About this site

  • Comments
    When you submit a comment, it won't be published until approved. This is to cut down on comment spam. However, I will also edit or block comments that are profane or offensive.
  • No Legal Advice
    Although I may from time to time discuss legal issues on this blog, nothing that I post should be construed as legal advice, nor as creating an attorney-client relationship between you and me. In fact, there's a good chance I'm not licensed to practice law wherever you are. If you need legal advice, you should consult an attorney licensed in your jurisdiction.
  • Personal View
    This blog is neither affiliated with my employer nor hosted by it. It is maintained through TypePad, and I pay the hosting fees. Nothing that is posted here should be construed as anything other than the views of the particular author of the post.
  • Tung Yin's Recent Papers (SSRN)

April 2014

Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30      

Stats


  • Check Google Page Rank

« Men At Work | Main | Blog housekeeping, or it's all a matter of choice »

October 29, 2004

Comments

Kevin Jon Heller

For the record, A. Rickey's comment that I've never posted an anti-Bush statement that I didn't like is inaccurate. When Bush commented that we could never completely win the war on terror, I criticized the hysterical responses from Democrats in no uncertain terms.

A. Rickey

As you note, "never" is an overstatement. Perhaps I should revise it to something like "Prof. Heller frequently posts anti-Bush statements from rabidly partisan sources, or fantastically anti-Bush figures that demand critical review, without giving them sufficient, or even cursory, evaluation. This hurts his credibility." I figured that was a bit wordy, but as you point out, never is too absolute.

As for the argument from authority that was referenced, let me repeat your argument to poor Fitz-Hume:

I see -- I'm supposed to take your conclusory statement that 100,000 people haven't been killed over the heavily documented conclusions of one of the pre-eminent schools of public health in the world. Thanks, but given the absurdity of your analogy (the Bloomberg report deals with an 18-month period, and the Tokyo firebombing took place in one day), as well as your obvious disinterest in reading the article or the report itself (which expressly excluded deaths in Fallujah, because of the abnormally intense violence there), I'll pass.

But please, feel free to submit your polling method, so I can compare it to the one the Bloomberg team used.


(emphasis mine)

Now, as I point out above, what's there isn't "heavily documented." (Or rather, it probably is, but not publicly documented.) Unless you've got access to more than the eight-page report I read, you have a methodological summary that should cast some immediate doubts in your mind, even if you're a non-statistician. You have a number of deaths from airstrikes (at least 25,000, or roughly 46 people per day, every day, over the study period) that should at least raise an eyebrow. And the lack of the word "civilian" near the number of 100,000 in the report? That is pretty telling, even if you do no math at all. What you don't have is enough numbers to do a reasonable evaluation.

The point is, you wanted to believe, so you printed it. In the past, you've said you're "resigned" to such things, but that's a resignation that I simply don't think is acceptable in someone who is trying to provide commentary and analysis, and wants to remain credible. This isn't a specialist/non-specialist issue: if you really don't feel you can evaluate something, even on a low level, don't comment on it. But none of the above requires more than some critical thinking and some high-school math.

In the meantime, you fell back on the argument from authority. Again, I have to wonder: a report that is highly useful in a partisan fashion is rushed to press days before an election. You didn't even wonder that the "highly respected" establishment you were reprinting might have had some less-than meritorious motives?

Kevin Jon Heller

That's an interesting semantic trick -- subtlely implying (again) that the Bloomberg report is biased against Bush, despite having no evidence of that fact other than (I guess) paranoia that any information that could be used to criticize Bush must be motivated by anti-Bush zealotry. In your first post, you note that the author of the report has an "axe to grind" against airstrikes and use that fact -- despite not having any evidence to back it up -- to imply that the report is anti-Bush. It's not, of course; there's no reason to believe Kerry would be any less willing to use airstrikes than Bush. And now you mention that the report could be "useful in a partisan fashion" and use that fact to imply that the report is anti-Bush. That's a logical fallacy -- the fact that something could be used by others to hurt Bush in no way proves that the author of the report intended it to be used that way or harbors anti-Bush sentiment himself. (The same is true of the fact that the Lancet published the report, which also tells us nothing about the author of the report's motives.)

