Dimensions of Harm in Hate Speech Fueled by Public Health Crises: Evidence from Anti-Asian Hate During COVID-19 (Preprint)

BACKGROUND Public health crises and hate speech are linked by a bidirectional relationship: crises provoke public fear and uncertainty, driving hateful and blaming discourse that disproportionately targets marginalized communities; while the resulting hate speech produces its own health consequences, including psychological harm, physical violence, and erosion of healthcare trust. The COVID-19 pandemic is a prominent example with the surge in both hate speech and hate crime alongside the health emergency. Despite growing studies of hate speech and hate-motivated behaviors in public health contexts, existing approaches treat it as a binary phenomenon and tend to overlook the varying degrees of harm across instances and communities. OBJECTIVE This study aims to empirically capture how individuals perceive and evaluate harm in hate speech in case of anti-Asian hate speech during COVID-19. We develop a framework to capture the multi-dimensional nature of harm in hate-motivated behaviors to inform evidence-informed strategies for harm prevention and reduction. METHODS We surveyed a diverse group of 303 participants to evaluate a set of anti-Asian tweets posted during the COVID-19 pandemic. Participants rated the harmfulness of some random content and identified factors that made certain speech more damaging than others. Through coding of participants’ short answers, we inductively summarized key dimensions people emphasize in evaluating harm. RESULTS Our analysis revealed six interrelated dimensions of harm: harm reflected in expression, proposition, and dissemination, as well as harm caused to individuals, targeted communities, and society. Participants most often mentioned harm in expression (87.1%), followed by harm in proposition (60.1%), harm to individuals (52.1%), harm to targeted community (46.9%), harm to society (39.6%), and harm in dissemination (25.7%). Participants who had personally encountered hate speech mentioned marginalization (p=.031) and the ability of hate speech to unite hostility (p=.039) more often than others. These dimensions together show how the perceived harmfulness is not only shaped by what is said, but also by how it is communicated and perceived. CONCLUSIONS This study highlights that hate speech and public health crises are mutually reinforcing, and that binary categorization of hate speech fails to capture the full scope of harm this relationship produces. The proposed framework can inform more effective responses to hate-motivated behaviors, violence prevention, and platform governance to protect community safety.