Updated edition of my post for journo nerds! (Includes full text of editors’ email to newsroom on unnamed sources) https://t.co/kyJkzBE1hf
— Margaret Sullivan (@Sulliview) March 15, 2016
The New York Times finally has a new and (hopefully) improved process for handling stories using unnamed sources. The process is outlined in a memo from Executive Editor Dean Baquet, Deputy Executive Editor Matt Purdy and Standards Editor Phil Corbett, and was reported Tuesday by Public Editor Margaret Sullivan.
The memo outlines which editors need to sign off on different types of uses of unnamed sources. As a frequent critic of the Times’ overuse of unnamed sources, I applaud the effort to be more demanding of reporters seeking to use them. I agree with Sullivan’s assessment:
This is a sensible, moderate and necessary plan. The devil, of course, is in the enforcement. The Times often has not done an effective job of carrying out the policy it already has, one element of which states that anonymous sources may be used only as “a last resort.”
If the Times editors uphold high standards in approving use of unnamed sources, the new process will be a huge step forward, ending the frivolous and needless use of confidential sources while still leaving the Times positioned to deal with informed source who sometimes are the only way to tell important stories on such matters as national security and law enforcement.
I’m especially glad to see the memo acknowledge the threat to Times credibility caused by overusing unnamed sources, a danger Sullivan has highlighted in her AnonyWatch feature. I have blogged repeatedly about overuse of unnamed sources at the Times and other newsrooms (see the links at the end of this post). And the Twitter feed @NYTanon highlights Times use of unnamed sources. (It cited 10 examples yesterday.)
— NYT Anonymous (@NYTAnon) March 15, 2016
I will post the full Times memo below, with some notes. I would like to see four possible improvements to the Times process and standards (and any or all of them might be part of the conversations with editors required by the new process):
Any consideration of using unnamed sources needs to focus on efforts the reporters made to verify what they were told by the sources. For instance, one of the two front-page stories last year that prompted the new process announced Tuesday should have been verifiable (and shouldn’t have been published without verification).
The Times reported incorrectly in December that San Bernardino shooting suspect Tashfeen Malik passed immigration background checks despite talking “openly on social media about her views on violent jihad.” The Times’ discussion of this story should have focused on whether the reporters had verified the claims of their unnamed sources (they hadn’t).
If this story had been true, reporters could have and should have verified the information from unnamed sources one of two ways:
- If Malik’s “open” social media posts were still available online, the reporters should have been able to find them, and the story should have quoted the posts and included screenshots. Maybe a source could point you to the post by telling you which social platform(s) were used, what user name Malik used, what date(s) she posted about violent jihad. If the source wouldn’t help, the reporters needed to find the posts through their own searching. If she posted “openly,” as the story said, Times reporters should have been able to find the posts that immigration screeners supposedly missed.
- If the social media posts had been deleted, the reporters should have found sources who had captured screenshots of them (or else how did the sources know?). Even if the sources would not turn the screenshots over to the Times for publication, a Times reporter should insist on seeing the screen shot, or that story isn’t ready for publication.
By requiring these verification steps, the Times at least would have delayed publication of the story as they searched for the non-existent posts. More likely, a source would have revealed (or the reporters would have learned independently) the error in the Times reporting: Malik hadn’t posted openly about violent jihad, but communicated her views in private messages. For instance, private messages on Facebook and Twitter look considerably different from public posts. If a source shared screenshots of such messages, the reporter would quickly catch the huge error in the story.
Verification may be murkier in other cases, but part of the conversation between reporters and editors before publishing a story using unnamed sources must be how the reporters have verified the information from unnamed sources and what standard of verification is appropriate for this story.
In dealing with unnamed sources, reporters need to ask sources and editors need to ask reporters journalism’s most important questions: How do you know that? And how else do you know that?
How have you pushed for on-the-record sources?
Another important part of the conversation about using an unnamed source should focus on what efforts the reporters have made to get some or all of the information on the record. I can think of three primary ways of doing this:
- As outlined in yesterday’s post of tips from Eric Nalder, editors should review steps reporters have made to persuade the unnamed sources to go on the record about some or all of the information on the record.
- Editors should ask reporters what documentation they have, or have sought, from the sources. If you have documents to cite (such as the hypothetical screenshots mentioned above), you may not need to mention the unnamed sources at all, or you use them only for secondary information.
- Editors should ask how reporters have used information they gathered from unnamed sources to prod other sources to speak for the record or provide documentation, as Eric discussed yesterday and as I addressed in a 2013 post.
As with verification, I’d like to see the importance of pushing for documentation and on-the-record sources spelled out in the Times process. But if it’s a focus of the editor conversations at the Times, that’s what’s important.
