Anonymous sourcing problems at the New York Times

There are at least three major failures of the modern press worth writing about. The first is excessive coverage of trivial events of little or no public interest. Sex scandals, celebrity lifestyles, and manufactured scandals come to mind as examples. The second is the false choice (and often wrong choice) between balance and objectivity.

The third failure is the excessive use of anonymous sources. Glenn Greenwald wrote about this in early 2010 in a post worth reading every year to remind us of what happens in democracies when the guardians of public interest forget what their jobs are. No party outside of the government was more helpful in manufacturing public support for the invasion and occupation of Iraq than the pillars of print journalism, precisely because of the trust we have in them.

Clark Hoyt, the Times’ public editor in 2009, blasted his own paper for continuing to shield government officials from public scrutiny in violation of the Times’ “stylebook” section on anonymous sources. Margaret Sullivan, public editor since 2012, did it again on Tuesday, noting the indefensible use of anonymous sources even in stories that have nothing to do with the government.

Sullivan introduced a regular feature of her column where notable failures to abide by the Times’ policies would be called out in the hope that these practices will change. In supporting this effort, I’ve decided to go a little bit further than that by examining all failures across a set period of time.


I asked the Times if I could see the section of the stylebook on anonymous sources so that I’m not just making educated guesses here, and they kindly and quickly obliged. I wasn’t told if that text can be publicly shared and I didn’t think to ask, so for now you’ll have to ask them for it if you want to see it for yourself.

With the policy in hand, I boiled it down to a series of questions I can ask about each use of an anonymous quote or citation:

  1. Is the information already available from a public source?
  2. Was the source’s job or access to the information described in detail?
  3. Was it explained why the source wouldn’t speak without anonymity?
  4. Was it a blind attribution? (“experts said”, “sources said”)
  5. Was a press officer’s identity shielded?
  6. Was the comment quoted or paraphrased trivial?
  7. Was it a personal or partisan attack?

I took some liberty with the final question because the Times new policy makes no mention of partisan attacks, although the previous policy did. Not all of these questions will apply to every anonymous quote or attribution which rules out a simple scale, so instead I’ve grouped stories together by the types of violations they contain.

For this post I had intended to examine every front page story published on the New York Times website every day for an undecided length of time, but a lot of this work is redundant. Instead, I read through 26 stories published between 9:25 PM on March 18th, and 12:41 PM on March 19th. All stories appeared on the front page of the website and were not picked for any other reason.

Blind Attribution

Of the 26 stories I examined, more than half (14) had information from anonymous or unnamed sources in them in violation of the stylebook. The most common offense felt like blind attributions, such as this one in a story about the Illinois gubernatorial GOP primary:

The state’s financial struggle is certain to emerge as a central theme in the coming campaign. Illinois’s credit ratings sank as it was late paying bills and as experts said the state pension system had become underfunded by about $100 billion — among the worst in the nation.

Times’ writers like to use “experts said” a lot in their stories, even though the stylebook says to describe the person(s) in as much detail as possible to give readers some idea whether or not the experts know what they are talking about. In this case we don’t even know which field these experts exist in, much less what their personal reputations are, or what kind of education and experience they have.

Are they experts in private sector financial management from Wall Street? Are they academic experts on public pension funds? Have they advocated a change in political policy, and thus have a stake?

This specific use of a blind attribution was especially puzzling. While researching the claim that the Illinois state pension system was underfunded by $100 billion — like many readers I’m sure, I was curious where this figure came from, and the article doesn’t say — I found another reference to it from the same writer in an article she wrote four months ago:

[Governor Pat Quinn] said, without providing more details, that the deal met a crucial standard of eliminating the unfunded debt and fully stabilizing the system, which has an estimated $100 billion in unfunded liability.

The writer, Monica Davey, used blind attribution twice more in the story, once where “experts” say the Illinois gubernatorial election this fall will be “deeply competitive”, and again to say that Governor Quinn defied the expectations of some experts when he won the office in 2009. Because we’re not given details about these experts, it’s difficult just to tell them apart.

Are the experts predicting a tight race this fall also the same people that wrongly thought Pat Quinn would lose in 2009? Are they experts in predicting elections nationwide, or just in Illinois? Do they have a track record of correct predictions? What qualifies them as experts in the first place?

How can the times expect readers to place their trust in these experts — and the story based on their expertise — when we can’t even tell them apart? Why was it necessary to paraphrase the experts instead of printing exact quotes, and why weren’t any experts named in the story?

I don’t mean to pick on Monica Davey here; I don’t know anything about her other than that she wrote this story. She may be an outstanding good journalist, that’s not the point. The problem at the Times seems to be cultural and systemic, not just with print journalists but also their editors, who allow these stylebook violations and shoddy stories to go to press. And with no reasonable excuses for any of it.

