Purpose We gathered daily diary data over three consecutive weeks from 109 employees under 13 mid-level officers, yielding 950 day-level observations. We applied multilevel modeling to test fear's mediating role and proactive personality's moderating role. Design/methodology/approach This study examines how abusive supervision spills over to third-party observers, shaping defensive silence and defensive voice through fear and explores whether proactive personality moderates these effects. Findings Observing coworkers' abuse heightened bystanders' fear, which in turn led to self-protective communication: defensive silence or defensive voice. Fear fully accounted for why witnessing abuse drove these behaviors. Additionally, less proactive observers displayed stronger fear ? Silence links, whereas more proactive observers channeled fear into voice, though still defensively. Originality/value By focusing on third-party observers, this research extends the abusive supervision literature beyond direct victims and integrates emotional process and avoidance–approach frameworks to clarify how and why fear prompts defensive communication. Proactive personality emerges as a key boundary condition in fear-based responses.
Cao et al. (2026) studied this question.