Tracking NPR's Political Coverage

The campaign for president — including how NPR covers it — is clearly top of mind for many listeners and readers who write to the Ombudsman's office. The topic has far outweighed any other in the hundreds of emails we've received each month during the past year.

The issue is a priority for us, as well. No other story is likely to take up so much NPR airtime or online space in the coming months. NPR needs to get its coverage right. But what exactly does "right" sound and read like in an election year as unusual as this one?

When it comes to campaign coverage, NPR has a deep well of experience to draw on internally. But this cycle's primary races challenged the existing playbook. As for the general election, no recent campaign has had a presumptive nominee remotely like Republican Donald Trump. His campaign style has highlighted the reality that the usual norms for campaign coverage may not suffice — or at least will be sorely tested.

To take one example that listeners are raising regularly with my office: Trump refers often to his presumed Democratic opponent Hillary Clinton as "Crooked Hillary," bypassing at least the veneer of civility that presidential candidates have adhered to in the past. Should NPR quote him repeatedly, as it has done? In my view, yes. It is hugely important to hear directly from candidates themselves. When voters consider candidates, style and presentation count, and Trump's over-the-top rhetoric is a major part of who he is. Moreover, his statements are a matter of record.

But, as Tifton, Ga., listener John Tibbetts argued, does the constant repetition embed Trump's charge in listeners' consciousness so that "he doesn't have to prove the charge, he simply gets it repeated enough to become part of the conversation"? The answer to that question clearly is yes, as Trump obviously intends.

More broadly, there is the question of how to handle fact-checking of a candidate who far more often than past candidates (and more often than his presumptive main rival) makes statements that are simply false or are seriously misleading. Other news organizations are grappling with this, as well; one of NPR's specific challenges is limited time in its on-air reporting.

For example, after recent competing speeches on the economy, NPR enlisted teams of reporters for rapid annotations of the remarks by Trump and Clinton. That did not satisfy listener Carol B. Barry of Vandalia, Ohio, who wrote that the All Things Considered story "led off with successive audio clips of Donald Trump's speech without any correction of his half-the-story or false statements other than to say at the very end to go to for NPR's fact-checker assessments. Seeing as how facts are the most important thing to evaluate candidates running for office, the story should have been presented with the facts, followed by Trump's twists and distortions, i.e. the audio clips. But you, like all other media, gave Trump the prime time, what he was after and gets for free. If voters want facts, they'll have to find time in their busy days to go to your web site."

Questions about the tone of coverage arrive regularly, with some listeners believing that Trump is not getting a fair shake from NPR reporters. We have not detected that, but we are listening closely for it.

There also are questions of balance. A handful of listeners wrote to my office upset that Trump's press conference on his Scotland golf course in the wake of the United Kingdom's referendum vote to exit the European Union was getting much more play from NPR than Clinton's more measured, diplomatic response. I would argue that a presumptive candidate who talks about his own business interests in the wake of such a potentially globally disruptive move is so far outside the norm as to be the story deserving the most attention.

But I understand the frustrations of Clinton supporters who felt as though NPR was not playing fair with its website home page that Saturday featuring two stories about Trump and Brexit, and none about Clinton. NPR later added a home page story about Clinton's response (when Clinton herself responded more forcefully). That pattern has repeated itself throughout the campaign season so far, and it is one that NPR needs to come to terms with how to handle.

Likewise, some listeners want more coverage of third-party candidates, including the Libertarian Party's Gary Johnson and the Green Party's Jill Stein. In the past, NPR has tended to give such candidates limited attention, reflecting the limited interest of voters, based on polls. But with voters expressing unprecedented unhappiness with the two presumptive main candidates, those third-party candidates may find more support. As Lucasville, Ohio, listener Adam Sheets recently wrote my office: "Governor Johnson has only been interviewed once on NPR this election season and has rarely gotten any news coverage from you," despite relatively strong polling. "What is NPR's threshold to begin covering the highest-polling third-party campaign since Ross Perot?"

Tracking the Coverage

To bring some data to what has so far been a highly impressionistic discussion, about five weeks ago my office started tracking all of the coverage of the presidential race that appears on the morning and evening newsmagazines and online. Starting today, and continuing through the election, we will publish some of those results (every few weeks until the fall and then more often, if warranted).

We are tracking the stories using several criteria: The number of pieces devoted to each candidate. Whether each piece was an interview or a reported story, or whether the voices of the candidates themselves were heard. We also track the focus of each story and whether it included the voices of both supporters and detractors of each candidate. Did the story discuss poll results, which senior Washington editor Beth Donovan has — appropriately — told the newsroom to go easy on, to minimize horse-race coverage?

One of the most important questions is how many articles fall into the category that we call "best practices." Through conversations with the newsroom, we have agreed that our criteria for these stories include those that fact-check the claims being made by the candidates, as well as stories that look at specific policies that candidates are proposing and their stances on the issues. Listeners have said that this is the coverage they look to NPR for. NPR has already put forth solid stories in this category all throughout the primary season, but too often they got lost in the sheer volume of coverage, so this column will attempt to highlight some of the best.

Enterprise reporting falls into this category, as well: stories that go beyond covering the events of the campaigns, to dig deeper into candidates' backgrounds and whether their statements match reality. NPR has the reporting talent to produce such stories, but the newsroom has done relatively few such pieces this election season. I would hope to see many more, and I'm told we will. Stories that get outside the standard campaign reporters' bubble, such as Sarah McCammon's piece from inside Trump's meeting with evangelical leaders, will also be included, as will candidate profiles.