Blind attributions were also used in a story about Malaysian Flight 370. The story has been rewritten since I first read it, changing the focus away from the Malaysian government asking the Federal Bureau of Investigation for assistance, to possible plane wreckage discovered floating 2500 kilometers off the coast of Australia in the southern ocean. But the blind attribution remains:

Investigators have said the plane’s extraordinary diversion from its intended course was probably carried out by someone who had aviation experience. The Malaysian police, who found that Mr. Zaharie had built a flight simulator at his home, said Wednesday that some data had been erased from the simulator on Feb. 3, more than a month before the ill-fated flight.

The problems here are two-fold. The previous paragraph mentions both the United States and Malaysian governments, and without further details about who the “investigators” are, it’s unclear which country they belong to. Not that one can assume it’s one of those two, because China, Australia, and many other countries are involved in the search for Flight 370.

Are these investigators the National Transportation Safety Board from the US, or the Malaysian military? Are these people actively involved in the search for Flight 370, or are they retired investigators with decades of experience in this area? Are these privately run parallel investigations or with official ties? What is it about these investigators that they were consulted by the Times for this story, and other investigators who may disagree with them were not?

It’s hard enough keeping everything in your head on the subject of Flight 370 with media saturation flooding our senses, shouldn’t the New York Times and other respected outlets be clarifying the situation instead of muddying it?

There have been dozens if not hundreds of “experts” and retired investigators appearing on cable TV over the past two weeks offering every possibly scenario they could imagine, and every single one did so with their name attached to their theories and analysis. I see nothing notable about what these experts said to the Times that warranted paraphrasing and blind attribution, and I don’t think anyone was going to get fired or have their personal safety endangered because of what they said.

Anonymity should be reserved for information important to the public from sources that might suffer serious consequences for revealing that information. In light of that, the more trivial the experts’ comments are, the less defensible the blind attributions become.

Sensitivity isn’t enough; Triviality

A story in the Times about Toyota agreeing to a $1.2 billion fine to settle a criminal investigation by the US government is even worse, using a blind attribution for a comment as trivial and obvious as you can imagine: “Toyota is now eager to also put its legal woes behind it, analysts say.

In addition to urging against blind attributions, the stylebook forbids anonymous quotes for trivial information. It doesn’t get any more trivial than that.

Quoting directly from the New York Times stylebook on anonymous sources:

It provides little insight merely to say that a source “was not authorized” to speak, or sought anonymity because the topic was “sensitive.”

And this from the previously mentioned story about Malaysia Flight 370:

To speed its efforts, the F.B.I. will probably make copies of the simulator’s hard drive and have its contents digitally relayed back to agents and analysts in the United States who specialize in retrieving deleted computer files. “Right now, it’s the best chance we have of finding something,” said a senior law enforcement official, who spoke on the condition of anonymity because of the sensitivity of the investigation. Unless the pilot used extremely sophisticated technology to erase files, the F.B.I. is likely to be able to retrieve them.

On-going law enforcement investigations are tricky to handle. I imagine that police and federal law enforcement officers are forbidden from speaking to the press in order to protect the integrity of their investigations. Yet few things stir the imaginations and capture the attention of the public more than criminal and accident investigations. One almost expects insider information on criminal investigations to come from anonymous sources who could face criminal charges themselves for revealing privileged information.

That said, the New York Times does provide guidance on how to balance the needs of their sources to be anonymous and the needs of their readers to know as much as possible about those sources, in order to gauge their credibility. If possible, it should be explained why the person required anonymity, what their access to the information is, and what their interests are (are they biased?). And if that’s not possible, the need for absolute anonymity should be stated clearly in the article.

That basically never happens, based on the 26 stories I examined. The ambiguity of “because they needed it” actually occurred twice in the story about Flight 370.

Here is another example:

The United States has employed its constellation of spy satellites in the search since its earliest stages, and is now using the satellites’ ability to capture high-resolution images to help narrow down the search area, a senior American military official said. […] the official, who spoke on the condition of anonymity because of the sensitivity of the investigation.

It’s public knowledge that several countries — including the United States — are using military intelligence assets to search for Flight 370. China publicly released spy satellite photographs a few days ago, as did Australia.

That paragraph is a real whopper, first because it’s publicly available information which means it should never come from an anonymous source. Second, because it used the trite “because of the sensitivity of the investigation” cop-out. Third, because the official wasn’t described in any meaningful way (which may not have been possible), and lastly because no legitimate reason for the anonymity was given.

