Preliminary Conference Programme

I develop and test a model in which regulator-verified, facility-level GHG disclosure realigns corporate incentives, shifting firms from strategic "greenwashing" toward substantive emissions abatement. Exploiting staggered firm-level exposure to the U.S. EPA’s Greenhouse Gas Reporting Program among a stable sample of firms with continuous voluntary disclosure, I find that treated firms reduce total organizational emissions by 15% through improved efficiency. These reductions emerge following the public release of verified data and extend significantly beyond regulated facilities. I further provide evidence on three mechanisms. First, managerial attention to climate issues increases, as reflected in earnings conference calls. Second, stakeholders discipline firms with larger discrepancies between voluntary and verified disclosures through negative abnormal returns, ESG-driven capital outflows, and environmental rating downgrades. Third, the sensitivity of firm value to emissions strengthens following verification. Overall, I demonstrate how bottom-up climate disclosure reshapes firm behavior by fundamentally altering the incentives underlying reporting and operational decisions.

We study stock market reactions to corporate announcements of emissions reduction targets. Using generative AI to screen news headlines, we identify 2,417 emissions target announcements from 2006 to 2023 and document three main findings. First, target announcements are associated with negative stock price reactions on average but with significant variations. Second, firms with more credible disclosures—those with higher environmental scores, reasonable assurance of emissions data, and a longer history of emissions reporting—receive more negative reactions. Third, firms with more negative market reactions subsequently reduce their emissions intensity more, consistent with the market correctly anticipating firms that will engage in costly decarbonization efforts. We identify two reasons why firms set targets and follow through despite negative market reactions: (i) benefits are realized later, as seen in less negative market reactions when climate regulation materializes; and (ii) ESG-linked pay incentivizes target announcement and subsequent decarbonization. Our findings shed light on how market values voluntary climate commitments.

Measuring Scope 3 carbon emissions (S3E) is complex and data intensive. We examine whether, and how, artificial intelligence—specifically, machine learning (AI-ML)—enhances firms’ ability to measure S3E for disclosure purposes. Using a novel measure of firms’ AI-ML adoption intensity based on job postings, we find that AI-ML adoption intensity is positively associated with voluntary disclosure of S3E. An instrumental variable design exploiting plausibly exogenous variation in talent supply supports a causal interpretation. We also find that AI-ML adoption intensity is positively associated with the quality of S3E disclosures. Path analysis reveals that these results are partly mediated by firms’ environmental value chain management practices, but that these mediation effects are secondary. Overall, our findings highlight the role of technology in enabling the disclosure of complex and difficult-to-measure carbon accounting metrics, offering insights for managers, regulators, and other stakeholders engaged in corporate climate accountability.