Project escalation is known to frequently occur in the context of information systems (IS) projects. The reluctance to hear bad news—a phenomenon that has been labelled the “deaf effect”—has been suggested as a possible reason for why projects are allowed to escalate for as long as they sometimes do. The deaf effect response to bad news reporting has received little research attention, yet may account for many cases of project escalation. The research reported here provides a description of conditions under which the deaf effect is likely to occur. Hypotheses regarding factors involved in causing the deaf effect are articulated based on Miceli and Near's theory of whistle blowing effectiveness and further elaborated using insights from the cognitive psychology literature of decision-making. The hypotheses were then tested experimentally using a role-playing experiment. Results suggest that when a decision maker perceives a relevant message, s/he is willing to de-escalate the project. Bad news reporter credibility and the gender of the bad news reporter were found to be key factors in the determination of message relevance.