Can narrative journalism overcome the political divide?

No self-respecting liberal would trust anything written on Breitbart, and every self-respecting conservative knows that The New York Times is a liberal rag controlled by people with New York values. Combine that with the echo-chamber of social media, the decline of local news, and the general political atmosphere of 2016, and you get a divided country with a divided media.

Or so goes the prevailing wisdom. But a study conducted this spring by the Columbia Journalism Review and the George T. Delacorte Center for Magazine Journalism found that sometimes the content of a story matters more than its source, even to politically polarized readers.

For our study, we randomly assigned readers along the liberal-conservative spectrum to read this California Sunday Magazine feature about a plot to kill a police officer. The story appeared to have run in one of two fictional magazines: a purportedly conservative publication called The Patriot, or the The American Progressive, which we presented as a liberal magazine. 

We specifically chose a story whose topic wasn’t overtly polarizing but contained enough political valence to trigger different reactions depending on where readers stood on the political spectrum. The core tension of the story is whether law enforcement entrapped two misguided but harmless malcontents or eliminated a legitimate extremist threat. We expected readers’ political views on issues like law enforcement and domestic terrorism to influence how they judged the actions of the cops and the would-be cop killers; we also anticipated their level of trust in the story would change depending on the political affiliation of the publication where it appeared.

Sign up for CJR's daily email

Instead, the study found that readers were equally likely to trust the story no matter where it had been published. On average, readers of both publications and from both parties also rated the credibility of the reporter and her sources within a similar range. The results suggest that people across the political spectrum are equally likely to trust a long narrative story, regardless of whether they read it in a publication whose political leanings align with or differ from their own.

While there are several possible explanations, previous research on narrative persuasion suggests that a lengthy, compelling feature told through the eyes of a character or characters with whom we can empathize—possibly despite ourselves—may be one antidote to political polarization.

A gripping narrative overrides our natural tendency to challenge information we don’t agree with, says Michael Slater, a professor of communications at Ohio State University and a leading researcher of narrative persuasion. When you have a strong narrative that’s really absorbing, it tends to suppress counter-arguing,” says Slater. “It’s hard to suspend disbelief and counter-argue at the same time.”

The story we chose, Ashley Powers’s “The Vegas Plot,” captivated many study participants, who described it as “interesting,” “engaging,” and “riveting.” One reader called it a “[c]razy, crazy, crazy story!” Another commented: “The piece got more and more interesting as it progressed.”

Powers, a former Las Vegas bureau chief for the Los Angeles Times, says she, too, was drawn in. Even as she reported the story, digging through news reports, court documents, and literature on the sovereign movement, and watching hours of police video, she felt “torn” about the motivations of the alleged plotters, David and Devon.

“I felt this was a story that lived in a gray area,” Powers says. “I wanted the story to reflect that. I wanted people to make up their own minds as to whether they saw David and Devon as dangerous.”

 

Our study began with a question: Given how many people now find news online, often through social media or other third-party websites, do magazine brands still matter?

Instead of visiting magazine homepages, many news consumers make split-second decisions about whether to reject or accept hundreds of links streaming through real-time feeds and across social networks. In such an environment, we wondered how readers decide which stories to trust. Do they rely on the imprimatur of a publication’s brand, or judge the article on its merits, or some combination of the two?

We undertook two experiments to find out. In the first, we pitted an established magazine brand against a younger digital native. A single story was presented as if it had been published in The New Yorker, BuzzFeed, and a fictional magazine called The Review. Subjects who believed they had read the piece in The New Yorker found it more credible than the same piece that appeared to have been published by BuzzFeed, suggesting that even in this era of routine digital sharing, brands carry weight. You can read more about the results of that study here.

In this related study on magazine branding and political bias, we looked specifically at how political branding affects reader trust. We began with the established premise that as social creatures, humans tend to organize into groups, gravitating to others who align with them culturally, politically, or religiously. They then default to assessing members of their own group positively and members outside the group as at best unfamiliar, at worst dangerous.

