THE BENEFITS AND BURDENS OF HIGH REPUTATION DURING DISRUPTIONS: THE ROLE OF MEDIA REPUTATION, ORGANIZATIONAL IDENTIFICATION, AND DISRUPTION TYPE

Organizational researchers are increasingly interested in the role of social approval assets, such as reputation and celebrity, for financial success of organizations. In this three-essay dissertation I examine the role of these assets when an organization is involved in negative disruptive events. In Essay 1, I introduce four media generated organizational types: celebrity, infamous, peripheral, and unfamiliar organizations and develop a theoretical framework and propositions that examine how stakeholder decisions whether or not to transact with an organization after disruptions depend on the type of organization under examination. In Essay 2, I argue theoretically and find empirically that stakeholder reactions to disruptions depend on the level of organizational identification. On a sample of on-campus murders in U.S. colleges and universities in 2001-2009, I find that universities receive fewer applications after murders, and this effect is stronger for ranked universities. Additionally, percentage of alumni donating to schools increases after on-campus murders, but only in ranked universities. I test the robustness of these findings using different operationalizations of disruptions and stakeholder groups. The results indicate that reputation is a liability during disruptions when stakeholders under examination have low levels of organizational identification and reputation is a buffer for reactions by high-identification stakeholders. In Essay 3, I argue that the amplifying role of organizational reputation is due to differences in news coverage of disruptions in high-reputation compared to low-reputation organizations. The results of empirical analysis of news coverage of 106 on-campus murders indicate that even after controlling for the characteristics of the event, disruptions in high-reputation organizations receive more coverage. I further examine this finding using content analysis of articles that covered four pairs of similar murders that took place in