Anonymous Sources

This quote from a 1960 journalism textbook makes it clear that the anonymous sourcing wasn’t always an accepted practice in journalism. Famed journalist John Hohenberg said that in the recent past,

[e]ditors generally insisted that the sources of the news must be identified by name, whether or not they could be quoted directly. The presence of an anonymous figure, who could not be described in any way except in relation to what he represented, was almost an affront to many reporters and editors. (p. 227)

He lamented that anonymity was becoming more commonplace.

Hohenberg noted that journalists were using it routinely to cover pedestrian beats like city hall and the courts. “The practice has been accepted,” he wrote. “The press has limited itself” (p. 228).

The executive editor for the Raleigh News & Observer took a novel approach to a White House request for anonymity:

When U.S. Treasury Secretary Timothy Geithner visited recently, his staff initially requested that our interview be on background. We would not have been able to attribute his comments to him.

We insisted the conversation be on the record. Geithner agreed. We published a front-page article in which Geithner defended the Obama administration’s efforts to stimulate the economy.

Great job. Perhaps Washington correspondents should “just say no” a little more often. Some research has shown that many unnamed sources would go on the record if the reporters simply insisted.

Ben Bradlee, the former Washington Post editor, tried to put the kabash on “deep background” sessions in the 1970s. He told his reporters to walk out of the room whenever a White House official refused to go on-the-record. After a few days of getting scooped by the New York Times, Bradlee rescinded the ban.

Any reduction in unnamed sourcing can’t just be unilateral — either all the media outlets buy in, or it just won’t work.

Sources: Gassaway, B. M. (1988). Are Secret Sources In the News Media Really Necessary? Newspaper Research Journal, 9(3), 69-77; Bagdikian, B. H. (2005). When the Post Banned Anonymous Sources. American Journalism Review, 27(4), 33.

After buying BusinessWeek, Bloomberg’s editor took issue with the magazine’s lackadaisical attitude toward unnamed sourcing. Editor-in-chief Matthew Winkler issued the following memo to explain what’s wrong with the magazine’s approach:

News engenders the most trust and is most actionable when attribution for assertions and quotations is transparent.

Bloomberg News, since its inception, uses anonymous sourcing reluctantly only when the benefit of reporting something that is actionable outweighs the lack of definitive attribution.

We shun anonymous quotations and assertions that are negative because readers have no proof that they are more credible than hearsay. In an age when news increasingly is asserted, we must be fact-driven. For every assertion, there should be an example or anecdote and comment from a recognized authority.

These examples from Bloomberg BusinessWeek, some written by Bloomberg News reporters, don’t meet these standards.

Unattributed quotations:

Meeting with him, says one executive, was “a sphincter-tightening experience.”

Several GM executives say the uncertainty is hurting morale. “It’s almost like experience is a liability these days,” one says.

((Each instance allows the accuser to hide without giving a sense of their position, experience or relevance to the news.))

Imprecise or opaque sourcing:

Particularly worrisome to GM veterans are the actions of private equity executives Steve Girsky, David Bonderman and Daniel Akerson.

GM has long needed an independent board, but GM insiders fear that too much second-guessing could prevent the company from sticking to a clear strategy.

GM insiders complain that the trio has already brought to bear the private-equity obsession with cost cutting at a company that badly needs to deploy its government-financed cash hoard to develop new cars and technology.

((What’s a veteran or an insider? This is written for the sources, not for the readers.))

In another move, Bloomberg BusinessWeek has learned, Best Buy plans to launch its own advertising business early next year.

((How do we know this?))

This game of political chicken, observers say, is likely to produce only partial victory, with the Fed hanging onto most, but not all, of its vast power.

((What’s an observer? It’s part of the vast Greek chorus, without any explanation why that opinion matters.))

Abu Dhabi may go after some pieces in exchange for bailout money, say analysts.

While many analysts fear that Citi in particular isn’t ready to fly solo, the White House gave its blessing.”

((Not all analysts are created equal. Some are more insightful or accurate. Providing names and credentials would make their points more credible.))

More broadly, governance experts worry the new board is overreaching.

((While this sentence sets up the story, no governance experts are quoted as worrying.))

