Resolving content moderation dilemmas between free speech and harmful misinformation.

Area of Science:

• Social Psychology
• Communication Studies
• Digital Ethics

Background:

• Online content moderation balances freedom of expression with harm prevention.
• Public opinion on these moral dilemmas is largely unknown, hindering principled rule-making.

Purpose of the Study:

• To investigate public judgments and preferences regarding content moderation of harmful misinformation.
• To understand how factors like misinformation type, severity, user, and account features influence moderation decisions.

Main Methods:

• A conjoint survey experiment with 2,564 US respondents.
• Respondents evaluated scenarios involving election denial, antivaccination, Holocaust denial, and climate change denial posts.
• Decisions on post removal and account suspension were recorded, along with user and post characteristics.

Main Results:

• A majority favored removing harmful misinformation over protecting free speech.
• Post removal was preferred over account suspension; severity and repetition increased punitive action.
• Account-specific features (partisanship, followers) had minimal impact on decisions.
• Partisanship significantly affected moderation preferences, with Republicans less willing to moderate than Democrats or independents.

Conclusions:

• Public opinion generally supports stricter content moderation for harmful misinformation.
• Content moderation policies should consider the severity and recurrence of misinformation.
• Partisan divides in content moderation preferences necessitate careful consideration in policy design.