Abstract
Incivility in online commenting sections can create a hostile environment and result in the silencing of vulnerable voices. Accordingly, content websites and social media platforms have an ethical responsibility-one that aligns with their strategic interests-to minimize users' exposure to uncivil content. To this end, platforms invest great effort and budget in automatic and manual filtering mechanisms. Yet, these efforts create a competing ethical quandary, as they often come at the expense of free expression, particularly in cases where comments do not explicitly infringe on stated guidelines but might nevertheless be interpreted as offensive. In this paper, we examine an alternative moderation approach, based on comment reordering as opposed to deletion of uncivil comments. Specifically, we show that exposure to uncivil (vs. civil) comments located in the head or at the bottom of a list of comments increases subsequent commenters' likelihood of posting uncivil comments themselves. Exposure to uncivil comments in the middle of a list, however, does not significantly enhance commenters' likelihood of commenting uncivilly. These results offer new theoretical insight into how incivility is transferred between users in online environments. Our results also suggest a straightforward technological solution for mitigating online incivility, which is more ethical and practical than current industry standards. This involves placing civil comments at the beginning and end of the comment thread, with uncivil comments located in the middle.