Bret Stephens and Paul Krugman: What Should a Correction Look Like in the Digital Era?

In my presentation “On the Future of the Economics Blogosphere” I talk about how the websites of newspapers and magazines that began as creatures of print often do things on their websites that are not best practice for online journalism. Underusing links is one example. Another related example is the way corrections are handled. Bret Stephens’s misuse of nominal income statistics in his piece “Obama’s Envy Problem” and the manner of the correction he made provide a good case study. 

To help you get up to speed, let me say that this post is a followup to my posts

One reason I thought a followup would be worthwhile is that my post “The Wall Street Journal’s Quality-Control Failure: Bret Stephens’s Misleading Use of Nominal Income in His Editorial ‘Obama’s Envy Problem’” is at this moment my 3d most popular blog post ever.

In his January 8, 2014 post “Do Publications Have Any Responsibility to Screen Their Editorials? (Part 2)” R. Davis wondered why the original Bret Stephens article as it appears on the Wall Street Journal’s website has no link to Bret Stephens’s correction piece, and indeed no indication of any problem. Paul Krugman took Bret Stephens to task for the same thing in his post “The Undeserving Rich,” writing:

Oh, and for the record, at the time of writing this elementary error had not been corrected on The Journal’s website.

As this sentence stands, it is false, because it says “website” rather than “the version of the original article appearing on the website.” In addition to his correction on January 3, 2014, “About Those Income Inequality Statistics: An answer to Paul Krugman” on the Wall Street Journal's website, there was a formal January 5, 2014 acknowledgement in “Corrections & Amplifications.”

In his followup "Department of Corrections, and Not,“ Paul Krugman gives more precision, saying what counts in his book as a correction:

… you fix the error in the online version of the article, including an acknowledgement of the error; and you put another acknowledgement of the error in a prominent place, so that those who read it the first time are alerted. In the case of Times columnists, this means an embarrassing but necessary statement at the end of your next column.

I fully agree with Paul Krugman about how a correction should be done in the digital era. Arguably, "About Those Income Inequality Statistics: An answer to Paul Krugman,” counts as “the next column,” so Bret seems to have done that. And the Wall Street Journal did what would have been adequate in the paper era in "Corrections & Amplifications.“But in the digital era, that is inadequate. Whenever a correction of a serious error is made, it needs to be either made directly on the web version of the original article or addressed with a clearly signaled link. The Wall Street Journal neglected to do this. To the extent that this is based on a standard practice of the Wall Street Journal, that is the more inadequate, since it will affect many articles.  

In his article "Stephens: Krugman and the Ayatollahs” Bret expresses some pique at Paul’s accusation that the article had not been corrected on the website. He never focuses on the fact that the original article as presented on the website has no clue of the correction, emphasizing his correction on January 3, 2014, “About Those Income Inequality Statistics: An answer to Paul Krugman,” and the formal January 5, 2014 acknowledgement in “Corrections & Amplifications.” Bret should be agitating strongly for a correction to appear on the online version of his original article as well. I look forward to seeing that correction appearing on the website version of the original article–especially if it betokens a general change in the Wall Street Journal's standard procedures for representing corrections on its website.

By the way, according to the principle I am enunciating here, and under the circumstances of a dispute like this, I would urge Paul to make sure that the online version of his column "The Undeserving Rich“ gets a link added to ”Department of Corrections, and Not,“ which provides the necessary clarification for the on-its-face inaccurate statement in ”The Undeserving Rich“ that the Wall Street Journal website had no correction for the error in Bret’s original article.

One final note: I am perfectly comfortable with silent edits of posts if errors are caught before they have become the basis for other people’s posts. I discuss my own policy in that regard in my post "It Isn’t Easy to Figure Out How the World Works” (Larry Summers, 1984)“:

… I routinely allow myself silent edits in my blog posts without an update notification. I am not running for office, I am not on trial, and I already have tenure, so I don’t have to play the game of “gotcha.” In my case, it is my words that matter, not me, and the words that matter are the ones I am willing to stand behind in the end.

One should, of course, forthrightly admit that an error was made previously if anyone asks, but as long as an error is corrected, I see no duty to advertise that previous error. The number one duty is to get things right for the readers. Beyond that, it is a simple matter of honesty about previous errors should the issue come up.

It is only when errors become the basis of other people’s posts or other writings that one needs to clearly signal errors, since knowledge of the error is required for readers to understand what the discussion is all about. Errors that are the basis of comments on the original post are a gray area. If the value of the relevant part of the comment is just to point out the error, then correction of the error respects the point of that comment. On the other hand, if the comment has something brilliant to say that is stimulated by the error, then it would be good form to somehow preserve the record of the error in order to avoid destroying the record of the stimulus for that comment. But the record of the error need not be in the original location; after all, there are other objectives served by having a casual reader able to read something correct straightaway without necessarily going through the whole convoluted path.