Name
Linguistic Signals of Climate-related Disclosures and Corporate Enviornmental Misconduct
Date & Time
Monday, July 6, 2026, 9:50 AM - 10:15 AM
Description
This study examines whether the linguistic features of climate-related risk disclosures are associated with subsequent corporate environmental misconduct. Using a sample of S&P 500 firms from 2000 to 2024, we extract climate-related risk narratives from firms’ 10-K filings using a fine-tuned ClimateBERT framework and measure disclosure sentiment and readability. We find that both more optimistic sentiment and greater readability in climate-related risk disclosures are positively associated with one-year-ahead environmental misconduct, suggesting that such disclosures may function as symbolic communication rather than substantive transparency. Cross-sectional analyses show that this association is weakened by carbon assurance and R&D investment, consistent with these mechanisms constraining managerial discretion and enhancing substantive environment commitment. In contrast, the association is amplified among financially distressed firms, indicating stronger incentives for impression management under economic pressure. Overall, the findings suggest that the linguistic features of climate-related risk disclosures contain incremental information about disclosure credibility and future environmental conduct. This study contributes to the literature on climate-related disclosure, disclosure decoupling, and corporate misconduct, and informs ongoing regulatory debates on the effectiveness of voluntary climate reporting frameworks.
Kaiying Ji
Keywords
Climate-related risk disclosure · Corporate environmental misconduct · Textual analysis · Sentiment · Readability
Theme
CSR
Author 1
Kaiying Ji
Author 2
Jin Liu
Author 3
Baljit Sidhu
Author 4
Chuan Yu