NEW YORK: A few weeks ago, about 10,000 copies of a fake edition of The Washington Post newspaper were handed out in the US capital.
Under the eye-catching headline “Unpresidented — Trump hastily departs White House, ending crisis”, the story declared that the president had resigned after scribbling a departure note on a White House napkin.
This particular piece of fake news was not really intended to deceive: Once you read the “paper”, it became clear it was a prank (it was, for example, dated May 1, 2019 and every page was filled with anti-Trump stories). A media-activist group, the Yes Men, later admitted responsibility.
This episode was an exception, though. Whereas it is easy to see a “fake” physical newspaper being handed out, in cyberspace it is often impossible to assess the provenance, veracity or distribution of a story.
So what can be done to halt the proliferation of fake news online? Last month, two separate academic studies were published that investigated the effect of social media and fake news during the controversial 2016 US election campaign. The results are fascinating — and often surprising.
The first, and most important, conclusion from both sets of research is that far fewer people are consuming or sharing fake news than we think. Never mind that the US president keeps tossing the phrase around, or that many Democrats still believe the principal reason Hillary Clinton lost to Trump in 2016 was that Russian trolls (and far-right activists) circulated a succession of anti-Clinton stories.
A TINY MINORITY
The research team from Northeastern University tracked more than 16,000 Twitter accounts in the 2016 election campaign, matched to voter profiles, to see whether they shared items that were obviously fake (i.e. demonstrably false, rather than slightly in dispute).
They concluded that while “fake news accounted for 6 per cent of all (overall) news consumption”, this was deeply skewed by hyperactivity from a tiny minority. A mere 1 per cent of users accounted for 80 per cent of fake news consumption, and just 0.1 per cent were circulating this material.
The second team of academics, from New York University and Princeton, looked at Facebook accounts.
“Using unique behavioural data on Facebook activity linked to individual-level survey data, we find ... that sharing fake news was quite rare during the 2016 US election campaign,” the authors observed in the journal Science Advances.
This echoes yet another study released last year by researchers at Dartmouth, Princeton and Exeter universities, which found that, although “one in four Americans” visited a fake news site during the run-up to the US election, a mere 10 per cent accounted for 60 per cent of the visits, meaning that “fake news consumption was heavily concentrated among a small group”.
What was notable in both the Northeastern and New York/Princeton studies was the profile of that small group of fake-news consumers. Both teams found that those most affected by fake news tended to be politically conservative, with some bias towards being male and white.
More strikingly, they were overwhelmingly older.
OLDER PEOPLE SHARES MORE FAKE NEWS
These days it is fashionable to fret about how teenagers and millennials are being manipulated by social media.
But the study of Facebook usage adjusted to account for its higher age profile concluded that people over the age of 65 were “sharing nearly seven times as many articles from fake news domains on Facebook as those in the youngest age group, or about 2.3 times as many as those in the next-oldest age group, after adjusting for ideology, education and differing levels of social media usage”.
The Twitter analysis echoed this finding (perhaps offering some explanation for the behaviour of the 72-year-old tweetstorm that is Donald Trump).
What are the reasons for this? Last week, I listened to academics debating the issue at a Silicon Valley conference, where a number of theories were put forward. One was that the older generation has “reduced cognition” and/or more fixed political prejudices; another that older people, who are new to the internet, are less savvy — and less cynical — than digital natives.
Either way, the findings of the research studies have two key implications. First, if we want to counter the democratic threats posed by “fake news”, it may not be smart to use a mass-market campaign.
If everyone shrieks about the danger of fake news, this can cut trust in all news.
More effective might be to send targeted “nudges” to those known to be hyperactively consuming or sharing fake news, or to label these news items “dangerous”. “We should treat it like tobacco,” suggests one academic who wants tech companies to put “health warnings” on some sites (presuming, of course, that everyone can agree on what is “fake”).
Second, we need to rethink who we are trying to educate online. Yes, it is important to warn teenagers about cyber manipulation but we must remember to keep an eye on their grandparents too.
While silver surfers do not usually hog the limelight, they seem to matter when it comes to fake news. You cannot always blame the kids.
© 2019 The Financial Times Ltd. All rights reserved.