We wanted to test whether this kind of phenomenon extends to media consumption. Do news consumers’ group identity—particularly their political affiliation—influence how they assess an article’s credibility? We hypothesized that readers with pre-existing political views would be influenced by the political slant of the publication in which an article appeared, and that readers whose political views aligned with the views of the publication would find the article more credible than those with opposing political views.

This is supported by America’s news habits, which are heavily polarized along political lines. According to a 2014 Pew Research Center report, the most trusted news sources on the right are the most distrusted on the left, and vice versa. Close to 88 percent of respondents with consistently conservative views trusted Fox News, and 78 percent of all conservative-leaning respondents cited it as their main source of political news, while 81 percent of consistent liberals distrusted Fox. Conservatives overall were more likely than not to distrust MSNBC, The New York Times, and The Washington Post, among others. Media trust in general was much higher among liberals, according to the report, with news consumers on the left trusting most of the 36 news sources presented by Pew and consistent conservatives distrusting more than two thirds of them.

We wondered whether these results would hold true for a magazine story. Would a conservative immediately disregard a long, deeply reported article from an overtly liberal publication? Or would she evaluate it on its merits?

To test this, we worked with psychologist and cognitive neuroscientist Jenna Reinen, a postdoctoral researcher at Yale University, to design an online survey. Because our research involved human subjects, we submitted a detailed proposal to Columbia’s Institutional Review Board, an independent group that ensured the study was ethically sound.

Next, we used Amazon’s Mechanical Turk, an online workforce service, to recruit 161 participants. For expediency, we also used TurkPrime, an online service that works with Mechanical Turk, to ensure an even demographic and political breakdown in our sample. Half our participants were men and half were women, with an average age of 35. The average participant had at least some college education and an income between $30,000 and $40,000 a year.

Participants were asked several questions to identify their political affiliations. When asked with which “political position” they most identified, 30 percent of respondents chose Democrat and 30 percent Republican. Of the remainder, 12 percent self-identified as fiscal conservatives, 8 percent as Libertarian, 7 percent as social democrats, 5 percent social liberals, and 1 percent were split between social conservatives and fiscal liberals. Seven percent reported no political affiliation.

When asked in April for whom they planned to vote in the 2016 presidential election, 34 percent said Bernie Sanders, followed by 26 percent for Donald Trump, 16 percent for Hillary Clinton, 10 percent for Ted Cruz, and 8 percent for John Kasich. Five percent said they would not vote for any of the listed candidates. 

Next, participants were directed to one of two fictional magazines, the conservative Patriot  or the liberal American Progressive. The Patriot was described as having a staunch belief in free enterprise and defending America against encroaching radicalism. “No matter the consequences, The Patriot defies political correctness in pursuit of the truth,” the magazine description read. The American Progressive was described as a “longtime standard-bearer of progressive values, civil rights activism, and investigative journalism.”

wapo-ted-cruz-cartoon.jpg

“The Vegas Plot” as it was presented to study participants, in the wrapping of the fictional magazine The Patriot.

Visual and design cues reinforced the branding. The Patriot’s color scheme was red, white, and blue, and articles on the page included one titled “The Party of the Constitution.” In The American Progressive, the faux Web page carried an ad for MSNBC and the navigation menu included buttons marked “Donate” and “Take Action,” as well as links to the Democratic frontrunners, but not to Republican candidates.

wapo-ted-cruz-cartoon.jpg

“The Vegas Plot” as it was presented to study participants, in the wrapping of the fictional magazine The American Progressive.

Close to half our subjects were assigned to read the story in The American Progressive, and the other half in The Patriot. The text of the nearly 6,000-word piece was lightly edited so as not to provide any overt political cues. For example, we removed the original subtitle (“In the world of right-wing extremism, how do you tell who is dangerous?”) and replaced the term “extremist” in the body of the story with the term “activist.”

“The Vegas Plot” centers on David and Devon, a disenfranchised duo scraping by in Las Vegas. David hawks bottled water illegally on the streets; Devon survives on a social security check. Previously, David had begun to identify with the sovereign citizens movement, an anti-government subculture whose members believe they are not beholden to the rule of law. The two team up with some friends David met in prison and plot to kidnap a cop as a political statement. In one scene, David’s new friend Scott takes him out to the desert to teach him to shoot.

