Trump vs social media

Note: I originally wrote this for the daily newsletter at the Columbia Journalism Review, where I am the chief digital writer

Donald Trump’s war on social media escalated recently after Twitter added a warning label to two of his tweets, with a link to a fact-check of the information he posted, and then blocked a third tweet with a message about violent content. Within days, Trump issued an executive order calling on the FCC to investigate whether social-media companies should lose the protection of Section 230 of the Communications Decency Act, which gives online platforms immunity for the content they host, because of what he claims are biased decisions about content. In contrast, Facebook has done nothing about Trump’s comments, despite the fact that a number of staffers have walked out to protest its lack of action—an unprecedented show of dissent for the company—and some have even quit their jobs. The executive order is widely viewed as legally dubious, but it is a convenient stick with which to threaten the social platforms. Will it work? Is that why Facebook has declined to take any action? Which approach is the right one, Twitter’s labelling or Facebook’s hands-off strategy?

To address these and other related questions, we used CJR’s Galley platform to host a virtual discussion with a group of journalists, legal analysts and other experts. Parker Molloy, editor-at-large for Media Matters, said the executive order is just another attempt to deal with what conservatives feel is a liberal bias in social-media companies. “Is there any evidence of this? No. We’ve done study after study after study on this topic, and there’s honestly no reason to believe there’s some sort of liberal/progressive bias at social-media companies. Conservatives are really just trying to ‘work the refs’ as a way to push these companies into adopting a pro-conservative bias.” Eric Goldman, a law professor at Santa Clara University and co-director of the High Tech Law Institute, said the order was aimed at internet companies—to discourage them from moderating conservative content—but was also a diversionary tactic to get the media to stop focusing on all the people who died from COVID-19. Even if the order has no actual legal effect, Goldman says, Trump “has likely accomplished his goals.”

Bridget Barrett, a researcher at the University of North Carolina, says that Facebook “had an opportunity here to clearly communicate what it would and wouldn’t tolerate, and for Trump, it looks like almost everything will be allowed. As someone who has spent the past couple months digging into the policies that platforms set for their users, this is incredibly frustrating. More importantly, as someone who wants our democracy to work, this is incredibly worrying.” And Errin Haines, editor-at-large for The 19th, a nonprofit news entity focused on gender politics, said “as a major source of information for a majority of people around the world and in our country, both of these platforms do have a responsibility to do no harm.” Jonathan Zittrain, a professor at Harvard Law School, made the broader point that “it’s strange that people like Mark (Zuckerberg, chief executive of Facebook) and Jack (Dorsey, chief executive of Twitter) have as much power as they do—whether to promote or squelch speech across billions of posts and users per day, including the power to do nothing.”

Continue reading

Twitter fact-checks Trump, but will it do any good?

Note: I originally wrote this for the daily newsletter at the Columbia Journalism Review, where I am the chief digital writer

When Twitter said earlier this month that it was making some changes to “limit the spread of potentially harmful or misleading content” by adding warning labels to tweets, one of the most obvious questions was whether the company would apply the labels to Trump’s various misinformed tweets. On Tuesday, we got the answer, when labels were added to two tweets posted to Trump’s account on the topic of mail-in voting ballots. The labels appeared just below the text of the tweet, with a hyperlink that said “Get the real story on mail-in voting.” The link took users to a collection of tweets with facts about the topic, curated by Twitter staff into what the company calls a Moment. The move was greeted with cheers in some quarters, while Trump and his supporters — including his son Donald Jr. — angrily tweeted about how the company was clearly trying to manipulate public opinion in advance of the federal election. Trump even promised to “strongly regulate, or close them down,” despite an obvious lack of any federal ability to do this.

On the one hand, it’s a tiny victory in the ongoing battle to stamp out misinformation online. After years of being accused of spreading lies, propaganda, and other noxious substances through its network, and of doing little to stop it, Twitter finally seems to be taking some baby steps towards responsible curation (Facebook said it would not take similar action because it believes that “people should be able to have a robust debate about the electoral process”). But at the same time, Twitter’s move is like taking a tiny drop of poison from a very large ocean and putting a label on it saying “for more information about poison, click here.” In fact, Twitter’s decision to add this kind of warning label raises as many questions as it answers. To take just one example, social researchers have found that fact-checking can cause what is known as the “implied truth effect,” where users assume that because one specific statement has been fact-checked and found to be false, others that haven’t been fact-checked must be true.

As a number of observers pointed out after Twitter added the label, the way that the company chose to phrase it was also imperfect. Saying “Get the facts about mail-in ballots” could be interpreted by some as adding weight to the misinformation rather than debunking it. Researchers such as Whitney Phillips of Syracuse University and Joan Donovan of Harvard’s Shorenstein Center often talk about the risks inherent in calling attention to misinformation, in part because doing so can amplify the incorrect info and cause it to travel much farther than it would have otherwise. Many wondered whether Twitter users would even bother clicking on the link, let alone read the facts in the Twitter Moment. Activist Charlotte Clymer said that Twitter’s label “is the most mild form of accountability” imaginable. The warning doesn’t say Trump is wrong or misleading people, she noted, and most people would probably scroll by without even noticing. “It’s weak and cowardly.”

Continue reading