Department of Awful Statistics
Poor, poor DougJ:
Department of Awful Statistics: Why Ask Y? - Megan McArdle:
DougJBalloon 16 hours ago: I see you're revisiting the brilliance of your arguments for the Iraq War on twitter. Any interest in correcting some of the order of magnitude errors?
McMegan 15 hours ago: Do I have any interest further interest in debating whether the direct cost to the taxpayer is going to be $2 trillion? An argument that is going to devolve--since there is at this point vanishingly little possibility that it will grow to $2 trillion--into the other side trying to argue that things which aren't direct costs to the taxpayer are in fact direct costs to the taxpayer, or complaining that it's not fair to just look at direct costs to the taxpayer, or claiming that it's innumerate of me to compare stocks to stocks and flows to flows? Surprisingly, not really. But Merry Christmas!
DougJBalloon 15 hours ago: How about putting a strike through the 0.1% at least?
McMegan 15 hours ago: You're not factoring in growth. The correct equation is not 20 * $11tt, which is what my interlocutor--and I presume, you--are doing. That's not even right by some weird arguable metric; it's just wrong unless you presume that GDP is actually going to shrink farther and then stay there until 2023. I was using the CBO's standard growth rate of 3%, which I freely predict was in error for the past few years, because I didn't see this huge crash coming, but no way of knowing how big an error until 20203. However, not an order of magnitude as you claim. At the peak of spending, the cost of Iraq was maybe 1% of GDP, and spending has now fallen sharply; how could you think that it was going to be that high over a 20-year period, when we are supposed to have withdrawn for half of it?
DougJBalloon 15 hours ago: You wrote this in 2003. Let's estimate the 2003 GDP as 11 trillion, let's estimate 4 percent growth (that's on the high side but I am a very generous person). Then over 20 years, the total GDP would be 327 trillion dollars. That estimate is probably on the high side, by the way. You wrote "But it is not going to run us several trillion dollars (though even if it did, that would work out to less than 0.1% of GDP over the next 20 years.)" Let's take "several trillion". I would say that several means at least 3 and probably 4. I'm not talking about how much the actual Iraq War costs -- though that would be a pretty good estimate in fact -- I am talking about your use of the phrase "several trillion". If I take 3 trillion and divide by 327 trillion, I get slightly less than 1%. If I take 4 trillion (really the kindest interpretation of "several trillion" I can think of) and divide by 327 trillion, I get over 1%. You meant 1%, not 0.1%. Are you really so quantitatively inept that you cannot see this, even after I brought it up again? Are you really so nuts that you're going to bs me about rates of GDP growth and not just divide 3 by 300? God help us all.
McMegan 14 hours ago: Sigh. Okay, so we are now not discussing the actual cost of the war, but the hypothetical cost of the war as represented by the term "several trillion", which to me means any sum over $2 trillion, but YMMV. As I recall--it was, of course, eight years ago--I ran nominal figures out to 2023 with interest to account for the borrowing that we were doing. You could probably quibble with my methodology if I could remember it. It's certainly possible there was an error in my calculations, though it was a spreadsheet so not all that likely. Probably a lot easier to have had this conversation if you had raised the issue eight years ago, but at the time your coblogger was still egging on my less sane moments, IIRC. At any rate, if I have time tomorrow, I'll try to figure out what I did, and see if I still want to defend it.
DougJBalloon 12 hours ago: Hypothetical??? It was your hypothesis. You are utterly and completely insane. God have mercy on David Bradley's soul for what he has unleashed on an unsuspecting public. Bring on the apocalypse.
DougJ: This set down: Now, look. I know that not all of you are accountants or economists or mathematicians or engineers. But let’s suppose I said to you that in 2003 the United States GDP was around 11 trillion and that is was expected to grow at an average rate of 4% or less over the next 20 years. You could get out a spreadsheet and come up with a reasonable upper bound of $327 trillion for the total GDP over those 20 years, right? Many of you could probably even do the rough estimate in your head that on average over those 20 years, it couldn’t be more than $20 trillion a year so that the total would be $400 trillion or less.
Suppose someone said to you that:
The (Iraq) war will certainly cost more than the $60b and change that the President is asking for. But it is not going to run us several trillion dollars (though even if it did, that would work out to less than 0.1% of GDP over the next 20 years.)
You’d register that “several trillion” means something like 3-4 trillion or more and say “nope, you mean 1% not 0.1%”, right? And if you yourself had made the 0.1% estimate and someone told you, nope, you’re wrong, it’s 1% and then explained to you in painstaking detail why several trillion is about 1%, not 0.1%, of $327 trillion, you would understand, right?
Because you know how to follow a link and operate a f---ing calculator right?
Don’t be afraid to say that no, you couldn’t do any of this, that you can’t follow what it means to divide 4 by 400 and you have no fucking idea why 3-4 trillion is 1%, not 0.1%, of 400 trillion. Because if you can’t perform these simple calculations, then you too can be the Business and Economics Editor of the Atlantic, you too can earn close to 200K a year, and appear on shows like NPR’s Marketplace.
I realize this is my second post on this not-very-important topic. I’ve talked about this enough and I’ll shut up now.