One thing we will not be tracking is the exact number of minutes devoted to each candidate. A handful of listeners have asked for that and we considered it. But that data is not easily accumulated and takes into account only the on-air reporting. Mostly, though, it is simply not a very good way to judge coverage—just as constantly reporting on poll results does not help listeners evaluate the candidates.

As Michael Oreskes, NPR's head of news, told me, "It's a false equivalency to just say if we had the same number of minutes on each candidate that that's creating better coverage." Sometimes an unbalance can reflect campaign strategies, he said. "The Clinton campaign is clearly following an approach of doing the occasional significant moment that they want to be focused on and not trying to get a lot of other attention. I don't think they even mind" that Trump is getting more attention, he added. "In their view, I think, a lot of this attention doesn't do him any good."

Indeed, even the approach we have settled on — how many stories each candidate has received — can be misleading and misinterpreted. As evident in the first round of numbers, below, just because one candidate receives more coverage than another, it is not an indicator of favoritism; sometimes the stories are reporting on a campaign setback. But, as Oreskes told me, balance does become an issue if the "minute-to-minute measurement starts to get too far out of whack, you might get into a situation where you're actually not even getting enough about one of the candidates, and that would be a fair criticism."

Oreskes said he views the issue "not as minutes, but as information. Are we getting across to you the material you need about both candidates? That's the challenge. That's what we should be judged by. We may actually give you more about one than the other, but as long as we've given you enough about both then I think our job is getting done."

(That brings me back to the readers above who were upset when NPR's Brexit home page focused solely on Trump. They were right in that case, but it was largely a problem of presentation and packaging. NPR should not make it difficult to find reporting on both candidates, and Oreskes said the newsroom now has people assigned to monitor that.)

Other reporting that will not appear in the roundups (although may occasionally be highlighted): the reporting and interviews heard on the midday newsmagazine Here and Now (produced at Boston station WBUR, with NPR's support) and the programs NPR distributes but does not produce, including The Diane Rehm Show and On Point. The NPR Politics Podcast is a key part of the politics team's efforts, but we are not tracking that, either (although I am a big fan and encourage a listen).

Finally, we will not be documenting the campaign coverage that listeners hear in the newscasts. They are, of course, another important part of NPR's total output and for listeners who have their radios on all day, may well be the most often heard stories. But the sheer volume of newscasts makes tracking them unfeasible.

Initial results

On to some results from the first breakdown, for the period from Sunday, June 12, through Thursday, July 14: During that period, NPR had 97 stories (online and on air) that focused mainly on Trump, and 60 on Clinton. Another 49 stories examined both candidates or wider campaign issues. Democrat Bernie Sanders was the focus of eight stories (the tracking period started after Clinton became the presumptive nominee). Third-party candidates were discussed in three stories. (Oreskes told me the newsroom is "going to be watching them very carefully," a response that I am sure will not satisfy those who feel they deserve more attention right now.)

Why are the numbers so out of whack for Clinton and Trump? Mostly it comes down to the first two weeks of tracking, when Trump's reaction to the Orlando shooting, including his controversial remarks about President Obama, received lots of press; he also fired his campaign manager during that period. The two middle weeks were relatively balanced. This past week, Trump again had a slight edge, in the lead-up to the Republican convention.

On the question of horse race-type coverage: Six of 29 stories in the week of June 26 referenced poll results, including this interesting interactive digital story about the electoral college battle, as did five of 31 stories in week four and none in the abbreviated last week. That seems reasonable to me.

Several stories fell squarely into the "best practices" category, including Michele Kelemen's June 15 "Fact Check: Donald Trump And Syrian Refugees," this Don Gonyea piece talking to Detroit voters, and Sam Sanders' conversation with two divided Democratic delegates, which included a good amount of policy discussion.

Going Forward

Back to NPR's plan for covering the rest of the campaign. In addition to the day-to-day coverage, Oreskes told me two new initiatives will follow the national conventions, which start today with the Republicans.

One, whose details will be announced shortly, will apply an intense focus to the broad themes that have dominated the election season.

In addition, Oreskes said, NPR is working on a format where, every day online and regularly on air, it will play an uninterrupted excerpt of what the candidates are saying. "In my mind a lot of it will be literally just segments from speeches, or interviews that they do, where we just give you a block of it, to just hear what they say," he said. I'll have more on that when details are finalized.

Overall, Oreskes said, the goal is to give listeners numerous ways to evaluate the candidates: "You may or may not like our coverage," he said, although the "goal is to be fair and complete in our coverage. But whether we are or we're not, we'll have a separate channel of material available to you as a citizen, that hopefully will help you with your job of picking who to vote for."

Finally, a note about NPR's coverage of the conventions. NPR has teamed with PBS for a joint evening simulcast. The sharing of public media resources makes absolute sense, and the two organizations to my mind have a similar approach to the news. But the initiative will be an interesting experiment, given the different needs of radio (audio) and TV (video). I'm looking forward to hearing how the producers handle such convention staples as the video biographies of the candidates, which on the surface certainly play better on TV than on the radio.

Editorial researcher Annie Johnson contributed to this report as did intern Shane McKeon, who also compiled the tracking data.

Copyright 2020 NPR. To see more, visit https://www.npr.org.