When To Trust A Story That Uses Unnamed Sources
By Perry Bacon Jr., FiveThirtyEight, Filed under The Media,
Published July 18, 2017
The various investigations into the Trump administration and its alleged ties to Russia are hard to follow. The allegations are sometimes muddled, the probes are still ongoing, and all sides in the dispute are leaking information that favors their points of view. These stories are also hard to follow because few officials are willing to put their names behind their claims and comments, leading to a stream of stories rife with unnamed sources.
What’s a reader to do? Well, here’s a guide to unnamed sources in government/politics/Washington stories — who they are, how reporters use them, and how to tell if you should trust what they say. Having covered Congress, the White House, several presidential campaigns and briefly the Education and State departments, I have begged (usually unsuccessfully) many sources to allow me to use their names, written a fair number of stories with unnamed sources, and spent a lot of time trying to decode stories with unnamed sources written by other journalists. For this piece, I also consulted other journalists and political types who have served in senior staff roles on campaigns, on Capitol Hill and in presidential administrations.
This is part one of two. I’ll cover some general principles for reading anonymously sourced stories here and break down the different types of such sources in part two. I wrote this piece because of all the Trump-Russia stories, but the rules, terms and designations apply to other Washington stories as well.
This is not a story meant to condone or encourage the use of unnamed sources. While President Trump and his defenders have bashed the use of anonymous sources, some journalists themselves also say the practice is overused. They argue that using unnamed sources limits journalistic accountability, since readers and other reporters can’t easily check the accuracy of an account if they don’t know where it comes from. Unnamed sources are often a feature of stories that I would argue are more about reporters showing how savvy and in the know they are than truly informing and enlightening readers.
But major investigative stories, both in Washington and outside of it, are often impossible to write without unnamed sources. The alternative to stories with unnamed sources is often not having the story published at all, rather than the same story with names. Sources have a wide range of motives for not going public. Some reasons are noble (whistleblowers may face retribution for leaking details to a reporter). Some are not (White House aides, both in the Trump administration and previous ones, sometimes don’t like one another and complain anonymously about their colleagues to the press).
Either way, there are many news outlets and often very few people who know the details of White House deliberations or the state of the Russia investigation. So the sources have the power to set the terms with the journalists, and one of those terms is often, “don’t use my name.”
5 tips for reading stories with unnamed sources
1. Multiple sources add up.
When an outlet says “six White House officials” or “seven Department of Justice officials,” it’s providing a level of precision that makes me more likely to trust the story. This does not necessarily mean that the story is correct. But it does suggest it was thoroughly reported.
A recent New York Times story, for example, described something top White House adviser Jared Kushner was saying in private meetings, according to “six West Wing aides.” Six people are less likely to be wrong than one — and this also indicates that the reporter was cautious and diligent enough to seek confirmation with more than one person. CNN’s recently retracted report that Congress was looking into a Russian investment firm with potential ties to people in Trumpworld, meanwhile, cited only a single anonymous source.
And, obviously, if multiple reputable news organizations or reporters report out and confirm the same story (not just link to another outlet’s reporting), that’s a reason to assume it is accurate. Conversely, if another major publication is casting doubts on a story you read that quotes lots of unnamed sources, that should heighten your skepticism. For example, In the Iraq War era, even as The New York Times and other outlets often echoed the Bush administration’s assertion that Iraq possessed large numbers of weapons of mass destruction, there were publications like McClatchy that were more skeptical of the government’s claims.
2. Unverifiable predictions are suspicious.
Trust a source who says something happened; distrust a source who says something might happen.
Axios and Politico, two publications targeted at political junkies, in particular often float “scoops” predicting that something will happen that never does. An April piece in Axios quoted “aides and advisers” to Trump who suggested that White House chief of staff Reince Priebus and chief strategist Steve Bannon could soon be pushed out by Trump, with House Majority Leader Kevin McCarthy potentially replacing Priebus. This would have been a huge shift of power on both Capitol Hill and in the White House and more than three months later has not happened. As early as February, Politico highlighted the potential of a broad shake-up of White House officials that included Priebus. Politico recently suggested that Priebus would be out of his job around July 4; that didn’t happen. The story was carefully hedged, of course, noting that Trump might not follow through on the idea of dumping his chief of staff.
I’m more dubious of stories that claim insider knowledge about future events, for three reasons.
First, they are almost impossible to disprove in any way. In the Priebus example, the reporter or news outlet (Politico) can always claim that Trump intended to fire his chief of staff around July 4 but then changed his mind.
A second concern, related to the first, is that the nebulous nature of these speculative stories creates an incentive for reporters to write them. If Trump had fired his chief of staff on July 3 or July 5, Politico would have looked very prescient. The firing did not happen, and the reporter can claim that Trump just didn’t follow through without suffering any loss of credibility. This kind of story “gives [the journalist] a provocative scoop that cannot be readily disproven, since it purports to reflect someone’s state of mind, which can always change, as opposed to an actual thing that has occurred,” said Brian Fallon, a Democrat who has served in top communications roles on Capitol Hill, as well as for the Department of Justice and Hillary Clinton’s 2016 presidential campaign.
