The Impact of Tightly Contested Governance Proposals on Firms' Narrative Disclosures: Evidence from a Regression-Discontinuity Design (RDD)

Abstract

Corporate governance and firm disclosure are endogenously determined. We exploit locally exogenous variations in corporate governance created by “close-call” governance-related shareholder proposals, using a fuzzy RDD and text analytics to examine whether better corporate governance causally affects the narratives in corporate disclosures. We find that although better corporate governance in firms leads to more disclosure in their 10-K filings, the passage of “close-call” governance proposals also significantly increases the complexity and the boilerplate nature of such disclosures. These results are robust to several robustness tests, alternative RDD bandwidths, and different specifications, and amplified when the investors are undistracted.

Keywords: Corporate governance, Textual Disclosure, Endogeneity, Regression Discontinuity Design (RDD), Computational Linguistics, Investor Distraction

JEL Classification: G23, G30, G34