Everyone knows headlines are broken. Here's how news organizations can start fixing them.

There’s a longstanding journalistic tradition of treating presidential utterances as inherently newsworthy -- so newsworthy, in fact, that the “president says” construct is a staple of headline writing. But how should journalists cover those remarks when a president is 10,000-plus false or misleading claims into his tenure?

Last fall, for instance, President Donald Trump claimed that he and congressional Republicans were working “around the clock” on a plan to implement “a major tax cut” for the middle class prior to the 2018 midterms, then just weeks away.

The claim caught the Republicans Trump said he’d been working with by surprise. The truth was that there was no plan and legislators were on recess. The boast was a fairly transparent play to help Republican chances in the elections.

Even so, CNN ran with it, posting the headline “Trump says GOP working on tax plan for middle class.” The Associated Press, Reuters, ABC News, The Washington Post, Axios, Politico, and others also touted his claim in their headlines, giving Trump exactly the press he wanted.

If you read beyond the headlines, you would have learned some important context. CNN noted in its piece that “it’s unclear what tax proposal Trump was referring to.” And the Post’s story mentioned that such a plan “appears highly unlikely.” But many readers don’t read the actual articles, and they might have been left thinking a 10% tax cut was coming their way.

Media Matters has been considering how headlines and tweets can inadvertently spread misinformation for months. We’ve documented some egregious cases of news outlets repeating lies and other misstatements in headlines or tweets. We also examined 32 Twitter feeds from some of the country’s biggest news outlets, cross-referencing tweets about Trump comments with The Washington Post’s database of false or misleading claims Trump made. We found that, all together, these accounts shared demonstrably false information from the president an average of 19 times a day; 65% of the tweets did not include any context indicating the statement was false.

That approach hardly serves to inform readers. The reality is that when information is repeated -- regardless of veracity -- it forms a lasting impression. Add to that readers’ tendency not to click beyond headlines and it becomes clear that the old rules for headline writing haven’t held up well.

So we set out to find some new ones. We consulted journalists, media ethicists, and other experts and examined published scholarship on the issue to guide us as we put together a set of best practices. Here’s what we found.

Why don’t the old rules work?

Journalists have written that headlines and tweets that quote misinformation without contextualizing it can do “a disservice,” “deliberately misinform,” “perpetuate false narratives,” “undermin[e] otherwise good work,” be “dangerous,” give Trump “the headlines and quote-tweets he wants,” and spread “a species of fake news.”

Part of the problem is that few readers have the time or luxury to read newspapers cover to cover or to click through every story they see on social media. A 2014 study by the American Press Institute found that 60% of Americans read only headlines in the week before they were surveyed. And in 2016, Columbia University and French National Institute researchers found that 59% of articles retweeted on Twitter hadn’t actually been read by the people sharing them.

“Readers, and I count myself as one of them, are lazy and busy, and that’s a terrible combination when trying to get someone to understand what something is,” Aimee Rinehart, director of development and partnerships of First Draft News, a nonprofit focused on digital media issues, told Media Matters.

And these repeated lies can stick. Steve Rathje of Psychology Today writes that repeating a false statement -- even in an effort to correct it -- can “fuel false claims even more. … A claim feels more ‘true’ if it comes to mind with fluency and ease, leading us to mistake what is familiar for what is true. This is called the ‘illusory truth effect.’”

George Lakoff, director of the Center for the Neural Mind & Society and a former professor of cognitive science and linguistics at the University of California, Berkeley, told us that the “way your brain works is that you take the things you hear first and understand other things in terms of that.” Thus a headline that quotes something that’s not true -- even if to counter it, Lakoff would argue -- can leave readers believing it.

A 2014 study published in the Journal of Experimental Psychology further explains how the brain processes information, noting that the impact of headline misinformation can persist even when readers do continue through to the article.

“Readers use available information to constrain further information processing. This means that any incoming evidence will always be weighted and interpreted in light of information already received, and a headline can thus serve to bias processing towards or away from a specific interpretation,” the study notes. And “correcting the misinformation conveyed by a misleading headline is a difficult task. Particularly in cases of non-obvious misdirection, readers may not be aware of an inconsistency, and may thus not initiate any corrective updating.”

So how can news outlets get it right from the start?

Does the quote belong in a headline? Is there news value in tweeting it?

A first step is to consider the newsworthiness of the statement and decide whether it belongs in a headline at all.

“The same ethical standards that exist for the front page of your newspaper or the lead story of your 5 p.m. newscast should apply in all circumstances,” Society of Professional Journalists Ethics Chair Lynn Walsh told us in an interview.

