It’s no secret that the president of the United States has a tumultuous relationship with the very concept of objective truth. Whether you call them lies or falsehoods, or simply chalk them up to his businessman bravado, anyone working in media knows by now that Trump’s statements should probably get a bit of scrutiny before being blasted out to the world as breaking news. Still, despite knowing this, news media continue falling into the same trap of publishing what the president said without providing relevant context. This week’s news around his expressed desire to end birthright citizenship (something that he campaigned on beginning in 2015) provided a prime example of how not to cover him.

Across the country, headlines like “Trump says he will sign executive order banning birthright citizenship” from The Hill and “Trump Says He Will Void Birthright Citizenship Law Through Executive Order” from NPR spread like some sort of digital cancer extending from the web to the chyrons of news shows. The most egregious example probably came from Axios CEO and co-founder Jim VandeHei, who tweeted, “Exclusive: Trump to terminate birthright citizenship.”

Another unfortunate tweet came from The Associated Press. The since-deleted tweet read:“Trump wants to order the end of birthright citizenship: ‘We’re the only country in the world where a person comes in and has a baby, and that baby is essentially a citizen of the United States.” And while it’s true that he did say that, it was a false statement in itself. At least 30 other countries grant citizenship to anyone born within their borders. Not including that crucial bit of context in the tweet only amplifies Trump’s lie.

Perhaps past presidents could be trusted to tell the truth about their own policies. It's not clear whether Trump even knows what's in his own.

In spring 2017, as Republicans pushed their “Obamacare repeal and replace” bill, the American Health Care Act (AHCA), through Congress, Trump gave an interview to CBS’s John Dickerson in which he claimed, “Pre-existing conditions are in the bill. And I just watched another network than yours, and they were saying, ‘Pre-existing is not covered.’ Pre-existing conditions are in the bill. And I mandate it. I said, ‘Has to be.’”

The AHCA, like all GOP health care bills, would have weakened the Affordable Care Act’s pre-existing conditions protections and allowed insurers to price people out of the market or exclude coverage of specific conditions. What Trump said simply wasn’t true, yet headlines that simply reported what he said were standard fare across political media: “Trump on health care: Pre-existing conditions will be covered” Politico wrote, “Trump guarantees coverage for people with pre-existing conditions in health care bill” CBS declared, and “Trump says coverage of pre-existing conditions will be in healthcare plan” The Hill stated.

There are dozens of similar examples of headlines on any number of topics Trump has commented on since he took office. While the headlines themselves may technically be true -- since they report what Trump actually said -- omitting the fact that what he said was not true makes them misleading.

Journalists need to grapple with an uncomfortable truth: Most people read only the headlines.

Many of the examples mentioned above do note within the articles that Trump’s claim mentioned in the headline is false. Unfortunately for journalists, the overwhelming majority of people who see a headline won’t actually click through to read the article (a single-digit click rate is pretty standard on Twitter). Just think about how often you’ve scrolled past an article shared on social media, not clicking it, but still registering what it said. Sure enough, a 2016 study estimated that 59 percent of the time someone retweets an article, they actually never clicked on it first.

Now, one could argue that it’s the responsibility of news consumers to better vet their sources and actually read articles. That’s totally fair! Still, there’s a question of why news producers should make it harder to get accurate information. Maybe omitting a tiny bit of context in the headline will create a “curiosity gap,” a deliberate effort to boost clicks by not telling the whole story in a headline. The curiosity gap was popularized by Upworthy (full disclosure: I worked at Upworthy between 2015 and earlier this year), but it was meant to draw people in to stories they otherwise wouldn’t have been interested in, not for use in straight reporting.

For more insight on headlines and where things went wrong, I reached out to Eli Pariser. Pariser helped popularize the idea of “filter bubbles” in his 2011 book, The Filter Bubble: How the New Personalized Web Is Changing What We Read and How We Think. Currently a fellow at the New America Foundation, Pariser co-founded Upworthy in 2012.

Part of the issue, Pariser suggests, is the medium itself: “Unlike with a physical product, the headline [of an online news article] lives separately from the content and needs to be evaluated on its own.”

Trying to distill a complex breaking news story into about 100 characters is a big challenge, leading Pariser to suggest “rethink[ing] headlining in general.”

Algorithms rule the media world, but they have a history of rewarding the wrong things.

Earlier this year, former Snopes Managing Editor Brooke Binkowski gave me a valuable tip during an interview.

“If you're reading, viewing, or listening to a story that's flooding you with high emotion, negative or positive — whether it's fear, rage, schadenfreude, amusement at how gullible everyone else is — check your sources,” she told me.

