone, it is wrong to say that he "contradicted" himself. And far from reducing confidence in the person you misunderstood, it reduces it in you.

Roberts supervised the 2004 survey -- he says that they did a 30 house cluster in about 3 hours. Which means that the 2006 survey's 40 house clusters would take roughly four hours. Why should I ever take someone seriously who can't understand this?

Let's change the domain here to explain why so many people think that this survey is questionable.

Look at doorknocking for political campaigns. How long does it take per door knocked, in an apartment building? How long does it take per person talked to in an apartment building? Here you are talking to people and maybe asking if they'd like to talk to the candidate. Many responses will be "no, your candidate is horrible because of X".

5-10 minutes per person met would be a very good average. This isn't asking for a census of the household or to see certificates, just here's my short spiel and talk to the candidate for a 2 minute pitch and 3 questions you may have. You're going faster in a warzone, where people are suspicious and in a culture where fairly elaborate hospitality is required.

Arab culture is renowned for hospitality rituals as well as for how long it takes to get to the point of a meeting. Never mind asking, for some random scientists, the composition of your family and any deaths. 1 or maybe 2 houses per hour would be credible.

But these researchers are able to make penetrating questions faster than you could do this in the West (with no transit time, bathroom breaks, or any other issues). Hmmmm.

None of this is proof. We're just asking for more data and some stronger proof of your extraordinary claims. For this, our integrity is being impugned and proof of misconduct is being demanded. This isn't what a well conducted study would do - a well conducted study would swamp us with data. Here's exactly how we were able to do this, here's the audits, we were surprised about the responses too and thanks to x and y they were enabled...

But sure, keep defending this study. Some great scientists we've got here!

So the productivity of a team of Baghdad doctors seems rather high to a London journalist? Yes, I suppose it would.

"Need I point out that if Davies is right, and Burnham et. al. are right, then we should be seeing massive floods of refugees?"

Wow, I don't think I've ever seen someone's credibility evaporate so quickly. Unless you're using some super-secret personal definition of 'massive' which means 'more than millions', I don't know how you can ask that question and expect to be taken seriously.

But what I wanted to blog about is a somewhat related phenomenon, which is the systematic human tendency to understimate how long things take. This was driven home to me rather poignantly when I went up against Spencer Ackerman in Blogging Chefs, and tried to estimate just how much I could do in 90 minutes. Then I tested how long it actually took to, say, cook macaroni and cheese.

THere were a few things I overestimated a little: it only takes four minutes, not five, to slice two boxes of grape tomatoes. (Well . . . I extrapolated from half a box. But I wasn't getting noticeably faster at it.) But mostly I turned out to have underestimated various steps by a factor of about three. The reason is that I imagined only the highlights: operating the food processor, stirring the sauce, etc. I forgot about all the little things that slow you down: opening the wrapper, finding the spoon, having the cheese get caught in the blade, etc. Even though I'd given myself a generous fudge factor in estimates, I was still off by a factor of two or more.

I think you should put this on the Economist blog - it's about of the same quality as the rubbish they're printing nowadays.

I also think that if you're going to insult people, you ought to have the guts to do it to their face.

Other than that, since your only actual argument depends on an assumption that there haven't been hundreds of thousands of refugees fleeing Iraq, I think all I have to do is point out that there have.

Hey, double d; I was able to determine within five minutes last night that my bartender likes Maker's Mark when he drinks bourbon, Laphroaig for scotch, Guinness for beer, Bombay for gin, and Tio's Texas Handmade for vodka, so what was done in Bagdhad is eminently credible, right?

Hey, double d; I was able to determine within five minutes last night that my bartender likes Maker's Mark when he drinks bourbon, Laphroaig for scotch, Guinness for beer, Bombay for gin, and Tio's Texas Handmade for vodka, so what was done in Bagdhad is eminently credible, right?

"I also think that if you're going to insult people, you ought to have the guts to do it to their face."

Because of course Daniel would beat Megan up if she insulted him to his face.

Daniel, blogging critiques of your Lancet commentary is hardly the same thing as whispering about you behind your back in prep. I was fully confident that you would be able to find this post, and lo, my faith in you has been fully justified. But if you want to fly to Washington DC, we can step outside and settle this like gentlemen.

As for your point about the refugees, you and the commenter above you seem not to have understood what was, I thought, a very simple point. If there have been so many killed, there should be even more refugees than a few hundred thousand. If there are even more refugees, they should be showing up as household figures that need to be (time consumingly) accounted for.

