How the Public Form Judgements of Online Hate Speech

• In Event: Free Speech, Censorship, & Self-Censorship

Abstract

What should be allowed to say about others on online platforms? Increasingly, major tech companies face pressures to self-regulate their platforms, while some governments introduce laws to counter (online) hate speech. Indeed, social media platforms struggle to consistently determine whether content is a legitimate post, or extremism inciting hatred and violence, which is amplified by the lack of consistency in legal frameworks across countries. One point of agreement, however, is that online political discussions are plagued by hate.

Yet what does qualify as hate in online speech? As a term, online hate speech continues to be used as a generic term in everyday language confounding incivility and the tone of debate with intolerant discourse outside the legal boundaries of freedom of expression. When assessing public demands for regulation of hate speech, researchers focus on abstract attitudes free of context, while real restrictions on freedoms of expression involves concrete judgements of who is being targeted, what is being said, and where the speech takes place. Understanding how people form judgements of online hate speech is key to understand how people perceive, navigate and react to other people online.

In this paper, I approach the debate on freedom of expression and regulation of online environments from a public opinion perspective and offer a comprehensive test of how ordinary citizens form judgements of online hate speech? I employ a paired vignette conjoint experimental design in two representative cross-national samples (n~3000) in low and high politically polarized countries (Denmark and United States respectively). In the experimental design, I ask respondents to rate the offensiveness of fourteen statements, whether the statement is, and ought to be illegal. The statements are generated varying seven attributes by random assignment, which makes it possible to gauge the relative causal effect of each attribute on whether a statement is perceived as offensive. These attributes are derived from legal and scholarly definitions concerning freedom of expression and regulation of online platforms. Specifically, I test how the target group, the type of target, severity, moral political issue, the sender’s gender and ethnicity, and whether the statement is expressed in a public or private setting affect judgements. The contribution of this paper is to show how ordinary citizens form judgements when they encounter controversial political statements on social media. Discussions on free speech and hate speech in online environments tend to focus on the balance between tech-companies and legislators which neglect the crucial role ordinary citizens on online platforms. Thus, this paper sheds new light on the public opinion and tolerance literature on free speech by moving from abstract to concrete judgements by ordinary citizens on what should be allowed to express about who on social media platforms.

Author

• Jesper Rasmussen, Aarhus University