Politics & Policy

Two Out of Three Ain’T Bad?

Worse -- Krugman's recount claim is a lie.

Chalk up a double win for the Krugman Truth Squad. We’ve forced Paul Krugman to publish a double-correction appended to his New York Times column last Friday. He’s had to publicly take responsibility for not one but two of the outlandish claims he made about the 2000 and 2004 presidential elections in his August 19 column. And there’s more to come — we’re going for a triple.

Krugman makes corrections very rarely and very “reluctantly“, and has said he finds them “humiliating.” The first one in Friday’s double-correction column is humiliating to the max. Savor this slice of humble pie: America’s most dangerous liberal pundit has had to confess that he cited erroneous information sourced from the Democratic party without corroborating it with official records.

The second of the two corrections is not literally a correction — it’s more a clarification of a previous correction that had not been properly labeled as such (what former Times “public editor” Dan Okrent called a “rowback“). But in the process of making this clarification Krugman actually repeats an error, and takes the opportunity to repeat the lie that media recounts of the 2000 election showed Al Gore the winner.

Here’s the entire text of the double-correction that was appended to his column last Friday:

Corrections: In my column last Friday, I cited an inaccurate number (given by the Conyers report) for turnout in Ohio’s Miami County last year: 98.5 percent. I should have checked the official state site, which reports a reasonable 72.2 percent. Also, the public editor says, rightly, that I should acknowledge initially misstating the results of the 2000 Florida election study by a media consortium led by The Miami Herald. Unlike a more definitive study by a larger consortium that included The New York Times, an analysis that showed Al Gore winning all statewide manual recounts, the earlier study showed him winning two out of three.

The first matter — that Krugman’s false allegation of voter turnout in Ohio’s Miami County in the 2004 presidential election was a suspiciously high 98.5 percent — is quite straightforward. The “Conyers report” to which Krugman refers is a document about the 2004 election released last January by the Democratic members of the House Judiciary Committee. The committee’s ranking member is Angry Leftist congressmen John Conyers of Michigan (whose more recent shenanigans include leading the Washington feeding frenzy about the so-called “Downing Street Memo”). The Democratic report’s lie that Krugman unquestioningly relied upon appears on page 58. The truth was unearthed by the Power Line blog’s John Hinderaker, who discovered it sitting there for all the world to see on the official website of the Ohio secretary of state.

The second matter, concerning the Miami Herald’s consortium study of the 2000 election, is a little more complicated. This story begins with a statement Krugman made in his August 19 column:

Two different news media consortiums reviewed Florida’s ballots; both found that a full manual recount would have given the election to Mr. Gore.

As I and many other observers have pointed out, this statement is deceptive. Both consortiums, in fact, found that under the most likely ways of conducting the recount — based on what the Gore campaign had asked for or what the Florida supreme court had ordered — George W. Bush would have been the winner. Even Krugman’s own New York Times reported it this way.

Responding to such criticisms, Krugman wrote in his August 22 column that he had not been talking about what the consortiums had found when they conducted the limited recounts requested by the Gore campaign and ordered by the Florida supreme court. Rather, he was talking about hypothetical “full” recounts in all 67 Florida counties — including those that had already completed their own manual recounts at the time of the court’s order.

In that August 22 column, Krugman said that in the Herald’s consortium study, “Two out of three hypothetical statewide counts would have given the election to Mr. Gore.” The correction that ran last Friday didn’t correct or modify that statement — indeed, it repeated it. Friday’s correction was only an “acknowledgment” that Krugman should have said all along that the Herald consortium gave the election to Gore only in two out of three hypothetical counts, not unambiguously (as his initial statement had suggested).

Thus, Krugman must think that his initial statement about how “both” consortiums “found” Gore the winner has been vindicated, even if it was somewhat softened in the process. Wrong. There’s another correction coming, because Krugman’s claim that Gore won “two out of three hypothetical statewide counts” is based on false information — just as false as the lies in the Conyers report.