All in all, an excellent job of trying to impeach a non-partisan report -- which may or may not be correct in its conclusions -- by making unsubstantiated assumptions about its motivations. That seems far less intellectually honest than making use of a rebuttable presumption that conclusions reached by a highly-respected, non-partisan academic institution are more likely to be true than, say, the conclusions of a Cato Institute or Center for American Progress.

The Bush administration, which makes a living attacking the messenger in order to divert attention from the message, would be proud.

A. Rickey

All in all, an excellent job of trying to impeach a non-partisan report -- which may or may not be correct in its conclusions -- by making unsubstantiated assumptions about its motivations.

Come, now, Prof. Heller. Let's argue about your term "unsubstantiated." First, let's look at the fact that this did not go through the normal peer-review process, but was "expedited." Secondly, let's cut the crap about "non-partisan" institutions. While an institution may not be partisan, it can be biased. Rushing a report out in front of an election does seem to be an attempt at influence.

And wow... two minutes of further research points at further hints that your "impartiality" may be missing. Let's see what the lead author of the report has to say for himself:

"I emailed it in on Sept. 30 under the condition that it came out before the election," Roberts told The Associated Press. "My motive in doing that was not to skew the election. My motive was that if this came out during the campaign, both candidates would be forced to pledge to protect civilian lives in Iraq.

Nope. No bias there. Completely standard process to want to rush peer-review to make sure something appears before an election, I'm sure. And the fact that it will be seized upon by anti-Bush partisans--such as you, Prof. Heller, or am I impugning non-partisan academics here?--should raise no suspicions whatsoever.

But again, we get the argument from authority. You're a law professor, Heller, so let's ask: why is the truth of a statement--that 100,000 "excess" civilian deaths--presumptively true, and its my responsibility to rebut them? Why is it not reasonable to say, "This deviates from a standard procedure, admittedly because of an election? That's an indication of bias?" The fact that the author admits that he was opposed to the war (same article as above) doesn't case suspicions of partisanship? But no. It's a "highly respected, non-partisan academic institution," so it must just be that crazy A. Rickey again.

Prof. Heller, I've put forward an argument for why the statistics don't hold water, to which you're perfectly welcome to reply. Furthermore, I've backed it both with research and quotations from the work itself, and from the press. You, on the other hand, have fallen back upon your normal tactics of accusing your opponents of smears, of maligning something sacrosant--last time it was your "academic integrity", if I recall correctly--and otherwise refusing to answer the question.

You've got the report in front of you. On the first page is the author's email address, to which you could email the author and ask for a full set of statistical data. (You, being a professor, are much more likely to get a response.) If you wish to say that 100,000 "excess" "civilian" casualties have occurred as a result of the war, check the data and see if it gives you the stomach to back it. But stop saying that the data is reliable because the institution has a certain reputation, or that there's some rebuttable presumption that exists for anyone who provides evidence that suits your own political prejudices.

I have never reprinted data on my site that I felt was "too good to be true" without looking into it, simply because I feel it hurts my credibility: inevitably it will be shown to be full of wholes. Stop saying that your failure to do so is somehow a failure of your critics to share your allocation of "rebuttable presumptions." Stop saying that so doing is "attacking the messenger." For god's sake, stop being "resigned."

A. Rickey

Ach. That's "full of holes" not "full of wholes."

A. Rickey

Oh yes, incidentally:

In your first post, you note that the author of the report has an "axe to grind" against airstrikes and use that fact -- despite not having any evidence to back it up -- to imply that the report is anti-Bush.

The mind boggles. My "evidence to back it up" comes from the Lancet report itself, the very last sentence:

In the interim, civility and enlightened self-interest demand a re-evaluation of the consequences of weaponry now used by coalition forces in populated areas.

Or how about from the first page?
Our results need further verification and should lead to changes to reduce noncombatant deaths from air strikes.