And Tim, I generally respect your work on this stuff, but I'm afraid I don't find the most parsimonious explanation for *every* discrepancy to be that the journalist is so thick that they are unable to ask coherent questions, or parse single-syllable answers. Your story is that Nature asked him about the 2006 survey, but he, for some reason, responded about the 2004 one.

Something he didn't make clear to the writer until confronted with a denial by one of the field workers.

The article is about the 2006 survey, not the 2004 survey, and the narrative of the questioning seems to make it clear that the focus was entirely on 2006. There was absolutely no reason for Roberts to think he was being asked about 2004, but not 2006, particularly since this has been one of the biggest points of contention over the current figures. Either Les Roberts was mistaken (to put it charitably) about the methods by which the surveys were collected; or he is a complete moron who does not understand what has been going on for the last few months, and simply happily babbles about whatever enters his head when he is asked questions about his work. Neither exactly fills one with breathtaking confidence in the results.

Jim wrote, "Need I point out that if Davies is right, and Burnham et. al. are right, then we should be seeing massive floods of refugees?" Wow, I don't think I've ever seen someone's credibility evaporate so quickly.

Yep.

Jane Galt wrote, If there have been so many killed, there should be even more refugees than a few hundred thousand.

Most of the figures I see cited in the news these days are of the order of millions having fled Iraq, and millions displaced within Iraq.

then we should be seeing massive floods of refugees?

Jane, I get the impression that you're very selective as to what you choose to see.

The Iraqis middle class is running for their lives, anyone who can get out is. That is all anyone needs to know about the situation in Iraq.

If Roberts was asked a question about whether there was enough time to do the interviews in the 2006 survey, it is completely relevant to talk about the 2004 survey because the questions were the same. If in 2004 you can survey 30 houses in 3 hours, in the 2006 survey, you can do 40 in roughly 4 hours. I don't have Roberts' reply to Nature, but I do have his reply to the BBC when they asked the same question:

During my DRC surveys I planned on interviewers each interviewing 20 houses a day, and taking about 7 minutes per house. Most of the time in a day was spent on travel and finding the randomly selected household. In Iraq in 2004, the surveys took about twice as long and it usually took a two-person team about three hours to interview a 30-house cluster. I remember one rural cluster that took about six hours and we got back after dark. Nonetheless, Dr. Hicks' concerns are not valid as many days one team interviewed two clusters in 2004.

You're the one happily babbling about whatever enters their head, not Roberts.

And as for refugees, the UN estimates that about two million people have fled Iraq. How come you are unaware of this?

WANKER!

Jane, I'm sure your macaroni is tasty. But please write about the topics about which you know (like macaroni), rather than writing the idiotic post above and your similarly stupid defenses of it.

How long would it take to get an estimate of the number of Iraqi refugees? Wow, my estimate of under 2 minutes was correct. Google is fantastic. The answer is that there are as many as 2 million currently outside Iraq, and another 1.7 million displaced within Iraq. There will be an estimated 2.7 million Iraqi refugees outside the country by the end of 2007. More than 40,000 Iraqis flee per day. Pretty massive.

http://www.realcities.com/mld/krwashington/news/columnists/warren_p_strobel/16646194.htm

And no, you needn't point out that IF Davies and Burnham are right then we should see massive flows of refugees. Just post more recipes. Macaroni is tasty.

More macaroni blogging. Less Iraq blogging. Leave the Iraq stuff to people who know things, like Spackerman.

Tee-hee!

I nominate Kevin's 3:39am commet for Most Snarky, and tomboy's first sentence at 1:18 ("Jane, I'm sure your macaroni is tasty.") for Smuttiest Sounding When Taken Out of Context. I do have to disagree with Jim about:
"Wow, I don't think I've ever seen someone's credibility evaporate so quickly."
Megan lasted, like, eight whole paragraphs before that last credibility-busting one. I've seen plenty of people destroy any credibility they might have had within their first sentence.
(More to the point of what she was trying to say: well, given that there are apparently-credible estimates of perhaps 2 million external and 1.7 million internal refugees, then that in itself suggests that Iraq is having a violent breakdown of substantial magnitude. If, as LizardBreath points out, folks are dealing with 'where they here before/after the war started, and did they die here', remembering details for folks who left or joined the household as war-related refugees will probably be pretty easy. (Additionally, entire households may be fleeing.))

Take me: in the last three years, I've lived in four places.

Shorter Jane: all arguments begin with me-me-me-me-me.

Unless you're using some super-secret personal definition of 'massive' which means 'more than millions', I don't know how you can ask that question and expect to be taken seriously.

Remember, in Jane-land, 'massive' has a different meaning when you're talking about death tolls and refugees. But since Iraqi refugees aren't engaging in internet cook-offs, that sorta passes her by.

