Sunday, October 24, 2004

Learning Math with Paul Krugman

Steve Sailor, who's been studying old SATs and Officer Qualification Tests and such things, says that the available evidence indicates that John Kerry's IQ is probably lower than George Bush's.

Now somebody needs to find out if Paul Krugman, one of the canned brains from the New York Times, is smart enough to operate his own keyboard - or if he has to dictate the junk that appears under his name. Yesterday he (or someone) wrote:
If the U.S. presidential election were held today and the votes were counted fairly, Senator John Kerry would probably win. But the votes won't be counted fairly, and the disenfranchisement of minority voters may determine the outcome.
Nobody knows what Krugman means when he looks into the future and states flat-out that "the votes won't be counted fairly", unless he's complaining about the fact that people who vote for Bush will have their votes counted, too. But this is the familiar "It's only fair if my guy wins" whine that we've been hearing for weeks now. Let's get on with the math:
Recent national poll results range from a 3-percentage-point Kerry lead in the AP-Ipsos poll released Thursday to an 8-point Bush lead in the Gallup poll. But if you line up the polls released this week from the most to the least favorable to President George W. Bush, the polls in the middle show a tie at about 47 percent.
Could you do that one on the blackboard, professor? Those of us who can freaking count are confused.

Krugman wrote this article on October 23rd, at the end of a week in which 15 major Bush/Kerry/Nader polls were released. Kerry showed a lead in only one of them, the AP-Ipsos poll. Bush led in 12 polls, with two ties. No matter what order you line them up in, the average of these polls is 48-46 in Bush's favor. In a single poll you could fairly fudge that to a tie of "about 47 percent", but not when you're comparing 15 polls. Only 6 of these polls showed an advantage that was outside of the poll's margin of error, and Bush won all six.

During the same week (10/16 to 10/23) there were 8 major head-to-head polls released. These polls excluded Ralph Nader, whose presence is a presumable disadvantage to Kerry. But Bush won in all 8 of these polls, with an average of 48.8% Bush, 45.2% Kerry.

People question the validity of polls, which is understandable. While any single poll must be suspect, an aggregate result from a dozen polls is pretty good, and has served us very well in the past. And if we pretend that polls mean nothing, then there is no point speculating about the election at all, because there is no other real basis on which to form a judgment. Polls aren't just the best evidence we have, they're the only evidence we have.

The only evidence that sane people have, I should say. People like Krugman are beginning to display another kind of creepy gnostic certainty: the idea that they are entitled to win, no matter what, and if they lose it can only be because they were cheated. In fact, a loss is prima facie proof that they were cheated - even if they know that they weren't.

Sentiments like Krugman's are growing, and such thinking is absolutely deadly to democracy. It might also cause gambling addiction, but I can't prove that mathematically.