WASHINGTON—Sharing false information on Facebook is old.
People over 65 and ultra conservatives shared about seven times more fake information masquerading as news on the social media site than younger adults, moderates, and super liberals during the 2016 U.S. election season, a new study found.
The first major study to look at who is sharing links from debunked sites found that not many people are doing it.
On average, only 8.5 percent of those studied—about one person out of 12—shared false information during the 2016 campaign, according to the study in yesterday's journal, Science Advances.
But those doing it tend to be older and more conservative.
“For something to be viral, you've got to know who shares it,” said study co-author Jonathan Nagler, a politics professor and co-director of the Social Media and Political Participation Lab at New York University.
“Wow, old people are much more likely than young people to do this,” he noted.
Facebook and other social media companies were caught off-guard in 2016 when Russian agents exploited their platforms to meddle with the U.S. presidential election by spreading fake news, impersonating Americans, and running targeted advertisements to try to sway votes.
Since then, the companies have thrown millions of dollars—and thousands of people—into fighting false information.
Researchers at Princeton University and NYU in 2016 interviewed 2,711 people who used Facebook. Of those, nearly half agreed to share all their postings with the professors.
The researchers used three different lists of false information sites—one compiled by BuzzFeed and two others from academic research teams—and counted how often people shared from those sites.
Then to double check, they looked at 897 specific articles that had been found false by fact-checkers and saw how often those were spread.
All those lists showed similar trends.
When other demographic factors and overall posting tendencies are factored in, the average person older than 65 shared seven times more false information than those aged 18-29.
The seniors shared more than twice as many fake stories as people aged 45-64 and more than three times that of people in the 30- to 44-year-old range, said lead study author Andrew Guess, a politics professor at Princeton.
The simplest theory for why older people share more false information is a lack of “digital literacy,” said study co-author Joshua Tucker, also co-director of the NYU social media political lab.
Senior citizens may not tell truth from lies on social networks as easily as others, the researchers indicated.
Harvard public policy and communication professor Matthew Baum, who was not part of the study but praised it, said he thinks sharing false information is “less about beliefs in the facts of a story than about signalling one's partisan identity.”
That's why efforts to correct fakery don't really change attitudes and one reason why few people share false information, he noted.
When other demographics and posting practices are factored in, people who called themselves very conservative shared the most false information, a bit more than those who identify themselves as conservative.
The very conservatives shared misinformation 6.8 times more often than the very liberals and 6.7 times more than moderates.
People who called themselves liberals essentially shared no fake stories, Guess noted.
Nagler said he was not surprised that conservatives in 2016 shared more fake information, but he and his colleagues said that doesn't necessarily mean that conservatives are, by nature, more gullible when it comes to false stories.
It simply could reflect that there was much more pro-Trump and anti-Clinton false information in circulation in 2016 that it drove the numbers for sharing, they said.
However, Baum said in an e-mail that conservatives post more false information because they tend to be more extreme, with less ideological variation than their liberal counterparts.
And they take their lead from U.S. President Donald Trump, who “advocates, supports, shares, and produces fake news/misinformation on a regular basis.”
The researchers looked at differences in gender, race, and income but could not find any statistically significant differences in sharing of false information.