Updated: NPR's Two-Ways: A Practice That Should Be Reconsidered

Every so often, another news organization will break a story that NPR isn't able to match by air-time. So an NPR news show invites the reporter who did the story to talk on air about the piece. It's known as a "two-way."

The journalist is paid about $100 or $150 depending on the show.

It's an unusual practice. You'd never find the Washington Post using a New York Times reporter to tell the news to Post readers. Nor would any television network do this. It's hard to imagine NBC interviewing ABC's Brian Ross about an investigative scoop.

But NPR has done two-ways with journalists from other news organizations for decades. The practice started in the early 1970s when NPR was a fledgling news operation and didn't have nearly the reporter power it has today.

"When a story happens with another newspaper or broadcaster, we try to match the story," said Christopher Turpin, All Things Considered's executive producer. "But given our tight 4 p.m. deadline, that can be a challenge and so we often go to the reporters who broke the story. We're the original aggregator. We have no problem with showcasing great work done by other organizations."

While this kind of two-way largely works well, it also has the potential of backfiring. That happened recently when ATC host Michele Norris interviewed Wall Street Journal reporter Mike Ramsey about his July 14 story on the latest news about Toyota's problems.

Read More

Ramsey's story – based on anonymous sources – discussed early findings from data recorders in Toyota cars involved in accidents. The findings, wrote Ramsey, showed those accidents were caused by driver error and not by sudden acceleration problems with the vehicles.

The ATC interview, which gave some highlights from Ramsey's story, brought a quick complaint from the National Highway Traffic Safety Administration (NHTSA), which is investigating 3,000 complaints alleging unintended acceleration in Toyotas since 2000. Half came after Toyota's mass car recall last October.

"The Journal story read like it was official government data. It wasn't," Olivia Alair, NHSTA's press secretary told me.

In the ATC interview, Ramsey told Norris "several dozen" incidents "were found to have the brake not depressed and the accelerator wide open." In other words, the drivers in these cases appeared to have hit the accelerator, rather than the brake, and must have caused what the drivers claimed was sudden, unintended acceleration.

NHTSA and Toyota each have access to the same data but are independently analyzing it. The government's findings are not due for a few months. The Journal story relied on "people familiar with the findings." We don't know if that's Toyota, someone from NHTSA or others. Toyota has the most to gain by making these findings public.

"In reality, the Journal used unnamed sources that purport to know what NHTSA has found," said Alair. "We can't confirm anything in their report because our investigation is continuing. We weren't contacted by NPR. They just interviewed the guy who did the story. From our position, his story was inaccurate. Had we been contacted by NPR we would have made our concerns known and made it clear his story isn't government data."

But that's not what you would learn from ATC's interview. Here's how the two-way 'intro' began:

MICHELE NORRIS: Now, an update on the investigation into the rash of Toyota and Lexus accidents that were initially blamed on sudden acceleration problems. After analyzing dozens of data recorders in Toyota vehicles, U.S. safety regulators have found that some drivers who claimed their car suddenly surged out of control may have mistakenly floored the accelerator when they meant to hit the brakes. The National Highway Traffic Safety Administration's findings are preliminary, and they do not clear the car maker.

That isn't quite accurate.

Hearing this, a listener would reasonably assume that the information definitely came from the government, particularly the statements: "U.S. safety regulators have found . . ." and "The National Highway Traffic Safety Administration's findings are preliminary, and they do not clear the car maker." Listeners were not told, and should have been, that the information in the interview came from anonymous sources.

Contrary to the impression given by the Journal story and the NPR interview, NTHSA has drawn no conclusions and released no data, said Alair.

It would have been better for ATC to frame the story as: This is what we know, and this is what we don't know. Then the listener can decide. But if it's presented as government data, when in fact it's not, that hurts NPR's credibility. NPR should have called NHTSA.

The basic problem is the Wall Street Journal story was thinly sourced, and therefore, so was NPR's. [The Journal takes issue with this statement. See editor's note below.] But even the Journal's story had more context and details than what ATC offered.