If in 2004 you can survey 30 houses in 3 hours, in the 2006 survey, you can do 40 in roughly 4 hours.

Here, according to the study itself, is what was supposedly accomplished in that amount of time:

In every cluster, the numbers of
households where no-one was at home or where
participation was refused were recorded. In every cluster, queries were made about any household that had been present during the survey period that had ceased to exist because all members had died or left. Empty houses or those that refused to participate were passed over until 40 households had been interviewed in all locations. The survey purpose was explained to the head of household or spouse, and oral consent was obtained. Participants were assured that no unique identifiers would be gathered. No incentives were provided. The survey listed current household members by sex, and asked who had lived in this household on January 1, 2002.

The interviewers then asked about births, deaths, and in-migration and out-migration, and confirmed that the reported inflow and exit of residents explained the differences in composition between the start and end of the recall period. Separation of combatant from non-combatant deaths during interviews was not attempted, since such information would probably be concealed by household informants, and to ask about this could put interviewers at risk. Deaths were recorded only if the decedent had lived in the household continuously for 3 months before the event. Additional probing was done to establish the cause and circumstances of deaths to the extent feasible, taking into account family sensitivities. At the conclusion of household interviews where deaths were reported, surveyors requested to see a copy of any death certificate and its presence was recorded. Where differences between the household account and the cause mentioned on the certificate existed, further discussions were sometimes needed to establish the primary cause of death.So even assuming that the teams were divided in two (Roberts' newfound claim), the teams would have to accomplish all of the following in 12 minutes per household:

1. Locate and travel to the home.
2. Knock on the door.
3. Pass over homes where no one answered, and record that no answer was received.
4. Find another home.
5. Knock on the door again.
6. Await an answer.
7. Introduce themselves.
8. Explain purpose of survey.
9. Get oral consent; if no consent given, start over at step 4 (also recording information about lack of consent).
10. Ask whether any nearby households had "ceased to exist" entirely during the survey period.
11. Assure participants of anonymity.
12. List members by sex.
13. Ask about occupancy since 2002.
14. Ask about births.
15. Ask about deaths.
16. Ask about in-migration and out-migration.
17. Confirm that explanation is consistent with any differences in household composition.
18. Find out if any deaths involved persons who had lived there continuously for three months.
19. Engage in "additional probing to establish the cause and circumstances of deaths."
20. Request copy of death certificate and record pertinent information.
21. Further discussions if necessary to inquire into cause-of-death listed on certificate.

If you can believe this, I've got a great story about a team of people who can all run a mile in one minute. And how dare you insinuate that my story is fishy just because you can't imagine running that fast yourself!

One more try:

If in 2004 you can survey 30 houses in 3 hours, in the 2006 survey, you can do 40 in roughly 4 hours.

Here, according to the study itself, is what was supposedly accomplished in that amount of time:

In every cluster, the numbers of households where no-one was at home or where participation was refused were recorded. In every cluster, queries were made about any household that had been present during the survey period that had ceased to exist because all members had died or left. Empty houses or those that refused to participate were passed over until 40 households had been interviewed in all locations. The survey purpose was explained to the head of household or spouse, and oral consent was obtained. Participants were assured that no unique identifiers would be gathered. No incentives were provided. The survey listed current household members by sex, and asked who had lived in this household on January 1, 2002.

The interviewers then asked about births, deaths, and in-migration and out-migration, and confirmed that the reported inflow and exit of residents explained the differences in composition between the start and end of the recall period. Separation of combatant from non-combatant deaths during interviews was not attempted, since such information would probably be concealed by household informants, and to ask about this could put interviewers at risk. Deaths were recorded only if the decedent had lived in the household continuously for 3 months before the event. Additional probing was done to establish the cause and circumstances of deaths to the extent feasible, taking into account family sensitivities. At the conclusion of household interviews where deaths were reported, surveyors requested to see a copy of any death certificate and its presence was recorded. Where differences between the household account and the cause mentioned on the certificate existed, further discussions were sometimes needed to establish the primary cause of death.

1. Locate and travel to the home.
2. Knock on the door.
3. Pass over homes where no one answered, and record that no answer was received.
4. Find another home.
5. Knock on the door again.
6. Await an answer.
7. Introduce themselves.
8. Explain purpose of survey.
9. Get oral consent; if no consent given, start over at step 4 (also recording information about lack of consent).
10. Ask whether any nearby households had "ceased to exist" entirely during the survey period.
11. Assure participants of anonymity.
12. List members by sex.
13. Ask about occupancy since 2002.
14. Ask about births.
15. Ask about deaths.
16. Ask about in-migration and out-migration.
17. Confirm that explanation is consistent with any differences in household composition.
18. Find out if any deaths involved persons who had lived there continuously for three months.
19. Engage in "additional probing to establish the cause and circumstances of deaths."
20. Request copy of death certificate and record pertinent information.
21. Further discussions if necessary to inquire into cause-of-death listed on certificate.

