Bad News Bear(er)s: FDA Inspection Outcomes and Managers' Voluntary Disclosure Choices

Abstract

This study examines the association between bad news and managers' disclosure choices. It is empirically challenging to investigate this relationship because bad news resides in managers' private information sets, which are often unobservable and difficult to measure. To overcome this obstacle, I use a proprietary dataset that documents the dates and outcomes of FDA inspections. With the ability to detect the existence and content of news, I find that bad inspection outcomes are associated with a higher probability of disclosure and higher quantities of disclosure in the following three months. These results are stronger when inspection outcomes are more material or more severe. Further, the relation between bad news and disclosure disappears when the FDA begins publicly disclosing outcomes on a monthly basis. My timeliness tests suggest that managers accelerate, as opposed to delay, the disclosure of bad news. I also explore incentives that may influence managers' choices and provide direct evidence supporting the importance of litigation risk in managers' disclosure decisions; however, I do not find any evidence of managerial self-dealing. Collectively, these results document a significant link between private information and voluntary disclosure: managers disclose – rather than withhold – bad news, and litigation risk functions as a key motivating mechanism.