Stories like these often get tons of buzz and attention, but reporters (both those who write these pieces and those who read them) know these stories are often just speculation. I’m not sure readers do.
Thirdly, sources have an incentive to encourage these kinds of speculative stories. If you are someone in the White House who does not like Priebus or you want to take his job, anonymously leaking that Trump is considering replacing Priebus is a great tactic. Trump did not publicly commit to keeping Priebus on, and now you (the Priebus rival) have put in the minds of everyone in Washington that the White House chief of staff is on notice. “In instances like these, the anonymous report serves the source’s interests,” Fallon said. “It allows them to float a trial balloon without being accountable for it.”
On the other hand, I would be more likely to trust a piece reporting, even with unnamed sources, that Trump was considering sending 20,000 troops to Syria. Why? Because the stakes for this claim are much higher. Staffers leave administrations all the time, while the U.S. does not as regularly deploy thousands of troops abroad.
3. Specifics matter.
What information does the story give you about its sources? The more, the better. For example, trust “Department of Justice officials” more than “administration officials.” If a story includes claims from unnamed officials from the Justice Department, those claims are typically run by the department’s press office. I would interpret a story sourced to “Department of Justice officials” without a denial from the press team there to be accurate — and perhaps even leaked by the department’s press team itself. An “administration official,” on the other hand, covers a much bigger group of people with disparate interests and points of view. It’s easy for other reporters to call the Justice Department and verify the story, while it’s much harder to confirm a story attributed to administration officials, which could mean any agency or the White House.
Then there are the descriptions of anonymous sources that essentially tell you nothing. A recent Washington Post story cited “U.S. officials briefed on intelligence reports” — that could be almost anyone. Or, worse still: “people familiar with the investigation.” Broadly speaking, you, dear reader, are “familiar with the investigation” — you’re reading about it after all.
4. Consider the outlet and the reporters.
If, say, Nate Silver, Harry Enten and I co-write a story with unnamed sources about Hillary Clinton’s campaign decisions in 2016, there are reasons for readers to trust that story. All three of us have long records covering electoral politics. If the three of us wrote an article claiming that Kushner had a secret meeting with a Russian oligarch, full of unnamed sources, you should be more skeptical, since we are not regularly breaking news about Kushner’s activities.
There are valid reasons to question the practices of the “establishment media,” but at least for now, I’m more comfortable with stories using unnamed sources, particularly about major national security or intelligence issues, that come from outlets and reporters who have a history of covering these issues, such as The Wall Street Journal, The New York Times and The Washington Post. (And, yes, I’m fully aware of the big blunders of the major papers and networks, such as CNN’s bungled Russia report.)
“Certain reporters have well-deserved reputations for being careful and not getting it wrong,” Fallon said. “Think Pete Williams at NBC in a breaking news situation when there is a major law-enforcement story.”
The big outlets also have another advantage in terms of unnamed sources: Important people come to them. If a tiny blog or a reporter you’ve never heard of breaks a story on some kind of major White House policy shift, one reason to be skeptical is that most administrations would rather leak big news to the Times or the Post than a more obscure publication.
“Top aides on campaigns or highly connected sources inside the law-enforcement community are just not spending a lot of time trusting some partisan blog or smaller digital outlet with sensitive information,” said Kevin Madden, a Republican who has served in senior press roles on Capitol Hill, in the Justice Department and on Mitt Romney’s presidential campaigns.
5. Watch for vague or imprecise “denials” of these kinds of stories. That often means they are accurate.
Another thing to make you trust a story: When an official spokesperson offers a “denial” that really isn’t a denial. Remember when the Post published a story in May, vaguely attributed to “current and former U.S. officials,” suggesting that Trump had disclosed “highly classified information” in a meeting with Russian officials? Responding to the story, national security adviser H.R. McMaster told the paper, “At no time were any intelligence sources or methods discussed, and no military operations were disclosed that were not already known publicly.” But the story had not actually claimed that the president had disclosed sources, methods or operations — only information, which McMaster did not deny. (Trump essentially confirmed that he had disclosed the information soon after the story ran.)
“If the person implicated in the report is unable to outright deny it, that’s a sign it can be trusted, even if the sources are anonymous,” Fallon said.
In conclusion, we think you should continue to read stories with unnamed sources, but carefully and cautiously. Even major outlets like CNN and The New York Times occasionally get things wrong when relying on unnamed sources. On the other hand, this article and its follow-up should help you understand why everyone in Washington knew that in February, then-national security adviser Michael Flynn was in deep trouble. He was accused of something that either happened or did not — a factual claim (talking on the phone with the Russian ambassador to the U.S. and discussing sanctions imposed by the U.S. against Russia) — in a story in a traditionally reliable outlet (The Washington Post) that was written by reporters known for covering national security and intelligence issues (Greg Miller, Adam Entous and Ellen Nakashima), with multiple unnamed sources making the claim (“nine current and former officials”).
A Flynn spokesman, asked to comment on the story, told the Post that Flynn “indicated that while he had no recollection of discussing sanctions, he couldn’t be certain that the topic never came up.” That response was well short of, “no, sanctions were not discussed.”
Flynn resigned from his job within a week.
Perry Bacon Jr. is a senior writer for FiveThirtyEight.