Twitter Says Changes Curbed Spread of Election Misinformation
(Bloomberg) -- Twitter Inc. said product changes made ahead of the U.S. election, including one that made it more complicated to share other people’s posts, diminished the spread of misinformation on the social-media service.
The San Francisco-based company on Thursday said it put labels on 300,000 user posts from Oct. 27 to Nov. 11 for violating its rules around election misinformation. Twitter also said changes to its retweet feature ahead of the Nov. 3 vote were useful in limiting the spread of confusing or misleading posts.
Before users could pass along a message, the new tool asked them to provide their own commentary, a process known as quote tweeting. Twitter said quote tweets, aimed at encouraging people to share more context on posts, were up 26% as a result of the prompt, and regular retweets were down 23%. Overall, quote tweets and retweets combined were down 20% in total, the company said in a blog post.
“This change slowed the spread of misleading information by virtue of an overall reduction in the amount of sharing on the service,” Twitter said, though it didn’t provide specifics on how much misinformation was prevented from proliferating. The company plans to leave the prompt in place while it takes “more time to study and fully understand the impact.”
Both Twitter and Facebook Inc. were under pressure to curb the spread of misinformation on their services during the election. Both social networks labeled misleading posts about the election outcome, and created rules that forbid users from trying to delegitimize election results.
They also both implemented product changes to slow down viral misinformation. In addition to nudging people to quote tweet instead of retweet, Twitter stopped putting recommendations for tweets and users to follow directly into people’s feeds. It said that effort didn’t lead to a “statistically significant difference in misinformation prevalence,” and will be reversed beginning Thursday.
Facebook made changes as well, including a temporary political advertising ban immediately after the polls closed on election night. That ban has been extended for as much as a month to prevent candidates from taking out ads intended to dispute the results, which President Donald Trump has been doing regularly. Twitter has long disallowed political or issue ads.
Twitter said that while it saw record levels of election-related discussions on its platform, the 300,000 posts it labeled for misinformation accounted for just 0.2% of all election-related posts on the service between Oct. 27 and Nov. 11. Facebook said this week that political content made up just 6% of all content people see on the service.
©2020 Bloomberg L.P.