In a story about the death of fashion designer L’Wren Scott, the Times quoted anonymous law enforcement officials to gain access to (what may or may not be) a first person account of (what may or may not be) a crime scene. In this case, it’s “investigators”:

Ms. Scott texted her assistant at 8:30 a.m. on Monday with instructions to come by her apartment in Chelsea, according to a law enforcement official who spoke on the condition of anonymity. The assistant arrived around 10 a.m. and found Ms. Scott hanged with a scarf attached to a door in the living room area, the official said.

Investigators did not find a suicide note, and there did not appear to be anything in that final text message that would have suggested that Ms. Scott planned to kill herself, the official said.

The violations in these two paragraphs were minor enough that it’s puzzling why they were allowed by the editor. The stylebook requires writers to explain as best they can why a person is being given anonymity, but that didn’t happen here. It could have been taken care of with a simple disclaimer: “The official spoke on the condition of anonymity because department policy forbids non-spokespeople from revealing details of an on-going investigation to the press or public.

It may not be possible to reveal any more details about the source than that without compromising their identity, but such ambiguity leaves a lot of questions unanswered about the credibility of the source, and thus the value of the story to readers.

Was this official personally inside Scott’s apartment when the body was found, or is this a third person recounting of events? Is this official a seasoned detective with years of experience, or a rookie who doesn’t necessarily understand what they were looking at? Is this a police department official who understands how to view a crime scene, or someone from the District Attorney’s office who has never seen one in their life until now?

Including details about the anonymous law enforcement official would have added credibility to a story that really needs it in order to be useful.

Anonymous Spokeswoman

Disclaimer: I follow Julie Bosman on Twitter, but otherwise have no contact with her and don’t know her personally or professionally.

Julie Bosman wrote a story on a new book publishing deal for John LeFevre. LeFevre had pretended to be an anonymous Goldman Sachs insider dishing out secrets about his employer on Twitter, but was outed has having never worked for the company which subsequently cost him a book deal with Simon & Schuster earlier this year.

Bosman wrote:

Grove Atlantic, the independent publishing house that has built a reputation for quality literary fiction, said on Wednesday that it had acquired “Straight to Hell: True Tales of Deviance and Excess in the World of Investment Banking,” a book that drew from the Twitter account that was purportedly based on conversations overheard in the Goldman Sachs elevator.

The publisher agreed to pay Mr. LeFevre a six-figure advance, a spokeswoman said. The book will be released in November.

The stylebook forbids anonymous official spokespersons:

Anonymity should not shield a press officer whose job is to be publicly accountable.

It’s hard to square Bosman giving anonymity to a person whose job it is to say what they are saying, with the official policies of the New York Times on anonymous sources. Such can be said of almost every anonymous quote and blind attribution that appeared on the Times front page that day.

Most violations seemed innocuous and lazy. A good example of that comes from a story about a recently released report documenting security failures at Los Angeles International Airport that resulted in the fatal shooting of one Transportation Security Administration officer and the wounding of others:

A copy of the report was presented to airport commissioners on Tuesday, and officials said they would respond quickly to the criticism. In particular, commissioners said, more emergency training programs would be developed for airport employees. Union officials have said that employees have not been properly trained to deal with emergencies and were essentially relying on their own judgment as passengers panicked in the airport.

Although none of the violations are serious, there are still numerous violations of the Times’ policy on anonymous sources in this paragraph. The airport commissioner “officials” should have been named because their comments were so trivial that they were paraphrased. Given their position and what they said to the Times, it’s unreasonable to expect that the officials feared for their jobs by stating that they would “respond quickly” to criticism.

The statements from “Union officials” were critical of airport security, which is of interest to the public. But again, there doesn’t seem to be any need for those officials to be anonymous as they’ve likely already made similar comments publicly. It’s also doubtful that those officials would fear for their jobs when they are accountable only to their union membership, and their paraphrased claims are pushing an issue important to their members and to their benefit.

Neither source were described in any meaningful detail and there was no explanation why any the sources required anonymity or were not named. While naming these officials may not make the story any more credible than it already is, that single paragraph violated four of the five policies that apply to this story.

All told, I found problems in 14 of the 26 front-page stories I read that day, which I think may have accounted for not even half the content produced. Some violations of the anonymous source policy were more serious than others, to the point of confusing the reader. It’s not anonymous government officials talking about aluminum tubes and uranium yellowcake in Iraq, but it’s not nothing either.

Perhaps it’s worse. Arguments can be made that some information is too valuable to obtain to worry about the internal policies on anonymous sourcing, but handing anonymity out like candy in trivial situations has no defense at all.

If the New York Times allows anonymous sourcing for trivial stories of no national consequence on a regular basis — and they provably do — it’s almost certain that they’re still doing it with important topics like national elections and foreign conflicts when readers need that information the most.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.