“Is that a real gun?” asked David, who hadn’t touched one since he was a kid.

Scott nodded.

“Now I know you’re not full of shit. You wonder about people, man.”

“We’re as real as you can be.” Scott hoisted the Rambo gun. “You want your picture?”

David hugged the gun awkwardly. Gave it a kiss.

Just before they are to execute their plan, David and Devon are arrested. It turns out that the “friends” who’d helped them, including Scott, were undercover cops, and the plot was actually a sting to neutralize the sovereign threat in the area. 

We wondered how readers of different political stripes would respond to a story rife with ambiguity. Was David a dangerous extremist or a hapless fool caught in a government trap? Did the cops abuse their power, or fulfill their duty to keep citizens safe?

We hypothesized that readers would find the story biased if they read it in a publication whose political views differed from their own. The story gives a picture of David and Devon’s hardscrabble lives and almost comical attempts at pulling off a high-level crime. Only at the end does the reader discover that much of the detailed information is based on law enforcement video recordings of their operation.

Overall, the data showed that readers of both publications found the writer’s sources more credible than not, and did not think the article was biased. However, when asked specifically whether they thought the author was biased against the police, readers across the political spectrum slightly agreed.

The author’s portrayal of law enforcement was of special interest, since the story subtly questions whether David and Devon posed a genuine threat. Sovereigns have killed police officers before, and the FBI considers them a domestic terrorist group. On the other hand, they can seem more like an amorphous band of agitators than an organized group with a committed platform, ideology, or leadership.

Some study subjects noted that author bias could be difficult to discern. A study participant from South Carolina who identified as Libertarian felt the story was more forthcoming about police overreach than most mainstream media accounts. “However,” he commented, “I’m not sure if that is a reflection of bias on the part of the author or if it was a fair approach which may appear to be biased when contrasted with mainstream media reporting.”

“I really did not see any kind of political leaning to the author,” wrote a 27-year-old woman from Illinois. “It just seemed like (pretty) good journalism, showing the different perspectives on the situation and leaving things open to speculation in the end. Most anyone could have seen both the scariness of the sovereigns and the possible unfairness of the police officers, I think.”

At the end of the study, we told our subjects that the story had actually run in The California Sunday Magazine, and provided a link to the original piece.

 

In 2014, the online magazine Matter published a story with a provocative title: “You’re 16. You’re a pedophile. You don’t want to hurt anyone. What do you do?” The story opened with the main character, Adam, downloading child porn. It described a brutal video that had compelled him to quit looking at child porn, seek help, and ultimately start an online support group for others like him. It compelled readers—probably despite themselves—to see Adam as a teenager struggling with a dark and terrible addiction, rather than as a monster.

Despite its unsettling subject, the story won wide acclaim and was a finalist for a National Magazine Award in Public Service. “It’s uncomfortable to deal with the consequences of that story,” says Mark Lotto, former editor-in-chief of Matter and co-founder of Matter Studios, but it was impossible to ignore.

That’s because a compelling, character-driven story like the Matter piece or the California Sunday article we chose for our study can be much harder to disregard than articles whose arguments are comprised of numbers and facts. Readers interact with stories differently than they do with information or news. They’re reading narrative pieces primarily to be entertained, not informed.

Lotto, who edited the Matter piece, says its power came from focusing on the characters. Early drafts were “buried in science,” but over many months, it became entirely about the people. “Knowing those guys, really seeing them up close, makes the argument infinitely better,” says Lotto.

This is supported by the work of Melanie Green, a media studies researcher and professor of communications at the University of Buffalo. Her work has focused on “transportation,” a term she and her co-author, Timothy Brock, introduced in a foundational 2000 study to describe what happens when we’re absorbed in a story. We’re in a suspended state, a sort of alternate reality where the events of the tale and the lives of the characters are more immediate than our own experience. That helps explain why, while reading stories, we’re willing to engage ideas, norms, or values that conflict with our actual views.

Did readers in our study put their political views aside because they got caught up in the narrative? Subjects reported being engrossed. “The piece….seems like a story that would make an interesting movie,” one participant wrote. Said another: “Long article, but it was interesting while figuring out where it was going.”