There weren’t three “statewide manual recounts” in the Herald consortium’s study — there were four. The consortium didn’t show Gore “winning two out of three.” They showed him winning two out of four — leaving Krugman’s initial statement that “both” consortiums “found” Gore the winner completely false. The Herald consortium didn’t find either candidate to be a definitive winner. It was a split decision.

The source of Krugman’s “two out of three” claim has been a bit of a mystery over the last week among Krugman’s critics in the blogosphere (an account appears on the Just One Minute blog). But now the mystery appears solved. Directly or indirectly, the claim must surely come from an April 4, 2001, Miami Herald story by Martin Merzer, headlined “Review Shows Ballots Say Bush.” (I can’t provide a free link to the story, but those interested can obtain it for a small fee from the Herald’s online archives). Merzer’s story, as it appears in the Herald’s online archives, presented the Herald consortium’s “statewide manual recounts” calculated against three different standards:

‐ “a loose standard where every dimple, hanging chad and pinprick was a vote” — by this standard, Gore would win by 393 votes;

‐ “a tougher standard where a dimple counted only if the ballot had other dimples on it” — by this standard, Gore would win by 299 votes;

‐ “a still tougher standard where a ballot had to have chads detached by at least two corners to count as a vote” — by this standard, Bush would win by 352 votes


At first blush, from this it would seem that Krugman is correct about Gore “winning two out of three.” But wait. The Herald is a Knight Ridder paper, and a Knight Ridder wire-service story written by the same Martin Merzer the very same day precisely repeated the results of the first three standards as given in the Herald version, but included a fourth as well. The Washington Post picked up the wire story under the headline “In Ballot Audit, Bush Prevails” (again, no free link, but it’s available for a small fee from the Post’s online archives), and the fourth standard was described as follows:

‐ “A vote was counted only when a hole was cleanly punched, the most restrictive standard” — by this standard, Bush would win by 416 votes.

Mark Seibel, the former managing editor who ran the ballot-review process for the Miami Herald and is now international managing editor for Knight Ridder’s Washington bureau, reviewed his original source material and personally confirmed to me that the fourth “clean punch” standard under which Bush wins Florida by 416 votes was indeed part of the Herald consortium’s findings.

Neither Seibel nor Merzer have any idea why the fourth standard was omitted from the Herald’s story — both in its original print rendition and in the online archives. I’ve discovered that a 2001 book edited by Merzer, The Miami Herald Report: Democracy Held Hostage, also omitted the fourth standard. Seibel told me that he and Merzer had been unaware of it all these years, and speculated that it might have been a production error. But Seibel told me, “I’m absolutely positive that the fourth clean-punch standard described in our wire story is correct.”

Krugman’s twice-repeated statement about Gore “winning two out of three” is an error that must now be corrected. Further, Krugman’s initial statement that “both” consortiums “found” Gore to be the winner in a “full” recount must now be corrected. The truth is that only one did. For the other, it was a split decision.

Here’s my suggested text for a new correction:

Corrections: In my column of August 22, and again in a correction appended to my column last Friday, I misstated the results of the 2000 Florida election study by a media consortium led by The Miami Herald, saying it showed Al Gore winning two out of three statewide manual recounts. In fact it showed him winning only two out of four. In light of this, the public editor says, rightly, that I must correct the statement in my column of August 19 that both of two media consortiums’ studies found Gore to be the winner of a full recount. In fact only the consortium that included The New York Times did so. I regret that I let my Angry Left agenda get in the way of the truth.

Okay, that last sentence isn’t gonna happen. As to the rest of it, “public editor” Byron Calame hasn’t responded to my request for comment. But I’ve made him aware of all the facts, and I know that he has communicated directly with Knight Ridder’s Seibel. Now it’s simply a matter of setting the record straight, and if the New York Times has the slightest shred of integrity, such a correction will be run.

Think of it as a full recount. Isn’t that what Krugman says he wants?

– Donald Luskin is chief investment officer of Trend Macrolytics LLC, an independent economics and investment-research firm. He welcomes your visit to his blog and your comments at don@trendmacro.com.


The Latest