It is a truism now that our information channels are tainted -- sometimes poisoned -- with the false or misleading. Enter fact-checking, or the efforts of news outlets and organizations dedicated to setting the record straight. But does it work? Can fact-checking pull the poison from the wound and stop mis- or disinformation from wreaking havoc on the body politic?
There is no doubt the fact-checking movement is growing. According to the International Fact-Checking Network (IFCN) -- a unit of the U.S.-based Poynter Institute for Media Studies -- just 44 fact-checking initiatives existed worldwide in 2014. Numbers spiked to 188 in 60 countries by 2019, and 290 in 83 countries by 2020, with professional news organizations making up a large chunk of new entrants -- a surge attributed to "global concerns about the spread of misinformation." Japan has FactCheck Initiative Japan (FIJ), which is not an IFCN signatory but does cooperate with the organization to promote local fact-checking.
Despite this enthusiasm, recent research suggests there is no black-and-white answer to how effective all this is. Formal fact-checking and debunking generally focuses on public speech and information from major figures like politicians, and on claims tearing around on social media (the topic of Part 2 of this series). And though it helps, it also runs up against our desire to have preexisting views confirmed, while information gatekeepers like journalists need to be careful about exactly what and how they fact-check.
Japan, for one, appears to have a fact-checking deficit. Dedicated fact-checking is growing as a movement, as exemplified by this and other newspapers' membership in FIJ, but it is playing catch-up with the best practices of more seasoned and established fact-checking organizations overseas.
"First of all, fact-checking needs to be done," said Hiroyuki Fujishiro, a professor of journalism at Hosei University in Tokyo. He called the global surge in fact-checking a positive development, "something that should be done as a counter to fake news." But, he emphasized, "we need to be careful about how it's done ... or (fact-checks) can be used in attacks" by political partisans.
Fujishiro and two co-authors conducted a study on the impact of local newspaper fact-checking during the 2018 Okinawa gubernatorial election, by analyzing the Twitter response to the articles. While their data collection technique meant the sample size was somewhat small, the results suggested that the fact-check articles on candidates' statements and election materials were not picked up by the public as information correctives, but rather as cudgels against opponents on social media.
A different study conducted in the United States reached similar conclusions. The 2019 meta-analysis of 30 experimental studies on the effectiveness of fact-checking in correcting political misinformation, with a total sample size of nearly 21,000, strongly suggested that fact-checks had the biggest impact when debunking the views of one's opponents. While not zero, factual corrections of people's own beliefs were far less effective -- an example of "motivated reasoning."
"There's absolutely no doubt that, keeping everything else equal, exposure to fact-checking helps," says the study's lead Nathan Walter, an assistant professor in the Department of Communication Studies at Northwestern University in Chicago. "It makes people's beliefs more accurate. How much more accurate? Not very much, but it helps."
On the other hand, Walter says, people want to be consistent in their political beliefs. "Motivated reasoning is all about belonging, about affirming that you're part of the group, affirming that your ingroup is good and the outgroup is bad."
Another study led by Walter manipulated the uncertainty of information, based on the hypothesis that this would prompt people to turn to fact-checkers more. "But what we found was that uncertainty has no effect. What really influences people's intent to fact-check is when they think that it will support their position. So, people fact-check mainly to affirm (their views) as opposed to confirm or assess the veracity of something."
Complicating matters is that fact-checking is not as easy as dividing the world into true and false. Fact-checkers are limited to what is checkable, Walter points out. When it comes to a political speech, for example, checkable nuggets may make up only a small part of a much longer presentation, interwoven with opinion. Furthermore, the claims may be only partially false, or partially true, and fact-checking's effectiveness falls in these less clear-cut cases.
Furthermore, what Walter says he found "troubling" in his team's meta-analysis was that the more the design of the experiments covered in the paper resembled real-world encounters with fact-checks, the less impact those checks had on the subjects, and therefore the less power they had to dispel misinformation.
Is it the responsibility of professional fact-checking organizations and journalists to address this? Not necessarily, says Walter. It depends on how they view their role. If journalists see their job as simply keeping an accurate record, then how their efforts are received could be of less concern.
Trust in fact-checking also hinges, however, on how it's done. The IFCN has a Code of Principles that commits its signatories to non-partisanship and fairness, set standards, transparency of sources, funding, organization, and methodology, as well as an open corrections policy. Japan's FIJ follows the same code.
Fujishiro's Okinawa paper notes that the local newspapers were not following IFCN best practices, for example by checking the gubernatorial election manifesto of the Liberal Democratic Party-backed candidate Atsushi Sakima, but not that of his opponent who ultimately won, Denny Tamaki. Tamaki won on a platform of opposing the relocation of a U.S. Marine Corps air station within Okinawa, a move the national LDP insists must happen and which the LDP-backed candidate avoided discussing during the campaign.
This kind of focus, says Fujishiro, can lead people to "view the news media as politicized," adding, "Partisan fact-checking links directly to distrust in the media. So news outlets need to take a good step back, even if they like or dislike a politician, and do a properly balanced job."
He also noted that "Japan has got into fact-checking late," and that there are no fact-checking organizations in the country with the same experience and track record as some overseas, like Snopes.com or FactCheck.org in the U.S. "There isn't the same know-how here," he said. Japan's news outlets face a steep learning curve.
To boost trust and improve selection of what gets fact-checked, Northwestern's Walter suggests that the practice needs to be more interactive and willing to accept suggestions from the public who checkers aim to serve, especially in our social media-saturated world.
However, journalists and news organizations have a lot of experience fact-checking their own content, he adds, "so they need to be part of the solution (for misinformation). I'm just not sure that fact-checking the way that we're doing it right now is the solution."
(By Robert Sakai-Irvine, The Mainichi Staff Writer)
This is Part I of a two-part series. Next: The perils of debunking social media misinformation.
Disclosure: Robert Sakai-Irvine is an adjunct lecturer in journalism and media literacy at Hosei University. However, he is not involved in any way with associate professor Fujishiro's work, and had no relationship with him before reporting for this article.