The Impact of Twitter Labels on Misinformation Spread and User Engagement: Lessons from Trump's Election Tweets

Social media platforms are performing "soft moderation"by attaching warning labels to misinformation to reduce dissemination of, and engagement with, such content. This study investigates the warning labels that Twitter placed on Donald Trump's false tweets about the 2020 US Presidential election. It specifically studies their relation to misinformation spread, and the magnitude and nature of user engagement. We categorize the warning labels by type -"veracity labels"calling out falsity and "contextual labels"providing more information. In addition, we categorize labels by their rebuttal strength and textual overlap (linguistic, topical) with the underlying tweet. We look at user interactions (liking, retweeting, quote tweeting, and replying), the content of user replies, and the type of user involved (partisanship and Twitter activity level) according to various standard metrics. Using appropriate statistical tools, we find that, overall, label placement did not change the propensity of users to share and engage with labeled content, but the falsity of content did. However, we show that the presence of textual overlap in labels did reduce user interactions, while stronger rebuttals reduced the toxicity in comments. We also find that users were more likely to discuss their positions on the underlying tweets in replies when the labels contained rebuttals. When false content was labeled, results show that liberals engaged more than conservatives. Labels also increased the engagement of more passive Twitter users. This case study has direct implications for the design of effective soft moderation and related policies.