If you can believe this, I've got a great story about a team of people who can all run a mile in one minute. And how dare you insinuate that my story is fishy just because you can't imagine running that fast yourself!

As to interviewing in a war zone, I've done it before (in the former Yugoslavia) and you can get a lot of information, particularly if you know the language, customs, and culture (which I did and do--and I presume the Iraqi interviewers did as well) in a very short period of time.
And Lizardbreath's comment is spot on. After having interviewed thousands of individuals, I became an expert on the exact dates/times of troop movements and when a particular group's army swept into a particular area. So much so that I would correct those being interviewed (who had actually been there at the time) regarding dates and times of events.
Thus, if you asked "was your husband at home on the night of April 29th", for example, the interviewee would most likely have no clue. So you didn't ask the question in that manner. You asked "was your husband at home the night the Serb tanks rolled into the village [which happened to be April 29th]" the interviewee would immediately respond "yes" or "no".
As to the whole debate taking place, why doesn't Jane, along with her libertarian friends, get some money together and do their own survey?

> As to the whole debate taking place, why doesn't Jane, along with her libertarian friends, get some money together and do their own survey?

Pish posh.

Let the invisible hand take care of it.

Jane,

Just wondering.

What sort of level of violence would be necessary to make 7% of Americans quit their jobs, pull their kids from school, abandon their lives and leave the country (moving to Canada and Mexico)?

Now, add to that the level of additional violence necessary to get 6% of folks from New York, Vermont, Massachusetts, and Connecticut to quit their jobs, pull their kids from school, abandon their lives, and move to Alabama, Georgia, Arkansas and South Carolina. And vice versa with folks from Alabama, rural Georgia, Oklahoma, and Mississippi moving to Los Angeles, San Francisco, Boston, New York, etc.

I'm sure that this comparison isn't really all that apt. But it could make one think about the cosmopolitan types who lived in Baghdad having to flee their homes to live in what they consider bumfuck rural areas. And vice versa, with REAL Iraqis, religious god-fearing country folks seeking refuge in what they previously considered to be dens of immorality, violence, and obscenity: Baghdad, Mosul, and Basra. What level of violence could cause the greatest population shift in the middle east since the founding of the state of Israel, an event which will have repercussions for centuries to come?

Now, I'm no expert in doing war-time demographic surveys. I'm told that the people at Lancet are, with other surveys that the have done being well received. But I'm pretty sure that those who try to minimize the level of violence that has occurred in Iraq earn the term "WANKER", especially while demonstrating complete ignorance about the plight of the Iraqi people and the size of the refugee flows.

Please remember that estimates of the current number of Iraqi refugees includes a large number that fled in the Saddam era.

"Since the war began four years ago this month, up to 2-million Iraqis have fled their violent, chaotic country."
-St. Petersburg [FL] Times, March 4, 2007

I have no reason to doubt the UN numbers on refugees. I have substantial reason to doubt the Lancet numbers on casualties.

The sectarian violence in Iraq has been hellish, but in the end, 700,000 casualties is an incredibly large number. Healthy skepticism should be anyone's first response.

"What sort of level of violence would be necessary to make 7% of Americans quit their jobs, pull their kids from school, abandon their lives and leave the country (moving to Canada and Mexico)?" [plus 6% moving within the country.]

Blow up the school, the water plant, and the power plant (which will shut most jobs down), and I'd be extremely surprised if only 13% left the area. (They'd move to another state rather than out of the country, but Iraq is more like one of our larger states than like the whole USA; it's only a few hundred miles to the border, and Arabic is the primary language across the borders to the north, west, and south, and is also spoken by all educated Iranis to the east.)

But you don't have to blow anything up to get that many Americans to pack up and leave. Close down a major employer in an area, and over 13% of the families are likely to be leaving. Furthermore, just look at how often Americans move across the country for a better job when the economy is strong. 13% per year is only moving every 7 years. I don't think long moves (clear out of the state and the metropolitan area) are quite that common, but moves between suburban school districts are probably more frequent.

Don't forget that atleast seven households consider Mustafa to have been member.

But if you want to fly to Washington DC, we can step outside and settle this like gentlemen.

Jane, I think you might be two men short.

I nominate Thorley Winston's last for comment of the week.

Comments are Closed.