“President Trump is someone that is in a position of authority,” Walsh explained. “It is someone, as journalists, that we do cover and should cover. But if what he is saying is incorrect, ... I think that is absolutely newsworthy to include in the story. But do we need to make it the headline? And do we need to make it what we are sharing specifically on Twitter or on Facebook?”

In the hyper-competitive digital news marketplace, it can be tempting to tweet and post quotes as soon as possible. But while that may help boost one’s profile and feed the online-news click beast, how is it benefiting readers?

Lewis Raven Wallace, a former reporter on the public radio show Marketplace and creator of an upcoming podcast that will explore the idea of objectivity in journalism, told us the competition among journalists and news outlets to be first to report a story ends up benefiting “propaganda and misinformation.” “But that whole structure isn't actually serving the public, at this point,” Wallace added.

“The rush to be the first or the grabbiest is good for no one except the broken business model/bottom line, and the owners of platforms like Twitter,” he said. “But I can't see how it's serving the public to ever distribute tweets and headlines that mislead.”

Bobby Magill, president of the Society of Environmental Journalists, told us repetition of an inaccurate statement can be damaging -- and it often isn’t even particularly newsworthy.

“If Trump tweets that climate change is a hoax for the 10 millionth time, we already know that Trump thinks that climate change is a hoax,” he said. “We don't need to read about it again. If you keep printing that, if you hear an untruth so many times, it eventually becomes the truth. You start believing it.”

Others said journalists should consider who benefits from repeating misinformation in a headline: Is it the readers -- or the speaker?

“I think the first reaction to a lie or a misleading statement should be to wonder why he has said it, why he might want it repeated, and whose interests are being served by covering it,” said Wallace. “Every outlet is weighing what to cover on limited space. Again, is the public served by a press that follows Trump in a herd, covering everything he's said? I doubt it.”

Lakoff has written that our president is a “super salesman” who knows how to work the media, getting the headlines and articles he wants.

“Trump knows the press has a strong instinct to repeat his most outrageous claims, and this allows him to put the press to work as a marketing agency for his ideas,” he writes. “His lies reach millions of people through constant repetition in the press and social media. This poses an existential threat to democracy.”

“I don’t think he’s stupid,” Lakoff told us in an interview. “He knows that he can get his point of view out there.”

Outlets should consider these options for addressing misinformation in headlines.

The Associated Press’ vice president for standards, John Daniszewski, writes that reporters have “a duty not to simply relay, channel or amplify” tweets and live statements. “Even in our news alerts, headlines and tweets about a public figure’s tweets, we need to show that we are not just re-transmitting the spin.”

“When we know a statement is factually inaccurate, or highly debatable,” he concludes, “we need to say so in some way.”

But there is more than one way to, as Wallace put it, “call a spade a spade.”

When a statement is blatantly false, the headline should say as much.

“It is not that much more difficult to write a headline that contradicts a lie than it is to quote tweet and perpetuate lies that way,” said Brooke Binkowski, managing editor at TruthOrFiction.com (and formerly Snopes.com). “If Joe Smith says ‘humans can breathe water,’ a bad and irresponsible headline would be something like, ‘Joe Smith: 'Humans Can Breathe Water.’ How do you get around that? Easily. ‘Joe Smith Lies, Claims Humans Can Breathe Water.’ Or, ‘People Can Breathe Water, Joe Smith Falsely Claims.’”

When there’s not such a straightforward debunk but the claim is still counter to common knowledge, another option is to use a term like “baseless” or cite a lack of existing evidence.

Writing in Vox last fall, Aaron Rupar highlights examples of both problematic and useful headlines about Trump’s November claim that ballots in Florida’s primary contest were “massively infected.” ABC, for instance, simply quoted the president, while The Guardian wrote that Trump repeated a “baseless voter fraud claim.”

Other experts and journalists question the value of phrases like “said without evidence” -- for one thing, it can seem to imply that evidence exists but simply wasn’t cited, while some view it as a “partisan slide” that “identifies the journalist as almost rolling their eyes at the comment,” as Rinehart put it. She recommends including the truth right away.

“It’s important to emphasize in a headline and tweet what is true, and not what something is not,” said Rinehart. “A classic example is the headline ‘Obama is not a Muslim.’ What two words stick out to readers who are scrolling through a newsfeed on a 3-inch screen? It’s likely to be ‘Obama’ and ‘Muslim,’ which reinforces to a scanning reader that myth.”

“A better headline is: ‘Obama is a Christian,’” she continued, “and begin the story that way, too. … All the information contained in about 200 characters needs to be true.”