I’m not sure if that is a reflection of bias on the part of the author or if it was a fair approach which may appear to be biased when contrasted with mainstream media reporting.

Research also shows that readers who are engaged in a story are more likely to disregard its source. In the Green and Brock study mentioned above, participants read a dramatic story about a murder at a mall. Half had been told beforehand that the story was true; the other half had been told that it was an excerpt from a novel. Afterwards, most of the participants could not recall whether the story was fact or fiction. “Our findings suggested that once a reader is rolling along with a compelling narrative the source has diminishing influence,” Green and Brock wrote in their analysis. 

It’s plausible that something similar happened in our study. Although participants would have initially noticed the brand of the publication, as they progressed through the story, they may have been transported into the world of David and Devon. The characters’ progress and plight may have been foremost in their minds, not the political identity of the publication they were reading. In fact, although our story was presented as a magazine feature, at least one participant thought it was fiction. “Great story,” she commented. “I looked it up at the end, and was shocked to see it’s actually true.”

Slater agrees that our readers could have been so absorbed by the story as to be inattentive to credibility cues, though he cautioned that if the story had been more overtly political, it probably would have jarred readers out of their suspended reality. “If the persuasive content and intent is so obvious as to become more salient during processing than the narrative itself,” writes Slater, “the narrative may fail, and so should the persuasive effort.”

In a 2006 study in the Journal of Communication, Slater and several co-authors found that watching a TV drama about a politically charged issue like gay marriage or the death penalty was enough to influence viewers’ support or opposition to those issues, at least in the short term. Participants were first graded on where they fell along the political spectrum based on a questionnaire. Then half the group watched a Law & Order episode about a convicted murderer and rapist who could be extradited to a state with the death penalty. The second half watched a show about a woman who lost her home after her gay partner of many decades died because they weren’t recognized as a legal couple.

Afterwards, participants were questioned about public policy on the death penalty and gay marriage. Results showed that among those identified as liberals, participants who viewed the Law & Order episode were more supportive of the death penalty than those who hadn’t. Conservative viewers’ opinions on gay marriage remained constant whether they watched the show about the gay couple or not, but those who had seen it didn’t challenge the facts of the story, either.

Narrative can be persuasive in part because it awakens our emotions, and research has shown that emotions are critical to decision-making. “If you’re involved with a character and they go through something, it’s not just the experience you see them go through, it’s the feelings you’re having with them that affects the way you think about it,” says Slater.

 

Our findings come with a few caveats. Since our magazines were fictional, they did not have the brand resonance of familiar outlets like Fox News or MSNBC. Brand research has shown that previous experience with a brand is a large part of its effect, as are the subliminal associations attached to a brand. It’s plausible that simply telling someone that a publication is conservative or liberal would not trigger the kind of familiarity or suspicion that a real brand would.

Our story also differed from real-world experience in that our readers didn’t select it of their own volition. From research on media selection, we know that a lot of filtering happens at the stage when people are choosing what to read. According to what’s known as “selective exposure theory,” readers seek out sources that will reinforce existing beliefs, and are especially likely to do so along partisan lines. That means that even if they were to believe a story in a politically adverse publication, as our study suggests, they might never read it in the first place. In addition, by choosing to invest in a story, readers have already indicated a willingness to either trust its source or disregard it, while our readers were making those judgements retroactively.

A final caveat, but one that supports the notion that longform can reach across political lines, comes from Lotto of Matter Studios. “People are super credulous readers of longform,” he says. “I think people assume that someone who has written that many words has done a huge amount of work and has been backed up by fact-checking and reporting and thinking and editing.”

We’re all familiar with suspension of disbelief in fiction. For the duration of a movie or a book chapter, we agree to live in a world where we’ve colonized space, dogs can talk, or a boy with Muggle blood can save the world.

Our study suggests that this same principle extends to nonfiction stories that bend the rules, not of the physical world, but of our political worldviews. If it’s gripping enough, we’re willing to suspend judgement, if only for a little while.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Danny Funt, Chava Gourarie, and Jack Murtha are CJR Delacorte Fellows.