Pandit’s hoping to hash out a plan in the next week or so, say people familiar with the situation.

((Adding the number of people and how they are aware of the plans would make this more credible. The most accurate people are those taking part in the action.))

These quotes add little that would require anonymity and would gain credibility by adding the person’s name and credentials:

“We were dark for a very long time,” says a senior executive who is helping find a new CMO.

“Why would I buy AOL?” asks a large media investor. “It would largely be a bet on Tim, given what he was able to do at Google.”

These attributions meet our standards:

The board batted the proposal back, say three people familiar with the meeting.

One of the three private equity directors asked if GM could buy an engine from someone else, say two people with knowledge of the meeting.

For more on sourcing and attribution, see the Ethics chapter in the Bloomberg Way.

Good stuff. Note that he doesn’t think unnamed sources should be avoided altogether. Rather, he insists that reporters take their credibility seriously when using veiled attribution. Note also that both of the instances that are acceptable involve at least two independent sources.

A Web developer has created a clever anonymous source tracker. It looks for various phrases (e.g., “a source close to” or “a person familiar with”) and posts the offending passage. It also tallies the number of unnamed sourcing in various news outlets. The No. 1 offender? BusinessWeek. That outlet has used unnamed sourcing 100 times in the past few days — fully three times more than the next offender, Reuters, with 31 instances. The numbers update continuously, so it will be interesting to watch them over time. Impressive bit of software coding.

Great column by Glenn Greenwald about how unnamed sourcing often turns out to be wrong:

Aside from falsity — and the fact that they become irreversibly lodged in our political culture as fact — what do all of these deceitful reports have in common? They’re all the by-product of granting anonymity to people and then repeating what they claim as fact, with the falsehood-disseminators protected by “journalists” from any and all accountability for their falsehoods. It’s exactly the same process that caused our leading media outlets to tell Americans about Iraq’s massive WMD program and Al Qaeda connections; Jessica Lynch’s heroic firefight with inhumane Iraqi devils and her “rescue” by our Marines; Pat Tillman’s death at the hands of Al Qaeda monsters; and government tests that “confirmed” the presence of bentonite in the anthrax used to attack the U.S., which meant it was likely that Saddam was behind the attacks.

Unjustified anonymity — especially when mindlessly repeating what shielded government sources claim in secret — is the single greatest enabler of false and deceitful “reporting.” Despite (or, really, because of) its unparalelled record of producing lies, it will never stop, because agreeing to it is how “journalists” end up being selected as favored message-carrying servants for the powerful. This falsehood-producing method isn’t ancillary to American journalism but central to it; the book which is occupying the attention of America’s political and media class is based exclusively on unattributed, shielded sources, and that seems to bother none of them.

The problem is determining when anonymity is justified or unjustified. Every journalist thinks his anonymously sourced report is totally justified — but clearly they can’t all be.

Good piece in Foreign Affairs about the long history of news leaks. The author would like to see less of them:

Some leaks are meant to be a straightforward pitch for or against a public policy. Leakers often pass along closely held documents, like that “confidential” McChrystal assessment of troop needs for Afghanistan, which found its way into Washington Post journalist Bob Woodward’s eager hands. Other leaks are meant to settle grudges. But most simply satisfy the leaker’s sense of self-importance. Washington is full of people who love to pass along a good tale.

Yet annoying as they may be, it is rarely worth the effort to plug the leaks. Reagan learned that the hard way after the leaks in his administration turned into a flood. “Reagan Ordered Sweeping FBI Probe of Staff for Source of Leak,” the Washington Post’s front page read on Nov. 23, 1983. But just a few weeks later on Dec. 13, the Post reported, “Justice Probe Fails to Disclose Source of Leak.” What went wrong? Turns out, many presidents are more leaked for than about. The latter piece quotes a White House official noting, “there is no evidence that reporters were told anything we didn’t want them to know.”

Some go as far as to say that leaks are beneficial. Political scientist Richard Neustadt, an advisor to John F. Kennedy’s and Lyndon B. Johnson’s administrations, elevated leaks to “a vital role in the functioning of our democracy.” After all, aren’t these “informal communications” through the media a faster, more nuanced way for the secretary of state to let the secretary of defense know what’s really bugging her? It’s certainly simpler than negotiating through the bureaucracy.