There’s even some evidence that the stronger the emotion a headline evokes in the reader, the more likely people are to share it with others. So while more straightforward headlines from sources like The Associated Press or Reuters might best help convey what actually happened in any given situation, it’s sites like Breitbart, The Daily Caller, Gateway Pundit, and The Federalist that tap into readers’ emotions. In turn, those sites get more shares, and as a result, get further boosted by algorithms designed to reward active engagement.

So long as social media algorithms reward engagement, it’s hard to fault any profit-driven media outlet for playing along. Seeing a headline that contains an obvious lie or omitted context, such as The Hill’s “Trump says migrant caravans are ‘larger’ than reported” can tap into those sharable emotions (anger, rage, disgust, and so on), continuing to incentivize flawed coverage.

Getting outlets to publish accurate headlines without sensationalist flourishes could be tough.

As an industry, media is in a really rough spot. Pew Research Center found that newsroom employment dropped by 23 percent between 2008 and 2017. In a highly competitive industry, publishers need to find whatever edge they can get or risk going out of business. Optimizing headlines and photo choices for the most clicks, shares, and engagements possible is a big part of that. Going against the grain by trying to restore trust in media -- which dropped from 54 percent in 2003 to 41 percent in 2017 -- could destroy media outlets as a business before their impact is felt.

I suggested to Pariser that maybe a large outlet with enough funding to stay afloat indefinitely, perhaps a paper like The Washington Post, can take the lead in setting new best practices for headlines. Or perhaps the Twitters and Facebooks of the world could be convinced to better factor in whether a headline is omitting key information or leaning into sensationalism when making changes to their algorithms.

Pariser’s not sure of the best response to deceptive headline practices, but he thinks the focus should be on establishing and encouraging industry-wide norms: “My guess is that this has more to do with encouraging greater adherence to norms and journalistic culture than to any particular set of rules, which will always be game-able and have exceptions.”

Trump's rise has shown media are ill-prepared to handle a serial liar as president.

Have past presidents lied? Absolutely. But none have lied as frequently about topics big and small as Trump has. This calls for a new approach to journalism, top to bottom. In a September 2017 post to his PressThink blog, New York University journalism professor Jay Rosen explained that journalists have failed in acting as a check on the presidency, normalizing Trump’s abnormality in the process.

Maybe journalists have been afraid to hold Trump to the same standards they held past presidents to because it would paint the picture of someone in way over his head with little understanding of, or desire to do, the job he was elected to do. That type of coverage, while honest, might sound biased against him, and so they give him a little extra leeway than they would have with, say, a President Clinton or President Cruz.

Rosen has his own theory:

If nothing the president says can be trusted, reporting what the president says becomes absurd. You can still do it, but it’s hard to respect what you are doing. If the president doesn’t know anything, the solemnity of the presidency becomes a joke. That’s painful. If they can, people flee that kind of pain. In political journalism there is enough room for interpretive maneuver to do just that.

In other words, he believes media treat Trump with kid gloves because their job is only as prestigious as the office they cover -- and in the Trump administration, the realistic assessment of that prestige level could be “not very.” I emailed Rosen to ask him whether it even matters what the motivation is behind padded coverage. Here’s what he had to say:

The motivation matters if we want to understand why it's happening and what might bring a change. For example, if the pattern is just about placating conservatives, well, conservatives are never going to be placated, so we may as well consider the pattern permanent. But if, as I believe, the press is attached to an image of the presidency in which they have a kind of psychological investment, it's possible that they can realize this investment is leading them astray in the case of the Trump presidency. I am not saying that is likely, or that it is happening now. But you asked, does the motivation matter? I think it does.

Echoing a point Pariser made about the disconnect between print and digital, Rosen said he believes sloppy headlines to be the result of a “cultural lag,” with digital publications hanging onto the tabloid tradition of going a bit over the top for effect, to get people’s attention, or to sound clever. He points to the famous “Headless body in topless bar” as an example of using a headline to draw a reader in:  

It was probably never harmless, but it was certainly thought harmless, and generations of reporters learned to say, ‘Hey, I don't write the headlines’ when there was criticism. There was always a wink, wink, nudge, nudge element to this practice. When newspapers went online it was ported over to the new cultural space without a lot of thought. That was not smart.

Changing these old habits might be hard, but in the meantime, Rosen suggests encouraging use of the “truth sandwich” approach to social media and headlines: Say what’s true, then say what’s false, and then say what’s true once again.

It may not be perfect, but it’s a start.