The New York Times Public Editor is highly annoyed at Paul Krugman:

Two weeks have passed since my previous post spelled out the errors made by columnist Paul Krugman in writing about news media recounts of the 2000 Florida vote for president. Mr. Krugman still hasn’t been required to comply with the policy by publishing a formal correction. Ms. Collins hasn’t offered any explanation.

As a result, readers of nytimes.com who simply search for “Krugman” won’t find any indication that there are uncorrected errors in the columns the query turns up. Nor will those who access Mr. Krugman’s columns in an electronic database such as Nexis or Factiva. Corrections would have been appended in all those places if Mr. Krugman had complied with Ms. Collins’ policy and corrected the errors in his column in the print version of The Times. (Essentially, to become part of the official archive of The Times, material has to have been published in the print paper.)

All Mr. Krugman has offered so far is a faux correction. Each Op-Ed columnist has a page in nytimes.com that includes his or her past columns and biographical information. Mr. Krugman has been allowed to post a note on his page that acknowledges his initial error, but doesn’t explain that his initial correction of that error was also wrong. Since it hasn’t been officially published, that posting doesn’t cause the correction to be appended to any of the relevant columns.
. . . .

A bottom-line question: Does a corrections policy not enforced damage The Times’s credibility more than having no policy at all?

(H/t: Regret.)

UPDATE: If you use the general link for the Public Editor’s column, you can scroll down for the story of that September 2 correction (he doesn’t seem to understand permalinks). Here are the highlights:

Opinions expressed on the editorial and Op-Ed pages of The New York Times aren’t part of the public editor’s mandate. But the facts are. And so are corrections of any misstatements.

So when I discovered on Aug. 19 that Paul Krugman’s Op-Ed column that morning contained a sweeping assertion that was wrong in at least one respect, a formal correction was my sole concern. . . .

But Mr. Krugman has been reluctant to formally correct his misstatement, starting when I raised the issue with Gail Collins, editor of the editorial page, on the day his column appeared. He wanted to use his Aug. 22 column, it seemed to me, to explain the misstatement without admitting any errors. He focused on the consortium led by The Miami Herald, and he acknowledged that Mr. Bush had won one of three statewide manual recount scenarios it conducted. But, absent a formal correction, the information didn’t get appended to his flawed Aug. 19 column.

When I pressed Mr. Krugman to do a formal correction after his Aug. 22 column, he agreed to run one at the bottom of his Friday, Aug. 26, column. In that correction, he reiterated that two of the Miami Herald’s three statewide recounts had shown Mr. Gore to be the winner. He also formally corrected an erroneous 2004 Ohio voter turnout percentage that a Times reader had brought to my attention two days earlier.

After the formal correction was published, I started checking out comments I had picked up in discussions earlier in that week with puzzled newspaper editors who had been involved in the two recount projects.

There were two problems with the formal correction about the recounts, I discovered. It was wrong on the results of the Miami Herald statewide manual recounts. And it didn’t deal with the fact that the original Aug. 19 generalization, the Aug. 22 column and the formal correction all erred in describing the findings of the other news media consortium (in which The Times was a participant).

. . . .

In passing the details on the statewide manual recounts to Mr. Krugman and Ms. Collins Monday, Aug. 29, I urged them to run a formal correction to clear up the whole tangle. “My first reaction,” Mr. Krugman responded by e-mail, “is that we’re really down to small points, which have no bearing on the original point of my remark about recounts—which was, after all, that the election was so close that even modest vote suppression was crucial.” As for Mr. Bush winning one of the six recounts done by the other news media consortium, Mr. Krugman said in another e-mail, “I thought that was a minor detail—frankly I can’t believe that anyone really thinks it’s important….”

Ironically, Mr. Krugman can make—and has made—a case that he was misled by the Miami Herald’s failure to detect and correct an omission in its April 4, 2001, article on the recounts conducted by its consortium. . . . . But if the Miami Herald had caught and corrected its omission back in 2001, Mr. Krugman might have been spared at least some of the tangle in which he finds himself now. One would think that possibility would give him some appreciation for what a formal correction could mean to readers of his column.

Mr. Krugman could also figure some of this out by . . . reading blogs.