I’d love to buy this, but I can’t. My own case studies show leaks to be too episodic and subject to multiple interpretations to be a useful way of communicating from one agency or individual to another. Still, we do learn things that we might not have otherwise known (and there are times when what we imagine are leaks are merely clever reporters carefully putting the various pieces of a story together).

The good news for presidents is that there is little evidence that leaks have endangered U.S. national security. Mostly, they cause embarrassment. Even Kissinger, not a man to shrug off embarrassment, ultimately concluded, “Most of the leaks — if you are philosophical about it — go away. I mean, they’re unpleasant, but so what? If you ignore them, most of them are not of that huge significance.” Obama must surely hope as much.

The Associated Press Stylebook offers concrete guidance on anonymous source use. Surprisingly, the stylebook did not dedicate an entry on the subject until 2004. That year’s entry reads:

Use anonymous attribution only when essential and even then provide the most specific possible identification of the source. Simply quoting “a source,” unmodified, is almost always prohibited. Do not attribute information to sources – anonymous or otherwise – when it is obvious, common sense or well-known. The basic guidelines for use of anonymous sources: The material must not be available except under the conditions of anonymity imposed by the source. In some cases, it may be appropriate to say why the source requested anonymity. The source must be in a position to have accurate information, and, to the best of the reporter’s ability to determine, must be understood to be reliable. Be sure to seek more than one source for the story. (p. 17)

⁠
The AP entry offers specific guidance on normative journalistic behavior. The entry suggests reporters only use anonymous attribution when “essential,” although it doesn’t offer any criteria for determining such a condition. The entry prohibits the anonymous attribution of “common sense” information, although few critics complained that such reporting was part of the problem. The entry is also riddled with room for journalistic rationalization. Reporters should “almost always” avoid anonymous attribution without any description of the source’s identity. Journalists should “seek” a second source, but the entry doesn’t prohibit reporting if the journalist can’t find one. The book offers normative guidelines, but they don’t seem strict enough to affect the decision-making process.

The 2009 Associated Press Stylebook revised the entry substantially with changes removing some of the wiggle room from the previous entry, although in one case the standards were loosened. It reads:

Whenever possible, we pursue information on the record. When a source insists on background of off-the-record ground rules, we must adhere to a strict set of guidelines. Under AP’s rules material from anonymous sources may be used only if: [a] The material is information and is not opinion or speculation, and is vital to the news report. [b] The information is not available except under the conditions of anonymity imposed by the source. [c] The source is reliable, and in a position to have accurate information. Reporters who intend to use material from anonymous sources must get approval from their news managers. Explain in the story why the sources requested anonymity. And, when it’s relevant, describe the source’s motive for disclosing the information. The story also must provide attribution that establishes the source’s credibility; simply quoting “a source” is not allowed. Be as descriptive as possible.

In a tightening of protocol, reporters are now told they must explain why the source requested anonymity instead of reserving that action for “some cases.” Journalists are also instructed to consider disclosing the motive for the source’s actions, a recommendation left out of the 2004 version. Reporters must now always provide a description of the source to help with credibility not just in some cases. And reporters must also seek the approval of their boss to use anonymous information.

These changes represent palpable, inflexible restrictions on anonymous source use. However, in one important respect, the AP loosened the rules. The suggestion to seek a second source has been removed entirely. The journalistic tradition of verifying anonymous information is no longer required by the Associated Press.

Here’s the best argument I’ve read for why we should just shut up and accept anonymously sourced trade rumors in sports:

Now, obvious problem #1 is that, if that were to truly be adopted, we’d have no trade rumors ever reported. We’d have a lot less reporting on evaluation of players. There would be much, much less information out there for sports fans.

Would that be a good thing, really? If we eliminated everything but what Ambinder classifies as “On the Record,” then not only are we losing the “an A.L. front office exec says the Blue Jays want the cloning rights to Albert Pujols in exchange for Roy Halladay” stuff, but also more nuanced stories. When Evan Grant or T.R. Sullivan writes, “The Rangers are interested in Jermaine Dye,” there is no specific source that is named or alluded to…but at the end of the day, Grant or Sullivan is relying on anonymous sources just as much as the Pujols clone story guy. And that eliminates much of the value that the beat reporters can offer the fan.

