Misinformation interventions are common, divisive, and poorly understood

Abstract

Social media platforms label, remove, or otherwise intervene on thousands of posts containing misleading or inaccurate information every day. Who encounters these interventions, and how do they react? A demographically representative survey of 1,207 Americans reveals that 49% have been exposed to some form of online misinformation intervention. However, most are not well-informed about what kinds of systems, both algorithmic and human, are applying these interventions: 40% believe that content is mostly or all checked, and 17.5% are not sure, with errors attributed to biased judgment more than any other cause, across political parties. Although support for interventions differs considerably by political party, other distinct traits predict support, including trust in institutions, frequent social media usage, and exposure to “appropriate” interventions.