That approach comports with the advice Lakoff has advocated. He advises a “truth sandwich” technique “to frame things in terms of what’s true” first so the accurate information will stick.

Eli Pariser, chief executive of Upworthy (disclosure: Co-author Molloy previously worked at Upworthy), pushed a similar approach.

“Famously, saying ‘Saddam Hussein had nothing to do with 9/11’ only convinced people there was some kind of relationship between Saddam and 9/11,” he said. “But saying ‘The 9/11 hijackers were from Saudi Arabia, Egypt, and UAE’ offers a complete story about the hijackers and actually makes it hard to work Saddam in.”

When it isn’t possible to lead with contradictory information, you can still flag a misleading statement for readers, said Binkowski.

“If you must use the claim in the tweet or hed, explain briefly that it's a problem, even if you can't explain why,” she said. “Then people will either understand that it is problematic and move on, or click on the story and read more.”

Wallace agreed, saying it’s better to use a word like “dubious” than to not flag the quote as problematic at all.

“If it's dubious for reasons you can factually provide, call it dubious in the damn headline,” he said. “If it seems to deliberately mislead, state your reasons for thinking so. … Better to be unsure and closer to the truth than to be 100% sure of something that turns out to be wrong or misleading.”

On the other end of the spectrum, particularly egregious falsehoods could warrant an entirely different story: one centered on the fact of the lie.

Daniel Dale, the Toronto Star’s Washington correspondent who is known for fact-checking Trump on Twitter, says that when the lying is the bulk of the topic covered, it should be the story.

“If a car salesman told you 36 untrue things in 75 minutes, that would probably be the first thing you told your friends about your trip to the dealership,” he writes. “It should have been the first thing we all told our readers about Trump’s August rally in Wilkes-Barre, Pa.”

And finally, there's the option of just refusing to report on the misinformation at all.

Walsh said people working in newsrooms should “question whether or not we do need to report on every single response that the president has,” because when the information is incorrect, “it can be actually more detrimental to our reporting.”

And Binkowski said media should “stop treating people who use 4chan and memes to inform their policy as though they are serious policymakers. They are lifting propaganda videos from Reddit and passing them off as their own. They lie with impunity. They push white supremacist memes and dog whistles. They call us ‘the enemy of the people.’ Why should we give someone like that free PR?”

Whatever the approach, truth must be central -- even if it means missing out on a few clicks.

“You cannot let your need for profit undermine your core values,” said Poynter Institute Senior Vice President Kelly McBride, who also serves as ethics chair at Poynter’s Craig Newmark Center for Ethics and Leadership. “And if truth is a core value, then you can’t let distribution techniques or tricks undermine that.”

Another consideration: Live coverage and Twitter

Many of the issues involved in headline writing also pertain to Twitter: partly because journalists often tweet out headlines and similarly pithy phrases when sharing their work, but also because of the phenomenon of live-tweeting. But what’s the value in such work?

“I think live-tweeting a speech in general is silly and noisy and not generally helpful for the public,” said McBride. “The kind of live-tweeting that I like is when you bring more context to something.”

That context, she and others said, can come via fact-checking.

“In this age of information and with all the tools we have available, there's no excuse to not be fact-checking the type of big, bold, misinformation claims that we see during political rallies,” Max Kamin-Cross, a data analytics expert and chief operating officer for FactSquared, said in an interview.

In discussing how to tweet about misinformation, multiple experts referenced the Toronto Star’s Dale, who writes that he’s “made it my mission to fact-check every word Donald Trump utters as president.”

“I think the work Daniel Dale at the Toronto Star does is a great example of one way to handle it,” said Kamin-Cross. “He differentiates between lies and dishonest statements.”

“See the Toronto Star’s Daniel Dale’s excellent Twitter feed,” said Binkowski. “He does it perfectly.”

For his part, Dale has written that his work is easier than many imagine: “People sometimes ask in response how I can blast out these corrections so quickly. But I have no special talent. My secret is that Trump tells the same lies over and over.”

The takeaway: Headlines and tweets must be able to stand on their own

The Atlantic’s Derek Thompson writes that the “news media cannot kill the virus” of Trump’s endless misinformation. “But by refusing to host it,” he says, “they can at least limit the spread.”

Whether journalists do that by not reporting on a falsehood, immediately highlighting contradictory facts, making the lie itself the story, putting contradictory evidence in a headline, or simply flagging a statement as dubious right off the bat will depend on the circumstances.

But whatever the case, one rule must hold true: Reporters should write tweets and headlines as though they’re the only part of a story that people will see. Often, they are.

Guidelines graphic created by Sarah Wasko.