How powerful or eager is the source?
If a source is powerful or initiates contact with the reporter, editors should be much less willing to use information from that source than a source who is less powerful or is approached by a reporter and will speak only if granted confidentiality. Journalists should be doubly skeptical of eager or powerful sources.
For instance, I think the Times editorial board was absolutely wrong to conduct an off-the-record interview with Donald Trump and Times columnists were absolutely wrong to agree to an off-the-record session with President Barack Obama. Presidents and presidential candidates should always be on the record when meeting with journalists. In fact, those sessions should have been livestreamed on video, live-tweeted or both.
(The Times editorial and news departments are separate, and the new process announced Tuesday probably doesn’t apply to Times opinion journalists.)
No opinions from unnamed sources
I was pleased to see the Times memo says editors will only rarely approve the use of unnamed sources providing “speculation or interpretation.” I would add opinion to that list of near-prohibitions. As I wrote more than a decade ago, unnamed sources should have unpublished opinions. Without a name, opinions (like speculation and interpretation) are worthless.
The Times memo
Here is the memo from Baquet, Purdy and Corbett, spelling out the new Times process for approving the use of unnamed sources. The boldface passages are key points in the Times memo and were boldface in Sullivan’s post (and, I presume, in the email to the staff). Italic paragraphs are my comments on the memo:
The use of anonymous sources is sometimes crucial to our journalistic mission. But it also puts a strain on our most valuable and delicate asset: our trust with readers.
At best, granting anonymity allows us to reveal the atrocities of terror groups, government abuses or other situations where sources may risk their lives, freedom or careers by talking to us. In sensitive areas like national security reporting, it can be unavoidable. But in other cases, readers question whether anonymity allows unnamed people to skew a story in favor of their own agenda. In rare cases, we have published information from anonymous sources without enough questions or skepticism — and it has turned out to be wrong.
The use of anonymous sources presents the greatest risk in our most consequential, exclusive stories. But the appearance of anonymous sources in routine government and political stories, as well as many other enterprise and feature stories, also tests our credibility with readers. They routinely cite anonymous sources as one of their greatest concerns about The Times’s journalism.
After consulting with a number of our most experienced reporters and editors, we have decided to take several steps to raise the bar and provide added scrutiny for our use of anonymous sources. These new guidelines require top editors to approve the use of anonymity. But it is incumbent on everyone producing journalism throughout the newsroom to share the responsibility.
Our basic, longstanding criteria remain unchanged: Anonymity should be, as our stylebook entry says, “a last resort, for situations in which The Times could not otherwise publish information it considers newsworthy and reliable.” That standard should be taken seriously and applied rigorously. Material from anonymous sources should be “information,” not just spin or speculation. It should be “newsworthy,” not just color or embellishment. And it should be information we consider “reliable” — ideally because we have additional corroboration, or because we know that the source has first-hand, direct knowledge. Our level of skepticism should be high and our questions pointed. Without a named source, readers may see The Times as vouching for the information unequivocally — or, worse, as carrying water for someone else’s agenda. As far as possible, we should explain the source’s motivation and how he or she knows the information.
Buttry note: I applaud the standard that unnamed sources cited in stories should have “first-hand, direct knowledge.” Reporters can and should grant confidentiality to sources with second-hand knowledge in order to get interviews. But they should learn what the source knows first-hand and use those sources to reach first-hand sources, not cite them in stories.
We recognize that in today’s hypercompetitive news environment, the tighter guidelines below inevitably mean that we will occasionally be beaten on a story. We have no intention of reducing our urgency in getting news to our readers. But we are prepared to pay the price of losing an occasional scoop in order to protect our precious credibility.
Buttry note: This will be the true test of the Times’ commitment to the standards (or any newsroom’s commitment to reducing its reliance on unnamed sources): your willingness to be beaten. This is especially true in stories that are going to come out soon anyway, such as the nomination of a Supreme Court justice or the hiring or firing of a coach or public official. I want to be first on those stories, but it’s more important to be right than to be first, and if you only have the information from unnamed sources, sometimes you need to let someone else get that story first, or wait until it’s announced.
Update: As John Robinson noted on Twitter (below), the Times did use unnamed sources (“White House officials”) in its story on the Supreme Court nomination:
Moments before I got an anonymously sourced NYTimes alert on Obama's nominee, I got @stevebuttry's piece on the Times policy on sourcing.
— John Robinson (@johnrobinson) March 16, 2016
And back to the Times memo:
This is not an easy balance to strike, and these new guidelines may be just the starting point. We will review these steps in the coming months and make adjustments if necessary. For now, we want to adopt these new procedures, starting immediately in all departments:
Buttry note: I also applaud the plans to review and adjust the guidelines. Newsroom rules don’t need to be cast in stone, and getting them right is more important than sticking with a policy you worked hard on (as the Times clearly did here). Stories in the coming months will test this policy, and I’m pleased to see the Times planning to assess how it’s working and adjust as needed.