"Whenever you do a radio interview, you are essentially condensing down drastically," said Turpin. "What happened here is you had a fairly complicated story and in condensing for the (time) slot, we ended up losing a little bit of context that would have been helpful to the listener."

Why didn't ATC use NPR's own automotive reporter, Frank Langfitt?

They tried. But Langfitt wasn't able to nail down Ramsey's unsourced story in time for ATC's 4 p.m. start time.

The next day, Langfitt did a 2-minute piece for Morning Edition providing similar information but relying on, and citing, Toyota's findings as its source.

Unlike the ATC interview, Langfitt's piece included information challenging the finding's reliability. Langfitt quoted Sean Kane of the Auto Safety Advocate, who works with plaintiff's lawyers and who pointed out that the data recorders aren't foolproof.

Langfitt also noted that even if some drivers were pressing the wrong pedal, there are still hundreds of other complaints alleging that sticky accelerator pedals and faulty floor mats may account for sudden acceleration. Toyota's problems are far from over.

That kind of context wasn't in ATC's two-way.

Turpin is "not very happy" with this two-way. He said the fault lies with how the introduction to the story was written and that it wasn't allotted enough time in the show.

"Normally, we would say the Wall Street Journal is reporting in today's paper that NTHSA has [what was reported]," said Turpin. "So we make it clear that the essence of this story is about the reporting by another publication. This is the kind of interview that would have probably benefited from another minute or so to explain some background information that would have contextualized it a little more."

One question Turpin now realizes should have been asked of Ramsey: How did you get your information?

I asked Turpin if, when ATC is planning a two-way with a non-NPR reporter, whether they try to corroborate or do their own reporting to gather new information for the story?

"Yes, frequently," he said. "I wish we had in this case. I wish we had called NHTSA. Very often in stories like this good news organizations have gone to a lot of trouble to make sure the people that they are reporting on get their say in the story. The response of the institution is inherent in the interview you do with the reporter from the newspaper."

Turpin said most two-ways with other journalists work really well but sometimes they just don't quite come together, as in this case. "We have excellent hosts and editors, and I don't' think this should reflect on them," he said. "It's a good learning experience and a good reminder for us to be really skeptical."

NPR also conducts two-ways between hosts and NPR reporters, especially on breaking stories when there's not enough time to put together a full news story. This also happens for some stories about national security or other sensitive topics when it's difficult to get people to speak on tape. In addition, the shows regularly interviews authors, experts, people making the news, and regular folks.

I am not a fan of using journalists from other news organizations to tell stories. But I understand that it benefits NPR's audience if it shares exclusive reporting from another news outlet, on say, an investigative project that has taken months. It also can be useful to interview other reporters at the scene of a breaking story if NPR can't get someone there right away.

But the Toyota story and another last week with a Washington Post reporter are stories NPR would have been better off matching when it could rather than relying on a competitor – and thus finding itself relying on someone else's reporting.

NPR has recently beefed up its newscasts, staffed an investigative unit and is pouring money into its digital operations. All these are signs of a strong news organization. But if NPR wants to be considered one of the nation's top-flight news organizations, it should be more judicious in using this kind of two-way.

EDITOR'S NOTE:

I regret using the word "thinly," but the story was based on anonymous sources. It is true that I do not know who those sources were. My issue was not with the Journal's story, but rather NPR's practices.

Below is a statement from the Journal:

"As NPR's ombudsman, Ms. Shepard is not in a position to know about the sourcing of articles in The Wall Street Journal. Her comments that the Journal's article was 'unsourced' and 'thinly sourced' are erroneous, and we were surprised that she didn't give us an opportunity to set her straight or comment before she published those strong words. The Journal's article was based on solid reporting and was appropriately sourced by people familiar with the findings of the Department of Transportations analyses. In addition, the National Highway Traffic Safety Administration was given the opportunity to comment on our report, but chose to decline to do so. The NHTSA hasn't contacted us to allege any errors in our coverage."

- Ashley S. Huston
Senior Director of Corporate Communications
Dow Jones & Company