Given that he attributes (at least) 25% of the deaths to airstrikes, and that he doesn't mention much else in the way of weaponry now used in populated areas, I think it's safe to say he's arguing for a reduced role for air strikes. That's pretty clear upon reading the report, so I didn't feel the need to cite one of the reports major conclusions directly. After all, you'd read it as well.

Fed.No.84

I'm trying to find out what the argument is.

Is the argument that Kevin Heller's biases interfere with his analysis of political issues. Thus, others are presenting data contradicting the Bloomberg report to prove this proposition?

Or is that argument that although 100,000 dead civilians is bad, if Bloomberg is really wrong, only 25,000 (or some other smaller number) civilians died. While 100,000 dead is bad, 25,000 is not that big of a deal.

I think that Kevin's initial point was, "Look at all these dead people. What do you think of the Iraq war now." I'm not sure that we can make a principled arguement why 25K is acceptable, but 100K is not. IOW, even if Bloomberg is really wrong (re: their number is off by 75%), Kevin's argument remains strong.

Tung Yin

Fed. No. 84, I should let Tony speak for himself, but it seems to me there are a couple of points here.

First, I'm in general agreement with Kevin that the credibility of a source is relevant to assessing how much investigation one must do of assertions reported by that source. I'm not familiar with the Lancet to be able to say anything about it, and I'd like to think that a major university like Johns Hopkins should be presumptively credible. At the same time, the Rathergate fiasco has demonstrated that even presumptively credible sources aren't. From herein out, I will not trust any story that is reported by Dan Rather on "60 Minutes," because Rather told us he was as confident as he has ever been on a story about the authenticity of the Killian memo, and well, we know how that turned out. Kevin hasn't said anything either way about Rather's credibility -- and I should quickly note, Kevin is under no obligation to do so -- however, I'd like to think that Kevin too agrees with me about Rather's utter unreliability right now.

(Again, the underlying substance of the story may be true -- Bush probably did get special treatment -- but there are two separate stories there: the political story and the media story.)

Second, the difference between 25,000 and 100,000 may or may not be significant -- though note that if Tony is right, it could be as few as 8,000 (within the 95 percent confidence level). Again, 8,000 is nothing to sneeze at, but the comparative difference between what Saddam plus the sanctions would have killed versus what our airstrikes supposedly have is the key issue. And it seems to me much easier to make the point that Saddam plus the sanctions may well have killed more than 8,000 people during the time that we're measuring.

Needless to say, each of those deaths is tragic, but if we're making policy arguments, we should have some reasonably solid grasp of the numbers we're talking about.

A. Rickey

Fed. 84:

So, let's recap. It doesn't matter if one posts highball headline numbers without a cursory review of the data. You're not really worried about that.

And of course, 100,000 is bad, and 25,000 is bad. So let's ask: what number isn't bad? What is the acceptable number of civilian casualties--which are relatively inevitable in any war, though one can do ones best to reduce them--at which Kevin's argument is no longer strong?

Now, if we really wanted to do that analysis, we'd have to make some kind of cost-benefit analysis. And that might very well cover more than 18 months post- (and pre-) invasion, might take into account other factors... whatever. The point is that your argument above either suggests that there is some number of acceptable casualties (below, it seems, 25,000), or no casualties are acceptable. In which case no military action is ever justifiable, which is a position, but you'd better take it without hesitation.

Now, once you have your number established--because, you know, Fed, we should know what we're aiming at here--the next question would be has that number been exceeded. And in order to do that, we'd need a study of more than dubious accuracy, at least if you wanted to make a solid policy argument.

That, of course, would deal with your point. My point is that when it comes to Prof. Heller, if he puts up a big headline number from a "reputable" source, you ought to check it, because he's pretty well "resigned" not to doing so himself. Indeed, he'll presume (charitably, that presumption is rebuttable) that it's true. That's all.

Fed.No.84

From herein out, I will not trust any story that is reported by Dan Rather on "60 Minutes".

Are you overstating here, or has that memo really done that much damage to Rather's credibility?

The comments to this entry are closed.