Still, it seems like that these “softer” unsourced reports that clearly rely on sources cause less controversy, and I doubt that even the most dogmatic of the “don’t use anonymous sources” folks really want Jeff Wilson to avoid saying in a column that, for example, the Rangers are still looking for catching help unless he states that he got that information directly from Daniels or Thad Levine or Nolan Ryan or whomever. It is when the information is “sourced” in an indirect way that folks get bent out of shape.

But why? What harm is there in, for example, a writer quoting “an N.L. West scout” as saying that Martin Perez has the best fastball in the Texas League? And as a corollary, why can’t people understand why people quoted anonymously in these sorts of articles can’t go on the record? Scouts are going to get fired if they go on the record with their evaluations, front office types can be fined for tampering and probably don’t want their individual opinions to be public knowledge, and expecting A.J. Preller or whomever to be willing to have his name used in connection with some tidbit like that is unrealistic.

So what is really acceptable on these sorts of sourcing issues? To me, it is just a matter of knowing who the writer in question is, and judging the reliability on that…because at the end of the day, a writer who runs with bad info is ultimately going to lose credibility.

But, here’s the rub — can anyone really keep up with who’s information is good and who’s information is bad? All I know is that a lot of times the stuff attributed to anonymous sources on sports pages turn out to be wrong.

Prominent reports based on anonymous sources have sometimes proven to be incorrect:

Much of the O.J. Simpson reporting from unnamed sources was later deemed inaccurate.^[2]

Newsweek retracted a story about a Qur’an being flushed down a toilet that led to riots in the Middle East; the Qur’an desecration controversy of 2005 was based upon one unnamed military source.^[3]

The L.A. Times retracted an article that implicated Sean “Diddy” Combs in the beating of Tupac Shakur.^[4]

The original article was based on documents and a large assortment of unnamed sources. When reporting on the original story, the Associated Press noted that “[n]one of the sources was named.”^[5]

– I wrote this portion of the Wikipedia entry on Journalism Sourcing

The New York Times Public Editor points out that the Times doesn’t always follow its internal guidelines on not publishing anonymous personal allegations. If the subject is accused of a crime, the chances are far greater:

It was no surprise that the case against Robert Joel Halderman, accused of blackmailing David Letterman, involved money and sex, a former colleague at CBS said. Halderman “lived on the edge” and had “a bit of a checkered love life.”

Was it proper for The Times to report such statements from people who would not stand behind them? The paper has a policy that says anonymous sources should not be used lightly, but as a last resort, and should not be allowed to engage in personal attack or speculation. These tidbits seemed at best like gossip and at worst unfair suggestions of motive or guilt.

Reece Pendleton, a reader from Chicago, asked “why on earth” The Times would grant anonymity to someone making “prejudicial comments — nothing more than smears” about Clark, who had been identified as a “person of interest” but had not yet been charged. The paper just wanted to “spice up” the story, he thought. Ellen Comisar, a former Times graphics editor who lives in Rochester, N.Y., wondered “what possible justification could there have been” for the anonymous comments about Halderman.

It is a common scenario: A sensational crime is committed or alleged, and reporters — nine of them on the Annie Le story — scramble to learn everything they can about a suspect. They aim for a balanced portrait, but readers can take comments that a person was hard to get along with or “lived on the edge” as not-so-subtle clues to the question they really wonder about: Did he do it?

When the comments are anonymous, the potential unfairness can be compounded because readers have no way to evaluate the motives or credibility of those doing the talking. Pendleton, for example, wondered if the remarks about Clark could have been payback by a co-worker for a past run-in.

One NY Times reporter and editor defended their anonymous allegation by arguing they were doing the reader a service in conveying this information. It’s an understandable argument.

But, I would ask whether the paper gains more in releasing that anonymous information or adhering to its own guidelines — rules that were created so that the paper would gain credibility in its fair treatment of all the people it covers.

I’ll go with the latter.