Disclosure Substitution

Abstract

This study develops and tests a simple model of voluntary disclosure in which managers can choose to withhold (i.e., redact) certain elements from mandatory disclosure. We consider a setting in which mandatory disclosure is a disaggregated disclosure (e.g., a financial statement), voluntary disclosure is an aggregate disclosure (e.g., an earnings forecast), and the costs of each type of disclosure are distinct. In this setting, we show that managers endogenously substitute between the two types of disclosure; managers that choose to withhold information from mandatory disclosure are more likely to provide voluntary disclosure. We test our predictions using a comprehensive sample of mandatory disclosures in which the SEC allows the firm to redact information that would otherwise jeopardize its competitive position. Consistent with our predictions, we find strong evidence that redacted mandatory disclosure is associated with greater voluntary disclosure. This paper was accepted by Suraj Srinivasan, accounting.