This is a shorter version of my recent rant about why Paul Krugman’s likely defense of his recent “Gore won two out of three undervote recounts” is still dishonest. I thought a quick summary might be a nice idea, since some readers said my earlier post was just too long to slog through.
I am assuming your familiarity with the controversy, and with Tom Maguire’s suggestion as to Krugman’s likely defense. Briefly, in an April 4, 2005 Miami Herald article that I found, Maguire found language that seems on its face to support the “two out of three” claim — if a full statewide recount were conducted, including new recounts in counties that had already done recounts. But in truth, the article doesn’t help Krugman — it just reveals new deceptions on his part.
To get the full background, or the detail regarding any of my arguments, please read my earlier comprehensive post on the topic. This is just the skeletal version.
Why is Krugman still wrong, even given the article found by Tom Maguire? Several reasons:
1) Krugman pretends that an unrealistic afterthought buried within the articles is really the study’s major finding.
2) Krugman claims that the study shows what would have happened if all undervotes were counted, but the study itself acknowledges that it didn’t necessarily examine all the undervotes.
3) Krugman collapses two standards where Bush was the winner into one, using cleverly deceptive language. The original study actually examined four standards, not three — and Bush won two, not one.
4) Krugman claims that the standard under which Bush won was unreliable and “almost certainly wouldn’t have been used in a statewide recount.” But the Miami Herald described it as “the standard most commonly used nationally.” Krugman claims this standard “would have discarded some ballots on which the intended vote was clear.” But his reference is to dimpled ballots, and dimpled ballots are far from clear — as I point out in my comprehensive post with numerous quotes, which are certain to bring up painful flashbacks of the 2000 post-election circus.
Interestingly, Don Luskin has independently picked up on my point number 3 above, and did some legwork of his own to confirm that there were indeed four standards, not three, even in Krugman’s hypothetical statewide recount. Krugman clearly owes yet another correction.