Consensus on community guidelines: an experimental study on the legitimacy of content removal in social media

Abstract

The popularization of social media has led to a considerable increase in the importance of discursive expressions of violence, especially when directed at vulnerable communities. While social media platforms have created rules to regulate such expressions, little information is available on the perception of the legitimacy of these rules in the general population, regardless of the importance of the former for the latter. It is therefore the objective of this study to analyze the perception of the seriousness of such content and the degree to which the population has established a consensus on the withdrawal of restricted discursive behaviour on three major social media platforms (Facebook, Instagram and Twitter). For this purpose, 918 participants were immersed in an experimental paradigm in three different groups (n1 = 302; n2 = 301; n3 = 315). Each was presented with stimuli containing discursive behaviour that is banned by community guidelines. The stimuli were presented differently to each group (i.e., description of the banned behaviour, description and accompanying example, example only). Our experimental data reveals that the degree of consensus on the need to remove content is quite high, regardless of the style of presentation. It furthermore suggests that the behaviour in question is perceived as very serious, due to the harm that our participants presume it to cause. These results have important implications for the debate on freedom of expression on the Internet and its regulation by private actors.