1. Special rules apply when the lead of a story — that is, the primary news element — is based entirely on one or more anonymous sources.
Buttry note: I like that the memo sets out different processes for different uses of unnamed stories. A story that’s entirely based on unnamed sources is a significantly different case from a story that quotes a single unnamed source in the body of the story about a single point. They should be handled differently, and the memo spells out different ways of handling them. Also, the “Susan” mentioned below is Deputy Executive Editor Susan Chira.
Any such story must be presented in advance by the relevant department head to Dean, Matt or Susan. They should be told explicitly why their approval is being sought — that is, the story’s main news element depends on anonymous sourcing. The department head should be prepared to discuss the details of the sourcing and other reporting, including the identity of the source, if asked.
This conversation or email exchange should not be part of a routine discussion of multiple stories. Sending a batch of summaries or simply passing along a copy without comment is not enough. This should be a dedicated conversation, focusing entirely on the sourcing issue of this one story.
Buttry note: Another thing I like about this: Requiring a conversation focused on the sourcing. No possibility of sneaking the approval through quickly in a conversation on other topics. And I like the points in the next paragraph, too.
If it sounds as though this will slow down the process — that’s part of the point. A story that hangs entirely on anonymous sourcing should always get special scrutiny. If, for any reason, you have not received specific approval, the story should be held.
On rare occasions when all three of those editors will be unavailable, they will designate Phil or another masthead editor to grant these approvals. A note on the story should specify which masthead editor approved the sourcing.
Buttry note: I presume the “note on the story” is an internal note, which should suffice, but I wouldn’t mind a note at the end of stories explaining the use of unnamed sources and why it was necessary to the story.
2. Every other use of anonymous sourcing anywhere in any story must be personally approved in advance by the department head or deputy.
A note on the story should indicate that the sourcing has been approved, and by whom. Slot editors, copy editors and producers should not publish a story with any anonymous sourcing that does not have a note indicating that the department head or deputy has approved the sourcing.
Buttry note: Last year, Sullivan and I both faulted a ridiculous use of unnamed sources in a Times story on awards-show red carpets. It’s hard to imagine such a quote getting past a department head, slot editors, copy editors and producers if the Times staff takes these guidelines seriously.
3. Direct quotes from anonymous sources will be allowed only in rare instances and with the approval of the department head or deputy.
Such quotes are generally used to add color — but by definition, merely adding color does not normally clear the bar of newsworthiness that justifies anonymous sourcing. If the substance of the quote is newsworthy, it can be paraphrased, and must be approved under the procedures above.
Sources who demand anonymity give up the opportunity to have their speculation or interpretation reflected in our stories, and such quotes will no longer be allowed except in the rare instances when the direct quote is pivotal to a story. Other exceptions might include ordinary individuals who are sharing personal details in difficult circumstances and whose voices are worth capturing — for instance, immigrants discussing their ordeal with smugglers, or patients sharing their medical histories. In all these cases, direct quotes from anonymous sources must be approved by the department head or deputy.
Buttry note: As I have noted before, personal stories that are private and sometimes painful are one of the best and most justifiable uses of unnamed sources. I would add survivors of sexual assault, domestic violence and other crimes to the examples cited of immigrants and patients.
4. As a reminder, it continues to be a hard-and-fast rule that at least one editor must know the specific identity of any anonymous source before publication.
Departments should set up regular procedures to make sure this rule is followed consistently.
I will note, but not belabor, my dislike of the phrase “anonymous sources” to describe sources known to the reporter (and that would be everyone covered by the Times policy, since reporters have to identify them to editors). To me, the phrase only applies to people unknown to the reporter (a caller who won’t identify herself or an email or social media post whose author is not identified and won’t identify himself). I have explained further in earlier posts, but feel the need to note here since the Times memo used the phrase extensively.
Congratulations to Margaret Sullivan
I am pleased that the Times adopted these new guidelines during Sullivan’s tenure. She has been a demanding and tireless critic of the Times on this issue, and performed an important service to the Times and journalism. She leaves soon to join the Washington Post as a media columnist. With my former TBD colleague Erik Wemple, she will give the Post two of the strongest voices (and best reporters) in media criticism.
I think Sullivan was hands-down the best Public Editor in Times history (and the others have been good). She will be difficult to replace.
Earlier posts on unnamed sources
I have blogged a lot here about unnamed sources. Here are earlier posts about the Times’ use of unnamed sources:
And here are general posts about using unnamed sources: